If you need assistance, please send an email to forum at 4hv dot org. To ensure your email is not marked as spam, please include the phrase "4hv help" in the subject line. You can also find assistance via IRC, at irc.shadowworld.net, room #hvcomm.
Support 4hv.org!
Donate:
4hv.org is hosted on a dedicated server. Unfortunately, this server costs and we rely on the help of site members to keep 4hv.org running. Please consider donating. We will place your name on the thanks list and you'll be helping to keep 4hv.org alive and free for everyone. Members whose names appear in red bold have donated recently. Green bold denotes those who have recently donated to keep the server carbon neutral.
Special Thanks To:
Aaron Holmes
Aaron Wheeler
Adam Horden
Alan Scrimgeour
Andre
Andrew Haynes
Anonymous000
asabase
Austin Weil
barney
Barry
Bert Hickman
Bill Kukowski
Blitzorn
Brandon Paradelas
Bruce Bowling
BubeeMike
Byong Park
Cesiumsponge
Chris F.
Chris Hooper
Corey Worthington
Derek Woodroffe
Dalus
Dan Strother
Daniel Davis
Daniel Uhrenholt
datasheetarchive
Dave Billington
Dave Marshall
David F.
Dennis Rogers
drelectrix
Dr. John Gudenas
Dr. Spark
E.TexasTesla
eastvoltresearch
Eirik Taylor
Erik Dyakov
Erlend^SE
Finn Hammer
Firebug24k
GalliumMan
Gary Peterson
George Slade
GhostNull
Gordon Mcknight
Graham Armitage
Grant
GreySoul
Henry H
IamSmooth
In memory of Leo Powning
Jacob Cash
James Howells
James Pawson
Jeff Greenfield
Jeff Thomas
Jesse Frost
Jim Mitchell
jlr134
Joe Mastroianni
John Forcina
John Oberg
John Willcutt
Jon Newcomb
klugesmith
Leslie Wright
Lutz Hoffman
Mads Barnkob
Martin King
Mats Karlsson
Matt Gibson
Matthew Guidry
mbd
Michael D'Angelo
Mikkel
mileswaldron
mister_rf
Neil Foster
Nick de Smith
Nick Soroka
nicklenorp
Nik
Norman Stanley
Patrick Coleman
Paul Brodie
Paul Jordan
Paul Montgomery
Ped
Peter Krogen
Peter Terren
PhilGood
Richard Feldman
Robert Bush
Royce Bailey
Scott Fusare
Scott Newman
smiffy
Stella
Steven Busic
Steve Conner
Steve Jones
Steve Ward
Sulaiman
Thomas Coyle
Thomas A. Wallace
Thomas W
Timo
Torch
Ulf Jonsson
vasil
Vaxian
vladi mazzilli
wastehl
Weston
William Kim
William N.
William Stehl
Wesley Venis
The aforementioned have contributed financially to the continuing triumph of 4hv.org. They are deserving of my most heartfelt thanks.
Well, after an absence of about 15 years I have fallen head on back into coiling. My last project in about 2001 was a pig powered spark gap beast, but now I am venturing into the realm of SSTCs, with their compact size and low noise.
The secondary I have wound for the project is 12" diameter PVC with about 1000 turns of 0.5mm wire (24AWG). It resonates at about 83kHz in use with a small corona ring.
I am using a H bridge made from IRG4PF50WDs, which are rated at 900v and use out to 150kHz. I have two 400V 1500W TVSs in series across each IGBT and a 17v TVS on the gates. The ultimate plan is to run using 586VDC from a rectified 3 phase supply (we have 415VAC between phases here in oz).
Currently I am using a very simple antenna feedback scheme with no PLL or interrupter. It works very well but dies when I draw arcs (as expected). I need to make a more sophisticated driver, but for now I am happy enough just watching the coil arc freely to air. I am using TO220 12A driver ICs to drive the GDT.
Anyway, I fired it up for the second time yesterday using a lashed primary of about 5 turns of 2.5mm2 wire. After initial low power testing I was able to wind up the variac to about 260VAC (~360VDC) at which point I was getting snarling thick arcs probably out to 18" or more. It was very impressive and scary, the corona ring had 3 simultaneous branches coming off it and swirling around all over the place. it was not like any SSTC video I have seen and was quite like a SGTC. Very frantic.
I have no idea what the power level was except that my 10A breaker was holding up! I was running full wave rectified 260VAC with about 100uF of capacitance to prevent the voltage from dropping to zero between cycles and causing the feedback scheme to stop working. It was working very nicely!
Anyway, at this point there was a big bang and a blue flash from near one of the IGBTs. The bridge is dead, but I can't see any visible damage to any of the devices. I will have to dissect the bridge this weekend and do some testing. This is the second bridge I have killed and neither showed any visible damage which is not in line with what people normally say about IGBT failure, so I suspect the TVSs have failed. Any thoughts?
So, progress has been good. Now some questions.
1) Is it a reasonable guess that I may have killed the TVS? The IGBTs were hardly getting warm in use.
2) If I have killed a TVS, any idea why? Perhaps I need some better decoupling across the bridge to adsorb transients (currently not using any other than the main smoothing caps in the PS)?
3) Are the IGBTs I have chosen up to the job of switching 10kW? Can anyone suggest a good alternative? I am not sure which specs to check other than perhaps rise and fall times and gate charge. Any advice on selecting a suitable IGBT other than of course the current and voltage ratings?
4) What are the benefits of using a transformer on the secondary ground lead rather than an antenna for feedback?
5) What exactly is happening when one draws arcs and causes the feedback system to stop working?
Registered Member #152
Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384
Hi,
First of all, congratulations to building something else than a DRSSTC which everyone today does. I also like the brutal power of CW or half wave coils.
There are just too many things to go wrong.
First of all, the TVS diodes are of no use. They can clamp a current of maximum around 3 amps while the voltage rises to 600 volts. All they can do is short out and take the rest of the bridge with them. Check the TVS diodes; if any have failed this is the most probable cause of your problem. Next time just don't use them, or only use them on the supply rails.
The most important thing in preventing voltage spikes is bridge layout. All connections must be as short and thick as possible, this holds true for the connection between the transistors and the supply rf bypass or snubber capacitor (I hope you used one?) Edit: I checked you did not !!! You absolutely must use this cap.
Then, of course you could have fried the transistors by overheating. Did you check the switching losses and calculate with thermal resistances and impedances? Do you use insulating pads between transistors and the heatsink? This all goes with the must to measure primary current waveform, and identifying the current at which switching transitions occur. Overheating of the anti-parallel diode in the transistor is also possible, as it has usually lower current rating than the IGBT.
Another thing you must have right are the gate waveforms. I believe you checked those.
Last but not least, control circuit glitch is also possible. These are the worst types of failures to debug.
Registered Member #15
Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068
Check for over-heating. CW type coils dissipate an ENORMOUS amount of heat as they are operating at 100% duty cycle (vs. low duty cycles of a DRSSTC etc....)
Check gate drivers (they can get smoking hot during 100% operation) as well as other switching power electronics.
I agree with the above! Outside of a few SGTCs all of my ~9 coils have been CW SSTC ones (3kW being my largest). Bridge layout is primary suspect. Remove the TVS diodes from everything except the gates. Place a GOOD snubber cap across each half bridge. Bring the bus capacitor(s) as close to the transistors as you can. Shorten the length between components as much as possible, including the bridge rectifier to the caps.
Antiparallel diodes are another suspect, though I generally work with MOSFETs where it is an inferior body diode. I always include ultrafast high current diodes parallel to the antiparallel intrinsic/included diodes for good measure, and in doing so have noticed significant reduction in the thermal dissipation of the transistors.
Using CT feedback eliminates the loss of feedback when you draw an arc or place a heavy load on the output. You could probably experiment with antenna placement to prevent it from dropping out as well; my coils all use antenna feedback and do not drop out when drawing an arc. I have found that shorter, closer antennae work far better than longer more distant ones.
I eventually switched over to a PLL based driver since at higher power levels I found RF feedback was causing interference patterns on the drive signal resulting in a harsh tone in the output and effects similar to SGTC output. This also eliminated the need for a minimum bus capacitance (to prevent feedback dropout at zero crossings). The unfiltered input results in longer sparks and lower power dissipation. You can then add bus capacitance as you see fit while keeping an eye on thermal dangers.
Finally, CW SSTCs tend to like high coupling factors, experiment with adding primary turns. Doing so increases the input impedance of the coil and serves to limit current. You may find that it works fine now on 370V but on 567V it may draw so much current that thermal destruction is inevitable. The primary coil itself can rack up a rather large I2R loss at the currents involved and overheat. I can only run my 3kW coil for short periods of time (<1min) because the primary will overheat, and I'm using 10ga wire on it (which I think is thicker than yours). You can also experiment with lowering the coupling to reduce current draw, but it is my experience that a higher coupling + more turns works better than lower coupling with less turns (unless you're working in Class-E in which case all that matters is the switching current!).
Registered Member #56
Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445
I would try to move away from the antenna feedback for a coil of this size, you may even have better luck running at a fixed frequency (with the huge streamer loading of these coils they actually become quite easy to run with a fixed frequency). I run all of my coils from a 4046 pll chip, with the PLL part disabled so that it is just a VCO connected to a knob on the coil somewhere.
You should also really make an effort to scope the bridge and make sure the ringing is not too bad, I have found that with large mosfets it is often necessary to add in snubbers to keep the ringing in check.
Registered Member #146
Joined: Sun Feb 12 2006, 04:21AM
Location: Austin Tx
Posts: 1055
Id bet all my money on poor thermal management... aka, those IGBTs got too hot!
Those IGBTs can do 200W with a perfect cooling system, not very impressive for a TO-247 to begin with (ive seen up to 600W advertised on this package).
If you use "sil-pad" or any other insulator between the IGBTs and heatsinks, then you are cutting its power dissipation down to a small fraction of that 200W, maybe just 10-20W depending on the material type (some are really bad!!). With individual small heatsinks (no insulator pads) and a fan, you might get 50W out of each IGBT, realistically.
I hear some people report "my IGBTs blew but the heatsink isnt even warm!". This is a good indication of having too much thermal resistance between IGBT chip and heatsink. In this case, you are not transferring the heat quickly enough and the little tiny IGBT chip (probably no more than 7mm on a side) overheats before the big aluminum heatsink feels warm.
That being said... a better controller/driver can reduce your IGBT losses and perhaps make it possible to operate with less cooling.
Ground arcing is hard on most TCs for various reasons. For the antenna feedback SSTC its hard because there is a disruption in the feedback signal and in the oscillation of the TC. Some SSTCs regain oscillation at the right frequency, some dont. The ones that dont will sometimes oscillate at some other mode, usually much higher frequency, due to the shorted secondary. Othertimes, however, the oscillations can slow down (before stopping altogether usually) and this can be very dangerous for the switches since the primary impedance can start to look very small at frequencies well below resonance of the TC. This can lead to large, hard-switched currents that can instantly pop an IGBT if outside of the SOA (safe operating area).
Many thanks to everyone for taking the time to write such informative answers. I agree that the SSTC really is cool with its high level of continuous RF. It is surprisingly easy to get rf burns just by touching anything metal in the area. I wonder how good being exposed to that rf field is for our health? The sparks a just a completely different beast to my old SGTCs.
Anyway, this weekend I will revise my bridge layout and do some scoping. I was actually pretty amazed that it worked as well as it did because everything is pretty lashed together. I think the low 83kHz operating frequency is cutting me some slack and is perhaps a bit more forgiving of layout than most!
I will remove the TVS diodes across the IGBTs and just put a few across the PS rails. It also sounds like I need some decoupling.
I agree that a small antenna seems to work well, currently I am not even using one! I don't think my igbts were getting particularly hot, only a tiny bit warm after a 15 sec run. However I would like some bigger ones if I can get them. So, some more questions:
1) My bridge is currently floating. Should it be grounded? Does it matter?
2) How large should my decoupling caps be? Is 1uF good?
3) Any ideas on a better IGBTto use?
4) I am interested in trying CT feedback. Can I clamp the CT voltage with normal diodes eg 1n4148 or will that cause phase problems? I remember back in the day people were using 1N60s due to their low capacitance. Is that still the case? I was hoping to use something similar to the attached image. Is there any reason for all the paralleled not gates after the initial feedback not gate? They seem superfluous. I was just hoping to run a simple non inverting Schmitt trigger directly into the driver ICs and no interrupter. Any thoughts? Does anyone have any good circuits they can share? I'm trying to avoid a PLL circuit for the moment but will use one if necessary!
Registered Member #146
Joined: Sun Feb 12 2006, 04:21AM
Location: Austin Tx
Posts: 1055
3) Any ideas on a better IGBTto use?
If you can link to a vendor that will supply you parts, i could browse around. You didnt comment on your heatsinking setup, i still strongly recommend individual heatsinks for each IGBT so that you dont need electrical isolation. This does mean that 2 of the heatsinks will be at 360V, and 2 of them will be jumping between the 2 DC rails, so you need to be careful mounting the heatsinks to avoid inadvertent short circuits.
2) How large should my decoupling caps be? Is 1uF good?
It really depends on the particular SSTC, but 1uF is a decent start. It will at a minumum help with really fast transient voltages that are produced by the IGBT switching through some stray inductance in the bridge layout. You should (if possible) scope the DC voltage at the IGBTs and look for how much its bouncing around... more caps is usually better. Keep them high quality PP film if possible.
4) I am interested in trying CT feedback. Can I clamp the CT voltage with normal diodes eg 1n4148 or will that cause phase problems? I remember back in the day people were using 1N60s due to their low capacitance. Is that still the case? I was hoping to use something similar to the attached image. Is there any reason for all the paralleled not gates after the initial feedback not gate? They seem superfluous. I was just hoping to run a simple non inverting Schmitt trigger directly into the driver ICs and no interrupter. Any thoughts? Does anyone have any good circuits they can share? I'm trying to avoid a PLL circuit for the moment but will use one if necessary!
Yeah, the 1n60 was used for its low capacitance, but i found that 1n4148 or 1n914 diodes have sufficiently low capacitance (just a few pF) to work fine with antenna or CT input clamping. Keep in input current to about 25-50mA on the feedback input as to not cook the diodes. Your SSTC sounds like it probably runs with about 1.5-3A of base current, so if you do try a CT, id wind it with a ratio of 1:100 or so.
Some of my old circuits use a CT feedback with a large series resistance... this was because i had little idea what it was i was doing and its probably a bad idea. Just running the CT outputs between ground and the diode-clamped input should do the trick.
Even with CT feedback it certainly does not tolerate ground sparking since the secondary resonance is usually severely disturbed by this. This was why the DRSSTC became so appealing, it could use primary side feedback and the primary coil will continue to oscillate at a "good" frequency during ground arcs so the system can maintain operation. DRSSTCs, however, introduce tuning challenges which can result in running more primary amps than would otherwise be necessary for the same performance on a (SR)SSTC.
My bridge did unfortunately use sil pads and all four igbts were mounted on a large 0.5 deg/w heatsink left over from an amplifier project. I will post some photos tonight. It's a very crude setup, amazing it worked well out to the voltage it did!
I am using MIC4451/2 12A TO220 driver ICs. They barely get warm with no heatsink at 83khz. I wholeheartedly recommend them. Not cheap though!
So for the CT feedback, no series resistor. Would you still use a series capacitor? Perhaps say 1-10nF?
I am tempted to try running the diode clamped CT straight into the GDT driver ICs. Do you think that would work? One day I will add an interrupter but for now CW is the goal.
Thanks for your advice! Hope to try the CT this weekend.
Don't laugh, but here is a photo of my bridge. I know it's pretty rough but it was never meant to be a final design, just a proof of concept. I wanted to see if I could build a working SSTC and I did! The final version will be much more compact and neat. The TVSs are hidden from view, being underneath the pins.
This site is powered by e107, which is released under the GNU GPL License. All work on this site, except where otherwise noted, is licensed under a Creative Commons Attribution-ShareAlike 2.5 License. By submitting any information to this site, you agree that anything submitted will be so licensed. Please read our Disclaimer and Policies page for information on your rights and responsibilities regarding this site.
Fired up my 10kW (to be) CWSSTC on the weekend. Big sparks then bang!....questions!
