Author Topic: Problem with SSTC  (Read 10808 times)

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2827
  • Karma: +134/-2
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Problem with SSTC
« Reply #20 on: June 18, 2022, 04:27:13 AM »
Quote
The only acquisition modes available are Auto, Normal and Single.
Those are usually called "trigger modes".  Perhaps your scope does not have any selection for acquisition modes.

Quote
But in fact I did it as you can see from the first post. But since the igbt's continued to burn, I thought of carrying out this test.
Thank you for the reminder.  Yes, I see.  Would guess the JavaTC results to be more reliable than the meter's.

Quote
while on pin 2 and therefore also on 3 and 4 I see nothing, high frequency noise only.
Pin 2/3 and 4 should have square waves from 0 to 5V.  Is the high-frequency noise 0-5V?
Is your scope probe 1X or 10X?  1X probes have much higher capacitance and low bandwidth.  If 1X, probe capacitance is likely significantly higher than antenna etc. capacitance.  You may need a significantly larger resistor value.  The square wave on pin 2/3 and 4 will be somewhat noisy, as the antenna is still picking up random signals from adjacent electronics (light fixtures etc.).  Pick a resistor that gets average around operating frequency.  (If probe attenuation can be switched, use 10X for all your probing.  Be sure to adjust probe compensation using a ~1kHz square wave signal, usually provided by the scope for that purpose.)

Quote
Yes, unused inputs are connected together.
I presume they each input is connected to output of another inverter.  If an inverter input is connected to its own output, it will oscillate.

Quote
I tried to redo the tests by applying to the antenna a 200kHz square wave generated with an arduino.
Now on the gate of the first igbt I see this:
This looks like roughly +-25V gate drive.  Are you sure your GDT is 8:12 and not 8:16 or some other such 1:2 ratio?  Or, did you increase driver supply voltage above 12V?  Or perhaps you are using 10X scope probe that has not been adjusted to match scope input capacitance.
Also, do you know if your arduino output is square (50% duty cycle)?  Or, does it match the duty cycle of gate signals?  If the driver circuit is causing the duty cycle change shown, that indicates some problem.
David Knierim

Offline Nunu00

  • High Voltage Enthusiast
  • *
  • Posts: 26
  • Karma: +1/-0
    • View Profile
Re: Problem with SSTC
« Reply #21 on: June 18, 2022, 11:56:14 AM »
Quote
Pin 2/3 and 4 should have square waves from 0 to 5V.  Is the high-frequency noise 0-5V?
Is your scope probe 1X or 10X?  1X probes have much higher capacitance and low bandwidth.  If 1X, probe capacitance is likely significantly higher than antenna etc. capacitance.  You may need a significantly larger resistor value.  The square wave on pin 2/3 and 4 will be somewhat noisy, as the antenna is still picking up random signals from adjacent electronics (light fixtures etc.).  Pick a resistor that gets average around operating frequency.  (If probe attenuation can be switched, use 10X for all your probing.  Be sure to adjust probe compensation using a ~1kHz square wave signal, usually provided by the scope for that purpose.)
No, the noise has a lower voltage.
My probe can switch to 10X but if I do it on pin 1 of the HC14 I don't see anything anymore. In addition, as soon as the probe touches pin 1, the GDT stops emitting a kind of buzzing sound.
Another problem is that my oscilloscope does not have the 1kHz output to compensate the probe so the 10X measurements I don't think are very reliable.

Quote
I presume they each input is connected to output of another inverter.  If an inverter input is connected to its own output, it will oscillate.
No, as you can see from the photo only the unused inputs are connected together.


Quote
This looks like roughly +-25V gate drive.  Are you sure your GDT is 8:12 and not 8:16 or some other such 1:2 ratio?  Or, did you increase driver supply voltage above 12V?  Or perhaps you are using 10X scope probe that has not been adjusted to match scope input capacitance.
Also, do you know if your arduino output is square (50% duty cycle)?  Or, does it match the duty cycle of gate signals?  If the driver circuit is causing the duty cycle change shown, that indicates some problem.
The GDT should be wound correctly, I think the problem is that the probe is not compensated correctly.
The output of the arduino is not perfectly at 50%, connecting it to the antenna, on pin 1 I see this

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2827
  • Karma: +134/-2
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Problem with SSTC
« Reply #22 on: June 18, 2022, 08:25:55 PM »
Quote
No, the noise has a lower voltage.
My probe can switch to 10X but if I do it on pin 1 of the HC14 I don't see anything anymore. In addition, as soon as the probe touches pin 1, the GDT stops emitting a kind of buzzing sound.
Another problem is that my oscilloscope does not have the 1kHz output to compensate the probe so the 10X measurements I don't think are very reliable.
Lower voltage may be due to limited bandwidth of 1x probe.  Can you make your arduino output a 1kHz (approximately) square (approximately) wave for adjusting the probe in 10x mode?  Scoping will be much more helpful with a 10x probe.

What is your scope's vertical sensitivity range (min and max volts/division)?  Or, if you have a link to a data sheet or specifications for your scope, I'll take a look.  Perhaps volts/division doesn't go low enough to see pin1 signal with 10x probe.  Still, 10x will be MUCH better for almost all scoping:  much lower loading on circuit nodes and much higher bandwidth.

Quote
No, as you can see from the photo only the unused inputs are connected together.
Are these four inputs also wired to ground or +5V on another layer?  If wired only to each other, they can still oscillate and cause issues.

Quote
The GDT should be wound correctly, I think the problem is that the probe is not compensated correctly.
The output of the arduino is not perfectly at 50%, connecting it to the antenna, on pin 1 I see this
That explains the duty cycle on gate voltage.  No issue there.  Gate voltage of +-25V is still above 1.5 x +-10V (or whatever the driver manages to generate on 12V supply, certainly not more than +-12V).  If you can get 10x working properly, I recommend probing GDT primary leads and then GDT output (Vge) with the same scope settings.  Then it will be clear if there is a GDT ratio issue or some other problem with scoping.
David Knierim

Offline Nunu00

  • High Voltage Enthusiast
  • *
  • Posts: 26
  • Karma: +1/-0
    • View Profile
Re: Problem with SSTC
« Reply #23 on: June 18, 2022, 09:55:23 PM »
Quote
Lower voltage may be due to limited bandwidth of 1x probe.  Can you make your arduino output a 1kHz (approximately) square (approximately) wave for adjusting the probe in 10x mode?  Scoping will be much more helpful with a 10x probe.

What is your scope's vertical sensitivity range (min and max volts/division)?  Or, if you have a link to a data sheet or specifications for your scope, I'll take a look.  Perhaps volts/division doesn't go low enough to see pin1 signal with 10x probe.  Still, 10x will be MUCH better for almost all scoping:  much lower loading on circuit nodes and much higher bandwidth.
The oscilloscope datasheet is this:
 [ You are not allowed to view attachments ] (I don't know why i get this error)

I tried to compensate the probe with a 1 kHz signal generated by the arduino but I can't do better than this:
 [ You are not allowed to view attachments ]

However trying this way on pin 1 I see this:
 [ You are not allowed to view attachments ]

While if I set the probe on 1X I see the wave that I had posted a few messages ago.

Quote
Are these four inputs also wired to ground or +5V on another layer?  If wired only to each other, they can still oscillate and cause issues.
No, they are just connected to each other.

Quote
That explains the duty cycle on gate voltage.  No issue there.  Gate voltage of +-25V is still above 1.5 x +-10V (or whatever the driver manages to generate on 12V supply, certainly not more than +-12V).  If you can get 10x working properly, I recommend probing GDT primary leads and then GDT output (Vge) with the same scope settings.  Then it will be clear if there is a GDT ratio issue or some other problem with scoping.
I tried again to carry out these measurements and on the GDT primary I see this:


While on the respective secondaries I see this:



« Last Edit: June 18, 2022, 10:04:37 PM by Nunu00 »

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2827
  • Karma: +134/-2
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Problem with SSTC
« Reply #24 on: June 18, 2022, 11:07:46 PM »
Quote
The oscilloscope datasheet is this:
 [ You are not allowed to view attachments ] (I don't know why i get this error)
I've gotten the same error once or twice.  Thought perhaps it was an issue of file extension being JPG instead of jpg, but that doesn't appear to be the issue.  So, I don't know why either.

Quote
I tried to compensate the probe with a 1 kHz signal generated by the arduino but I can't do better than this:
That is unfortunate!  Looks like the probe is designed for a scope with higher input capacitance than yours.  But, it also looks like arduino is putting out only 1.5V square wave, which seems too low.  Do you know what the arduino output voltage should be?  Perhaps the real issue is with the probe, internal 9meg resistor has failed to ~20-30meg.  That would explain both the inability to compensate and the low signal amplitude at low frequency.  Do you have a DMM?  I'd suggest measuring across scope probe input, in both 1x and 10x mode.  Should be 1meg in 1x mode and 10meg in 10x mode.  If 10meg is way off, can you find another 10x probe somewhere?  Even this cheap scope will be very valuable as long as it is functioning correctly.
If you know what voltage the arduino is outputting (typically same as supply voltage, either 3.3V or 5V), you could adjust the probe to give the correct amplitude at high frequency (100kHz and up).  Then 10x mode would be useful for measuring AC signals at TC operating frequency.

Quote
However trying this way on pin 1 I see this:
Yes, that makes sense.  Self-oscillating resistor value will need to be higher to get frequency reasonable without scope probe loading.  As I'd mentioned, it will be noisy due to antenna picking up random signals.

Quote
No, they are just connected to each other.
That should be fixed.  You could add a solder-bridge between pins 13 and 14.  Then all unused inputs are connected to +5V.

Quote
I tried again to carry out these measurements and on the GDT primary I see this:
Ratio looks reasonably close to 1:1.5.  Given probe compensation failure, I'll ignore absolute amplitudes.
David Knierim

Offline Nunu00

  • High Voltage Enthusiast
  • *
  • Posts: 26
  • Karma: +1/-0
    • View Profile
Re: Problem with SSTC
« Reply #25 on: June 19, 2022, 01:56:08 AM »
I measured the resistance of the probe and in 1X mode I read 340 ohm while in 10X mode I read 9 Mega ohm, i get the same results also on another identical probe.
I also have the "probe" that comes with the oscilloscope which is simply a pair of alligator clips and the resistance of this one is zero ohm (but I can't switch it to 10X mode).

The input resistance of the oscilloscope is about 400k ohm even if 1 mega is written on the datasheet (this is also confirmed by a video I saw on youtube).
« Last Edit: June 19, 2022, 02:04:52 AM by Nunu00 »

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2827
  • Karma: +134/-2
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Problem with SSTC
« Reply #26 on: June 19, 2022, 03:42:14 AM »
Quote
I measured the resistance of the probe and in 1X mode I read 340 ohm while in 10X mode I read 9 Mega ohm, i get the same results also on another identical probe.
I also have the "probe" that comes with the oscilloscope which is simply a pair of alligator clips and the resistance of this one is zero ohm (but I can't switch it to 10X mode).
Sorry, I wasn't clear.  I was referring to measuring probe tip to probe ground with probe connected to scope and scope turned on.  But I think your measurements are sufficient to understand the issue given comment quoted below.

Quote
The input resistance of the oscilloscope is about 400k ohm even if 1 mega is written on the datasheet (this is also confirmed by a video I saw on youtube).
Is this measured with the scope turned on?  If so, that explains why 10x probes do not work correctly.  Since the scope specifications both say "1 meg" and define input capability when using 10x probes, this would be blatantly dishonest advertising.

This is definitely limiting on scope usefulness :(  Probably best to stay with 1x and probe only low-impedance nodes (chip outputs in general).  For adjusting self-oscillation frequency, I'd probe HC14 pin 4.  That way scope load doesn't affect delay of pin1-2 inverter, as scope probe capacitance may affect even the output terminal somewhat.  Or, probe output of MCP14E5 to ground.  (Better not to probe across GDT input terminals.  Even with an isolated battery-powered scope, connecting scope "ground" lead to an active signal adds a large antenna to that node, which will couple to your SSTC feedback antenna and change circuit operation.)  Always ground scope probe to ECB ground, or to low-side IGBT emitter (or high-side emitter when half-bridge is not powered) when measuring gate voltages.

Back to possible general issues:  Besides connecting unused HC14 inputs to +5V or ground, the antenna lead wiring may need shielding.  Goal is to minimize unwanted feedback from half-bridge or other primary-side signals.  Some designs use coax cable or other shielded cable to bring the antenna connection from driver board to outside the box above primary coil.  Then the antenna is picking up mostly secondary voltage.  Shield is connected to driver board ground.  (Such coax or other shielded wiring will add capacitance.  Best to make that change before adjusting self-oscillation feedback resistor value.  Should reduce noise too.)
« Last Edit: June 19, 2022, 03:45:15 AM by davekni »
David Knierim

Offline Nunu00

  • High Voltage Enthusiast
  • *
  • Posts: 26
  • Karma: +1/-0
    • View Profile
Re: Problem with SSTC
« Reply #27 on: June 19, 2022, 11:48:17 AM »
Quote
This is definitely limiting on scope usefulness :(  Probably best to stay with 1x and probe only low-impedance nodes (chip outputs in general).  For adjusting self-oscillation frequency, I'd probe HC14 pin 4.  That way scope load doesn't affect delay of pin1-2 inverter, as scope probe capacitance may affect even the output terminal somewhat.  Or, probe output of MCP14E5 to ground.  (Better not to probe across GDT input terminals.  Even with an isolated battery-powered scope, connecting scope "ground" lead to an active signal adds a large antenna to that node, which will couple to your SSTC feedback antenna and change circuit operation.)  Always ground scope probe to ECB ground, or to low-side IGBT emitter (or high-side emitter when half-bridge is not powered) when measuring gate voltages.
I connected the unused inputs of the HC14 to +5V and then I did several tests (all with the probe in 1X) with different resistors but the only thing I get is the frequency variation on pin 1. On pin 4 and also on the MCP inputs I only see noise . The only way to get something is to touch the antenna with the hand, in this way the frequency obviously changes but increases the amplitude of the signal on pin 1 up to over 1 Vpp and on pin 4 i get a quite good square wave.

Quote
Back to possible general issues:  Besides connecting unused HC14 inputs to +5V or ground, the antenna lead wiring may need shielding.  Goal is to minimize unwanted feedback from half-bridge or other primary-side signals.  Some designs use coax cable or other shielded cable to bring the antenna connection from driver board to outside the box above primary coil.  Then the antenna is picking up mostly secondary voltage.  Shield is connected to driver board ground.  (Such coax or other shielded wiring will add capacitance.  Best to make that change before adjusting self-oscillation feedback resistor value.  Should reduce noise too.)
I could do this test but the only type of shielded cable at my disposal is the one used for the satellite TV signal. Could it work well?

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2827
  • Karma: +134/-2
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Problem with SSTC
« Reply #28 on: June 19, 2022, 06:49:26 PM »
Quote
I could do this test but the only type of shielded cable at my disposal is the one used for the satellite TV signal. Could it work well?
Yes, that cable will be great electrically.  Hope it isn't physically too stiff to work with.

Quote
On pin 4 and also on the MCP inputs I only see noise.
Only possible explanation I can come up with is that antenna is coupling to GDT wires or other circuitry, causing oscillations that are too fast for your scope to measure cleanly.  I'd suggest scoping HC14 pin 4 one more time after changing to shielded cable for connection from antenna to circuit board.  If it is still noise, I'm at a loss for ideas.  Any very-nearby radio stations or other sources of strong RF signals being picked up by the antenna??  Not too likely.
David Knierim

Offline Nunu00

  • High Voltage Enthusiast
  • *
  • Posts: 26
  • Karma: +1/-0
    • View Profile
Re: Problem with SSTC
« Reply #29 on: June 19, 2022, 07:50:17 PM »
I replaced the antenna cable with the coaxial one but nothing has changed. On pin 1 I always see the same signal from about 600 mVpp but on pins 2/3 and 4 I see nothing. I might think the chip has some problem but applying an external signal it works fine so I don't know.

Quote
Any very-nearby radio stations or other sources of strong RF signals being picked up by the antenna??
No, the only repeater near home is for mobile 4G but I don't think it can give me problems. Maybe it could be the switching power supply I'm using to power the circuit while doing the tests?

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2827
  • Karma: +134/-2
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Problem with SSTC
« Reply #30 on: June 19, 2022, 08:58:16 PM »
Quote
On pin 1 I always see the same signal from about 600 mVpp but on pins 2/3 and 4 I see nothing.
Do you mean "nothing" or is there noise centered around 2.5V (middle between 0V and logic supply of 5V)?  Perhaps it is worthwhile to check voltages with your DMM (presumably 10meg input impedance) on HC14 pin1, 2/3, and 4.  I wonder if one of the antenna input diodes or HC14 is partially damaged, pulling pin1 close to 5V, and only the scope's 400k to ground brings it into operating range.  Seems quite unlikely, but I'm running out of ideas.

Quote
No, the only repeater near home is for mobile 4G but I don't think it can give me problems. Maybe it could be the switching power supply I'm using to power the circuit while doing the tests?
4G should be much too high frequency to cause issues.  If it was the switching supply, shielded antenna cable should have helped at least some.  Cable shield is connected to circuit board ground, correct?
David Knierim

Offline Mads Barnkob

  • Administrator
  • Executive Board Member
  • *****
  • Posts: 2301
  • Karma: +73/-0
  • Denmark
    • View Profile
    • Kaizer Power Electronics
Re: Problem with SSTC
« Reply #31 on: June 19, 2022, 09:42:02 PM »
Quote
The input resistance of the oscilloscope is about 400k ohm even if 1 mega is written on the datasheet (this is also confirmed by a video I saw on youtube).
Is this measured with the scope turned on?  If so, that explains why 10x probes do not work correctly.  Since the scope specifications both say "1 meg" and define input capability when using 10x probes, this would be blatantly dishonest advertising.

This is definitely limiting on scope usefulness :(

These cheap "DSO" oscilloscopes are grossly over advertised. It is downright lies and slander, its the "normal" Chinese exaggerated branding crap like tasers with 5MV output.

I would not trust any of these scopes at frequencies over 1/10th of their advertised band width. This is at least my own experience.

/>
With dodgy input "impedance" given as 1M instead of a input resistance and capacitance, its worthless. It only makes sense to talk about impedance at a specific frequency.

Nunu00: Spend the money on a proper cheap DSO from Rigol, Hantek or similar. Even a old analog 20 MHz oscilloscope from a flea-market will be better. It might not show frequency and voltages, but at least you can trust its measurements :)
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 Nunu00

  • High Voltage Enthusiast
  • *
  • Posts: 26
  • Karma: +1/-0
    • View Profile
Re: Problem with SSTC
« Reply #32 on: June 19, 2022, 11:00:41 PM »
Quote
Do you mean "nothing" or is there noise centered around 2.5V (middle between 0V and logic supply of 5V)?  Perhaps it is worthwhile to check voltages with your DMM (presumably 10meg input impedance) on HC14 pin1, 2/3, and 4.  I wonder if one of the antenna input diodes or HC14 is partially damaged, pulling pin1 close to 5V, and only the scope's 400k to ground brings it into operating range.  Seems quite unlikely, but I'm running out of ideas.
I made the measurements with the DMM (according to the datasheet it has an input impedance of 10M and it shouldn't be crap like the oscilloscope) and on pins 1, 2/3 and 4 it detects about + 2.3V.
I also repeated the measurements with the oscilloscope in AC mode and on pin 1 I see this (I know the frequency is wrong but I was doing some tests with higher resistance values):
 [ You are not allowed to view attachments ]

On pin 4 in AC mode I see this:
 [ You are not allowed to view attachments ]

While in DC mode I see this:
 [ You are not allowed to view attachments ]

The measurements seem to me consistent between the DMM and the oscilloscope.


Quote
Spend the money on a proper cheap DSO from Rigol, Hantek or similar. Even a old analog 20 MHz oscilloscope from a flea-market will be better. It might not show frequency and voltages, but at least you can trust its measurements :)
Unfortunately I know, in fact I am planning to buy a Rigol DS1102Z.
« Last Edit: June 19, 2022, 11:57:49 PM by Mads Barnkob »

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2827
  • Karma: +134/-2
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Problem with SSTC
« Reply #33 on: June 20, 2022, 07:06:54 PM »
Quote
I also repeated the measurements with the oscilloscope in AC mode and on pin 1 I see this (I know the frequency is wrong but I was doing some tests with higher resistance values):
That may be about the correct resistance, as frequency will increase when probe is removed.  No way to know until probing pin 4 works correctly.

Quote
On pin 4 in AC mode I see this:
It is clear now that the HC14 is oscillating at 9 or 11MHz (aliasing with scope's 20Msps sample rate).  Amplitude is low because this is well beyond even the scope's specified 5MHz bandwidth.  This high-frequency oscillation is likely present only when loaded by the scope probe.  However, I think it may indicate a more general issue.  Looking back at images of the circuit board, ground appears to be routed in traces that don't form any sort of grid or other approximation of a ground plane.  In particular for this 9MHz oscillation, ground trace from HC14-7 to 0.1uF bypass capacitor is rather long.  I'd suggest hand-adding wires directly between ground points on the back of the board.  Or, even better, use copper foil to add ground connections.  A small capacitor directly between HC14 pins 7 and 14 would also help.  (Other chips would likely benefit from short-leaded bypass capacitors on the back too.  I haven't examined the overall layout in detail.)  Just adding the back-side HC14 bypass capacitor would likely be enough to stop the 9MHz oscillation.  A complete ground grid will prevent other issues.  This specific 9MHz oscillation may not be the exact problem causing IGBT frying.  It may be some other parasitic oscillation caused by ground (or supply) wiring inductance.

Quote
Unfortunately I know, in fact I am planning to buy a Rigol DS1102Z.
A good scope will definitely help with debug, especially allowing proper use of 10x probing.  In the mean time, you could probe at 1x with a ~1k resistor between probe tip and HC14-4 or whatever pin you are probing.  That will likely isolate probe capacitance from HC14-4 enough to prevent oscillation, but will further reduce measurement bandwidth.
David Knierim

Offline Nunu00

  • High Voltage Enthusiast
  • *
  • Posts: 26
  • Karma: +1/-0
    • View Profile
Re: Problem with SSTC
« Reply #34 on: June 21, 2022, 04:44:13 PM »
Ok, I think that until I will buy a decent oscilloscope I will stop testing because I feel like I'm wasting time and I'm wasting your time too.
Thanks, see you soon.

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2827
  • Karma: +134/-2
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Problem with SSTC
« Reply #35 on: June 22, 2022, 03:57:54 AM »
Quote
Ok, I think that until I will buy a decent oscilloscope I will stop testing because I feel like I'm wasting time and I'm wasting your time too.
While you wait for a new scope, you could still be enhancing ground connections and bypass caps on your board.  That will clean up signal quality and make debugging even easier.  Might even fix the real problem.

Yes, a new scope will make debugging more efficient.  Still, it has been a somewhat fun challenge to decipher what is happening based on constrained measurements.  The 9-11MHz oscillation when probing HC14-4 is a key clue to ground interconnect and chip power bypassing being insufficient.

Good luck going forward!
David Knierim

Offline Nunu00

  • High Voltage Enthusiast
  • *
  • Posts: 26
  • Karma: +1/-0
    • View Profile
Re: Problem with SSTC
« Reply #36 on: April 17, 2023, 03:16:40 PM »
Finally after almost a year I was able to buy a decent oscilloscope. So I am again available to carry out any test you recommend to try to definitively solve the problems with this circuit.
In the meantime I was thinking of redoing the pcb of the circuit since the ground plane is totally missing and some decoupling capacitors are also missing.  I also accept advice on possible improvements to be made to the circuit.
« Last Edit: April 18, 2023, 04:04:58 PM by Nunu00 »

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2827
  • Karma: +134/-2
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Problem with SSTC
« Reply #37 on: April 19, 2023, 04:17:44 AM »
Quote
I also accept advice on possible improvements to be made to the circuit.
I'd suggest two changes:

1) Change to a 1:1:1 GDT.  Use 15V for existing driver chip.  Better, change to a higher-voltage driver chip such as:
NCP81071
IX4340
UCC27624
1EDN751x or 1EDN851x single-channel SOT23-5/6 pin package
IVCR2401
and run at 18 or 19V.  1:1:1 GDT will have much lower leakage inductance than 1:1.5:1.5.  Even better, 1:1:1:1 with paralleled primaries as in:
    https://highvoltageforum.net/index.php?topic=1854.msg13949#msg13949

2: Lower priority, but I'd add a resistor from 74HC14-2 to right side of 1kR input resistor.  That will keep 74HC14-1 input close to threshold voltage, so make startup more reliable.  For especially reliable startup, adjust the value of this resistor until 74HC14 oscillates close to your operating frequency when interrupter input is off.

For layout, I start with ground plane covering entire back side of ECB.  Attempt to place parts so that all non-ground routing can be done on top layer only.  Where that isn't possible, make only short cross-over connections on back side (which create holes in ground plane).  Avoid long traces on back side (long gaps in ground).  If absolutely necessary to have a long gap, make ground ties across gap on top side.  Place bypass caps near power pins.  Ground should be low inductance, so less critical to have cap near ground pins.
David Knierim

Offline Nunu00

  • High Voltage Enthusiast
  • *
  • Posts: 26
  • Karma: +1/-0
    • View Profile
Re: Problem with SSTC
« Reply #38 on: April 19, 2023, 05:31:14 PM »
I wanted to redo the pcb by reusing the components of the old one to save money but the new recommended drivers are all in the smd package. At this point I wonder if it would make sense to redo the whole board in the smd version or just replace the driver and use all the other components in through hole package.

Regarding the second advice, I can try to make all the connections only on the top layer but I think that in the power area that goes from the GDT to the IGBTs it is not possible.
 
The last doubts are whether the plane ground will have to be only on the bottom layer and whether it will also have to cover the IGBT area.
 
Then i wonder if it is also better to put the mains voltage rectification stage on this board with the two large electrolytics which are currently on a other board.

Furthermore, the size of the traces in the old pcb seems to me really exaggerated, in fact the largest are 6mm. Do they have to be that big or can I make them smaller?

Offline davekni

  • Executive Board Member
  • *******
  • Posts: 2827
  • Karma: +134/-2
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Problem with SSTC
« Reply #39 on: April 20, 2023, 04:11:13 AM »
All the options have advantages and disadvantages.  Just helped a friend debug his EVR Mirrobrute kit.  It uses 15V drivers and 1:1:1 GDT.  Earlier versions used higher GDT ratios.  Going above 15V helps if you want to push IGBT current higher.

6mm traces are great for IC power connections.  Even for those connections, thinner traces are usually fine as long as bypass capacitors are very close to power pins.  For logic connections such as to/from 74HC14, thin traces are preferred to minimize stray capacitance.

The EVR kit I just saw integrates IGBTs with control.  Most designs don't.  My personal preference is to keep those separate.  Power board can often be made by dremel-tool cuts in raw (unetched) copper clad board, at least of gate signals are kept separate, as in this thread:
    https://highvoltageforum.net/index.php?topic=1324.msg9795#msg9795
« Last Edit: April 20, 2023, 06:14:51 AM by davekni »
David Knierim

High Voltage Forum

Re: Problem with SSTC
« Reply #39 on: April 20, 2023, 04:11:13 AM »

 


* Recent Topics and Posts

post Re: UD+ debugging (randomly skipping pulses)
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
Today at 04:12:52 AM
post Re: UD+ debugging (randomly skipping pulses)
[Dual Resonant Solid State Tesla coils (DRSSTC)]
AstRii
Today at 03:22:21 AM
post Re: Success!!!
[Spark Gap Tesla Coils (SGTC)]
MRMILSTAR
June 15, 2024, 09:03:06 PM
post Success!!!
[Spark Gap Tesla Coils (SGTC)]
Terry
June 15, 2024, 08:29:43 PM
post Re: Benjamin's DRSSTC3
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Benjamin Lockhart
June 15, 2024, 05:38:36 AM
post Re: Benjamin's DRSSTC3
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Jj
June 15, 2024, 03:09:41 AM
post Re: Are TVS diodes necessary for full bridge?
[Solid State Tesla Coils (SSTC)]
Simranjit
June 15, 2024, 01:53:07 AM
post Re: UD+ debugging (randomly skipping pulses)
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
June 15, 2024, 01:07:30 AM
post Re: UD+ debugging (randomly skipping pulses)
[Dual Resonant Solid State Tesla coils (DRSSTC)]
AstRii
June 15, 2024, 12:45:22 AM
post Re: UD+ debugging (randomly skipping pulses)
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
June 14, 2024, 10:41:03 PM
post UD+ debugging (randomly skipping pulses)
[Dual Resonant Solid State Tesla coils (DRSSTC)]
AstRii
June 14, 2024, 10:20:00 PM
post Re: Are TVS diodes necessary for full bridge?
[Solid State Tesla Coils (SSTC)]
davekni
June 14, 2024, 09:47:48 PM
post Fronius IG Plus 2LT-2 6kW Solar Inverter Teardown
[Electronic Circuits]
Mads Barnkob
June 14, 2024, 09:40:33 PM
post Re: Are TVS diodes necessary for full bridge?
[Solid State Tesla Coils (SSTC)]
Simranjit
June 14, 2024, 07:44:12 AM
post Re: Are TVS diodes necessary for full bridge?
[Solid State Tesla Coils (SSTC)]
davekni
June 14, 2024, 05:21:55 AM
post Re: Are TVS diodes necessary for full bridge?
[Solid State Tesla Coils (SSTC)]
Simranjit
June 14, 2024, 03:24:43 AM
post Re: Are TVS diodes necessary for full bridge?
[Solid State Tesla Coils (SSTC)]
davekni
June 14, 2024, 03:13:44 AM
post Re: Are TVS diodes necessary for full bridge?
[Solid State Tesla Coils (SSTC)]
Simranjit
June 14, 2024, 02:27:06 AM
post Re: Are TVS diodes necessary for full bridge?
[Solid State Tesla Coils (SSTC)]
davekni
June 14, 2024, 01:56:58 AM
post Are TVS diodes necessary for full bridge?
[Solid State Tesla Coils (SSTC)]
Simranjit
June 13, 2024, 11:21:46 PM
post Re: Bigger sparks?...
[Vacuum Tube Tesla Coils (VTTC)]
Teslaventje
June 13, 2024, 08:50:56 PM
post Re: Steve Ward's DRSSTC Interrupter stops working when adjusting pulse width
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mads Barnkob
June 13, 2024, 07:43:37 PM
post Re: Steve Ward's DRSSTC Interrupter stops working when adjusting pulse width
[Dual Resonant Solid State Tesla coils (DRSSTC)]
petespaco
June 13, 2024, 04:52:08 PM
post Re: Turning SSTC into DRSSTC
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Simranjit
June 13, 2024, 02:14:17 AM
post Re: Turning SSTC into DRSSTC
[Dual Resonant Solid State Tesla coils (DRSSTC)]
flyingperson23
June 12, 2024, 11:04:57 PM
post Turning SSTC into DRSSTC
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Simranjit
June 12, 2024, 10:23:08 PM
post Re: UD2.7C constantly in OCP despite having signal
[Dual Resonant Solid State Tesla coils (DRSSTC)]
verliebt_in_neukölln17
June 12, 2024, 08:41:06 PM
post Re: UD2.7C constantly in OCP despite having signal
[Dual Resonant Solid State Tesla coils (DRSSTC)]
flyingperson23
June 12, 2024, 07:53:31 PM
post Re: Van de Graaf Voltage measurement with voltage sensor and voltage divider
[Static Electricity]
haversin
June 12, 2024, 06:57:13 PM
post UD2.7C constantly in OCP despite having signal
[Dual Resonant Solid State Tesla coils (DRSSTC)]
verliebt_in_neukölln17
June 12, 2024, 06:54:49 PM
post Re: Van de Graaf Voltage measurement with voltage sensor and voltage divider
[Static Electricity]
Mads Barnkob
June 12, 2024, 03:47:28 PM
post Van de Graaf Voltage measurement with voltage sensor and voltage divider
[Static Electricity]
vK
June 12, 2024, 03:42:18 AM
post Re: Bigger sparks?...
[Vacuum Tube Tesla Coils (VTTC)]
davekni
June 12, 2024, 03:32:22 AM
post Re: Bigger sparks?...
[Vacuum Tube Tesla Coils (VTTC)]
Teslaventje
June 11, 2024, 11:05:49 PM
post Re: Bigger sparks?...
[Vacuum Tube Tesla Coils (VTTC)]
MRMILSTAR
June 11, 2024, 05:36:46 AM
post Re: Sling Psychrometer
[Laboratories, Equipment and Tools]
Twospoons
June 11, 2024, 12:29:31 AM
post Bigger sparks?...
[Vacuum Tube Tesla Coils (VTTC)]
Teslaventje
June 10, 2024, 11:05:55 PM
post Re: Syntherrupter - A Feature-rich, Polyphonic Interrupter
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Max
June 10, 2024, 10:57:51 PM
post Re: Benjamin's DRSSTC3
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Benjamin Lockhart
June 10, 2024, 08:08:04 PM
post Re: Plasma Torid - Class E Self Resonant Dual/Stereo - Plasma Torid Build
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Hi-Deff
June 10, 2024, 06:04:09 PM
post Sling Psychrometer
[Laboratories, Equipment and Tools]
klugesmith
June 10, 2024, 04:43:59 PM
post Re: Benjamin's DRSSTC3
[Dual Resonant Solid State Tesla coils (DRSSTC)]
flyingperson23
June 10, 2024, 03:35:28 PM
post Re: Benjamin's DRSSTC3
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Jj
June 10, 2024, 01:36:41 PM
post Re: Is a 60UF 800VDC Dawncap good for an induction heater?
[General Chat]
Anders Mikkelsen
June 09, 2024, 05:44:29 PM
post Is a 60UF 800VDC Dawncap good for an induction heater?
[General Chat]
yourboi
June 09, 2024, 03:31:30 PM
post Re: Steve Ward's DRSSTC Interrupter stops working when adjusting pulse width
[Dual Resonant Solid State Tesla coils (DRSSTC)]
NyaaX_X
June 09, 2024, 01:34:49 PM
post Steve Ward's DRSSTC Interrupter stops working when adjusting pulse width
[Dual Resonant Solid State Tesla coils (DRSSTC)]
verliebt_in_neukölln17
June 09, 2024, 01:56:51 AM
post Re: Return of Electronics Flea Market in "Silicon Valley"
[Sell / Buy / Trade]
klugesmith
June 08, 2024, 06:06:25 AM
post Re: Coulometric hourmeters
[Laboratories, Equipment and Tools]
klugesmith
June 08, 2024, 05:38:14 AM
post Interrupter for SSTC
[Beginners]
Simranjit
June 08, 2024, 12:47:47 AM
post Re: Adding capacitor to primary in SSTC
[Solid State Tesla Coils (SSTC)]
Michelle_
June 07, 2024, 03:59:57 PM
post Re: Adding capacitor to primary in SSTC
[Solid State Tesla Coils (SSTC)]
davekni
June 07, 2024, 06:42:26 AM
post Re: Space-X Starship test this week
[Science, Research and News In Other Fields Than Electronics]
MRMILSTAR
June 07, 2024, 05:08:00 AM
post Re: Half bridge sstc different waveform on two mosfets
[Solid State Tesla Coils (SSTC)]
davekni
June 07, 2024, 03:16:31 AM
post Re: Space-X Starship test this week
[Science, Research and News In Other Fields Than Electronics]
klugesmith
June 06, 2024, 11:37:20 PM
post Adding capacitor to primary in SSTC
[Solid State Tesla Coils (SSTC)]
Simranjit
June 06, 2024, 11:20:23 PM
post Re: Plasma Torid - Class E Self Resonant Dual/Stereo - Plasma Torid Build
[Dual Resonant Solid State Tesla coils (DRSSTC)]
alan sailer
June 06, 2024, 10:54:48 PM
post Re: Half bridge sstc different waveform on two mosfets
[Solid State Tesla Coils (SSTC)]
ZakW
June 06, 2024, 05:17:02 AM
post Re: Half bridge sstc different waveform on two mosfets
[Solid State Tesla Coils (SSTC)]
Simranjit
June 06, 2024, 04:50:00 AM
post Re: Half bridge sstc different waveform on two mosfets
[Solid State Tesla Coils (SSTC)]
davekni
June 06, 2024, 04:00:52 AM
post Re: Half bridge sstc different waveform on two mosfets
[Solid State Tesla Coils (SSTC)]
ZakW
June 06, 2024, 01:43:41 AM
post Re: Half bridge sstc different waveform on two mosfets
[Solid State Tesla Coils (SSTC)]
Simranjit
June 05, 2024, 11:11:17 PM
post Re: Half bridge sstc different waveform on two mosfets
[Solid State Tesla Coils (SSTC)]
Michelle_
June 05, 2024, 10:18:57 PM
post Re: Half bridge sstc different waveform on two mosfets
[Solid State Tesla Coils (SSTC)]
Simranjit
June 05, 2024, 10:13:03 PM
post Re: Half bridge sstc different waveform on two mosfets
[Solid State Tesla Coils (SSTC)]
ZakW
June 05, 2024, 09:29:18 PM
post Half bridge sstc different waveform on two mosfets
[Solid State Tesla Coils (SSTC)]
Simranjit
June 05, 2024, 08:46:53 PM
post Re: Want to build HV HF oscilloscope probe.
[Electronic Circuits]
Twospoons
June 05, 2024, 07:04:45 AM
post Re: Want to build HV HF oscilloscope probe.
[Electronic Circuits]
Domo
June 05, 2024, 04:58:35 AM
post Re: Pearson current monitor model 1330 teardown (100kA)
[Laboratories, Equipment and Tools]
klugesmith
June 05, 2024, 01:42:04 AM
post Re: Coulometric hourmeters
[Laboratories, Equipment and Tools]
klugesmith
June 04, 2024, 10:31:58 PM
post Re: My completed 14-stage Cockroft-Walton voltage multiplier
[Voltage Multipliers]
pete_dl
June 04, 2024, 08:08:28 PM
post Re: Want to build HV HF oscilloscope probe.
[Electronic Circuits]
Twospoons
June 03, 2024, 11:13:26 PM
post Re: First DRSSTC questions
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Benjamin Lockhart
June 03, 2024, 10:35:57 PM
post Re: First DRSSTC questions
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Hi-Deff
June 03, 2024, 05:51:05 PM
post Re: Want to build HV HF oscilloscope probe.
[Electronic Circuits]
Domo
June 03, 2024, 07:41:41 AM
post Re: First DRSSTC questions
[Dual Resonant Solid State Tesla coils (DRSSTC)]
flyingperson23
June 03, 2024, 05:28:44 AM
post Re: RDRSSTC - Project Build
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
June 03, 2024, 05:02:19 AM
post Re: First DRSSTC questions
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Benjamin Lockhart
June 03, 2024, 03:47:33 AM
post Re: First DRSSTC questions
[Dual Resonant Solid State Tesla coils (DRSSTC)]
flyingperson23
June 03, 2024, 02:23:58 AM
post Re: First DRSSTC questions
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Hi-Deff
June 03, 2024, 01:25:24 AM
post Re: RDRSSTC - Project Build
[Dual Resonant Solid State Tesla coils (DRSSTC)]
ZakW
June 03, 2024, 01:08:47 AM
post Re: RDRSSTC - Project Build
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
June 02, 2024, 10:29:02 PM
post Re: Steve Ward SSTC5 build
[Solid State Tesla Coils (SSTC)]
Michelle_
June 02, 2024, 10:02:45 PM
post Re: Racing sparks
[Spark Gap Tesla Coils (SGTC)]
davekni
June 02, 2024, 08:25:37 PM
post Re: Steve Ward SSTC5 build
[Solid State Tesla Coils (SSTC)]
Michelle_
June 02, 2024, 04:51:58 PM
post Re: RDRSSTC - Project Build
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Michelle_
June 02, 2024, 04:44:30 PM
post Re: My completed 14-stage Cockroft-Walton voltage multiplier
[Voltage Multipliers]
MRMILSTAR
June 02, 2024, 03:25:11 PM
post APC Smart-UPS 1500VA Teardown
[Electronic Circuits]
Mads Barnkob
June 02, 2024, 01:36:11 PM
post Re: My completed 14-stage Cockroft-Walton voltage multiplier
[Voltage Multipliers]
pete_dl
June 02, 2024, 11:43:40 AM
post Re: RDRSSTC - Project Build
[Dual Resonant Solid State Tesla coils (DRSSTC)]
ZakW
June 02, 2024, 08:49:11 AM
post Re: RDRSSTC - Project Build
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Benjamin Lockhart
June 02, 2024, 08:36:07 AM
post Re: RDRSSTC - Project Build
[Dual Resonant Solid State Tesla coils (DRSSTC)]
ZakW
June 02, 2024, 12:41:50 AM
post Re: Steve Ward SSTC5 build
[Solid State Tesla Coils (SSTC)]
Benjamin Lockhart
June 01, 2024, 08:24:35 PM
post Pearson current monitor model 1330 teardown (100kA)
[Laboratories, Equipment and Tools]
Mads Barnkob
June 01, 2024, 08:22:28 PM
post Re: Steve Ward SSTC5 build
[Solid State Tesla Coils (SSTC)]
Michelle_
June 01, 2024, 03:56:53 AM
post Re: Racing sparks
[Spark Gap Tesla Coils (SGTC)]
Terry
June 01, 2024, 12:03:50 AM
post Re: First DRSSTC questions
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Hi-Deff
May 31, 2024, 11:46:20 PM
post Re: Steve Ward SSTC5 build
[Solid State Tesla Coils (SSTC)]
Michelle_
May 31, 2024, 06:13:59 PM
post Re: First DRSSTC questions
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
May 31, 2024, 01:55:22 AM
post Re: First DRSSTC questions
[Dual Resonant Solid State Tesla coils (DRSSTC)]
flyingperson23
May 30, 2024, 11:57:49 PM

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