Author Topic: Royer oscillator (ZVS) driven Jacob's ladder, E80 core transformer  (Read 1798 times)

Offline davekni

  • High Voltage Technician
  • ***
  • Posts: 131
  • Karma: +5/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Inspired by Phoenix' huge ferrite transformer arcs:
    https://highvoltageforum.net/index.php?topic=433.msg2609#msg2609
I decided to make a small version powered from standard US 120V outlet.  It uses a ZVS Royer oscillator directly from the rectified line voltage input, driving an 8-turn center-tapped winding on one E80 core half.  The other core half has 102 turns.  Halves joined with 0.55mm spacing for K=0.83 coupling factor.  K < 0.86 allows the oscillator to run over the full range of arc load resistance, at higher frequency under lower-impedance load.







Most of the work involved exploring how to get the ZVS oscillator to start cleanly without a huge inrush current spike, which is followed by an oscillator voltage spike due to the large energy stored in the input inductor (L4 in the following schematic).  My final solution was to bypass the power switch with a resistor R2 (actually a small incandescent light bulb).  This starts the oscillator at low power.  Once oscillating, it transitions to high power more smoothly.



Switch S1 is actually an electronic switch with isolated LV control and over-current shutdown.  Internal details of S1 aren't shown here.

Output voltage before an arc forms is +-5.5kV.  Short-circuit current (very short initial arc) is +-2.7A.  Below is a scope-capture of voltage (1kV/div with 1000x probe) and current (1A/div using 10ohm low-inductance sense resistor).  This capture is at the end of a rising arc, just as it is breaking up, where power is highest.  Second image is a zoom into the middle of the first.




To get a longer time view of arc characteristics, I switched rectified sensing.  Below are DC signals of average current and voltage.  Full-wave rectification with low-pass filtering, but no cap at the diodes (to avoid getting peak voltage).  Voltage is 450/div, or about 500V/div RMS if sine-wave is assumed.  Current is 0.5A/div average, or about 0.56A/div RMS for a sine wave.  First image is a 2-second overview, followed by two zoom-ins.





At the left voltage goes off-screen for 12ms, from turn-on until the initial arc strike.  There's much more instability in arc impedance than I'd expected, as the arc momentarily sticks to rough spots (or whatever surface characteristics make a favorable arc point).  The final arc break-up at the end of the scope capture triggers my DC input over-current limit.  Voltage doesn't drop to 0 because of the switch bypass resistor R2.

The 8ms period modulation of the current and voltage waveforms is due to rectified 60Hz line power.  The "470uF", which measures 415uF, input cap doesn't hold voltage well between line half-cycles.  That's intentional to keep power factor somewhat reasonable.  I'd initially started with just 3uF to filter HF components.  However, the arc goes out much too easily without some current flowing continuously.  Adding the 415uF DC bus capacitor made a big improvement in overall performance.

Finally, a couple videos:
/>
/>
David Knierim

Offline Weston

  • High Voltage Enthusiast
  • *
  • Posts: 27
  • Karma: +2/-0
    • View Profile
Re: Royer oscillator (ZVS) driven Jacob's ladder, E80 core transformer
« Reply #1 on: November 11, 2019, 10:20:39 AM »
Cool design! I have not kept track of the variations in ZVS type oscillators for driving high voltage transformers over the last few years but I have not seen the use of a mosfet for controlling the power device gate like that before. It seems advantageous in ensuring ensuring the gate voltage goes below the turnoff threshold and in reducing power used to drive the gate.

Is there some oscillation criteria you used to calculate the maximal coupling such that it stays in oscillation when the output is shorted, or did you back that out of simulation?

Offline davekni

  • High Voltage Technician
  • ***
  • Posts: 131
  • Karma: +5/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Royer oscillator (ZVS) driven Jacob's ladder, E80 core transformer
« Reply #2 on: November 12, 2019, 06:02:30 AM »
Thank you for the compliment!  I'd came across the conventional diode and pull-up resistor ZVS circuit almost two years ago, but didn't like the wasted gate-drive power.  So, I spent a couple months working on improvements with simulations and prototypes.  Ended up designing two reasonably-simple versions.  This one with the small HV FETs for gate-drive is my favorite.  The other variation is closer to the original, just adding small LV PFETS in series with the pull-up resistors.  PFET gates wire to the opposite side's primary NFET/IGBT gates.  That way the pull-up resistors draw power only briefly at the zero-voltage transitions, allowing use of lower-ohms for faster switching.  Here's a somewhat messy schematic.  Can't find any of the cleaner versions on this computer.


The K<0.86 threshold comes from simulation.  Couldn't think of any way to calculate it directly.  Above 0.86, it will oscillate with high and low impedance loads, but not at intermediate loads.  The Q is high at either load extreme.  Low load impedance is all leakage-inductance oscillation (high frequency).  High load impedance is normal inductance oscillation.  Only in the middle does the load lower the Q enough to end oscillation (and cause input current to ramp up indefinitely) if the K is above ~0.86.

I found roughly the same K threshold for H-Bridge ZCS series-resonant drive.  Used that ZCS topology in a 6kW PFC feeding my DRSSTC Vbus.
David Knierim

Offline Weston

  • High Voltage Enthusiast
  • *
  • Posts: 27
  • Karma: +2/-0
    • View Profile
Re: Royer oscillator (ZVS) driven Jacob's ladder, E80 core transformer
« Reply #3 on: November 12, 2019, 07:39:00 PM »
I think that the biggest advantage with the N-channel version is that you eliminate the voltage drop of the feedback diode. I don't recall many previous ZVS oscillator variants with IGBTs that have worked well and I think this is largely because the saturation voltage of the IGBT + the voltage drop of the feedback diode is very close to the threshold voltage of the IGBT, so it does not easily turn off.

With the N-channel fet replacing the feedback diode you only have the channel resistance instead of a voltage drop so you get another ~1V of turn-off margin on the gate.

I don't remember you mentioning the PFC in your thread about your tesla coil. Do you have any more info on that? How are you achieving variable conversion ratio and constant resistance across the line cycle

Offline davekni

  • High Voltage Technician
  • ***
  • Posts: 131
  • Karma: +5/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Royer oscillator (ZVS) driven Jacob's ladder, E80 core transformer
« Reply #4 on: November 13, 2019, 05:40:55 AM »
I'd thought of the diode-drop improvement as a secondary advantage to my design, but it does help.  The case where it seems most useful is with higher voltage FETs for power switches.  Their gate threshold voltage is typically lower than that of IGBTs, and the Vds drop higher that IGBTs at max current.  Lower-voltage FETs (as in the Chinese 1-2kW 36-48V induction circuits) have lower voltage drop at max current, more like IGBTs.  The issue I saw, at least in simulation, was gate capacitance.  The circuits I'd seen used 100 ohm gate pull-up resistors.  I wanted to parallel more FETs for efficiency, but then they turned on too slowly.

My FET gate drive topology does have one notable drawback.  The gate voltage varies from one diode drop above the control voltage down to one gate-threshold-voltage below the control voltage.  For driving IGBT gates to 15V minimum, I use a 19V-20V control voltage, but that pushes the control FET's Vgs right to their 20V maximum limit.

This Jacob's ladder circuit is my first ZVS using IGBTs.  Initially the gate-drive FETs kept frying.  I had to add the 2.8-ohm series resistors to the drains of the gate-drive FETs, and the IGBT gate-to-ground diodes to fix the frying (R8, R9, D2, D3).  I think the issue is slow forward-recovery time of the IGBT diodes causing negative collector spikes.  Not certain, but adding those four parts fixed the frying.  My previous all-FET ZVS circuits don't have those four parts.

Sometime I need to post my "PFC" design, although I'm not sure what category would be appropiate.  It is crude.  My primary goal is reasonable power factor at full load, to get the most from a given power line circuit.  The power factor gets worse at lighter load.  Even at full load it's perhaps ~95%.  Input current is closer to constant (opposite of bridge directly to bulk caps), dropping to zero when the line voltage is near zero (under ~30% of peak voltage).  Current does rise some towards the center of each line half-cycle.

The circuit is a bit like a DRSSTC driver feeding a series-resonant circuit - cap feeding primary of a K=0.85 transformer.  Transformer secondary is bridge-rectified into bulk capacitors, becoming the isolated PFC output.  It handles charging the output bulk capacitors from 0V without needing any inrush-current limiting thermistors.  That part is like this Jacob's ladder - the resonant frequency is higher with low output voltage.  (Power factor isn't as clean during charging.)
David Knierim

Offline Mads Barnkob

  • Administrator
  • High Voltage Expert
  • *****
  • Posts: 1180
  • Karma: +17/-0
  • Denmark
    • View Profile
    • Kaizer Power Electronics
Re: Royer oscillator (ZVS) driven Jacob's ladder, E80 core transformer
« Reply #5 on: November 13, 2019, 05:04:30 PM »
So how much higher voltage rated MOSFETs could be used? Or have you tested?

The offline (230VAC~325VDC) fed ZVS drivers have always been a widely sought topology but it almost always ends up in flames :)
http://www.kaizerpowerelectronics.dk - Tesla coils, high voltage, pulse power, audio and general electronics

Offline profdc9

  • High Voltage Engineer
  • ****
  • Posts: 232
  • Karma: +11/-0
    • View Profile
Re: Royer oscillator (ZVS) driven Jacob's ladder, E80 core transformer
« Reply #6 on: November 13, 2019, 05:51:27 PM »
So how much higher voltage rated MOSFETs could be used? Or have you tested?

The offline (230VAC~325VDC) fed ZVS drivers have always been a widely sought topology but it almost always ends up in flames :)

Yeah, the peak voltage on the transistors is very high.  Perhaps 1200 volt transistors would work for ZVS from mains, but half/full bridge can be run with ZVS easily with an inductive load in series with the primary coil.  For me the high-voltage ZVS is a dead end.  The one advantage of the low voltage ZVS is that it is somewhat safer for induction heaters if no isolation transformer is used to the work coil, but probably isolation is needed for any reasonable safety.


Offline davekni

  • High Voltage Technician
  • ***
  • Posts: 131
  • Karma: +5/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Royer oscillator (ZVS) driven Jacob's ladder, E80 core transformer
« Reply #7 on: November 14, 2019, 04:32:50 AM »
I haven't personally tested above 120VAC/170VDC.  A friend in Germany prototyped a ZVS induction heater running on 230VAC (325V peak) using 1200V SiC FETs.  He's using my older circuit topology with small low-voltage PFETs to reduce gate-drive power and allow stiffer pull-up resistors (shorter turn-on time).  He's planning to make a commercial product, so likely doesn't want details shared here.

Even 400VAC (566V peak) should be possible using 2500V IGBTs.  It would likely require gate-driver ICs and more complex control.  I think of simplicity as the key advantage of ZVS Royer oscillators.  Once the control gets more detailed, is there a reason for ZVS over bridge ZCS?

profdc9:   Is there a bridge ZVS topology?  I'd love to learn about that.  Or, did you mean ZCS for the bridge?

BTW, 1200V FETs at 325V peak input has very little headroom.  Roughly the same ratio as my 170V input here using 600V IGBTs.  Success with such tight margins required two design features:  Lots of TVS protection and controlled oscillation startup.  The latter is the reason for my low-power idle oscillation before the main switch is closed.
David Knierim

Offline Uspring

  • High Voltage Technician
  • ***
  • Posts: 122
  • Karma: +9/-0
  • Physicist
    • View Profile
Re: Royer oscillator (ZVS) driven Jacob's ladder, E80 core transformer
« Reply #8 on: November 14, 2019, 11:45:15 AM »
For a fast power up of the Royer, the tank voltages can rise up very high, i.e. 2*pi*input voltage. See here: https://highvoltageforum.net/index.php?topic=46.msg245#msg245
That's from theoretical considerations omitting any losses and assuming an instantaneous rise of the DC input voltage. In reality the DC caps will prevent this and the rise up timing has to be seen relative to the Royer oscillating period.

Offline T3sl4co1l

  • High Voltage Experimenter
  • **
  • Posts: 91
  • Karma: +8/-0
    • View Profile
Re: Royer oscillator (ZVS) driven Jacob's ladder, E80 core transformer
« Reply #9 on: November 14, 2019, 01:50:30 PM »
The "ZVS circuit" isn't worth using beyond, say, a kW or so.  IMHO, it's not worth using beyond 10W or so.  At high power, the added cost of support components more than justifies the modest cost of gate drivers, protection and control circuits, for a conventional half or full bridge resonant design.

Tim

Online John123

  • High Voltage Enthusiast
  • *
  • Posts: 18
  • Karma: +0/-0
    • View Profile
Re: Royer oscillator (ZVS) driven Jacob's ladder, E80 core transformer
« Reply #10 on: December 11, 2019, 02:22:10 PM »
Very nice looking unit! Does that soft start switch help prevent the latch up problems some people have experienced with the ZVS driver? I've never experienced it myself but some people think it has to do with power supplies taking too long to supply the initial "kick" to get the oscillation going.

Offline davekni

  • High Voltage Technician
  • ***
  • Posts: 131
  • Karma: +5/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Royer oscillator (ZVS) driven Jacob's ladder, E80 core transformer
« Reply #11 on: December 12, 2019, 04:51:08 AM »
Yes.  I've built 5 or 6 ZVS circuits and ran lots of simulations.  When power is applied, current in the power-feed inductor(s) from the supply ramps up fairly quickly.  Initially there's no oscillation, so both FET drains (or IGBT collectors) are at 1-gate-voltage above ground.  Eventually the positive feedback gets oscillation going.  By that time the power-feed inductor current is much higher than in normal operation, perhaps even saturating and/or shutting down the DC input supply.  If the supply doesn't shut down, the high current causes a separate issue.  All that stored energy causes the oscillation voltage to spike way above normal.  (The oscillation amplitude rings for several cycles.)  To handle this voltage, either the FETs need to be rated for much higher voltage than needed during normal use, or TVS diodes are needed to clamp the voltage.  I've done both.

Before this design I played with lots of ideas to kick-start the oscillation, from just unbalanced part values to extra circuitry to hold off one gate signal until power is up, or hold of both gates with staggered release.  At best, these got the oscillation voltage overshoot (and ring) down to 2x rather than 3-5x.  At this point, the only solution I've found is what's in this design:  Let the oscillation start slowly at low current, feed by bleed resistor, R2 in this circuit.  Once oscillation is going, applying full voltage isn't so problematic.  It takes much less time for the oscillation to ramp from the low amplitude than it does to start with no oscillation at all.  I still have TVS diodes, as I'm pushing the voltage, 600V rated IGBTs for a normal-operating peak of 533V (PI * sqrt(2) * 120V, for 120VAC line power input).

I'm about to build a small low-voltage ZVS with current-limited buck-regulator input, using the ZVS input inductor for double-duty as the buck-regulator inductor.  Of course, the buck-regulator itself isn't zero-voltage or zero-current switching, so somewhat defeats the point of ZVS.
David Knierim

Online John123

  • High Voltage Enthusiast
  • *
  • Posts: 18
  • Karma: +0/-0
    • View Profile
Re: Royer oscillator (ZVS) driven Jacob's ladder, E80 core transformer
« Reply #12 on: December 12, 2019, 02:42:19 PM »
It's weird how the starting-latchup thing only happens for some people and not others. I always used car batteries in series for this circuit, could the practically zero output impedance from those have been why it never failed to oscillate?

I'm glad it didn't anyway, a 48v battery bank shorting out on the dining room table would of put me in the doghouse!

Offline AndreiRS

  • High Voltage Experimenter
  • **
  • Posts: 73
  • Karma: +0/-0
    • View Profile
Re: Royer oscillator (ZVS) driven Jacob's ladder, E80 core transformer
« Reply #13 on: December 13, 2019, 03:59:49 AM »
By the different power supplies I tested. I think what makes the circuit fail to oscilate, is when the power supply can't hold current peaks without too much voltage drop. Maybe electric noise from bad psu too.

High Voltage Forum

Re: Royer oscillator (ZVS) driven Jacob's ladder, E80 core transformer
« Reply #13 on: December 13, 2019, 03:59:49 AM »

 


* Recent Topics and Posts

post Re: Unusual polarized non-electrolytic energy discharge capacitor
[Capacitor banks]
MRMILSTAR
Today at 05:19:51 AM
post Re: High voltage Christmas ornaments
[Light, lasers and optics]
klugesmith
Today at 04:57:27 AM
post Re: Unusual polarized non-electrolytic energy discharge capacitor
[Capacitor banks]
T3sl4co1l
Today at 04:04:35 AM
post Re: High voltage Christmas ornaments
[Light, lasers and optics]
John123
Today at 03:29:26 AM
post Re: Mineral oil breakdown voltage
[Voltage Multipliers]
klugesmith
Today at 01:14:06 AM
post High voltage Christmas ornaments
[Light, lasers and optics]
klugesmith
Today at 12:03:49 AM
post Small 65Kv X Ray Transformer hosuing build
[Transformer (iron core)]
DashApple
December 14, 2019, 10:52:49 PM
post Re: Mineral oil breakdown voltage
[Voltage Multipliers]
johnf
December 14, 2019, 09:23:27 PM
post Re: Mineral oil breakdown voltage
[Voltage Multipliers]
MRMILSTAR
December 14, 2019, 09:12:47 PM
post Re: Mineral oil breakdown voltage
[Voltage Multipliers]
johnf
December 14, 2019, 09:07:42 PM
post Re: Unusual polarized non-electrolytic energy discharge capacitor
[Capacitor banks]
MRMILSTAR
December 14, 2019, 09:04:49 PM
post Re: Mineral oil breakdown voltage
[Voltage Multipliers]
MRMILSTAR
December 14, 2019, 08:32:36 PM
post Re: Unusual polarized non-electrolytic energy discharge capacitor
[Capacitor banks]
klugesmith
December 14, 2019, 07:01:18 PM
post Re: Mineral oil breakdown voltage
[Voltage Multipliers]
klugesmith
December 14, 2019, 06:21:27 PM
post Mineral oil breakdown voltage
[Voltage Multipliers]
MRMILSTAR
December 14, 2019, 05:00:39 PM
post Re: QCW ramp generator
[Dual Resonant Solid State Tesla coils]
hammertone
December 14, 2019, 03:05:48 PM
post Re: Unusual polarized non-electrolytic energy discharge capacitor
[Capacitor banks]
T3sl4co1l
December 14, 2019, 05:45:54 AM
post Re: Modern go-to switching devices
[Beginners]
T3sl4co1l
December 14, 2019, 05:37:17 AM
post Re: SGTC MK1 - An Accomplishment in Progress
[Spark gap Tesla coils]
davekni
December 14, 2019, 05:26:41 AM
post Re: Modern go-to switching devices
[Beginners]
klugesmith
December 14, 2019, 04:54:15 AM
post Re: Duty cycle when driving a CRT TV flyback transformer
[Transformer (ferrite core)]
John123
December 14, 2019, 02:30:40 AM
post Re: Duty cycle when driving a CRT TV flyback transformer
[Transformer (ferrite core)]
jturnerkc
December 14, 2019, 02:21:19 AM
post Re: Aethra CCD Webcam Teardown - From Back When Streaming Was Expensive!
[Electronic circuits]
John123
December 14, 2019, 02:14:48 AM
post Modern go-to switching devices
[Beginners]
John123
December 14, 2019, 01:54:55 AM
post Re: Duty cycle when driving a CRT TV flyback transformer
[Transformer (ferrite core)]
John123
December 14, 2019, 01:37:03 AM
post Re: SGTC MK1 - An Accomplishment in Progress
[Spark gap Tesla coils]
jturnerkc
December 14, 2019, 12:57:15 AM
post Re: SGTC MK1 - An Accomplishment in Progress
[Spark gap Tesla coils]
MRMILSTAR
December 13, 2019, 11:48:12 PM
post Re: SGTC MK1 - An Accomplishment in Progress
[Spark gap Tesla coils]
jturnerkc
December 13, 2019, 11:19:27 PM
post Re: SGTC MK1 - An Accomplishment in Progress
[Spark gap Tesla coils]
John123
December 13, 2019, 11:00:18 PM
post Re: SGTC MK1 - An Accomplishment in Progress
[Spark gap Tesla coils]
jturnerkc
December 13, 2019, 10:50:03 PM
post Re: Unusual polarized non-electrolytic energy discharge capacitor
[Capacitor banks]
MRMILSTAR
December 13, 2019, 10:46:20 PM
post Re: Unusual polarized non-electrolytic energy discharge capacitor
[Capacitor banks]
klugesmith
December 13, 2019, 10:31:03 PM
post Re: Unusual polarized non-electrolytic energy discharge capacitor
[Capacitor banks]
AndreiRS
December 13, 2019, 10:00:04 PM
post Re: SGTC MK1 - An Accomplishment in Progress
[Spark gap Tesla coils]
John123
December 13, 2019, 09:08:28 PM
post Unusual polarized non-electrolytic energy discharge capacitor
[Capacitor banks]
MRMILSTAR
December 13, 2019, 09:06:47 PM
post Re: SGTC MK1 - An Accomplishment in Progress
[Spark gap Tesla coils]
MRMILSTAR
December 13, 2019, 08:56:27 PM
post Re: SGTC MK1 - An Accomplishment in Progress
[Spark gap Tesla coils]
jturnerkc
December 13, 2019, 05:05:55 PM
post Re: How do you keep in contact with forum people...?
[General chatting]
Mads Barnkob
December 13, 2019, 09:31:58 AM
post Re: My first DRSSTC on bricks
[Dual Resonant Solid State Tesla coils]
johnf
December 13, 2019, 07:56:39 AM
post Re: Royer oscillator (ZVS) driven Jacob's ladder, E80 core transformer
[Transformer (ferrite core)]
AndreiRS
December 13, 2019, 03:59:49 AM
post How do you keep in contact with forum people...?
[General chatting]
AndreiRS
December 13, 2019, 03:23:31 AM
post Re: Duty cycle when driving a CRT TV flyback transformer
[Transformer (ferrite core)]
klugesmith
December 13, 2019, 12:16:53 AM
post Re: My first DRSSTC on bricks
[Dual Resonant Solid State Tesla coils]
Laci
December 12, 2019, 08:35:24 PM
post Re: How To Make a Stencil For Spray Paint With Cheap Hand Tools
[General chatting]
johnf
December 12, 2019, 07:53:35 PM
post Re: Duty cycle when driving a CRT TV flyback transformer
[Transformer (ferrite core)]
John123
December 12, 2019, 02:52:58 PM
post Re: Royer oscillator (ZVS) driven Jacob's ladder, E80 core transformer
[Transformer (ferrite core)]
John123
December 12, 2019, 02:42:19 PM
post Re: Full bridge much "weaker" than half bridge?
[Solid state Tesla coils]
nick
December 12, 2019, 11:17:36 AM
post Re: Royer oscillator (ZVS) driven Jacob's ladder, E80 core transformer
[Transformer (ferrite core)]
davekni
December 12, 2019, 04:51:08 AM
post Re: Duty cycle when driving a CRT TV flyback transformer
[Transformer (ferrite core)]
davekni
December 12, 2019, 04:26:22 AM
post Re: How To Make a Stencil For Spray Paint With Cheap Hand Tools
[General chatting]
AndreiRS
December 12, 2019, 01:21:35 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