Author Topic: Some questions about my first drsstc.  (Read 4414 times)

Offline Mathieu thm

  • High Voltage Enthusiast
  • *
  • Posts: 43
  • Karma: +3/-0
    • View Profile
Some questions about my first drsstc.
« on: August 22, 2022, 10:27:32 AM »
Hello,
First of all sorry for my bad english, I am French
My name is Mathieu I'm 18 and I make my first drsstc, i wish i could tune my tesla coil better
and I have some questions.

First the setup:
     -controller : UD 2.7 rev c
     -IGBT : HGTG30N60A4D
     -OCD : 350 A (can I set it higher ?) 315 A
     -MMC : 10kv 60nf
     -CT : n87 (400:1 with two 20:1)
     -Secondary frequency : 210khz (10x30cm ; 0,25mm wire)
     -Primary frequency : 204khz 190khz
     -interrupter : up to 450hz and up to 100us (I ordered and am waiting for the famous Midistick!)


Here are some pictures :








During my first lightning test I obtained 50-60 cm of lightning (primary at 213 khz and secondary at 210.3khz) but the ocd at 350a was triggered at 100 Vac
I lower the frequency of the primary to 204 khz and I get 60-70cm of lightning and the ocd is now triggered at 150Vac.
I also noticed that the tesla coil draw less power at the input.
Do I have to lower the frequency of the primary much more? Are there any factors to consider?


In the next part the bus voltage is 30v which comes from a lab power supply because I don't have a differential probe

I tried to adjust phase lead as best as I can (7m3 153 inductor), here is the signal, in yellow the output of the bridge and purple the primary current:


What bothers me is that we can see that the beginning of the signal is quite good but towards the end there is a huge spike,
here is another screenshot:



And finally, here is a photo of the wires that go from the gdt to the igbt (around 20-30cm), are they a little too long?


« Last Edit: October 17, 2022, 06:35:19 PM by mthome4 »
Mathieu Thomé

Offline Mads Barnkob

  • Administrator
  • Executive Board Member
  • *****
  • Posts: 2085
  • Karma: +58/-0
  • Denmark
    • View Profile
    • Kaizer Power Electronics
Re: Some questions about my first drsstc.
« Reply #1 on: August 22, 2022, 01:23:53 PM »
Hi Mathieu  and welcome to HVF! Your English is fine, do not worry about that!

Did you run your numbers through JavaTC for tuning point? I usually takes what it suggests and detune the primary to a 5% lower frequency. That gives for the loaded secondary change in frequency.

Your ZCS looks fine and the huge spike is at your turn-off, you can see the phase reversal as the energy flows back into the DC bus.

Your GDT wires are not longer than what I have used before, so I do not see a problem there.

350A OCD is suitable for a full-bridge of TO-247 switches. I blew mine up at 500A :)
https://kaizerpowerelectronics.dk - Tesla coils, high voltage, pulse power, audio and general electronics
https://www.youtube.com/KaizerPowerElectronicsDk60/join - Please consider supporting the forum, websites and youtube channel!

Offline Mathieu thm

  • High Voltage Enthusiast
  • *
  • Posts: 43
  • Karma: +3/-0
    • View Profile
Re: Some questions about my first drsstc.
« Reply #2 on: August 22, 2022, 03:54:23 PM »
Quote
Hi Mathieu  and welcome to HVF! Your English is fine, do not worry about that!
Thanks ! ;D

Thank you for all this information, and thank you for making me understand about the spike, by the way the huge spike is not too dangerous?

Quote
Did you run your numbers through JavaTC for tuning point? I usually takes what it suggests and detune the primary to a 5% lower frequency. That gives for the loaded secondary change in frequency.

This is what javaTC says :


I detune by 5% compared to the primary frequency given by javaTC or compared to the secondary frequency given by javaTC?

What was your igbt that you exploded?

« Last Edit: August 22, 2022, 06:33:13 PM by mthome4 »
Mathieu Thomé

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2004
  • Karma: +103/-1
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Some questions about my first drsstc.
« Reply #3 on: August 23, 2022, 06:53:44 AM »
Looks like a great clean build.  Congratulations on getting it working.

Quote
OCD : 350 A (can I set it higher ?)
Looks like a good value for those IGBTs.  Does depend on Vge in your build (UD2.7 supply voltage) and on your tolerance for risk of fried IGBTs.  HGTG30N60A4D datasheet short-circuit current graph is continuing up at it's 15Vge endpoint.  So it might be possible to push beyond 350A for short bursts with 20 or 24 Vge.

Quote
Do I have to lower the frequency of the primary much more? Are there any factors to consider?
Test lower primary frequency to see how it works.  Scoping primary current is possible at full power, as is scoping secondary current.  That can show what's happening even w/o being able to scope H-bridge outputs.

Quote
What bothers me is that we can see that the beginning of the signal is quite good but towards the end there is a huge spike,
here is another screenshot:
Hard to say what the spike will look like at full H-bridge voltage.  However, I'd explore it's cause a bit more at 30V.  It is a very short spike.  Zoom in to see what it looks like in detail.  Also scope gate voltages.  HGTG30N60A4D are fast IGBTs (good), so may respond to short Vge glitches.

Spikes also depend on H-bridge layout and parasitic inductances.  Since your normal switching looks good, parasitic inductance is likely low enough.

Quote
And finally, here is a photo of the wires that go from the gdt to the igbt (around 20-30cm), are they a little too long?
Do you happen to have a photo of GDT itself?  Leads are very nicely twisted together.  I presume GDT primary leads are also twisted together.
David Knierim

Offline JCF

  • High Voltage Experimenter
  • **
  • Posts: 76
  • Karma: +2/-0
    • View Profile
Re: Some questions about my first drsstc.
« Reply #4 on: August 23, 2022, 10:27:26 AM »
Belle réalisation Mathieu ! bienvenue sur le forum

Offline Mathieu thm

  • High Voltage Enthusiast
  • *
  • Posts: 43
  • Karma: +3/-0
    • View Profile
Re: Some questions about my first drsstc.
« Reply #5 on: August 23, 2022, 05:45:22 PM »
Quote
Looks like a great clean build.  Congratulations on getting it working.
Thanks ! ;D
Quote
Belle réalisation Mathieu ! bienvenue sur le forum
Merci !  ;D

I tried to reduce the resonance frequency of the primary and I get more and more impressive results.
I will continue to reduce the frequency until I have the best possible results.
We can note that the better the drsstc is tune, the less the OCD is triggered, now I can go up to 190-200Vac without the ocd being triggered (with the same 350a OCD).
I still hadn't put a video so here is a video of the drsstc in operation ! :
/>And Some pictures:

lightning reaches 80 cm, maybe more

Quote
Does depend on Vge in your build (UD2.7 supply voltage) and on your tolerance for risk of fried IGBTs.
The us 2.7 is supplied with a 24V switching power supply.
What do you mean by "risk of fried IGBTs." ? I don't understand  :)

Quote
Zoom in to see what it looks like in detail.
I'll tell you all about it soon :)

Otherwise I noticed that the more I put a big dummy load, the more the spike shrinks, so I assume that with the secondary, and well adjusted, maybe the spike shrinks ?


« Last Edit: August 23, 2022, 05:54:43 PM by mthome4 »
Mathieu Thomé

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2004
  • Karma: +103/-1
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Some questions about my first drsstc.
« Reply #6 on: August 24, 2022, 04:17:15 AM »
Quote
What do you mean by "risk of fried IGBTs." ? I don't understand  :)
If you try 400A OCD, you may get higher performance.  Or IGBTs may fry (may be destroyed due to the high current).  At 24Vge, you may get away with 400A.  However, if you do not want to take the chance on needing to buy more IGBTs and replace damaged ones, then say at 350A.

Quote
Otherwise I noticed that the more I put a big dummy load, the more the spike shrinks, so I assume that with the secondary, and well adjusted, maybe the spike shrinks?
That sounds encouraging.  If the cause of the spike can be found, it may be possible to predict if it is likely to be an issue at full bus voltage or not.  I wouldn't count on loading fixing the spike.  Arc loading can come and go as arcs form and end.

Nice performance!  Clearly the spike isn't too problematic.  No IGBTs have died yet.

Quote
I will continue to reduce the frequency until I have the best possible results.
At some point the primary frequency will get too low to start an arc.  OCD will trip before arcs start.  Just above that frequency is likely to produce longest arcs.  Best music playing (MIDI) will be at higher primary frequency (less detuning) than for longest arcs.  Experiment to see what works best for your coil.
« Last Edit: August 24, 2022, 05:42:05 AM by davekni »
David Knierim

Offline Mathieu thm

  • High Voltage Enthusiast
  • *
  • Posts: 43
  • Karma: +3/-0
    • View Profile
Re: Some questions about my first drsstc.
« Reply #7 on: August 24, 2022, 07:51:44 PM »
Quote
Zoom in to see what it looks like in detail.
So I zoomed in, and here's what it looks like (at 30v)

 [ You are not allowed to view attachments ]
 [ You are not allowed to view attachments ]
 [ You are not allowed to view attachments ]
 
I thought of a solution for the spike, maybe increase / replace the snubber capacitors, I currently have two 941c 1200v 0.47uf in parallel, so 0.94uf in all, is that enough?

Quote
Experiment to see what works best for your coil.
I experimented with a little more detune but the arcs get too big and there are flashovers, so I went back to the previous setting which is the final setting.
I prefer the drsstc to be more reliable  than produce more arcs, I think this setting will be suitable for midi reading, I will tell you all that when I receive my Midistick :)


« Last Edit: August 24, 2022, 11:30:51 PM by mthome4 »
Mathieu Thomé

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2004
  • Karma: +103/-1
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Some questions about my first drsstc.
« Reply #8 on: August 25, 2022, 04:00:20 AM »
Quote
I thought of a solution for the spike, maybe increase / replace the snubber capacitors, I currently have two 941c 1200v 0.47uf in parallel, so 0.94uf in all, is that enough?
For a 20ns spike, 0.94uF is plenty, a tiny fraction of an ohm.  The spike is across snubber lead wires and other H-bridge parasitic interconnect inductance.  Amplitude could be reduced by reducing parasitic inductances.  However, there is a limit to that reduction given minimum of IGBT lead inductance.  Vge scope traces might show some glitch or ringing that is causing rapid current switching (hard switching).  Removing the cause of rapid current switching is better (if possible) than mitigating voltage amplitude by just reduced interconnect inductance.

In particular, I'm wondering why there's any H-bridge output voltage transition 500ns after the normal transition.  UD2.7 should disable within ~20ns of last normal transition, the delay of 74HC74.  That 20ns glitch (from transition to disable) may or may not get through gate driver and to gates.  Even if it does get through, doesn't usually cause issues.  All four IGBT gates should be off well before 500ns.  Thus I'm suspicious of some issue with either your UD2.7 or with GDT or R//D connections from GDT to IGBTs.  Would help to see scope traces of GDT inputs (UD2.7 outputs) and IGBT gates zoomed-in around the last normal cycle and first post-disable cycle.

Of course, since no IGBTs have failed yet, you don't necessarily need to find the cause of glitches.

BTW, Americans often call damaged or failed or dead electronic parts "fried".  It's another synonym within electronics field.  I'm guessing that was the cause of confusion with my initial post.

Quote
I experimented with a little more detune but the arcs get too big and there are flashovers, so I went back to the previous setting which is the final setting.
I prefer the drsstc to be more reliable  than produce more arcs, I think this setting will be suitable for midi reading, I will tell you all that when I receive my Midistick :)
Yes, flashovers or racing sparks or other such issues can limit arc length.  Sounds like a good plan to leave it tuned for likely-good MIDI performance.
David Knierim

Offline Mathieu thm

  • High Voltage Enthusiast
  • *
  • Posts: 43
  • Karma: +3/-0
    • View Profile
Re: Some questions about my first drsstc.
« Reply #9 on: August 25, 2022, 01:33:29 PM »
Quote
Thus I'm suspicious of some issue with either your UD2.7 or with GDT or R//D connections from GDT to IGBTs.
the gdt I use is a Chinese 18/18 gdt  but when i tested it with my function generator the signal was perfect.
maybe shortening the wires from the gdt to the IGBTs would help.
 [ You are not allowed to view attachments ]
I tried to make a new gdt with n87 ferrite, and here is the result:
 [ You are not allowed to view attachments ]
And here is the output of the gdt (with 50ohm load) with my generator connected to the input at 20v and 190khz:
 [ You are not allowed to view attachments ]
 [ You are not allowed to view attachments ]
I don't think replacing it will make a big difference.

Quote
Would help to see scope traces of GDT inputs (UD2.7 outputs) and IGBT gates zoomed-in around the last normal cycle and first post-disable cycle.

I will test all that and I will tell you again, I have already tested to scope the outputs of the gdt in operation on the drsstc and I remember that the signals were clean, but I will tell you again to be sure.

Edit:
 this is an output of the gdt scope on an igbt:

and that, an output of the ud2.7 (from ground to an output pin of the ud2.7):






« Last Edit: August 25, 2022, 04:01:19 PM by mthome4 »
Mathieu Thomé

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2004
  • Karma: +103/-1
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Some questions about my first drsstc.
« Reply #10 on: August 26, 2022, 04:40:22 AM »
Your new GDT looks great!  Old one may be fine, but not sure.

Quote
this is an output of the gdt scope on an igbt:
Looks a bit questionable in two respects.  First, Vge is ~+5V after burst ends.  Should be ~0V.  This may be the cause of the H-bridge output glitch.  The IGBT may not be completely off.  I'd suggest scoping that GDT output.  If GDT output is 0V but Vge isn't, then diode from GDT to gate must be open or reversed.  (Is the scope probe ground connected to IGBT emitter when scoping gate?  If grounded farther away, ~5V might be difference in ground to emitter voltage.)
Second, Vge is +-18-19V.  UD2.7 output is about 0 to 22V, so should make +-22V between the two UD2.7 outputs, and therefore should make +-22Vge.  Could be a GDT issue or an issue with R//D between GDT outputs and gates or an issue with the other UD2.7 output that isn't shown in a scope trace.

BTW, the 20ns glitch may never cause a real failure or problem.  Don't feel obligated to keep chasing the cause.  However, I'd suggest understanding the cause.  It could be something that eventually does cause a failure.
David Knierim

Offline Mads Barnkob

  • Administrator
  • Executive Board Member
  • *****
  • Posts: 2085
  • Karma: +58/-0
  • Denmark
    • View Profile
    • Kaizer Power Electronics
Re: Some questions about my first drsstc.
« Reply #11 on: August 26, 2022, 08:46:30 PM »
Congratulations on getting it to run and make sparks!

The glitch is a "feature" of the UD2.x designs. Steve Ward tried to get it removed, but I seem to recall that it was just overcomplication for something that does not pose a problem.

About testing at 30VDC on the bus, you will see much large spikes compared to the input voltage. This is due to the capacitances in the IGBT are all set in stone, no matter the voltages it sees. Try to up the DC bus voltage and take close note if the spikes get smaller in ratio to the DC bus or if it keeps growing as well.
https://kaizerpowerelectronics.dk - Tesla coils, high voltage, pulse power, audio and general electronics
https://www.youtube.com/KaizerPowerElectronicsDk60/join - Please consider supporting the forum, websites and youtube channel!

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2004
  • Karma: +103/-1
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Some questions about my first drsstc.
« Reply #12 on: August 26, 2022, 09:16:30 PM »
Quote
The glitch is a "feature" of the UD2.x designs. Steve Ward tried to get it removed, but I seem to recall that it was just overcomplication for something that does not pose a problem.
Agree, UD2.X generates an output glitch due to 74HC74 delay time.  It shows up on mthome4's scope capture of one UD2.7 output.
I don't think this UD2.7 output glitch is the cause of the negative glitch 500ns later on H-bridge output.  The one Vge scope capture shows Vge isn't being held close to 0V after burst ends.  It doesn't look like the normal ring-down of GDT inductance with UD2.7 output coupling capacitance.  Vge of other three IGBTs may look different.  I suspect an issue with R//D connections from GDT to IGBTs, or possibly some issue with GDT and/or UD2.7 output coupling caps.

Also agree that H-bridge output glitch may not grow with DC bus voltage.  Clearly it hasn't been a serious issue yet given great performance.  The glitch cause may be benign.  I do tend to obsess with any anomalies that aren't understood.  However, if Vge isn't being held close to 0 after bursts, higher temperature or other changes at full power might lead to thermal runaway.

One possibility just came to mind:  Are there 18V TVS or zener diodes on gates?  That could explain some of the symptoms.  If feeding UD2.7 with 24V, gate clamp diodes need to be 24V or higher.
« Last Edit: August 26, 2022, 09:38:28 PM by davekni »
David Knierim

Offline Mathieu thm

  • High Voltage Enthusiast
  • *
  • Posts: 43
  • Karma: +3/-0
    • View Profile
Re: Some questions about my first drsstc.
« Reply #13 on: August 26, 2022, 10:51:30 PM »
Quote
Congratulations on getting it to run and make sparks!
Thanks !
Quote
About testing at 30VDC on the bus, you will see much large spikes compared to the input voltage. This is due to the capacitances in the IGBT are all set in stone, no matter the voltages it sees. Try to up the DC bus voltage and take close note if the spikes get smaller in ratio to the DC bus or if it keeps growing as well.
Quote
The glitch is a "feature" of the UD2.x designs. Steve Ward tried to get it removed, but I seem to recall that it was just overcomplication for something that does not pose a problem.
I will try and thanks for this valuable information on the ud2.7.

Quote
(Is the scope probe ground connected to IGBT emitter when scoping gate?
Yes I scoped on the gate and emitter of the IGBT.

Quote
One possibility just came to mind:  Are there 18V TVS or zener diodes on gates?  That could explain some of the symptoms.  If feeding UD2.7 with 24V, gate clamp diodes need to be 24V or higher.
Oh :-\, it’s true that I use 18V TVS on IGBT gates....
Can you explain to me how that a problem?
If this is the problem I order 24v tvs to replace them.
« Last Edit: August 26, 2022, 11:14:32 PM by mthome4 »
Mathieu Thomé

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2004
  • Karma: +103/-1
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Some questions about my first drsstc.
« Reply #14 on: August 27, 2022, 02:23:05 AM »
Quote
Oh :-\, it’s true that I use 18V TVS on IGBT gates....
Can you explain to me how that a problem?
If this is the problem I order 24v tvs to replace them.
« Last Edit: August 26, 2022, 11:14:32 PM by mthome4 »
The problem is that UD2.7 is attempting to drive Vge to +-24V.  (Actually slightly lower due to IGBT gate capacitance, perhaps actually +-23V.  The 1V drop is mostly across UD2.7 output coupling capacitors.)  The 18V TVS diodes are clamping actual Vge to +-18V.  Draws large current from UD2.7 outputs and leaves excess voltage on UD2.7 output coupling capacitors, which is likely the cause of the remaining ~5Vge after enable ends.
Your options are to either reduce UD2.7 supply voltage to 19V or under, or to change TVS diodes.  Most DRSSTC coils use 27V-33V TVS diodes.  As Mads has posted before, actual IGBT gate oxide punch-through failure occurs around 80V.  TVS diodes gate-emitter are intended to clamp occasional spikes such as might be caused by flash-over or other unintended events.  That's why they are normally several volts above normal operating Vge.

Hopefully this will fix both your Vge waveforms and remove the short negative H-bridge output spike.  Needing to change something is a nuisance, but I'd suggest thinking of it as a good situation.  Now you know what the (hopefully only) issue is.  Your coil should be more reliable after fixing this.
« Last Edit: August 27, 2022, 02:30:23 AM by davekni »
David Knierim

Offline Mathieu thm

  • High Voltage Enthusiast
  • *
  • Posts: 43
  • Karma: +3/-0
    • View Profile
Re: Some questions about my first drsstc.
« Reply #15 on: August 27, 2022, 05:04:47 PM »
Thank you very much for explaining to me the problem with 18V tvs :)

I tested several things and the results are interesting:

Quote
Most DRSSTC coils use 27V-33V TVS diodes.
I tested to put 27v zener diodes instead of 18v TVS:


I also put my new gdt:


The results are interesting, first of all I no longer have a problem with the 5v voltage at the end of the cycle, the result is the same with the old gdt (so I will order 27v TVS because my zener are of poor quality, thanks for noticing that!)


And the second good news is that the spike seems to be getting shorter:



I still have a question about my new gdt, so I'm using an n87 ferrite core with a cat 5 cable, I did 10 turns.
However my cable has a metal shield:


The performance of the gdt can be better by removing this shielding?
Mathieu Thomé

Offline Mads Barnkob

  • Administrator
  • Executive Board Member
  • *****
  • Posts: 2085
  • Karma: +58/-0
  • Denmark
    • View Profile
    • Kaizer Power Electronics
Re: Some questions about my first drsstc.
« Reply #16 on: August 27, 2022, 07:38:26 PM »
If you put the shielding in parallel with the primary wires of the GDT, you get one of the lowest possible inductance GDTs and with the highest possible coupling. Compared to delta, sectional, circumferential and trifilar wound GDTs.
https://kaizerpowerelectronics.dk - Tesla coils, high voltage, pulse power, audio and general electronics
https://www.youtube.com/KaizerPowerElectronicsDk60/join - Please consider supporting the forum, websites and youtube channel!

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2004
  • Karma: +103/-1
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Some questions about my first drsstc.
« Reply #17 on: August 27, 2022, 08:31:46 PM »
Quote
The results are interesting, first of all I no longer have a problem with the 5v voltage at the end of the cycle, the result is the same with the old gdt (so I will order 27v TVS because my zener are of poor quality, thanks for noticing that!)
Yes, Vge waveform looks more normal now.  The undershoot on falling edges is normal due to diodes from GDT to gate.  Rising edges have some ring.  That is within normal range too.  Higher-value gate series resistors would reduce that ring, but might also require readjusting phase lead to be a bit more.

Quote
And the second good news is that the spike seems to be getting shorter:
That 330ns positive pulse is normal, the time from burst end until primary current reversal.  It is the very-short negative spike at the end of this positive pulse that might be of concern.  As Mads mentioned, that short ~20ns spike might not increase with higher bus voltage.  As long as the sum of spike voltage and DC bus voltage together doesn't exceed IGBT's 600Vce rating, there is no issue.  If that short negative spike does grow with bus voltage, then it might eventually lead to IGBT failure due to excess Vce.

Just realized that you are probably connecting scope probe ground lead to one H-bridge output and probing the other output.  Even with an isolated bench supply, connecting scope ground to an active signal can cause problems.  The bench supply has significant capacitance to ground.  I'd suggest scoping one H-bridge output at a time with scope probe ground connected to negative bus supply.  Scoping the two outputs individually also makes it clear which output has the spike.  Scoping both outputs at the same time using two probes can be helpful.  Also, two-channel captures showing both one output and one gate simultaneously is helpful.

Your decision on whether to continue scoping signals to figure out the 20ns negative spike, or just have fun enjoying your nice new coil.  If you do continue scoping, a couple more suggestions on scope use:  First, I'd suggest changing vertical channel mode to DC rather than AC.  Including DC information is often useful.  Second, use the lowest volts/division setting possible without spikes going off-screen.  Makes seeing subtle detail easier.  Vge is difficult to read at 50V/div.  10V/div would be more appropriate.  For chasing this specific burst-end spike, most useful horizontal scale would be about 1us/division, showing one cycle before and one cycle after burst end.

Do you happen to have any Y-caps around (capacitors rated for line to ground use)?  If you have one, or some ~2-10nF capacitor of sufficient voltage rating, it is possible to get useful scope captures even when running at full voltage.  Connect scope ground lead through capacitor to Vbus-.  That will be sufficient high-frequency grounding to allow scoping H-bridge output spikes to see if they grow to a problematic amplitude.  Use low burst repetition rate when scoping since many probes have low RMS voltage rating at high frequency.  Of course, use caution when working with line power.

Quote
I still have a question about my new gdt, so I'm using an n87 ferrite core with a cat 5 cable, I did 10 turns.
However my cable has a metal shield:
I haven't seen a GDT use shielded cable before.  Mads has a good idea of paralleling with primary wires.  Only disadvantage that comes to mind is increased capacitance between windings due to shield.  With your fast IGBTs and resulting fast output slew rates, lower inter-winding capacitance could also be advantageous.  Another thing that can help prevent issues resulting from fast output slew rates is to add ferrite beads around the two GDT outputs going to high-side IGBTs.  So, perhaps use Mads suggestion for shield and then add ferrite beads to high-side outputs to mitigate possible capacitive coupling issues.

Interesting project!  But, don't let my personal curiosity drive what you want for your project.  Proceed as you prefer.
David Knierim

Offline Mads Barnkob

  • Administrator
  • Executive Board Member
  • *****
  • Posts: 2085
  • Karma: +58/-0
  • Denmark
    • View Profile
    • Kaizer Power Electronics
Re: Some questions about my first drsstc.
« Reply #18 on: August 27, 2022, 11:24:02 PM »
I think the idea originally comes from using coaxial cables in GDTs, so not to see the shield as a shield, but solely as a primary winding. I got some measurements somewhere that should go in the not-yet-finished GDT part of the DRSSTC guide. Maybe some day its done :)

Quote
I still have a question about my new gdt, so I'm using an n87 ferrite core with a cat 5 cable, I did 10 turns.
However my cable has a metal shield:
I haven't seen a GDT use shielded cable before.  Mads has a good idea of paralleling with primary wires.  Only disadvantage that comes to mind is increased capacitance between windings due to shield.  With your fast IGBTs and resulting fast output slew rates, lower inter-winding capacitance could also be advantageous.  Another thing that can help prevent issues resulting from fast output slew rates is to add ferrite beads around the two GDT outputs going to high-side IGBTs.  So, perhaps use Mads suggestion for shield and then add ferrite beads to high-side outputs to mitigate possible capacitive coupling issues.

Interesting project!  But, don't let my personal curiosity drive what you want for your project.  Proceed as you prefer.
https://kaizerpowerelectronics.dk - Tesla coils, high voltage, pulse power, audio and general electronics
https://www.youtube.com/KaizerPowerElectronicsDk60/join - Please consider supporting the forum, websites and youtube channel!

Offline Mathieu thm

  • High Voltage Enthusiast
  • *
  • Posts: 43
  • Karma: +3/-0
    • View Profile
Re: Some questions about my first drsstc.
« Reply #19 on: September 02, 2022, 09:58:02 PM »
Bad news :(:
I tested the drsstc with the midistick, everything worked perfectly, I moved the coil outside to be able to test better.
when I turn it on, nothing happens, I look at the fuse and it burnt.
Infact two igbt are dead, I don't know why they burned, the phase lead was well adjusted.
I did the test at 100-120Vac max
small note: the heatsink where the igbts are was warmer than in fixed mode

Buying new IGBTs is not a problem, it's more the fact that why it burned?

otherwise this is how i installed the new gdt:



I will let you know when I have replaced the IGBTs, the TVS and maybe put higher gate resistors (10ohms)

EDIT:
I may have a hypothesis, to test I used midi files with a lot of notes, I saw the files of bytecode77 and they have a lot less notes, so I think with all these notes, on times overlapped and maybe it made hard switching.
By the way, I have 60N60SMD lying around, are they as good as the HGTG30N604AD?

Anyway I will still buy HGTG30N60A4D.
« Last Edit: September 02, 2022, 11:58:28 PM by mthome4 »
Mathieu Thomé

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2004
  • Karma: +103/-1
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Some questions about my first drsstc.
« Reply #20 on: September 03, 2022, 05:53:29 AM »
Quote
Bad news :(:
Sorry about your bad news of IGBTs frying.

Quote
otherwise this is how i installed the new gdt:
Looks well-done.

Quote
y the way, I have 60N60SMD lying around, are they as good as the HGTG30N604AD?
Presuming FGH60N60SMD, they are probably as good.  Quite a few differences in spec details, but overall FGH60N60SMD looks as good, perhaps slightly better.  May need to tweak phase-lead when changing IGBT part types.  10-ohm gate resistors might also be good, as typical gate charge is slightly lower.

Quote
I may have a hypothesis, to test I used midi files with a lot of notes, I saw the files of bytecode77 and they have a lot less notes, so I think with all these notes, on times overlapped and maybe it made hard switching.
Hard switching is a possibility, or just too-high a duty cycle.  The hard-switching issue can be fixed with the UD2.7 self-oscillation mod I've posted about if also combined with a tweak to the enable FF:  Instead of using preset input for enable, use the D input (from the high-true version of enable gate output).  That way enable always starts on a zero-current edge.  If no current, self-oscillation provides edges to clock in enable.  (This is the way my DRSSTC works, though not based on UD2.7.)

Good luck with the repairs!
David Knierim

Offline Mathieu thm

  • High Voltage Enthusiast
  • *
  • Posts: 43
  • Karma: +3/-0
    • View Profile
Re: Some questions about my first drsstc.
« Reply #21 on: September 09, 2022, 12:19:06 PM »
I received my TVS diodes and resistors, I also ordered other more powerful fans because the fans installed are not powerful, really not powerful. I still buy other HGTG30N60A4D just in case.

I installed 27v TVS diodes and 10ohm gate resistors, and the FGH60N60SMD.


Here is the output of the gdt on the igbt, Vge (30v on the bus):








And here is the output of the bridge, I used several lipo in series for a total of about 80v :





We can also see that the spike is smaller :







I set the OCD to 250a for the first test, I think maybe increase it to 300a later.

I will wait to receive the new fans to do some tests with more power, the other fans are really not powerful, so I had a hypothesis: maybe the fans are so not powerful that the IGBTs have overheated and have fried.
Mathieu Thomé

Offline Mathieu thm

  • High Voltage Enthusiast
  • *
  • Posts: 43
  • Karma: +3/-0
    • View Profile
Re: Some questions about my first drsstc.
« Reply #22 on: September 09, 2022, 03:09:31 PM »
Well...........

I tested in low power, 100Vac, two Igbt died.
It demoralizes me very much, I understand nothing.

I'll try to replace the gdt with the old one and if that doesn't change anything, maybe it's better to switch to skm100 IGBTs or something like that ?
« Last Edit: September 09, 2022, 07:46:36 PM by mthome4 »
Mathieu Thomé

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2004
  • Karma: +103/-1
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Some questions about my first drsstc.
« Reply #23 on: September 10, 2022, 06:16:36 AM »
Quote
tested in low power, 100Vac, two Igbt died.
It demoralizes me very much, I understand nothing.
Ouch!  Yes, failures are very frustrating.  Especially given how many different possible causes exist.  However, you are learning a lot!  (Sometimes it helps to take a break after a failure.  Both helps emotionally and allows time to think through possible issues and strategies for testing and/or design changes.)

One simple question:  I presume you remembered to disconnect scope ground before powering from AC.

Quote
Here is the output of the gdt on the igbt, Vge (30v on the bus):
There are short Vge spikes at each H-bridge output transition after enable ends.  Is this a low-side IGBT (emitter to Vbus-)?  If high-side, could be a scoping artifact.  Presuming low-side, could be a combination of GDT inter-winding capacitance (especially with shielded cable) and/or emitter lead inductance.

I can't quite tell from previous pictures:  How long are the IGBT leads from IGBT case to ECB?  Especially with your fast IGBTs, excess emitter lead or trace inductance causes Ice current transitions to induce spikes in the local Vge within IGBT devices.

Quote
I'll try to replace the gdt with the old one and if that doesn't change anything, maybe it's better to switch to skm100 IGBTs or something like that ?
The old GDT likely has lower inter-winding capacitance, so may reduce Vge spikes.  Or, add ferrite beads to high-side GDT outputs.

One possibility is to try slower TO247 IGBTs.  Phase-lead needs to increase to compensate for more delay.  Sensitivity to Vge spikes is less because slower IGBTs don't respond to such short glitches as much.  Switching loss will be a bit higher with slow IGBTs even under ZCS conditions.

Bricks are generally slower than TO247 IGBTs, but do have a few advantages.  One key advantage is separate emitter leads for GDT connection.  This mostly eliminates the issue of emitter lead inductance.  Ice current doesn't share the same interconnect path, so no common inductance with Vge connection.  TO247 IGBTs with single emitter leads require more careful ECB layout and lead length minimization to reduce the amount of emitter parasitic inductance that is common to both Ice and Vge.  Especially critical for fast IGBTs.

Another advantage of bricks is isolated case metal.  Avoids the need for insulating thermal pads externally, increases contact area with heatsink, and increases thermal mass within IGBT package.

A related advantage is that bricks have mounting holes on both ends.  Allows more consistent thermal contact with heatsink.  TO247 packages that are mounted by their screw hole have pressure applied at only that one point.  Power dissipation (IGBT silicon die) is between hole and pins.  Pressure (thermal contact) under die can be poor.  More screw torque can actually make thermal contact worse.  Screw torque compresses the thermal pad under the mounting hole, which can pivot the lead-end of the TO247 package away from the heatsink.  Most commercial TO247 mounting uses spring clips or other mechanical methods to apply pressure to the package epoxy directly over the die.  The hole often isn't used at all.  It is quite possible that this latest failure was due to insufficient thermal contact between die and heatsink.  (My DRSSTC uses 40 TO247 IGBTs, each clamped over the die, not using screw hole.  Also has very-short IGBT leads.)

Good luck!  Keep learning.  That leads to expertise.
David Knierim

Offline Mathieu thm

  • High Voltage Enthusiast
  • *
  • Posts: 43
  • Karma: +3/-0
    • View Profile
Re: Some questions about my first drsstc.
« Reply #24 on: September 10, 2022, 10:33:42 AM »
Quote
Ouch!  Yes, failures are very frustrating.  Especially given how many different possible causes exist.  However, you are learning a lot!  (Sometimes it helps to take a break after a failure.  Both helps emotionally and allows time to think through possible issues and strategies for testing and/or design changes.)

Yes it's true that we learn a lot!
It's true that taking breaks helps, it helps not to do anything stupid!

Quote
One simple question:  I presume you remembered to disconnect scope ground before powering from AC.

yes I disconnected the oscilloscope from the bridge, I always check several times before powering up.

Quote
There are short Vge spikes at each H-bridge output transition after enable ends.  Is this a low-side IGBT (emitter to Vbus-)?  If high-side, could be a scoping artifact.  Presuming low-side, could be a combination of GDT inter-winding capacitance (especially with shielded cable) and/or emitter lead inductance.

It was on the high side but I also tested on the low side and it's the same.
For the gdt I finally made a "classic" gdt without shielding because I couldn't do anything clean (the shielding is very difficult to solder)

Quote
I can't quite tell from previous pictures:  How long are the IGBT leads from IGBT case to ECB?  Especially with your fast IGBTs, excess emitter lead or trace inductance causes Ice current transitions to induce spikes in the local Vge within IGBT devices.

Here is a picture showing the leads of the igbt :



Quote
The old GDT likely has lower inter-winding capacitance, so may reduce Vge spikes.  Or, add ferrite beads to high-side GDT outputs.

I do not understand, why these spikes make the igbt die?

Quote
One possibility is to try slower TO247 IGBTs.

The problem is that the igbt are difficult to solder, that's why I asked if bricks like the skm100 could solve the problem.(btw thank you for this information on the bricks!)

So the bricks can be more adapted to my configuration, there would be less stress compared to the igbt to247 ?

Quote
Most commercial TO247 mounting uses spring clips or other mechanical methods to apply pressure to the package epoxy directly over the die.  The hole often isn't used at all.  It is quite possible that this latest failure was due to insufficient thermal contact between die and heatsink.  (My DRSSTC uses 40 TO247 IGBTs, each clamped over the die, not using screw hole.  Also has very-short IGBT leads.)

It's true that it can only be better but it's rather difficult to do on my configuration.
I saw that your coil with 40 igbt works so well, well done!

Quote
Good luck!  Keep learning.

Thanks a lot !
Mathieu Thomé

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2004
  • Karma: +103/-1
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Some questions about my first drsstc.
« Reply #25 on: September 11, 2022, 04:01:09 AM »
Quote
Here is a picture showing the leads of the igbt :
Looks great.  Nice short leads.

Quote
I do not understand, why these spikes make the igbt die?
May not be an issue even for fast IGBTs.  The only times I've experienced issues with short spikes due to GDT capacitance is with FETs, which are generally faster than even the fastest IGBTs.  In my FET cases, the spike sometimes gets large enough to turn FET on briefly, which causes another output voltage transition, which causes another spike, another transition, etc.  Result is high-frequency oscillation at each transition.  Causes excess power dissipation and damaged FETs.  Clearly you are not seeing such oscillation in scope captures.  Only possible concern is that such oscillation occurs at higher voltage where output voltage swings are larger and faster.  (Faster because IGBT capacitance is lower at higher voltage.)

Quote
The problem is that the igbt are difficult to solder, that's why I asked if bricks like the skm100 could solve the problem.
Soldering large devices such as IGBTs requires a large soldering iron.  Size and material of the tip are more important than wattage.  It is the stored heat within the tip that provides energy to solder one connection.  The heating element replaces that energy mostly between soldering operations.  Use a large tip and start with the tip a bit extra hot.  Then IGBTs will not be a problem to solder.

Quote
It's true that it can only be better but it's rather difficult to do on my configuration.
I saw that your coil with 40 igbt works so well, well done!
Here's one option that isn't too difficult.  Use a large washer (fender washer) bent to touch heatsink past IGBT case on one side and press on top of IGBT die at the other end.

Example parts for a single IGBT.  (This is actually an ancient BJT.  An old part that was easy to grab for a mechanical sample.)  The un-bent washer in the photo isn't used, just there to show the washer size used for the bent one.




Assembled to heat sink.  (Sorry for the very-fuzzy image.  Had a hard time getting my camera to focus close-up.)


Another variation is to add a plastic spacer to help define location of clamping force over IGBT and to soften contact line.  A short section of a nylon zip-tie works well.  Didn't cut the zip-tie for this picture.  Also a bit fuzzy.


My 40-IGBT build is a bit different and more complex.  Probably much more complicated than necessary.
« Last Edit: September 11, 2022, 04:28:37 AM by davekni »
David Knierim

Offline Mathieu thm

  • High Voltage Enthusiast
  • *
  • Posts: 43
  • Karma: +3/-0
    • View Profile
Re: Some questions about my first drsstc.
« Reply #26 on: September 11, 2022, 09:17:53 AM »
Thanks so much for all the advice!

Quote
Here's one option that isn't too difficult.  Use a large washer (fender washer) bent to touch heatsink past IGBT case on one side and press on top of IGBT die at the other end.

Thank you for this tip, I'll try all that in a few days, and I'll get back to you.
« Last Edit: September 11, 2022, 09:40:44 AM by mthome4 »
Mathieu Thomé

Offline Mathieu thm

  • High Voltage Enthusiast
  • *
  • Posts: 43
  • Karma: +3/-0
    • View Profile
Re: Some questions about my first drsstc.
« Reply #27 on: September 11, 2022, 07:39:19 PM »
I just realized that I forgot a question:

Quote
May not be an issue even for fast IGBTs.  The only times I've experienced issues with short spikes due to GDT capacitance is with FETs, which are generally faster than even the fastest IGBTs.  In my FET cases, the spike sometimes gets large enough to turn FET on briefly, which causes another output voltage transition, which causes another spike, another transition, etc.  Result is high-frequency oscillation at each transition.  Causes excess power dissipation and damaged FETs.  Clearly you are not seeing such oscillation in scope captures.  Only possible concern is that such oscillation occurs at higher voltage where output voltage swings are larger and faster.  (Faster because IGBT capacitance is lower at higher voltage.)

There is something that I do not understand, you say that the capacity of the gdt is not necessarily a problem even for fast igbt, with my old gdt (which probably has less capacity) I did not have fried igbt problem , but only after replacing the gdt, the igbt started to fried.
The waveform of the bridge is quite good (at least not worse than with the old gdt).
So I ask myself : maybe the problem comes from my new gdt? but I do not understand why, the difference is 10 turns instead of 18 turns and a capacity, which I suppose higher.
I'm not 100% sure but I think replacing the gdt with the old one might solve the problem.(because I had no fried igbt problem before gdt replacement.)
Maybe something escapes me but I do not understand how the new gdt burn the igbt's (if we based on my hypothesis of gdt problem)

Excuse me if my English has mistakes at times, I am not a pro in English :)
« Last Edit: September 11, 2022, 07:41:28 PM by mthome4 »
Mathieu Thomé

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2004
  • Karma: +103/-1
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Some questions about my first drsstc.
« Reply #28 on: September 11, 2022, 09:19:30 PM »
Quote
There is something that I do not understand, you say that the capacity of the gdt is not necessarily a problem even for fast igbt, with my old gdt (which probably has less capacity) I did not have fried igbt problem , but only after replacing the gdt, the igbt started to fried.
The waveform of the bridge is quite good (at least not worse than with the old gdt).
So I ask myself : maybe the problem comes from my new gdt? but I do not understand why, the difference is 10 turns instead of 18 turns and a capacity, which I suppose higher.
I'm not 100% sure but I think replacing the gdt with the old one might solve the problem.(because I had no fried igbt problem before gdt replacement.)
Maybe something escapes me but I do not understand how the new gdt burn the igbt's (if we based on my hypothesis of gdt problem)
Figuring out why an electronic part failed is usually difficult.  It is rare to have a scope capture showing the exact failure event.  The strategy is to look for all possible causes of failure, choose which ones seem most likely to have caused problems, and improve those parts of the circuit.  It is still possible that the Vge glitch is related to a failure mechanism, but perhaps not.  I've seen other scope traces posted on the forum with glitches that didn't cause failures.  I've personally experienced failures, but only on FET H-bridges.
It is also possible that the latest failure was due to one IGBT having poor thermal contact with heatsink.  Each time an IGBT is replaced, there's a chance that the new part isn't quite as flat on the bottom or that the thermal pad picked up some grit or didn't compress as well.  (Most thermal pad manufactures specify that the pad needs replacing every time a part is removed.  I think most hobbyists don't bother replacing pads.  Replacement is less critical if force is applied to IGBT over die area and if the pad is kept clean.)  Quality of thermal pad makes a difference too.
The IGBT failure cause may also be something I haven't guessed yet.  Perhaps at high power some signal within UD2.7 is getting disturbed from magnetic or electrostatic coupling to external fields.  I think this is less likely in your build.  However, many builds place driver board inside a grounded metal box to minimize chances of such issues.
Another possibility:  The original 18-turn GDT likely has much higher leakage inductance than home-built ones.  Generally GDT leakage inductance is undesirable.  However, leakage inductance would slow Vge edge slew rates.  Slower Vge transitions translate to somewhat slower Vce transitions, which reduce sensitivity to H-bridge interconnect parasitic inductance.  Some DRSSTC builds include a smaller resistor in series with gate diodes, say 3 ohms for turn-off in series with diode and 10 ohms for turn-on across diode.  That is another way to slow Vge edges.  (In an ideal world, fast Vge edges are best.  Minimizes IGBT power dissipation.  However, if there is significant H-bridge interconnect inductance, fast edges may cause excess Vce spikes that fry IGBTs.)

Quote
Excuse me if my English has mistakes at times, I am not a pro in English :)
Your English is fine.  If I don't understand something, I'll ask.  If I misunderstand and answer the wrong question, please ask again.  (I'm in USA and know no other languages.  We get spoiled by so much of the world being willing to communicate in English.)
David Knierim

Offline Maju

  • High Voltage Enthusiast
  • *
  • Posts: 22
  • Karma: +3/-0
    • View Profile
Re: Some questions about my first drsstc.
« Reply #29 on: September 12, 2022, 04:32:03 PM »
Hi

In my opinion your IGBT are still dying becouse of waaaaay to long pulses. TO247 really hate to be turned on for so long (based on your scopes) with so high currents.

To clarify, you should trying to achieve OCD goal ASAP - in just fiew cycles! With your frequency it should be possible in under 40-50us (lesser = better).
So if you still want experimenting with TO247, increase capacity of your resonant cap and decrise number of turns in primary (even by half or more) to achieve the same resonant freq. with lesser on time.

P.S. SKM100 are great if you're gonna done with TO247

Good luck!

Offline Mathieu thm

  • High Voltage Enthusiast
  • *
  • Posts: 43
  • Karma: +3/-0
    • View Profile
Re: Some questions about my first drsstc.
« Reply #30 on: September 12, 2022, 08:09:42 PM »
Hi

Quote
In my opinion your IGBT are still dying becouse of waaaaay to long pulses. TO247 really hate to be turned on for so long (based on your scopes) with so high currents.
Thank you for this information, but the problem is that before I replaced the gdt, the coil was working fine, with about 190Vac and 100us of on time.I will test with the old gdt to see if everything works again.

Quote
So if you still want experimenting with TO247, increase capacity of your resonant cap and decrise number of turns in primary (even by half or more) to achieve the same resonant freq. with lesser on time.
The problem is how i made the MMC it's hard to modify it to make something clean.

Quote
P.S. SKM100 are great if you're gonna done with TO247
If I can't do it with TO 247, I think I'll switch to skm100, it will be more reliable than with TO 247, from what I've seen there is also much less interconnect inductance on the H-bridge , which is much better I think.







Quote
It is also possible that the latest failure was due to one IGBT having poor thermal contact with heatsink.
it is likely that the problem comes from there, so I will do the washer technique (thank you for this technique ;D) and replace the gdt with the old one to see if there are any differences.

Quote
Perhaps at high power some signal within UD2.7 is getting disturbed from magnetic or electrostatic coupling to external fields.  I think this is less likely in your build.  However, many builds place driver board inside a grounded metal box to minimize chances of such issues.
I agree that the problem does not come from there, I put the ud2.7 in a metal box connected to the ground.


Quote
Another possibility:  The original 18-turn GDT likely has much higher leakage inductance than home-built ones.  Generally GDT leakage inductance is undesirable.  However, leakage inductance would slow Vge edge slew rates.  Slower Vge transitions translate to somewhat slower Vce transitions, which reduce sensitivity to H-bridge interconnect parasitic inductance.  Some DRSSTC builds include a smaller resistor in series with gate diodes, say 3 ohms for turn-off in series with diode and 10 ohms for turn-on across diode.  That is another way to slow Vge edges.  (In an ideal world, fast Vge edges are best.  Minimizes IGBT power dissipation.  However, if there is significant H-bridge interconnect inductance, fast edges may cause excess Vce spikes that fry IGBTs.)
Thank you for these explanations, it's always a pleasure to learn more.



Of course, I will keep you informed of the progress.

btw,I also received and installed the new fans which are much more powerful than the fans I got from a computer.


« Last Edit: September 12, 2022, 08:24:58 PM by mthome4 »
Mathieu Thomé

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2004
  • Karma: +103/-1
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Some questions about my first drsstc.
« Reply #31 on: September 13, 2022, 03:43:58 AM »
Quote
    In my opinion your IGBT are still dying becouse of waaaaay to long pulses. TO247 really hate to be turned on for so long (based on your scopes) with so high currents.

Thank you for this information, but the problem is that before I replaced the gdt, the coil was working fine, with about 190Vac and 100us of on time.I will test with the old gdt to see if everything works again.
The definition of too-long for TO247 depends on the power dissipation and thermal resistance.  Many (I think most) QCW coils use TO247 devices and run ~10ms on-time.  My TO247 DRSSTC typically runs 100-200us when playing music, and up to 1ms when tuned for longest arcs.

Quote
Of course, I will keep you informed of the progress.
Yes, please do.  Definitely interested.

Quote
btw,I also received and installed the new fans which are much more powerful than the fans I got from a computer.
Great.  Good thermal conductivity to heat sink and from heat sink to air is likely to make IGBTs more robust.  Of course, even perfect thermal management won't protect IGBTs if spikes exceed Vce rating for long or high-frequency oscillations occur at transitions.

Good luck with this round of repairs!
David Knierim

Offline Mathieu thm

  • High Voltage Enthusiast
  • *
  • Posts: 43
  • Karma: +3/-0
    • View Profile
Re: Some questions about my first drsstc.
« Reply #32 on: September 16, 2022, 11:05:46 PM »
After a little break, here I am again in the build of my drsstc.

I followed your tips of washers on the IGBTs but I did not take washers but pieces of metal.
Here is the result :






I also have a question, I have 400V TVS diode on each IGBT (connect to CE), I saw that it can be annoying, I understood
 that if a spike exceeds the voltage of the TVS, it goes shorted which also shorts an IGBT, the IGBT can short itself, and then short the other IGBT next to it.
My question is, do I remove the TVS diodes?
I could see several things on it but good and bad but often it is recommended to remove them, if I'm not mistaken.

Here is a pic of the TVS:



I am also seriously starting to look at skm100gb125, I think it could be more suitable, during my tests I could see that I could not go up to more than 190Vac, I would like to be able to supply it with 230v directly.
The problem is that I think that the 600v limit of IGBT TO-247 will be quickly reached whereas with the skm100 I would have much more margin.
I prefer to put 325 vdc in a skm 100 with more margin rather than in a to-247 which I think, with my configuration, is not powerful enough.
I have heard only good things about IGBT bricks and I would also like experiment with bricks :)
« Last Edit: September 16, 2022, 11:50:50 PM by mthome4 »
Mathieu Thomé

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2004
  • Karma: +103/-1
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Some questions about my first drsstc.
« Reply #33 on: September 17, 2022, 01:50:27 AM »
Quote
I followed your tips of washers on the IGBTs but I did not take washers but pieces of metal.
Here is the result :
Very nicely implemented!  I expect your mounting to work quite well.  Thermal resistance will be lower than before, perhaps much lower.

Your metal tabs would be a great example for some sort of TO247 tutorial page.  I'm not sure where to put it so it would get found by new coil builders looking for guides.  Perhaps somewhere in the "beginners" page.  Would love for your example to get more visibility than buried in the middle of this one thread.

Quote
I also have a question, I have 400V TVS diode on each IGBT (connect to CE), I saw that it can be annoying, I understood
 that if a spike exceeds the voltage of the TVS, it goes shorted which also shorts an IGBT, the IGBT can short itself, and then short the other IGBT next to it.
My question is, do I remove the TVS diodes?
I could see several things on it but good and bad but often it is recommended to remove them, if I'm not mistaken.
Difficult to know for sure in your situation.  HGTG30N60A4D has no specification for avalanche energy.  There is no way to know if IGBT avalanche energy capability is higher than TVS diode or not.  Hopefully you'll get other opinions.

My suggestion is to lift one end of each TVS diode and add a capacitor in series, with a resistor in parallel with the capacitor.  I used this configuration for my resonant isolated PFC.  Allows TVS diodes to clamp short transients without excess total power dissipation.  For values, something around 0.1uF paralleled with 1k.  (Avoid carbon film for 1k resistor.  In case of over-temperature, carbon film resistors tend to fail towards low resistance.)

Quote
I am also seriously starting to look at skm100gb125, I think it could be more suitable, during my tests I could see that I could not go up to more than 190Vac, I would like to be able to supply it with 230v directly.
The problem is that I think that the 600v limit of IGBT TO-247 will be quickly reached whereas with the skm100 I would have much more margin.
I prefer to put 325 vdc in a skm 100 with more margin rather than in a to-247 which I think, with my configuration, is not powerful enough.
I have heard only good things about IGBT bricks and I would also like experiment with bricks :)
If your H-bridge ECB layout has low parasitic inductance, then 600V IGBTs will be fine at 325Vbus.  I use 650V IGBTs for 450Vbus.  Here's a recent example of a great low-inductance TO247 H-bridge ECB layout:
https://highvoltageforum.net/index.php?topic=2183.msg16087#msg16087
And my tutorial on such layout, starting with a half-bridge version:
https://highvoltageforum.net/index.php?topic=1324.msg9795#msg9795
Parasitic inductance does not necessarily need to be that low.  However, HGTG30N60A4D are quite fast, making low inductance more important.

Yes, bricks are generally more robust, especially going to 1200V bricks.  Even this "fast" SKM100 brick is about half the speed and twice the gate charge of your HGTG30N60A4D IGBTs.  Shouldn't be an issue with your ~200kHz coil.  More of a concern with ~400kHz QCW coils, especially if phase-shift is used for ramping.
« Last Edit: September 17, 2022, 01:56:00 AM by davekni »
David Knierim

Offline Mathieu thm

  • High Voltage Enthusiast
  • *
  • Posts: 43
  • Karma: +3/-0
    • View Profile
Re: Some questions about my first drsstc.
« Reply #34 on: September 17, 2022, 11:56:26 AM »
Quote
Very nicely implemented!  I expect your mounting to work quite well.  Thermal resistance will be lower than before, perhaps much lower.

Your metal tabs would be a great example for some sort of TO247 tutorial page.  I'm not sure where to put it so it would get found by new coil builders looking for guides.  Perhaps somewhere in the "beginners" page.  Would love for your example to get more visibility than buried in the middle of this one thread.

Thanks ! If you want I can do a little tutorial on that, I also think the beginner page would be suitable.


Quote
Difficult to know for sure in your situation.  HGTG30N60A4D has no specification for avalanche energy.  There is no way to know if IGBT avalanche energy capability is higher than TVS diode or not.  Hopefully you'll get other opinions.

My suggestion is to lift one end of each TVS diode and add a capacitor in series, with a resistor in parallel with the capacitor.  I used this configuration for my resonant isolated PFC.  Allows TVS diodes to clamp short transients without excess total power dissipation.  For values, something around 0.1uF paralleled with 1k.  (Avoid carbon film for 1k resistor.  In case of over-temperature, carbon film resistors tend to fail towards low resistance.)

I found four capacitors in my stock but I'm not sure if they will fit, they are 0.1uf but they are different and have voltages of 250Vac and 275Vac.




Quote
If your H-bridge ECB layout has low parasitic inductance, then 600V IGBTs will be fine at 325Vbus.  I use 650V IGBTs for 450Vbus.  Here's a recent example of a great low-inductance TO247 H-bridge ECB layout:
https://highvoltageforum.net/index.php?topic=2183.msg16087#msg16087
And my tutorial on such layout, starting with a half-bridge version:
https://highvoltageforum.net/index.php?topic=1324.msg9795#msg9795
Parasitic inductance does not necessarily need to be that low.  However, HGTG30N60A4D are quite fast, making low inductance more important.

Yes, bricks are generally more robust, especially going to 1200V bricks.  Even this "fast" SKM100 brick is about half the speed and twice the gate charge of your HGTG30N60A4D IGBTs.  Shouldn't be an issue with your ~200kHz coil.  More of a concern with ~400kHz QCW coils, especially if phase-shift is used for ramping.

I am impressed by your gdt, he has so few turns!

I saw somewhere that Mads said that he had stopped with the TO-247 because he found them too fragile, I'm not far from the limit of the TO-247, what I like with the skm100 is that there are many advantages, as you said with the two dedicated wires for the gate drive, the better heat dissipation, the metal plate is isolated and many other advantages.
I also like that the primary current will have more margin, I saw the skm100 works well with 550a OCD,
if I put my OCD at 400a, I would have more margin.
I like the fact of not pushing the IGBTs to the maximum but keeping margin.
Mathieu Thomé

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2004
  • Karma: +103/-1
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Some questions about my first drsstc.
« Reply #35 on: September 17, 2022, 06:37:23 PM »
Quote
Thanks ! If you want I can do a little tutorial on that, I also think the beginner page would be suitable.
I think that would be great.  Even just a copy of those images moved to a beginner page would be useful.  Perhaps a new thread about TO247 mounting or some such title.  Whenever you get to that (no hurry), I'll add a link to your thread from my low-inductance bridge tutorial.

Quote
I found four capacitors in my stock but I'm not sure if they will fit, they are 0.1uf but they are different and have voltages of 250Vac and 275Vac.
Electrically fine.  Ceramic capacitors are likely smaller. 100V or so would probably suffice, though 200V would be prefered.  Down to 0.047uF would likely be fine, certainly no worse than leaving off TVS diodes.

Quote
I am impressed by your gdt, he has so few turns!
Most GDTs have more turns than necessary, which increases parasitic inductance in gate drive.  However, too few turns causes more serious issues than too many turns.  Thus most GDTs have extra turns just for margin.  With proper calculations and/or measurements, there is no need for extra turns.  Some day I want to add a section about such measurements and calculations to my GDT tutorial page.  Still need to figure out how to explain it simply and clearly.
David Knierim

Offline Mathieu thm

  • High Voltage Enthusiast
  • *
  • Posts: 43
  • Karma: +3/-0
    • View Profile
Re: Some questions about my first drsstc.
« Reply #36 on: September 19, 2022, 09:05:35 PM »
Quote
I think that would be great.  Even just a copy of those images moved to a beginner page would be useful.  Perhaps a new thread about TO247 mounting or some such title.  Whenever you get to that (no hurry), I'll add a link to your thread from my low-inductance bridge tutorial.
It's done :)

Quote
Electrically fine.  Ceramic capacitors are likely smaller. 100V or so would probably suffice, though 200V would be prefered.  Down to 0.047uF would likely be fine, certainly no worse than leaving off TVS diodes.

I found 0.047uf capacitors, I'll let you know soon

I just thought something, and if you can confirm what I think would be great.
When the drsstc was working, I noticed that with low on times it produces small streamers (normal), when I increase the on time, almost nothing happens until the streamers suddenly become large, I therefore think that it is due to the detune, when the streamers are too small, the frequency was too much detune and when the on time is large enough, the streamers grow to "better" be in tune with the primary, which means that the streamers suddenly becomes large.
Sorry if that's not fully understandable, but is that what i think?
« Last Edit: September 19, 2022, 09:10:51 PM by mthome4 »
Mathieu Thomé

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2004
  • Karma: +103/-1
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Some questions about my first drsstc.
« Reply #37 on: September 20, 2022, 05:20:07 AM »
Quote
I just thought something, and if you can confirm what I think would be great.
When the drsstc was working, I noticed that with low on times it produces small streamers (normal), when I increase the on time, almost nothing happens until the streamers suddenly become large, I therefore think that it is due to the detune, when the streamers are too small, the frequency was too much detune and when the on time is large enough, the streamers grow to "better" be in tune with the primary, which means that the streamers suddenly becomes large.
Sorry if that's not fully understandable, but is that what i think?
Perfectly understandable, and normal.  I believe your explanation is exactly correct.

That level of detuning is ideal for making longest-possible arcs.  A little less detuning is better for music (MIDI interrupter).  At least that has been my experience.  Provides more linear response of arc size to pulse width, so better sound reproduction.
David Knierim

Offline Mathieu thm

  • High Voltage Enthusiast
  • *
  • Posts: 43
  • Karma: +3/-0
    • View Profile
Re: Some questions about my first drsstc.
« Reply #38 on: September 23, 2022, 10:55:21 PM »
I soldered the 0.047uf capacitors and their 1k resistor :



I'll adjust and test it probably tomorrow.

Quote
That level of detuning is ideal for making longest-possible arcs.  A little less detuning is better for music (MIDI interrupter).  At least that has been my experience.  Provides more linear response of arc size to pulse width, so better sound reproduction.

Thank you for your test feedback, I will test to see what is the best.
Mathieu Thomé

Offline Mathieu thm

  • High Voltage Enthusiast
  • *
  • Posts: 43
  • Karma: +3/-0
    • View Profile
Re: Some questions about my first drsstc.
« Reply #39 on: September 25, 2022, 09:43:24 PM »
I installed the bridge and I can feel that with the new cooling and the IGBTs mounted better, cooling is much better.
I tested the bridge with lipo in series for a total of about 110Vdc on the bus, here is what it gives :


We agree that the max voltage Vce of a single IGBT is not the "peak to peak" value on the oscilloscope but the "max" value which is around 236v, in theory I can have 1200v peak to peak to have a max of 600v on each IGBT ?
« Last Edit: September 25, 2022, 09:47:44 PM by mthome4 »
Mathieu Thomé

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2004
  • Karma: +103/-1
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Some questions about my first drsstc.
« Reply #40 on: September 26, 2022, 05:10:13 AM »
Quote
I tested the bridge with lipo in series for a total of about 110Vdc on the bus, here is what it gives :
Common inductance of scope leads and power connections to H-bridge output is significantly improved.  Makes viewing easier.

Quote
We agree that the max voltage Vce of a single IGBT is not the "peak to peak" value on the oscilloscope but the "max" value which is around 236v, in theory I can have 1200v peak to peak to have a max of 600v on each IGBT ?
Yes, presuming output waveform is symmetric, which it generally is.  More accurate to look separately at positive peak voltage and negative peak voltage.  Neither should exceed +-600V.
David Knierim

Offline Mathieu thm

  • High Voltage Enthusiast
  • *
  • Posts: 43
  • Karma: +3/-0
    • View Profile
Re: Some questions about my first drsstc.
« Reply #41 on: October 02, 2022, 12:46:19 AM »
So I was able to test the drsstc in MIDI and here is the result! I tested it at around 110Vac:

/>
/>

IGBTs are cold, which is good news.
So I think I can test with 150Vac to see the behavior.
I really feel that the DRSSTC is in better "health"
« Last Edit: October 02, 2022, 12:50:20 AM by mthome4 »
Mathieu Thomé

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2004
  • Karma: +103/-1
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Some questions about my first drsstc.
« Reply #42 on: October 02, 2022, 04:48:00 AM »
Quote
IGBTs are cold, which is good news.
So I think I can test with 150Vac to see the behavior.
I really feel that the DRSSTC is in better "health"
Yes, looks good.  BTW, there are a few places in the videos, especially first one, where sparks and sound seem to fade out.  I presume that corresponds to quieter parts of the MIDI input and not some coil performance issue.
David Knierim

Offline Mathieu thm

  • High Voltage Enthusiast
  • *
  • Posts: 43
  • Karma: +3/-0
    • View Profile
Re: Some questions about my first drsstc.
« Reply #43 on: October 02, 2022, 02:24:19 PM »
Quote
Yes, looks good.  BTW, there are a few places in the videos, especially first one, where sparks and sound seem to fade out.  I presume that corresponds to quieter parts of the MIDI input and not some coil performance issue.

In fact it's me who modifies the on-time to see how the sparks react.
« Last Edit: October 02, 2022, 02:35:05 PM by mthome4 »
Mathieu Thomé

Offline Mathieu thm

  • High Voltage Enthusiast
  • *
  • Posts: 43
  • Karma: +3/-0
    • View Profile
Re: Some questions about my first drsstc.
« Reply #44 on: October 04, 2022, 10:28:54 PM »
I have a question for the phase lead, I adjust the phase lead so that there is as little peak as possible but it works on both sides, let me explain, I can either adjust the phase on the left compared to the primary current or right compered to the primary current (on the scop).

So adjust the phase before the primary current, this corresponds to the igbt turn on a little before the primary feedback, and adjust the phase after the primary current, this corresponds to the igbt turn on a little after the primary feedback ?

I don't know if it translates like that but, I have either spikes that appear from rising edges or from falling edges, here is a diagram that illustrates what i mean:

it's not exactly like that in reality, it's just to illustrate what I think I understand




I don't know if I have explained well and if I have the right approach, if you can explain it to me better it may be good, but my question is (if that's what i think i understand) , what is better between the  two solutions ?
« Last Edit: October 04, 2022, 10:50:27 PM by mthome4 »
Mathieu Thomé

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2004
  • Karma: +103/-1
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Some questions about my first drsstc.
« Reply #45 on: October 05, 2022, 06:22:38 AM »
Quote
So adjust the phase before the primary current, this corresponds to the igbt turn on a little before the primary feedback, and adjust the phase after the primary current, this corresponds to the igbt turn on a little after the primary feedback ?

I don't know if it translates like that but, I have either spikes that appear from rising edges or from falling edges, here is a diagram that illustrates what i mean:
Seeing actual scope captures will make it more clear, especially zoomed-in horizontally to see details of spikes.  The general answer is that IGBT switching should occur slightly before zero current times.  IGBT turn-off occurs first, then the short dead-time gap, then turn-on of opposite IGBTs, then current crossing zero to other polarity.  This is proper phase-lead adjustment.  Should minimize spikes and minimize IGBT switching power loss.
David Knierim

Offline Mathieu thm

  • High Voltage Enthusiast
  • *
  • Posts: 43
  • Karma: +3/-0
    • View Profile
Re: Some questions about my first drsstc.
« Reply #46 on: October 08, 2022, 07:56:12 PM »
Quote
Seeing actual scope captures will make it more clear, especially zoomed-in horizontally to see details of spikes.  The general answer is that IGBT switching should occur slightly before zero current times.  IGBT turn-off occurs first, then the short dead-time gap, then turn-on of opposite IGBTs, then current crossing zero to other polarity.  This is proper phase-lead adjustment.  Should minimize spikes and minimize IGBT switching power loss.
Thanks, I'll look at the waveforms next time.

As everything worked well with 110Vac, I tested the DRSSTC at around 150 Vac, after having it running for quite a long time, the IGBTs were cold (with peaks at 900-1000w on the highest notes), so I think the phase lead is pretty well tune.
Here are two videos of the tests at around 150 Vac :
The videos are a little blurry, I apologize, I only noticed that after. It's quite hard to focus in the dark.

/>
/>
I also noticed that the OCD never turn on at 130 Vac even on heavy ground strike.
At about 150 Vac with an OCD of about 280A (I'm not sure, I need to check it) the OCD rarely triggers even on heavy ground strike.
« Last Edit: October 08, 2022, 08:44:06 PM by mthome4 »
Mathieu Thomé

Offline Mathieu thm

  • High Voltage Enthusiast
  • *
  • Posts: 43
  • Karma: +3/-0
    • View Profile
Re: Some questions about my first drsstc.
« Reply #47 on: October 09, 2022, 10:08:27 PM »
I need to know something, it may be a bit of a silly question, but I would like to be sure. When I do my tests with a dummy load, at about 110Vac, with MIDI mode, I put the on time at 100us and I leave it constantly at 100us, so for about two minutes, there is 300A passing through the primary circuit and I draw around 7A on the mains on the highest notes (IGBTs are almost not lukewarm).
When I test with the secondary coil, at the same bus voltage, I draw around 6A on the highest notes, but on the highest notes I feel like it's going "heated" more compared to the test with a dummy load and I dare not pushing the on time too high.
I think it's all psychological, we agree that IGBTs cannot heat up significantly more with the secondary coil with the same primary current and the same (even a little less) mains current draw ?
« Last Edit: October 09, 2022, 10:31:11 PM by mthome4 »
Mathieu Thomé

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2004
  • Karma: +103/-1
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Some questions about my first drsstc.
« Reply #48 on: October 10, 2022, 04:19:10 AM »
Quote
I need to know something, it may be a bit of a silly question, but I would like to be sure. When I do my tests with a dummy load, at about 110Vac, with MIDI mode, I put the on time at 100us and I leave it constantly at 100us, so for about two minutes, there is 300A passing through the primary circuit and I draw around 7A on the mains on the highest notes.
When I test with the secondary coil, at the same bus voltage, I also draw around 6-7A on the highest notes, but on the highest notes I feel like it's going "heated" more compared tested with a dummy load.
I think it's all psychological, we agree that IGBTs cannot heat up significantly more with the secondary coil with the same primary current and the same mains current draw ?
If I understanding correctly, you are comparing equal mains current.  Equal mains current indicates equal total power input.  (Mains power factor is likely the same at the same current.)  Equal mains current does not imply equal primary coil current.  It may require significantly higher primary RMS current to cause similar power dissipation.  Depends on the Q of your dummy load compared to the system Q of your operating Tesla coil including arc loading.

There are other possible (but less likely) causes for H-bridge power dissipation (heat) to be higher in operation than with dummy load, including arcs slightly disrupting driver circuit operation (if not sufficiently shielded).
David Knierim

Offline Mathieu thm

  • High Voltage Enthusiast
  • *
  • Posts: 43
  • Karma: +3/-0
    • View Profile
Re: Some questions about my first drsstc.
« Reply #49 on: October 16, 2022, 12:12:57 PM »
Quote
Seeing actual scope captures will make it more clear, especially zoomed-in horizontally to see details of spikes.
I made scope captures with a bus voltage of about 100v:




I also did a test with MIDI at around 140Vac, I set the on-time at 100us (I did not reduce the on-time on the high note like the last time) and I set the OCD at 300a :
/>
For the first time, I completely closed the case with the air vent plate.



I'm happy, the IGBTs are cold after this run, I also did a test with MIDI with slightly higher notes and the IGBTs don't even heat up !
I think the frequency of the primary circuit is a little bit high, I would try to adjust the primary for a frequency of about 5-10khz lower.
(currently the primary is at 190 khz and the secondary at 213 khz)
I also think I can increase the bus voltage a little more, I think to adjust my OCD for a final value of 325a.


« Last Edit: October 16, 2022, 03:22:46 PM by mthome4 »
Mathieu Thomé

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2004
  • Karma: +103/-1
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Some questions about my first drsstc.
« Reply #50 on: October 16, 2022, 09:47:59 PM »
Quote
I made scope captures with a bus voltage of about 100v:
Looks good.  Spikes near the beginning of bursts are common.  UD2.7 phase lead and duty cycle are less accurate at low current (low feedback input voltage).  Also, initial low current may not be enough to cause H-bridge output voltage transitions fast enough during the dead-time between one pair of IGBTs switching off and the other pair switching on. Your switching looks nice and clean after the normal start-of-burst low current spikes.

Enjoy your coil!  Hopefully no more IGBT failures.
David Knierim

Offline Mathieu thm

  • High Voltage Enthusiast
  • *
  • Posts: 43
  • Karma: +3/-0
    • View Profile
Re: Some questions about my first drsstc.
« Reply #51 on: October 21, 2022, 08:00:09 PM »
Quote
Looks good.  Spikes near the beginning of bursts are common.  UD2.7 phase lead and duty cycle are less accurate at low current (low feedback input voltage).  Also, initial low current may not be enough to cause H-bridge output voltage transitions fast enough during the dead-time between one pair of IGBTs switching off and the other pair switching on. Your switching looks nice and clean after the normal start-of-burst low current spikes.

I don't understand why I couldn't remove the spikes at the beginning of the burst, thank you for explaining it.

Quote
Enjoy your coil!  Hopefully no more IGBT failures.

Thanks a lot! Now I reach a little over 70 cm of streamers, which I find rather impressive for a DRSSTC of this size, I also wanted to thank you, you made me understand a lot of things that I did not understand well  :D
« Last Edit: October 21, 2022, 10:44:31 PM by mthome4 »
Mathieu Thomé

Offline Mathieu thm

  • High Voltage Enthusiast
  • *
  • Posts: 43
  • Karma: +3/-0
    • View Profile
Re: Some questions about my first drsstc.
« Reply #52 on: November 20, 2022, 02:28:22 PM »
Hello everyone,

Since my last post, I made some changes to my drsstc.

I detune the primary a bit more for about 8-9%, I added an ammeter on the front panel of the tesla coil and made a small protection on the left side to protect the primary circuit.



My final OCD setting is 315A, with a little more detune I was able to enlarge the arcs a bit.
I used it quite a lot, almost every weekend for about 1h-1h30 and I no longer have any IGBT problems!

/>





« Last Edit: November 20, 2022, 02:39:25 PM by mthome4 »
Mathieu Thomé

High Voltage Forum

Re: Some questions about my first drsstc.
« Reply #52 on: November 20, 2022, 02:28:22 PM »

 


* Recent Topics and Posts

post Re: improving zvs driver
[General Chat]
davekni
Today at 03:42:13 AM
post Re: MidiStick V2.0: Next gen tesla coil interrupter
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
TMaxElectronics
Today at 01:01:56 AM
post Re: improving zvs driver
[General Chat]
x-rayLamp
Today at 12:04:20 AM
post Re: ud 2.7c poped caps
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
January 29, 2023, 07:15:43 PM
post Re: ud 2.7c poped caps
[Dual Resonant Solid State Tesla coils (DRSSTC)]
costas_p
January 29, 2023, 06:13:45 PM
post Re: ud 2.7c poped caps
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mike
January 29, 2023, 05:53:11 PM
post Please, help me (financially) build a DRSSTC
[Crowdfunding]
Pavol
January 29, 2023, 03:02:55 PM
post ud 2.7c poped caps
[Dual Resonant Solid State Tesla coils (DRSSTC)]
costas_p
January 29, 2023, 02:04:59 PM
post Showa SP-103 Standalone Xray Head Teardown
[Transformer (Iron Core)]
Mads Barnkob
January 29, 2023, 01:04:56 PM
post Re: Ramped SSTC - Power Supply Question + General Improvements
[Solid State Tesla Coils (SSTC)]
davekni
January 29, 2023, 04:24:43 AM
post Re: Ramped SSTC - Power Supply Question + General Improvements
[Solid State Tesla Coils (SSTC)]
ZakW
January 29, 2023, 12:54:13 AM
post Re: Welding a MOT
[Transformer (Iron Core)]
Alberto
January 28, 2023, 11:47:22 PM
post Re: Ramped SSTC - Power Supply Question + General Improvements
[Solid State Tesla Coils (SSTC)]
ZakW
January 28, 2023, 11:34:58 PM
post Re: Welding a MOT
[Transformer (Iron Core)]
Mads Barnkob
January 28, 2023, 10:14:54 PM
post Re: Welding a MOT
[Transformer (Iron Core)]
Alberto
January 28, 2023, 04:53:39 PM
post Re: improving zvs driver
[General Chat]
davekni
January 28, 2023, 04:59:05 AM
post Re: First test of the SSTC
[Solid State Tesla Coils (SSTC)]
davekni
January 28, 2023, 04:44:51 AM
post Re: Ramped SSTC - Power Supply Question + General Improvements
[Solid State Tesla Coils (SSTC)]
davekni
January 28, 2023, 04:35:22 AM
post Re: First test of the SSTC
[Solid State Tesla Coils (SSTC)]
LoOdaK
January 28, 2023, 03:29:09 AM
post Re: Ramped SSTC - Power Supply Question + General Improvements
[Solid State Tesla Coils (SSTC)]
ZakW
January 27, 2023, 11:15:03 PM
post Re: Welding a MOT
[Transformer (Iron Core)]
Mads Barnkob
January 27, 2023, 08:43:51 PM
post Re: Welding a MOT
[Transformer (Iron Core)]
Alberto
January 27, 2023, 08:17:06 PM
post Re: Welding a MOT
[Transformer (Iron Core)]
Mads Barnkob
January 27, 2023, 08:07:12 PM
post Re: Welding a MOT
[Transformer (Iron Core)]
Alberto
January 27, 2023, 01:30:02 PM
post Re: improving zvs driver
[General Chat]
x-rayLamp
January 27, 2023, 01:01:01 PM
post Re: Phase lead problems on ud2.7c
[Dual Resonant Solid State Tesla coils (DRSSTC)]
vince1
January 27, 2023, 12:49:46 PM
post Re: Ramped SSTC - Power Supply Question + General Improvements
[Solid State Tesla Coils (SSTC)]
davekni
January 27, 2023, 06:07:49 AM
post Re: Phase lead problems on ud2.7c
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
January 27, 2023, 05:22:46 AM
post Re: Welding a MOT
[Transformer (Iron Core)]
davekni
January 27, 2023, 05:10:14 AM
post Re: First test of the SSTC
[Solid State Tesla Coils (SSTC)]
davekni
January 27, 2023, 05:05:40 AM
post Re: improving zvs driver
[General Chat]
davekni
January 27, 2023, 04:44:24 AM
post Phase lead problems on ud2.7c
[Dual Resonant Solid State Tesla coils (DRSSTC)]
vince1
January 27, 2023, 01:26:20 AM
post Re: improving zvs driver
[General Chat]
x-rayLamp
January 26, 2023, 08:13:24 PM
post Re: First test of the SSTC
[Solid State Tesla Coils (SSTC)]
LoOdaK
January 26, 2023, 12:02:14 PM
post Re: Welding a MOT
[Transformer (Iron Core)]
Alberto
January 26, 2023, 12:01:48 PM
post Re: Ramped SSTC - Power Supply Issues
[Solid State Tesla Coils (SSTC)]
ZakW
January 26, 2023, 05:35:32 AM
post Re: My completed 14-stage Cockroft-Walton voltage multiplier
[Voltage Multipliers]
MRMILSTAR
January 26, 2023, 04:55:24 AM
post Re: improving zvs driver
[General Chat]
davekni
January 26, 2023, 04:14:49 AM
post Re: My completed 14-stage Cockroft-Walton voltage multiplier
[Voltage Multipliers]
davekni
January 26, 2023, 03:57:19 AM
post Re: Welding a MOT
[Transformer (Iron Core)]
davekni
January 26, 2023, 03:42:01 AM
post Re: Welding a MOT
[Transformer (Iron Core)]
Da_Stier
January 25, 2023, 10:21:43 PM
post Welding a MOT
[Transformer (Iron Core)]
Alberto
January 25, 2023, 09:37:27 PM
post Re: My completed 14-stage Cockroft-Walton voltage multiplier
[Voltage Multipliers]
MRMILSTAR
January 25, 2023, 09:36:22 PM
post Re: improving zvs driver
[General Chat]
SteveN87
January 25, 2023, 08:35:49 PM
post improving zvs driver
[General Chat]
x-rayLamp
January 25, 2023, 04:32:37 PM
post Re: MidiStick V2.0: Next gen tesla coil interrupter
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
nzoomed
January 25, 2023, 08:27:01 AM
post Re: Ramped SSTC - Power Supply Issues
[Solid State Tesla Coils (SSTC)]
davekni
January 25, 2023, 04:08:42 AM
post Re: Ramped SSTC - Power Supply Issues
[Solid State Tesla Coils (SSTC)]
ZakW
January 24, 2023, 06:37:39 PM
post Re: Ramped SSTC - Power Supply Issues
[Solid State Tesla Coils (SSTC)]
davekni
January 24, 2023, 05:21:49 AM
post Re: 8 kW Induction Cooktop Teardown and Reverse engineering
[Electronic Circuits]
davekni
January 24, 2023, 04:41:01 AM
post Ramped SSTC - Power Supply Question + General Improvements
[Solid State Tesla Coils (SSTC)]
ZakW
January 23, 2023, 10:48:17 PM
post Re: 8 kW Induction Cooktop Teardown and Reverse engineering
[Electronic Circuits]
rikkitikkitavi
January 23, 2023, 10:23:54 PM
post Re: 160mm DRSSTC project
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mathieu thm
January 23, 2023, 07:21:10 PM
post Re: 160mm DRSSTC project
[Dual Resonant Solid State Tesla coils (DRSSTC)]
futurist
January 23, 2023, 06:21:07 PM
post Re: 8 kW Induction Cooktop Teardown and Reverse engineering
[Electronic Circuits]
Mads Barnkob
January 23, 2023, 01:23:41 PM
post Re: First test of the SSTC
[Solid State Tesla Coils (SSTC)]
davekni
January 22, 2023, 07:17:25 PM
post Re: First test of the SSTC
[Solid State Tesla Coils (SSTC)]
LoOdaK
January 22, 2023, 12:53:01 PM
post Re: First test of the SSTC
[Solid State Tesla Coils (SSTC)]
davekni
January 22, 2023, 06:21:40 AM
post Re: Creating arcs off my variac lol, popped my igbts
[Transformer (Iron Core)]
klugesmith
January 22, 2023, 04:45:46 AM
post Re: First test of the SSTC
[Solid State Tesla Coils (SSTC)]
LoOdaK
January 21, 2023, 11:47:55 PM
post [WTS] (free except shipping) 115V 3600RPM sync. motor
[Sell / Buy / Trade]
davekni
January 21, 2023, 08:50:05 PM
post Re: First test of the SSTC
[Solid State Tesla Coils (SSTC)]
davekni
January 21, 2023, 08:19:31 PM
post Re: First test of the SSTC
[Solid State Tesla Coils (SSTC)]
LoOdaK
January 21, 2023, 03:01:21 PM
post Re: First test of the SSTC
[Solid State Tesla Coils (SSTC)]
davekni
January 21, 2023, 05:44:15 AM
post Re: Creating arcs off my variac lol, popped my igbts
[Transformer (Iron Core)]
Avenger
January 21, 2023, 04:44:09 AM
post Re: First test of the SSTC
[Solid State Tesla Coils (SSTC)]
LoOdaK
January 20, 2023, 10:52:40 PM
post Re: QCW questions
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
January 20, 2023, 06:20:32 AM
post Re: QCW questions
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Rafft
January 20, 2023, 06:03:20 AM
post Re: QCW questions
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
January 20, 2023, 04:49:50 AM
post Re: First test of the SSTC
[Solid State Tesla Coils (SSTC)]
davekni
January 20, 2023, 04:19:23 AM
post Re: First test of the SSTC
[Solid State Tesla Coils (SSTC)]
Rafft
January 20, 2023, 03:12:39 AM
post Re: MidiStick V2.0: Next gen tesla coil interrupter
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
TMaxElectronics
January 19, 2023, 11:39:37 PM
post Re: First test of the SSTC
[Solid State Tesla Coils (SSTC)]
LoOdaK
January 19, 2023, 10:54:51 PM
post Re: 160mm DRSSTC project
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mathieu thm
January 19, 2023, 09:19:19 PM
post Re: First test of the SSTC
[Solid State Tesla Coils (SSTC)]
Mads Barnkob
January 19, 2023, 08:33:18 PM
post Re: QCW questions
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Rafft
January 19, 2023, 03:48:02 PM
post Re: First test of the SSTC
[Solid State Tesla Coils (SSTC)]
LoOdaK
January 19, 2023, 11:09:39 AM
post Re: QCW questions
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
January 19, 2023, 04:30:49 AM
post Re: 8 kW Induction Cooktop Teardown and Reverse engineering
[Electronic Circuits]
klugesmith
January 19, 2023, 03:10:56 AM
post Re: 8 kW Induction Cooktop Teardown and Reverse engineering
[Electronic Circuits]
Mads Barnkob
January 18, 2023, 08:23:57 PM
post Re: Creating arcs off my variac lol, popped my igbts
[Transformer (Iron Core)]
Mads Barnkob
January 18, 2023, 08:13:28 PM
post Re: 160mm DRSSTC project
[Dual Resonant Solid State Tesla coils (DRSSTC)]
dbach
January 18, 2023, 05:56:11 PM
post Creating arcs off my variac lol, popped my igbts
[Transformer (Iron Core)]
breezetix
January 18, 2023, 08:24:15 AM
post Re: QCW questions
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Rafft
January 18, 2023, 06:40:30 AM
post Re: QCW questions
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
January 18, 2023, 06:29:45 AM
post Re: QCW questions
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Rafft
January 18, 2023, 04:56:33 AM
post Re: QCW questions
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
January 18, 2023, 04:39:55 AM
post Re: 8 kW Induction Cooktop Teardown and Reverse engineering
[Electronic Circuits]
davekni
January 18, 2023, 04:27:49 AM
post Re: 160mm DRSSTC project
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mathieu thm
January 17, 2023, 11:16:00 PM
post Re: 8 kW Induction Cooktop Teardown and Reverse engineering
[Electronic Circuits]
rikkitikkitavi
January 17, 2023, 09:37:16 PM
post Re: 8 kW Induction Cooktop Teardown and Reverse engineering
[Electronic Circuits]
klugesmith
January 17, 2023, 05:38:24 PM
post Re: I need a review of my circuit. Polarities specifically
[Electronic Circuits]
RoamingD
January 17, 2023, 03:35:47 PM
post Re: 8 kW Induction Cooktop Teardown and Reverse engineering
[Electronic Circuits]
rikkitikkitavi
January 17, 2023, 12:43:26 PM
post Re: 8 kW Induction Cooktop Teardown and Reverse engineering
[Electronic Circuits]
klugesmith
January 17, 2023, 07:51:05 AM
post Re: QCW questions
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Rafft
January 17, 2023, 07:02:37 AM
post Re: 8 kW Induction Cooktop Teardown and Reverse engineering
[Electronic Circuits]
Twospoons
January 17, 2023, 05:41:59 AM
post Re: QCW questions
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
January 17, 2023, 05:41:38 AM
post Re: QCW questions
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Rafft
January 17, 2023, 05:06:24 AM
post Re: 8 kW Induction Cooktop Teardown and Reverse engineering
[Electronic Circuits]
davekni
January 17, 2023, 05:04:21 AM
post Re: 8 kW Induction Cooktop Teardown and Reverse engineering
[Electronic Circuits]
klugesmith
January 17, 2023, 04:41:26 AM

Sitemap 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 
SimplePortal 2.3.6 © 2008-2014, SimplePortal