Author Topic: Strange behavior of ZVS driver  (Read 460 times)

Offline AstRii

  • High Voltage Experimenter
  • **
  • Posts: 81
  • Karma: +2/-0
  • Czech Technical University in Prague
    • View Profile
Strange behavior of ZVS driver
« on: September 07, 2020, 11:50:30 PM »
Hello guys :)
I think we've all seen the "Mazzilli ZVS flyback driver" schematics as it's very popular and very easy to build.
Just as a remainder, here is the schematic.



I've recreated the driver with following changes: 470ohm 2W gate resistors went to 330R 5W resistors.
Instead of 0.68uF resonant capacitor, i use 4x Tesla TC343 1000V 150nF impulse capacitors for total of 600nF.
Instead of IRFP250 MOSFETs, I am using FGA25N120 IGBTs (datasheet here: http://www.farnell.com/datasheets/1774719.pdf).
As the diodes i use FES16JT (datasheet here: https://www.gme.cz/data/attachments/dsh.221-055.1.pdf).
I also added 3900uF capacitor at the input for smoothing. I'm driving (pretty large i would say) ferrite core transformer for some huge arcs at high input voltages (Inspired by https://www.vn-experimenty.eu/index.php?stranka=zvs_driver_modified).
But my driver is acting strangely.

It oscillates well at 7kHz unloaded, and about 33kHz when drawing an arc. 
But when the input voltage reaches 12V+, when drawing an arc it stops oscillating after a while and both IGBT gates stay at 12V.
And i have no idea why.

Here are some videos so you can see it better. (On the oscilloscope i scoped both IGBT gates. You can see the voltage and current draw on the panel of the lab bench power supply).


Input voltage of 7V. Nothing is wrong here.


Input voltage of 11V. Still nothing wrong and the arc got bigger (so i suppose it works well)


Input voltage of 14V. Here's where the trouble begin :) (actually at 12V already:D)

I have no explanation of why this could be. Maybe the IGBTs are too slow? But it's only 33kHz... Maybe the leakage inductance of the wiring is doing some mess?.. I don't know.
It's rather suspicious that this happens when the voltage rises above 12V (as it's the zener voltage of the zener diodes)

Here are a few photos of my setup.





I will appreciate any help :)
Thank you
- Mark





« Last Edit: September 07, 2020, 11:54:11 PM by AstRii »

Offline SteveN87

  • High Voltage Enthusiast
  • *
  • Posts: 32
  • Karma: +1/-0
  • FPGA, Software Defined Radio, HV (mainly flyback)
    • View Profile
Re: Strange behavior of ZVS driver
« Reply #1 on: September 08, 2020, 12:22:22 AM »
Below 12V isn't really enough gate drive for an IGBT. Maybe as you go up in voltage and the gates are getting their 12V, those croc leads (particularly the tank connections) are struggling to handle the current...

Offline davekni

  • High Voltage Senior
  • *****
  • Posts: 556
  • Karma: +31/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Strange behavior of ZVS driver
« Reply #2 on: September 08, 2020, 02:14:01 AM »
As Steve said, change to 15V zener diodes.  And avoid clip leads at least for the connection from resonant caps to transformer primary.

The biggest issue is likely insufficient gap in the transformer ferrite core.  ZVS oscillators like this one fail if the resonant Q drops too low.  Amplitude gets too low to switch the gates.  It will work when lightly loaded, at the transformer's normal inductance, resonating at 7kHz in your case.  It also works well at high load (short circuit or short arc) at the transformer's leakage inductance, resonating at 33kHz in your case.  Between no load and short, there's a load range where Q drops too low, oscillation stops, leaving both transistors on.  (Use a current-limited supply and shut down quickly to avoid frying transistors when this happens.)

I've found through experiments and simulation that a transformer with coupling factor below 0.85 will result in Q staying sufficiently high across all output loads.  See:
https://highvoltageforum.net/index.php?topic=831.msg5491#msg5491
For my Jacob's ladder based on a similar IGBT ZVS oscillator running on rectified 120VAC.  One simple way to check coupling factor is by comparing open and shorted resonant frequencies.  A 2:1 ratio corresponds to 0.866 coupling.  So you need those two frequencies to be slightly closer than 2:1.  33:7
David Knierim

Offline AstRii

  • High Voltage Experimenter
  • **
  • Posts: 81
  • Karma: +2/-0
  • Czech Technical University in Prague
    • View Profile
Re: Strange behavior of ZVS driver
« Reply #3 on: September 08, 2020, 07:45:50 PM »
Hello again.
Thank you for valuable advice.
Unfortunately i'm still having the same problem.
I've replaced the primary croc connections with solid connections and i replaced the zener diodes.
I didn't have 15V zener diodes so i use 15V TVS.
I have also added some paper in between the two parts of the core to increase the inductance.
New resonant frequency at no load is 22kHz. (Still the same 33kHz at short)
But it still acts the same way as before.


Here on the oscilloscope you can kinda see the gate voltage of one of the IGBTs.



Any other ideas?



Offline AstRii

  • High Voltage Experimenter
  • **
  • Posts: 81
  • Karma: +2/-0
  • Czech Technical University in Prague
    • View Profile
Re: Strange behavior of ZVS driver
« Reply #4 on: September 08, 2020, 07:59:38 PM »
Sorry for the double-post. I've just noticed that the gate voltage somehow rises and does not go back to 0V.

On this video you can see the gate of one of the IGBTs, when unloaded, then short circuit and then drawing an arc. Input voltage is about 14V, at this voltage the oscillation doesn't stop yet.


Offline Mads Barnkob

  • Administrator
  • High Voltage Expert
  • *****
  • Posts: 1516
  • Karma: +27/-0
  • Denmark
    • View Profile
    • Kaizer Power Electronics
Re: Strange behavior of ZVS driver
« Reply #5 on: September 08, 2020, 08:18:58 PM »
Ringing on the gate drive waveform itself is normal due to a number of reasons, but ringing on the middle of the low signal... hmm anyone got an idea about that?

From what I remember about the ZVS and IRFP250N is that people had problems switching that MOSFET out to something else, and it had to do with putting in switches that had a very different gate charge specification (Qc)

http://www.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 AstRii

  • High Voltage Experimenter
  • **
  • Posts: 81
  • Karma: +2/-0
  • Czech Technical University in Prague
    • View Profile
Re: Strange behavior of ZVS driver
« Reply #6 on: September 08, 2020, 08:31:09 PM »
Maybe i should get some high voltage MOSFETs with low Rds(on). But i really see no reason why this should happen when using IGBTs.

Offline Teravolt

  • High Voltage Technician
  • ***
  • Posts: 124
  • Karma: +2/-0
    • View Profile
Re: Strange behavior of ZVS driver
« Reply #7 on: September 08, 2020, 09:07:14 PM »
if there is not any problem with your silicon I recomend that you use a old fashoned raw dc power source because your power supply may be folding under the needed current. your zener diodes look like transientt voltage supressers or transorbe, what are they? also a solder type bread board may help you trubleshoot your circuit. the transformer that you have should be able to put out a cuple of kilowatts and using a IGBT brick setup should help you achieve that. there a lot of examples hear at the forum

Offline AstRii

  • High Voltage Experimenter
  • **
  • Posts: 81
  • Karma: +2/-0
  • Czech Technical University in Prague
    • View Profile
Re: Strange behavior of ZVS driver
« Reply #8 on: September 08, 2020, 09:24:37 PM »
The power supply i'm using is rated 30V@35A, it should handle the low voltage (15V) testing fine. The new 15V diodes i use to suppress the voltage are 15V TVS (BZW0615) https://cz.mouser.com/datasheet/2/395/BZW06%20SERIES_J1602-972389.pdf
I was kinda hoping i could use this simple circuit to push a few kilowatts through :D

Offline Teravolt

  • High Voltage Technician
  • ***
  • Posts: 124
  • Karma: +2/-0
    • View Profile
Re: Strange behavior of ZVS driver
« Reply #9 on: September 08, 2020, 11:16:34 PM »
one thing I'm seeing on your transformer primary is the amount of windings. there should be about 4-6 turns of hevey wire like 14-10 awg. When the unit is in resonance the primary caps and secondary will have a lot of circulating current. the secondary should have the same resonance that will help with your power and voltage gain

Offline johnf

  • High Voltage Technician
  • ***
  • Posts: 135
  • Karma: +4/-0
    • View Profile
Re: Strange behavior of ZVS driver
« Reply #10 on: September 09, 2020, 03:43:05 AM »
Check your clip leads with a magnet
I bought some a couple of years ago and found they were poorly copper plated steel wire and only good for a few amps because of the resistance of them.
Only trust clip leads made by yourself

Offline davekni

  • High Voltage Senior
  • *****
  • Posts: 556
  • Karma: +31/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Strange behavior of ZVS driver
« Reply #11 on: September 09, 2020, 05:34:24 AM »
IMy guess is oscillation in the low-half of gate signal is due to wiring inductance and parasitic capacitance, especially from IGBTs to resonant caps.  It may be worse because the low gate voltage is rising to ~5V, probably because the high half of the other IGBT gate is only 12V, so not fully saturated.  If you have a laptop supply or other separate source of 15-20V, I'd suggest using that for gate-drive.  Either that or you need to keep your primary supply above 15V.

Looks like you have the transformer ferrite core gap set about right.
David Knierim

Offline AstRii

  • High Voltage Experimenter
  • **
  • Posts: 81
  • Karma: +2/-0
  • Czech Technical University in Prague
    • View Profile
Re: Strange behavior of ZVS driver
« Reply #12 on: September 09, 2020, 05:00:51 PM »
I could not find a problem with the IGBTs. I replaced them for very nice MOSFETs STW57N65M5-4 and it worked right away.
Unfortunately i'm dumb and managed to destroy them anyway :D

I was playing with arcs at 70-80V @15A input and the transistors were pretty cool (40-50C after some minutes) the same goes for the transformer core (30-40C) but after a while this happened.


I still find it interesting, these MOSFETs should handle that voltage and current without a doubt.
Maybe there was some cross-conduction which killed it.


Offline davekni

  • High Voltage Senior
  • *****
  • Posts: 556
  • Karma: +31/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Strange behavior of ZVS driver
« Reply #13 on: September 09, 2020, 06:41:00 PM »
If you want to keep playing with this, scope it at slightly lower voltage to see what's happening, especially just as an arc breaks at the end when power is highest.  You may have high-frequency oscillations still due to wiring inductance, or slow transitions due to gate capacitance.

I prefer using copper-clad board for the FET/IGBT and drive circuitry (diodes and resistors) to keep inductance low.  (A few dremel-tool cuts on one side of blank copper-clad can implement simple circuits.  Even if you still need a few wires, inductance is much lower when those wires are flat against copper.  I'd use the back side of the copper-clad as ground (source/emitter) plane.

Aside from high-frequency oscillation, there are a few other possible issues.  What is the inductance and saturation current of the inductor?  At high currents, its inductance may drop too low, causing high peak currents.  At 60A, Vds is high enough (~4V) that adding a diode-drop makes the other "off" FET start conducting noticeable current.  15A average doesn't prevent peaks of 60A.

Perhaps the most likely issue is what I find most problematic with ZVS oscillators.  When a load abruptly changes from high to zero (end of an arc), stored energy in the inductor causes voltage to spike.  I routinely saw 3x voltage spikes before adding many 1500W TVS diodes.  At 80Vdc input, normal peak Vds is 250V (PI * 80).  A 3x ring is enough to fry 650V FETs.
David Knierim

Offline AstRii

  • High Voltage Experimenter
  • **
  • Posts: 81
  • Karma: +2/-0
  • Czech Technical University in Prague
    • View Profile
Re: Strange behavior of ZVS driver
« Reply #14 on: September 09, 2020, 06:56:53 PM »
Actually i think i will give up this idea to drive the transformer with this circuit. I will probably make some full bridge driver for it :)
But it was fun anyway, i like how you can make this circuit literally in 5 minutes if you have the components for it :D

Yet i wonder, what actually happens when the arc breaks? When i powered the driver from my lab bench power supply, it actually maxed out 30V 35A, which is crazy.
« Last Edit: September 09, 2020, 07:00:56 PM by AstRii »

Offline Teravolt

  • High Voltage Technician
  • ***
  • Posts: 124
  • Karma: +2/-0
    • View Profile
Re: Strange behavior of ZVS driver
« Reply #15 on: September 10, 2020, 04:49:53 AM »
were any of the other devices bad. you could add some TVS or diodes across the drains and sources to help agents spikes.

High Voltage Forum

Re: Strange behavior of ZVS driver
« Reply #15 on: September 10, 2020, 04:49:53 AM »

 


* Recent Topics and Posts

post Re: Polyphonic Interrupter with STM32F103C8T6 (Bluepill)
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Max
Today at 11:48:25 PM
post Re: Half bridge sstc help with troubleshoot
[Solid State Tesla Coils (SSTC)]
costas_p
Today at 09:56:15 PM
post Re: Easy to build Polyphonic MIDI Interrupter
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
futurist
Today at 09:52:47 PM
post Re: Half bridge sstc help with troubleshoot
[Solid State Tesla Coils (SSTC)]
davekni
Today at 09:50:44 PM
post Re: Half bridge sstc help with troubleshoot
[Solid State Tesla Coils (SSTC)]
costas_p
Today at 09:16:53 PM
post Re: Half bridge sstc help with troubleshoot
[Solid State Tesla Coils (SSTC)]
Zipdox
Today at 09:09:46 PM
post Re: Half bridge sstc help with troubleshoot
[Solid State Tesla Coils (SSTC)]
costas_p
Today at 09:03:23 PM
post Re: Half bridge sstc help with troubleshoot
[Solid State Tesla Coils (SSTC)]
Zipdox
Today at 08:50:33 PM
post Re: Stroboscopic photography & motors
[Digital Compact Camera]
Zipdox
Today at 08:49:09 PM
post Re: Half bridge sstc help with troubleshoot
[Solid State Tesla Coils (SSTC)]
costas_p
Today at 08:40:11 PM
post Re: Stroboscopic photography & motors
[Digital Compact Camera]
klugesmith
Today at 07:36:45 PM
post Re: DRSSTC Tuning for Music vs Big Sparks?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Uspring
Today at 07:07:29 PM
post Re: Stroboscopic photography & motors
[Digital Compact Camera]
Zipdox
Today at 06:41:56 PM
post Re: Polyphonic Interrupter with STM32F103C8T6 (Bluepill)
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Zipdox
Today at 06:28:50 PM
post Re: Stroboscopic photography & motors
[Digital Compact Camera]
klugesmith
Today at 05:14:51 PM
post Re: Polyphonic Interrupter with STM32F103C8T6 (Bluepill)
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Max
Today at 04:44:56 PM
post Re: Stroboscopic photography & motors
[Digital Compact Camera]
TMaxElectronics
Today at 10:50:31 AM
post Re: Stroboscopic photography & motors
[Digital Compact Camera]
plasma
Today at 09:15:09 AM
post Re: Stroboscopic photography & motors
[Digital Compact Camera]
Zipdox
Today at 09:09:45 AM
post Re: type of driver
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mads Barnkob
Today at 08:39:30 AM
post Re: Stroboscopic photography & motors
[Digital Compact Camera]
klugesmith
Today at 08:16:46 AM
post Re: Polyphonic Interrupter with STM32F103C8T6 (Bluepill)
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Zipdox
Today at 01:27:07 AM
post Re: Polyphonic Interrupter with STM32F103C8T6 (Bluepill)
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
TMaxElectronics
Today at 01:14:46 AM
post Re: Rogowski current probe
[Laboratories, Equipment and Tools]
flyglas
Today at 12:30:58 AM
post Re: Rogowski current probe
[Laboratories, Equipment and Tools]
Hydron
September 27, 2020, 11:48:35 PM
post Polyphonic Interrupter with STM32F103C8T6 (Bluepill)
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Zipdox
September 27, 2020, 11:43:14 PM
post Re: Rogowski current probe
[Laboratories, Equipment and Tools]
flyglas
September 27, 2020, 11:01:13 PM
post type of driver
[Dual Resonant Solid State Tesla coils (DRSSTC)]
octopus1
September 27, 2020, 10:51:03 PM
post Re: Easy to build Polyphonic MIDI Interrupter
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Max
September 27, 2020, 10:44:43 PM
post Re: MIDI videos from my DRSSTC1/2/3 demonstrations/shows
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mads Barnkob
September 27, 2020, 09:44:58 PM
post Re: DRSSTC Tuning for Music vs Big Sparks?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
September 27, 2020, 09:20:32 PM
post Re: Easy to build Polyphonic MIDI Interrupter
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
futurist
September 27, 2020, 07:49:49 PM
post Stroboscopic photography & motors
[Digital Compact Camera]
klugesmith
September 27, 2020, 07:08:06 PM
post Re: Led Closeups with Mobile phone and lens .
[Smart Phones]
klugesmith
September 27, 2020, 06:43:46 PM
post Re: DRSSTC Tuning for Music vs Big Sparks?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mads Barnkob
September 27, 2020, 06:42:46 PM
post Re: Finding accelerometer location inside a phone
[Smart Phones]
klugesmith
September 27, 2020, 06:36:41 PM
post Re: steam engine
[Capacitor Banks]
plasma
September 27, 2020, 02:32:14 PM
post Re: Easy to build Polyphonic MIDI Interrupter
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Max
September 27, 2020, 02:02:53 PM
post Re: Easy to build Polyphonic MIDI Interrupter
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
futurist
September 27, 2020, 12:24:20 PM
post Re: Easy to build Polyphonic MIDI Interrupter
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Max
September 27, 2020, 12:52:36 AM
post Re: DRSSTC Tuning for Music vs Big Sparks?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
September 26, 2020, 11:06:04 PM
post Re: One NOS ac flyback is nice. Finding its identical twin 3 months later is awesome
[Transformer (Ferrite Core)]
profdc9
September 26, 2020, 10:48:41 PM
post Re: Rogowski current probe
[Laboratories, Equipment and Tools]
Hydron
September 26, 2020, 05:27:37 PM
post Re: DRSSTC Tuning for Music vs Big Sparks?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Zipdox
September 26, 2020, 04:32:53 PM
post Re: DRSSTC Tuning for Music vs Big Sparks?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mads Barnkob
September 26, 2020, 04:17:51 PM
post Re: DRSSTC Tuning for Music vs Big Sparks?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
GKnapp
September 26, 2020, 02:06:04 PM
post Re: Rogowski current probe
[Laboratories, Equipment and Tools]
flyglas
September 26, 2020, 10:57:58 AM
post Re: DRSSTC Tuning for Music vs Big Sparks?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mads Barnkob
September 26, 2020, 08:55:27 AM
post Re: DRSSTC Tuning for Music vs Big Sparks?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
September 26, 2020, 05:19:45 AM
post Re: Rogowski current probe
[Laboratories, Equipment and Tools]
klugesmith
September 26, 2020, 04:46:12 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