SSTC (Not) Working Correctly

[Laci]

A few things happened again.The GDT problem solved,because I made a halfbridge on a breadboard with IRFZ44N's and it worked pretty good,but with the IGBT's the signal was not even like square,probably due to the poor design and the BD139-140 output stage.
New components have arrived.
Ferrite toroids(3x): https://www.tme.eu/gb/details/tx42_26-3e25/ring-ferrites/ferroxcube/tx422613-3e25/
Capacitors for DRSSTC(3x in series): https://www.tme.eu/gb/details/phe450rd6220j/standard-foil-polypropylene-capacitors/kemet/phe450rd6220jr06l2/

And other chips.I put together a PLL based IR2109 full bridge driver circuit on a perfboard.Sadly,the output of the IR2109 was not enough for the IGBT's,so I bought 2x IR2184 and added them to the perfboard.I will test it tomorrow.

Yesterday I created an arduino based interrupter with bps+duty cycle control and SD card music play function.Since I don't have optical fiber,I will connect the interrupter to the main driver circuit via some kind of insulated cable .

Tomorrow I will try to make the CT and the GDT.What is your opinion about the toroids?Can I use them for both the CT and the GDT at 150kHz?(Since the secondary res.freq. is 150kHz)

[futurist]

3E25 toroids will be fine for both GDT and CTs. I use them too for CTs, and even worse N87 material for GDT

Why not invest in fiber optic transmitter/receiver? They are not that expensive:
https://www.ebay.com/itm/2x-AGILENT-HFBR-1414T-TX-FIBER-OPTIC-TRANSMITTER-160MBD-8-Pin/182851049308
https://www.aliexpress.com/item/HFBR-2412TZ-RECEIVER-FIBER-OPTIC-5MBD-TTL-ST/32709652010.html

Especially because arduino won't like EMI from the coil. I have one as fan speed controller and bus precharge controller, and it kept resetting on ground strikes before I added various protection to reset pin, etc.

[Laci]

I've made the GDT !What is your opinion about it?The first scopeshot shows the unloaded GDT,the other one the GDT with an IGBT connected to it,which probably loads down my scope's built in 50R impedence function generator.

[Mads Barnkob]

Remember to mark up the phasing of the GDT windings, before twisting it all together, makes it much easier to avoid mistakes later on.

For your next GDT, I would advise you to use a complete network cable, normally I parallel all the white/colour wires and use the 4 coloured pairs for the gates, this gives a nice and sturdy primary winding that can conduct some current, heat up less and has high coupling to the secondary windings.



Square wave looks fine and despite your signal generator might be weak, its actually doing a fair job driving that gate! But I would not recommend that you continue to use it for that, did you insert a 50 Ohm resistor in series with the primary to match the output impedance of the signal generator?

Looking forward for the coming updates

[Laci]

Updates again.
I hooked up the GDT to the IGBT full bridge.The driver circuit was the PLL circuit,with the UCC chips converted to another perfboard.
At the beginning,I experimented with the GDT phasing,but it was not working,so I made some changes and marked the cables from one end.Additionally,I have a question:is that a problem if I have single coil connected to the UCC's instead of two in parallel?
The phasings looks correct,the waveforms also OK i think,but it doesn't outputs anything to the primary.The chips heats up pretty fast,especially the UCC37321.The current starts from 2A and drops to around 1A in around 10 sec.,while the chips heats up uncomfortably hot.The old chips were died,as I expected,when I put them in the socked,they exploded like a firecracker...
I see a problem on the waveform:it's around 2V at 2V/div so 4V.It should be the same as the UCC's 12V rail right?And what is the current drawn of the UCC's in your circuits?

[Laci]

Photos:

[Laci]

Well...I'm about to give up this project...Nothing is working

After I removed the zener diodes from the IGBT's,the fluorescent lamp started to light up.
Since I'm an idiot,I directly shorted one of the UCC's 6 & 7 pins to ground.After I fixed it,the coil started to work pretty well.The UCC's didn't died and I made some changes to the fullbridge.
It was working nice on the 18V laptop battery charger.I still couldn't connect the zener diodes,because with them high current started to flow in the driver circuit,so I think that they shorted the transistors gate to emitter at such a high frequency.
Then I connected my 55V transformer to the fullbridge.I pressed the driver circuit's power button.Immediately very high current started to flow to the IGBT's.I turned off very fast everything.
Next I disconnected the transformer and turned on the driver circuit alone.Around 1.5A started to flow,and guess what?Two IGBT's were completly shorted.

Altought,the GDT worked pretty well,the signals were good,in phase.After fixing the wrong connection on the perfboard,the driver circuit drawed around 0.4A and the UCC's heated up slower.

As a last hope,I'm gonna check if the other two IGBT is good or not.If yes,then rebuild the whole thing as a halfbridge.If not then just use my remaining 2xIRFP460.
I'm really disappointed,but I should expected these fails to happen.Hopefully this forum will help other peoples to learn about my problems and they will not make the same mistakes as me...

A question is still existing:why the IGBT's blew up?And why the thing was not working with zener diodes?I think they died due to gate overvoltage.I've seen,that in other schematics two zeners are used in opposite direction.Did this caused the death of my IGBT's?

[Mads Barnkob]

A single wire primary on the GDT is not a problem unless it really heats up or has a high resistance. It all depends on the load the GDT has to drive, did you calculate how much driving power is needed? ( http://kaizerpowerelectronics.dk/calculators/igbt-gate-drive-calculator/ ) This will also give you an idea on how much power the UCCs draw.

UCC chips should NEVER be running without a load connected to their outputs, if you have this, it is almost guaranteed that they are destroyed. What happens is that the high output current capability of the IC can self oscillate with the stray inductance/capacitance at its connections, into MHz frequencies and it will just do this until it burns out.

Regarding the output voltage, are you sure you are not short-circuiting anything with those connections to the oscilloscope? You should not measure the primary side of the GDT, but the secondary side, this is where you have to check for your good looking gate switching waveform and adjust gate resistor accordingly to minimize ringing at turn-on.

Do not worry too much about small failures, we have all been there and I also used to have a mountain of burned silicon It takes exploding parts to build knowledge. Keep up the good work and experiments!

Edit: regarding exploding IGBTs/zener diodes, this is almost always two things, wrong phasing so that you are turning on two wrong switches at a time and they short circuit the power supply across themself and die or too much current.

Do not think that just because you test the bridge at 12V that it is switching correctly, that voltage is so low that even if its short circuiting the IGBTs could very well just dissipate the power, but with higher voltages they will go puufff!

[Laci]

Thank you for the reply!
I'm upset,because Friday I could show it on a physics presentation and things are not going very well.
I should start everything from the beginning and upgrade everything lineary.The only thing that worked very well,was the halfbridge with the IR2109 bootstrap.That worked more than 1 minutes :O But as always,that also died...

Quite a lot of SSTC's and DRSSTC's are made with halfbridge.I want to go back to that and make it working.Only if that is done,then upgrade.

The resonance capacitors are like addiction to drugs for me.Almost every time,when some arcs are coming out from the secondary,I connect them to the primary.This way failure is guarantied,which is a routine for me.

I send some pictures of the fullbridge.Tomorrow some pictures will be uploaded about the GDT waveforms.Maybe test some bootstrap circuits,remove the died two IGBT from the heatsink.

[Laci]

Hmmmm......
The full bridge got repaired,by removing the 2 dead IGBT and replacing them with 2x1uF MKP caps.Driver circuit was the same,PLL driven UCC's and as interrupter a simple 555 astable multivibrator,then Steve Ward's interrupter with burst mode.
It was working nice and good at 55V.Then I increased to mains rectified with 60W lamp in series.
Strange things started to happen...It was working sometimes,and sometimes not.The current into the halfbridge was measured with a multimeter.It started to randomly jump up,and reached my meter's limit.
While this was happening I heared clicking sounds coming under my workbench.I looked down,and saw an interesting thing:The HV was arcing from the mains ground(the bottom of the secondary)to the neutral.
Then I wake up in an other world and realised the big deal.I WAS RUNNING THE TESLA COIL ALL THE TIME WITHOUT GROUNDING THE SECONDARY.

The problem was my wall outlet's connector multiplier or splitter or whatever it's called,because it's cable doesn't contained the 3rd ground wire...

I fixed this problem quickly,then started running the thing interrupted.Steve's interrupter resulted awesome looking and sounding sparks,,but not really long ones.So I made the astable multivibrator for higher duty cycle.The result is shown in the video;around 20 cm sparcs.

One active problem is the feedback.When a put my hand closer to the coil,the arcs lenght increases.I already had this problem,and I thought it was repaired by a solder connection.Is this normal?


I'm gonna upload a picture of my planned CT to use.What is your opinion about it?

[Laci]

Video: https://mega.nz/#!MfollKQK!felQUB7hoZHNeaBLQouAqYp8BmTamvcU1xY57miKpZ4

[Laci]

And the sad thing comes again.I kinda finalized this SSTC,so it would be possible to show it to other peoples in a school presentation,but that ended this day,and I had to say to my physics teacher,that I couldn't make it....

[futurist]

Don't worry, I'm sure you'll have another chance of presenting your project and that your teacher understands that building SSTC isn't a piece of cake

[Laci]

Yes,it's not easy at least for me.The system is down again,due to my stupidity,because I disconnected the interrupter and the breaker popped.One of the IGBT's died.

Now I can relax a bit,and focus on other things.I'm proud,that I could make a pretty good SSTC.The driver circuit works great.The IGBT's gate waveform was really good(see the picture).The interrupter is OK,but without shielding anything,or optical fiber connection,it isn't stable.

I still have quite a lot things to upgrade.I'm planning a DRSSTC,but it wouldn't be a good idea to run it on a IRFP460 halfbridge...

Many thanks for this forum,and for everybody who helped me to get here!I will not leave,just take a break,until new transistors will arrive

[Mads Barnkob]

You have learned a lot of lessons and that is more valuable than having a kit work in the first go

You are absolutely right about IRFP460s not going to work in a DRSSTC, MOSFETs have no place in switching so high peak currents, just do a quick calculation with on-resistance and peak current so see that you are trying to dissipate the heat of the sun in a glass of water.

Have a nice break and certainly looking forward to help you build a DRSSTC

[Laci]

While I'm making the DRSSTC,I also rebuilt a fullbridge of 2xIRFP460 and 2xIRFP240,since I had these transistors.It still works very good.The power supply is around 120V,by putting some transformers in series,and it's still working after two days,which is a record for me
The interrupter is my scope's built in,connected to the UCC's enable pin,but I want to audio modulate it.The audio modulator in the Kazier SSTC 2 looks promising,but I would like to have cleaner sound,not the midi-interrupted type.Do you have any idea?

[Mads Barnkob]

While I'm making the DRSSTC,I also rebuilt a fullbridge of 2xIRFP460 and 2xIRFP240,since I had these transistors.It still works very good.The power supply is around 120V,by putting some transformers in series,and it's still working after two days,which is a record for me
The interrupter is my scope's built in,connected to the UCC's enable pin,but I want to audio modulate it.The audio modulator in the Kazier SSTC 2 looks promising,but I would like to have cleaner sound,not the midi-interrupted type.Do you have any idea?

The audio modulator I used in the SSTC2 is a midi-like interrupter as it does PWM and interrupted tones, there is not much difference in sound from this and a pure MIDI modulator, except the MIDI modulator has better and more precise tone control, which actually sounds better

You are properly asking for FM modulation like sound, like the one you hear in a Class E Tesla coil, you could take a look at the LTC6992 IC that Acid Byte uses in his SSTC, that might just be the thing for you: https://highvoltageforum.net/index.php?topic=164.0

[Laci]

I think this project is done
I've put everything together on Tuesday and everyhing was working.On the next day I showed it on a physics demonstration,it was working without problems for hours!
Yesterday I made a small modification:added the 3x220nF resonant caps to the primary.It was awesome!Around three times larger arcs than before,up to 40cm!Sadly,I today killed the mosfets,because I pulled an arc direcly from the topload by a long wire and probably a high peak current killed the mosfets.Otherwise it was working amazingly wel.I think this setup is much more stable than anytime before,but I really want to know the difference between the DC blocking cap and the resonant cap.Does just the driving circuit make the difference?

Anyways now I can continue the making of the DRSSTC.IGBT's will come soon,probably not the same ones as before,probably IGW60N60H3's or HGTG30N60A4D's...
A short question about that:what kind of caps can be used in the 1.3b driver's gdt driver section?

[Mads Barnkob]

The difference between a DC blocking capacitor in a SSTC and a resonant capacitor in a DRSSTC is actually not that big, they are most likely both MKP type and could be the same capacitance.

You can have a 1 uF DC blocking capacitor in a SSTC and also a 1 uF resonant capacitor in a DRSSTC, but what makes the difference is the inverter impedance, which is calculated from inverter voltage output divided by inverter output current. A SSTC has a low current compared to a DRSSTC.

The reactance Xc of a DC blocking capacitor should be many times smaller than the inverter impedance, where in a DRSSTC the reactance is much higher due to smaller capacitances in the MMC, but to go back to our example of 1 uF DC blocking vs. 1 uF resonant capacitor, a small SSTC would maybe be 320V/10A so 32 Ohm impedance but the large DRSSTC using 1 uF resonant capacitor could be 564V/2000A so ~0.25 Ohm impedance.

The reactance Xc = 1 / (2 * pi * frequency * capacitance), so the higher capacitance, the lower reactance at the same frequency.

[Laci]

So basically a capacitor in the primary circuit doesn't effects the primary resonant frequency,and the important thing is to drive it at the secondary's resonant frequency?