High Voltage Forum
High voltage => Transformer (Ferrite Core) => Topic started by: zytra on January 17, 2021, 05:55:11 AM
-
Hi,
I am creating a new thread before splitting off too deep from the original thread (link: https://highvoltageforum.net/index.php?topic=1374.msg10493#new )
I'll re-post my last message (and David's reply) as it gave a fairly detailed description:
The transformer is fully custom, I'll post pictures later tonight. It's made of 2x "C" shape ferrite bolted together. The secondary is roughly 26mm ID x 50 mm long (that includes 15 separations that are about 2mm thick). The core has an OD of 20mm, meaning 3mm of insulating PTFE. The segmented secondary was turned on a piece of PTFE. The core is not gapped.
For the primary I have 2x 2.5 turns, which is a bit on the small side which could explain the lack of success with the ZVS driver.
After running the transformer on the SSTC driver, I decided to make my own ZVS driver. And since I ended up using roughly the same components, results were the same outside of the capacitor that was film but with high ESR, it ended up heating a lot so I replaced it with a better quality capacitor with significantly lower ESR but also slightly higher capacitance. Results were immediately better, and after playing with the few caps I had available, found that 4.7 uF was giving me the best results. Note: I did go back to the commercial ZVS and put the same 4.7uF cap across the far ends taps of the primary and results were much better, but as good as mine oddly - perhaps due to the other caps still being there. I say "caps" because that driver used 2 for some reason. The ZVS circuit I am familiar with use 1 cap across the far ends of the primary, and this one uses two. I looked at the PCB and it looks like they use 2x ~0.330F but I am not sure how it is all wired together.
I proceeded with assembly of my 200mm plasma ball, flushed it with argon at 1atm and results were great. Voltage was probably a bit high because I can't run a vacuum on that cheap plastic globe (that has a flat bottom - I tried by the way, and it's not a good idea, haha). I used a grounded striker to test contact and it would definitely shock you as opposed to normal plasma ball.
The goal is to make a bigger one (glass, round, under vacuum) so I don't think I will dial down on the voltage yet, just to make sure I have enough for the larger one down the road.
I did experience a set back with a failure of the secondary. The last chamber, last few turns (hard to tell exactly) arc'ed through the PTFE to the ferrite. Not through the thickness which was 3mm but through the last separation which was only 2mm in my design for some reason.
I've made revisions to the design and will turn it this week, increased the thicknesses separating the windings from the ferrite to a total of 4mm. I had to reduce the thickness of chamber to chamber separation to compensate, but I think this will be fine, as I don't expect any arc to get through 1mm PTFE on two adjacent chamber. If it took 16 chambers to go through 2mm or so on the HV side where the potential difference is maximum, I think I should be fine with 1mm between 2 adjacent chamber.
Another setback today after rewinding a spare V1 core. I wound this one with only 2 differences: I didn't fill the last chamber (HV side), and I tried to compensate by adding 1mm more thickness on all of the other chambers. That one failed with an arc within one chamber, along the surface of the PTFE which darkened, but no penetration. I think what happened is that my layering wasn't great on this way, I kinda rushed it last night, the added layers didn't help as it increased the voltage difference within one chamber.
With all tests I found that I was lacking a direct voltage measurement on the HV side. I actually was able to use a HV probe on my early tests (that didn't work well) and measured 8kV. After tweaking the capacitor on my own ZVS I was able to get intense plasma at ambient air and discharges over 2" long, so there is no doubt that my 10kV probe wouldn't be able to work there.
And here's David's reply:
It would be interesting to see waveforms for this - primary voltage and secondary with just antenna pickup (reasonable phase, just no amplitude calibration). ZVS drivers usually aren't used with ungapped inductors. Inductance is high and saturation current low. I wonder if your system is running at a frequency determined by leakage inductance and the primary and secondary (intrawinding) capacitance.
Besides the HV end failure, the common tricky problem with segmented bobbins is the wire transition from the top of one segment to the bottom of the next. Unless there are slots or other accommodations in the separation walls, that wire to the bottom ends up adjacent every layer (including the top) of the new segment as it is wound. That leaves a full segment's voltage across two thicknesses of enamel insulation.
BTW, the two 0.33uF (1200Vdc, 630Vac) caps in the commercial ZVS are almost certainly in parallel. They are standard Chinese induction cooktop capacitors, the ones I use for my DRSSTC MMC. Some ZVS units have 6 or 8 in parallel.
And my response:
- The driver I am using is this one: https://www.amazon.com/gp/product/B07BNZ5HC1/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1
- The first HV failure was definitely a hole through the PTFE, extremely small. The second HV failure (today's failure) was exactly as you described. The wire coming from the previous chamber arc'ed with another one from the new chamber. This is a typical difficult to address mechanically. I initially tried to add a tiny hole at an angle through the separation, which was just too difficult to do without at least a 4th axis. Then I thought about just using a groove. I figured that oil fill up that groove, while this could work I think I haven't been able to machine this, it is entirely possible I just need time to actually setup in the CNC machine one of my spares, or even start with one of the damaged ones.
- I haven't hooked up the oscilloscope for relevant tracing yet, but I will do that on my next try. On one early try with the HV passive probe I measured 8kV and about 50kHz.
A few more things:
As you know me by now, I am not familiar with the technical jargon and I tend to describe things at length when in fact there is usually a simple technical term for it. I'm sorry about that; so you'll judge by my pictures but I'm pretty sure my ferrite core isn't gapped, just two "C's" forming a "O" with a centered-tap primary on one cylinder and a 3000 turns or so secondary. Secondary inductance was measured to several dozen henry, and its DC resistance to ~ 370 ohm.
I have very little footage with that first secondary design, I should have recorded more, it actually worked for a few hours before failing, it actually only failed because I hooked up a larger PSU (I was using a 60V capable PSU but limited to ~3A, so voltage input was around 13V @ 3.2A) and it failed around 17V and 10A or so. So here's the only video I have, sadly with all lights on. It doesn't show it "starting up", which at times can be tricky (I need to use a switch and perhaps use a large bulk cap). It worked pretty well overall and without the need for a sharp breakout point. Also the video was taken as I was experimenting with the Amazon ZVS with the 4.7uF which I found to be optimal with my home made ZVS. So the video shows a little bit less performance than I had with the home made ZVS.
/>
Specs:
- roughly 3000 turns of 34AWG enamel wire
- U shape cores I used: link (https://www.amazon.com/2pairs-Transformer-ferrite-3750nH-Isolator/dp/B074JXFYLC/ref=sr_1_57_sspa?dchild=1&keywords=ferrite+core+ex&qid=1610863998&sr=8-57-spons&psc=1&spLa=ZW5jcnlwdGVkUXVhbGlmaWVyPUEzSlVHRks5TlZDSVBGJmVuY3J5cHRlZElkPUEwOTAzNzMyMzQ3OEZES1ZNSE41TyZlbmNyeXB0ZWRBZElkPUEwMTQ1Njc4M0w3TURGMU8wUjJJMiZ3aWRnZXROYW1lPXNwX2J0ZiZhY3Rpb249Y2xpY2tSZWRpcmVjdCZkb05vdExvZ0NsaWNrPXRydWU=)
- a (unnecessarily) large oil container (I only had 1ft tubing and nothing to make a cut straight enough for a good seal)
- all parts for the container and globe base/globe seal were CNC'ed
- the globe itself is a plastic flat neck (not sure really the material) and sadly a flat bottom (not vacuum/pressure happy) - from amazon. Took this one for its flat neck, which would allow me to design/make a simple seal. The goal is use a 350mm glass globe eventually, that will of course not fail under vacuum.
- the globe base a bunch of fittings for gas fill up and vacuum (I can bring the pressure down by 0.1-0.15 bar safely but that's enough to see a significant improvement on the arcs/plasma).
- 1.5" Brass ball
What's next:
- Short term: rewind yet another secondary, this time I'll add one layer of kapton tape to stick each new wire coming in a new chamber against that separation and keep it there, all the way down. I'll wind a bit slower too, to ensure I layer those turns as evenly as possible. I'll also try to not wind all the way to the top like I did on my first attempt which was overall much better considering it failed at much higher voltage and with a failure mode that can be completely avoided by skipping that last chamber (like I did on the second attempt).
- I am making an improved PTFE core, with thicker ends, and thicker body around the ferrite.
- I am redesigning everything else as well, larger diameter container but also much shorter. I'll also get the HV ground to come down from the bottom of the container rather than the top. I am designing this around a much bigger 20L boiling flask (that was the initial idea), and will be using vacuum valves (2 of them) for fill up and vacuum. I am also going to be using a large test tube (25mm diameter, only those are long enough to reach the center of the flask) as HV electrode support. All of those 3 tubes will go through the rubber cap of the boiling flask.
- I'll probably stick with my home made ZVS driver, unless I can figure out a way to use my SSTC driver (Full bridge, TO247 IGBT's). But that would only possible if I can get anywhere close to soft switching. And if I do stick with the ZVS driver, I need to research more what David mentioned about ZVS driver working better with gapped transformer which is not my case.
- I'd also like to add some kind of interrupter to the system (which would be trivial if I used the SSTC driver).
Pictures:
- Secondary Rev1.0 before failure
- Secondary Rev1.0 after failure: the black area has a pinhole pierced by the arc that went from the HV'most enamel wire through the PTFE to the ferrite
- Secondary Rev1.1: freshly wound secondary, this time with the last chamber skipped. Notice how, out of greed I added more turns to compensate for one less chamber? haha
- Driver pictures: I CNC'ed single sided Clad (4 or 5 oz I think). I used SOT227 Mosfets mainly for practical reasons; also thought I'd easier to kill them, especially considering I'm using current limited PSU's. I need to upgrade the 470 ohm resistor to higher wattage.
-
If you come up with a good solution for segmented bobbin cross-over from segment to segment, we'd all like to hear about it! One option I've pondered is 3D printing to make an appropriate groove for cross-over. If the printed part had enough intentional porosity, oil could fill all the spaces.
Concerning startup issues, that is a common problem with the conventional ZVS circuits. A separate supply for gate pull-up resistors is one option. I've attempted to explain the issue in this thread:
https://highvoltageforum.net/index.php?topic=1227.msg8991#msg8991
-
I actually started with a 3D printed housing. I made 2-3 different designs which all failed but there is quite a bit more to experiment. One of the design was pretty satisfying in regards to the cross over, basically a groove coming from the top of the previous chamber, down to the bottom of the next one. This gives the advantage of spacing out that cross over wire from other wires.
The main reason for these failures is that I wasn't using oil, the number of turns was significantly lower than what I achieved with the turned housing and thought I would get around with it. Failures were due to arc'ing to the ferrite core:
- first try was with a 2-piece housing and it went through at the silicone filled seam
- the next one was a 1-piece housing with in-fill it went through the material
- the final failed in the exact same way despite 100% infill
I am pretty sure that 3D printing is the way to go, when I get my hands on a SLA printer, I will design an open cell unit for oil filling (and with vacuum degassing), with the groove mentioned above.
Last night I designed another turned PTFE unit, much larger as I found some UY30 ferrite. It won't be as compact, but generally speaking I think there are no substitute for size when it comes to this. The bigger the transformer can be, the easier it is to prevent arc'ing.
-
I read your reply in the thread you linked and I have a question. The schematics I used to build my circuit (link: http://4.bp.blogspot.com/-2okKc8gWuqQ/UNpmJYrPJbI/AAAAAAAAACs/pXKNUsBjP-s/s1600/mazzilli_zvs-1.png ) is only different than the one discussed in that thread in how the primary connection is done with the rest of the circuit. I'm pretty sure your suggestion of using a separate supply for the gates would work.
Would adding a 7212 regulator to feed the gates, and adding a switch before the inductance be sufficient? The way I was doing it was by having the PSU on and hot plugging the + banana plug. But I think the reasons why it still wasn't always sufficient (especially on my second attempt) was that it was probably arc'ing/corona somewhere.
I took a look at the ignition coil that came with the ZVS today. I could never get that ignition coil to work, but I didn't need it. Anyway, there was a gap under the primary coils. Probably under 1mm.
In your previous post you mentioned ZVS drivers aren't typically used with ungapped transformers. Could this explain why it actually performed (a lot) better when I used the SSTC driver (driver input controlled by a signal generator)?
I'll wind another secondary and get some scope action tomorrow.
-
The previous link I posted was for an explanation of why there is a startup issue (diodes not conducting so no positive feedback), not a new schematic. Atomillo had success with a separate gate supply, as have I. For two other circuit options, here's my ZVS-driven Jacob's ladder thread:
https://highvoltageforum.net/index.php?topic=831.msg5491#msg5491
Even without a regulator, the key is applying gate power before coil (drain) power. The advantage of using a lower voltage for the gate supply is that it allows for stiffer (lower resistance) gate pull-up resistors (faster switching) without as much added gate-resistor power. 100 ohms works well at 12V. Power would be too high at 40V. My second schematic from the above thread is an easy way to allow even stiffer gate pull-up with even lower power (two small PFETs to disable gate resistors when not needed).
The "ignition coil" in the kit you purchased appears to be a TV flyback transformer, mislabeled in the seller's listing.
Ungapped ferrite cores are used for transformers, but generally not for inductors. Inductance is high, but with correspondingly low saturation current. Inductance changes drastically with drive level. ZVS oscillators are resonant circuits counting on inductance of the load. However, in your case, I wonder if the ZVS oscillator can run well with the transformer's leakage inductance. That could be a useful topology for HV AC, something I hope to explore more, at least in simulation. Your success with ZVS (with 4.7uF added) proves that something is possible. That makes me interested in understanding your success.
-
It completely makes sense, the fact the gate resistor was a 2W actually made me wonder, particularly when I realized I didn't have one handy.
Good thing we talk about this because it reminds me of another interesting thing I picked up.
Because I didn't have a 2W resistor (but only 1/4W) I pulled the thermal camera as to not run the system too long. And well yeah these 1/4W resistors do heat up, as expected. However something else caught my attention:
The primary is wired with the positive DC coming straight into the center tap (after the inductor). As such I as expecting that wire to be hotter than the other 2 wires which should only see half the average current. However, the FLIR short revealed the exact opposite. I've attached a screenshot and you can see the 2 outer primary wires being significantly hotter than the center tap wire (it's right in between the 2 outer wires which appear yellow; note: you can see a 4th wire close to the right primary wire - that 4th wire is the secondary lead going to the ground).
That puzzled me, so I pulled the clamp meter (all the other oscilloscopes and probes were still hooked on another experiments) and found numbers that match my expectations!!! roughly 3.5A in the line going to the center tap and 1.6-17 A on both the other wires. All three are from the same spool, machine wire, 14AWG, stranded, and same length +/- 1/4".
I'm sure there's an explanation, but I didn't immediately see one. The clamp is not super accurate, I'd expect +/- 0.1A - measuring AC, which may not be appropriate, I can't tell without taking a look at the waveform first. But thermal images don't lie. These 2 outer wires were definitely hotter and I can't explain it.
On that test I was pulling around 60W, and the tank capacitor was a large 20 uF (the largest one I tried which gave similar performance to the 4.7 uF one), the shot is probably 1-2 mn after start up.
-
Nice FLIR image you got there.
Maybe the switched wires have more resistance because of skin effect.
They have lots more high frequency current than the center wire.
For heating from mixed AC and DC, you can simply add the I^2R from DC component
to the I^2R from AC component, RMS, possibly with significantly larger R value.
What is the skin depth in copper at your switching frequency?
-
Were you measuring AC or DC current with your clamp meter? The center wire will have mostly DC current. (That's the purpose of the series inductor - to block AC current.) The outer wires will have half the DC current, but much more AC current. RMS current is generally much higher in the outer wires. They are part of the LC resonant circuit, 5-turn primary as L and 4.7uF + 2 x 0.33uF as the C. If Q is high (ie. little arc load), that resonant current can easily be 10x the DC current.
You can calculate the RMS current in those outer wires if you know frequency and voltage (and capacitance = 5.36uF). RMS voltage will be roughly PI/sqrt(2) of your DC input voltage, a bit lower due to IGBT Vce drop.
-
Thanks guys, I will redo these measurements.
The numbers I gave you are AC.
I don't recall if I measured the center-tap line in DC but I did for the outer lines and they read 0A.
-
Perhaps the DC was swamped by the much-higher AC current. If current was being drawn from the supply (beyond just gate-resistor current), it must flow through the inductor, through the transformer primary, to the FET drains.
-
Just to make sure I understand that concept correctly; although the RMS current going in the center of the primary is greater than what flows either side, the heat generated is less because the frequency in this part of the circuit is much less than what it is coming out of the primary, and so the effective section is section (driving the R, and heat up)?
-
>>the frequency in this part of the circuit is much less than what it is coming out of the primary...
Close. The frequency is the same everywhere, but the magnitude of AC current is much smaller in the center wire.
When we combine the outer wire currents to get the center wire current,
typically their DC components will add while the AC components cancel. (not a political word!)
As Dave said, center wire AC current should be very small -- there's a choke inductor in series.
That means the outside wire AC currents have approximately equal magnitudes & opposite phases.
With clip-on probe, can you view combined DC and AC current waveform on a scope?
Or measure the clip-on output with a multimeter?
Good multimeters reject AC when set to DC mode, and reject DC when set to AC mode,
but I would be skeptical of the AC accuracy at tesla coil frequencies.
Still waiting for report of skin depth in copper at your switching frequency. Easy to find formulas, charts, even interactive calculators online.
-
Slight detail: The AC component of center wire current is at 2x oscillation frequency (plus harmonics of that 2x frequency). Voltage at the center looks like full-wave rectified AC, going from roughly 0V to roughly 0.5 * PI * DC input voltage. Current is the integral of that waveform minus the input DC voltage, since the input inductor current is the integral of voltage across it (scaled by 1/inductance). Even though that reduces skin depth, current is so much lower than in the outer wires that it still heats less.
Yes, good idea to look up skin-depth. You could also look up the similar "proximity effect" for inductor and transformer windings - the reason litz wire is sometimes needed.
-
Hi guys,
Well I thought I'd be reporting good news and while I was able to get some results, I'm afraid to report the transformer failed again. Not sure where yet, I'll drain it tomorrow.
It's surprising though, because I decided to rewind with a much more conservative figure of 150 turns per chamber, and still skip that last chamber. 150 turns per chamber would represent roughly 1200V per chamber if it was a simple step up transformer. (I used 2.5 turns for the calculation since technically, only "half the primary" is driving the secondary at a time. (my primary has a total of 5 turns exactly between the 2 ends). I hooked the counter this time and I wanted to do 100 turns (or 800V per chamber) but when I realized that I was probably running 300 turns per chamber before, I thought 150 turns should still be reasonable. Guess not.
So I reinstalled the freshly wound secondary, filled up with oil and went on with tuning. With roughly half the turns compared to before (DC resistance confirmed that at about 175 ohm) I found that 20 uF was way too much and I couldn't get any output. I tried 4.7 uF and it seemed pretty good so I stuck with cap for the rest of the tests. By the way this was with a sharp copper wire wrapped around the threaded rods, right before the brass ball. At this point I took the wire out, and filled up with argon. Everything looked pretty good. Voltage could be raised all the way to 30V this time before hitting the PSU current limit. Before I would be in CC at roughly 17-18V. I was expecting as much with the lower inductance secondary.
I then proceeded with moving the setup to the other side of the lab so I could use the 4 channel oscilloscope, and hooked up everything. I was able to capture a few screenshots and a video before severed arc'ing occured in the secondary. It was too bright to really tell, it could have been between primary and secondary (which I doubt considering the OD of the windings on the secondary is at least 6mm less, hence farther away from the primary, but I'll only know for sure when it's drained in the AM.
Screenshots: they're basically all the same, with a different time scale.
Channel 1 and 2 are gates: I don't understand why there is basically no signal on the channel 1 gate since the system was apparently working.
Channel 3 is a current probe on one of the outer wire coming back from the primary (actually on the Mosfet driven by the gate on channel 2)
Channel 4 is a current probe on the middle wire going to the center tap of the primary (they're not the same current probe, hence the different scales)
Also, don't pay attention to the "measurements" these were setup for the measurements I was taking on the SSTC.
We can clearly the strong DC (~10A) component on Ch4, and strong AC component Ch3.
Frequency's around 16.6 kHz.
Unless my understanding of how these simple ZVS drivers work was even more off, I don't think that gate on channel looks normal. Both channels 1 and 2 are differential probes and both are probing across the diodes in front of the gates.
And here's a video I took a minute or so before the failure (it failed on the next boot up) - sadly, the camera wasn't rolling when it failed.
/>
-
Are you sure the channel 1 probe wasn't accidentally switched to 500x mode? That would explain the low voltage display. It might be possible for the circuit to oscillate with only one FET, but that seems much less likely. Scoping drain voltage is generally more useful for these circuits.
That ~10A DC current is much too high for the +-8A current on one of the legs. Perhaps something isn't working correctly. Also, that 10A must be the sum of DC components of the two outer currents. The one probed shows very little DC current. If there isn't some measurement error, the other must have ~10A (ie. a shorted FET). They should roughly-equally share the DC current.
If working properly, at 30Vdc input, the drain voltages will peak around 94V, a bit less due to losses. That is 18V/turn peak, or 2700V peak per 150-turn segment. That would easily break down the enamel insulation if the top of a segment's winding touches it's cross-over to the bottom. (Actual secondary peak will be a bit lower due to coupling factor. And, this presumes running at the lower pole frequency. I haven't analyzed the case of leakage-inductance resonance yet. That would likely be well higher than 16kHz, so unlikely here.)
Good luck with repairs!
-
After taking those plots I verified correct physical connections, multipliers on probes and scope. So unless something was wrong with the probe or scope, that was it. It doesn't help that incidentally, I had the second current probe (Channel 3) on the mosfet that had correct gate plots, and of course I didn't have a third current probe to see them all at once. Tomorrow, I'll double check that once the secondary has been repaired, and reboot the oscilloscope just in case. And once that's resolved I'll switch to probing drains.
My calculations were wrong, which is great because it gives a good reason for that secondary to fail. There is no doubt that the new wire coming down to a new chamber will eventually make contact with one last 30 turns per chamber. The intent was to use Kapton tape to cover that one wire before layering a new chamber but I had a very difficult time getting the tape to stick... Not entirely surprising since I am using one of the best non-stick materials. I got a brand new roll of KT which hopefully will stick better. And since my numbers were off, I think this time I'll stick with 100 turns as I should have. Hopefully I'll find a capacitor that works well with the new secondary inductance.
-
If you can get some of the 3M Kapton tape with the high-temperature silicone adhesive, I find that better for adhesion. More expensive though. 3M 5413 is one example. I think there are others with silicone adhesive. The 3M 92 electrical tape with acrylic adhesive is better than knock-off brands, but silicone is better.
If you have anything else besides what you are building, plasma treatment will increase surface energy, making tape stick better.
-
I ordered 7648A42 from mcmaster, which I think is the one with the silicone adhesive you are talking about. It doesn't say 3M but that should be it.
I finally found a LCD resin with good dielectric strength (14kV/mm) so I'll give my original design a shot this week-end. I am pretty hopeful as it really addresses the cross-over wire issue by preventing any contact with any of the upper layers on the new chamber. Oil should have no problem filling up the space around the cross over.
Edit: I've attached an attempt at a (not finished/work in progress) schematic that would:
* add separate supply for the 2 mosfet gates (I need a 12-15V regulator the mosfet driver)
* add a bulk capacitor on the power supply
* add an interrupter circuit (not shown yet I wanted to see if I could get that circuit to simulate first) with a N channel mosfet + driver (I have some UCC37322P) driven by a signal generator on the input (the enable pin would be permanently pulled up)
I am not quite sure if that's a good idea for a ZVS driver generally speaking but I have feeling that it would 1. reduce the stress on the transformer as I'm figuring it out how to no arc, 2. potentially add some cool feature on plasma modulation
-
Hi,
I rewound the secondary with only 100 turns per chamber for all 16 chambers this time. (down from 150 turns and 15 chambers)
I then proceed with "tuning" the tank capacitor, and found that out of all the caps I have 2 uF gave the best results (2x 1uF in parallel).
I reinstalled the globe and added ~1atm of argon. Sadly, even at 30V I am not longer generating any discharges in the globe. With the voltage being lower by a whooping 33% I was afraid of that.
Technically, my mosfet being rated for 650V I could leave gate resistors on 12V (change the R to 100 ohm) and use a 60V PSU.
I checked everything again this morning regarding the gates and the discrepancy was still present measuring 30V off the PSU. Rebooting the scope fixed the issue.
As for yesterday's current waveform, scaling was off on one of the probes (the outer one i.e. ch 3 cyan, so current was in fact 10 times higher).
I'm attaching 2 screenshots:
- Gates + current (both outer lines)
- Drains + current (both outer lines)
Note that on the second screenshot (Drains + current) there is no corona or discharge (this is with the globe and 1atm of argon). On the first one (Gates + current), I had a sharp wire wrapped around the electrode with 10mm plasma shooting out of the tip.
-
Those new scope captures make things more clear (I think). It appears that the ferrite core is hard-saturated. That is likely why current pauses near zero (low enough to not saturate a non-gapped ferrite core), then peaks much higher. The saturated core means that much of the primary flux isn't flowing through the secondary.
The high-frequency bursts showing up in drain-voltage after each switching transition are likely caused by parasitic wiring inductance. I've seen that behavior frequently on my ZVS builds. It sometimes becomes the dominant oscillation. It is high frequency were the intended 2uF capacitance looks like a short circuit. Resonance is between the FET drain capacitance and inductance of wiring to the 2uF capacitor. (The other possibility is that it is oscillating with the transformer's leakage inductance. Wouldn't expect it to be that high of frequency though.)
-
I added a sharp wire, closed the globe and refilled with argon.
Here's a oscilloscope video showing a voltage ramp from 12 to 30V. Sorry the potentiometer on that PSU is a bit sticky, so that ramp wasn't super gradual. Hopefully it seems like it starts showing signs of saturation around 75V on the drains (~ 24V input). Perhaps it's time to start adding gaps.
/>
Does the zero crossing look normal to you though?
about the parasitic inductance, no doubt - right now, wiring' isn't cleaned up at all as I was figuring out capacitor size and such.
-
Zero-crossing of what? This is a ZVS. Switching occurs at zero voltage, which is close to current maximum.
Yes, that's when I see high-frequency oscillation bursts - when clamping on different caps to experiment. Is this what you mean by the zero-crossing looking questionable - the burst of high-frequency after each zero-voltage point?
-
Yes, those bursts make it difficult to see what's going on with the switching, but they're likely a consequence of the switching, not the other way around, so I wont' worry too much about them especially as I intend to clean this up a bit.
Too bad I couldn't get those shots yesterday before the failure. The output is significantly lower than yesterday, I really need to find a way to add turns reliably. And of course I need a suitable enclosure that I can pull vacuum from. I wonder if the later alone would be enough to make those discharges reach the inner surface of the globe.
With the reduced output, adding an interrupter doesn't feel as urgent.
edit: I'm attaching a plot of the center tap line in channel 4. the magnitude of the current in that wire is just plain small compared to what's going on in the other legs. That would seem to be the logical explanation for the thermal image recorded a couple days ago, wouldn't it? The average current on channel 4 roughly equates the numerical value displayed by the PSU (3A), and the currents in the 2 legs although with very large peak to peak value mostly cancel out and equate what's measured in that center tap line. The fet's are switching quite the current on that small circuit, I'm glad I went a bit overkill with those.
-
Yes, outer wire current is much higher, as I'd said in reply #7.
The FETs are not conducting that high current, only the center wire current. The high outer lead currents are conducted from the transformer to the capacitor, not through the FETs.
Definitely simulate any interrupter design before constructing. You will need some way to handle energy stored in the supply inductor at turn-off.
At higher frequency (ie. gapped core) you can get more volts/turn. However, secondary intra-winding capacitance becomes more problematic as frequency increases. Your UY30 ferrite will allow higher volts/turn at the same frequency.
-
Good detecting.
Meanwhile I couldn't stop doing the skin effect exercise.
Guessing a frequency of 100 kHz, found that an AC ampere in 14 AWG wire
produces almost as much heating power as a DC ampere in 18 AWG wire.
When your outside wires appeared bright in the thermal infrared image,
were they perceptibly warm to the touch?
-
Yes, outer wire current is much higher, as I'd said in reply #7.
The FETs are not conducting that high current, only the center wire current. The high outer lead currents are conducted from the transformer to the capacitor, not through the FETs.
Definitely simulate any interrupter design before constructing. You will need some way to handle energy stored in the supply inductor at turn-off.
At higher frequency (ie. gapped core) you can get more volts/turn. However, secondary intra-winding capacitance becomes more problematic as frequency increases. Your UY30 ferrite will allow higher volts/turn at the same frequency.
Yes you called it :) it really took the waveforms for me to understand it.
I'd like to keep running with this type of frequencies, so the UY30 ferrite will be welcome to continue pushing on this project.
The secondary wound with 100 turns per chamber has not failed, and none of the early signs of failure have appeared yet (bubbles forming in the secondary).
I installed LTSpice last week and I have to spend more time on it. I was trying to simulate the ZVS on circuit lab last night and couldn't; I tried a bunch of things to create a slight unbalance initially to allow the system to start oscillating but none worked, and it seems like I have the same issue with LTSpice; however with LTSpice if I try for example to change the value of the 470 ohm to try to generate an imbalance, the simulation converges immediately (as opposed to crunching for a long time with tiny time steps). edit: the option "skip initial operating point solution" fixed the problem :) - there must be some other problems still as it doesn't seem to oscillate correctly
Good detecting.
Meanwhile I couldn't stop doing the skin effect exercise.
Guessing a frequency of 100 kHz, found that an AC ampere in 14 AWG wire
produces almost as much heating power as a DC ampere in 18 AWG wire.
When your outside wires appeared bright in the thermal infrared image,
were they perceptibly warm to the touch?
Yes, the image was taken after one or two minute only which was more than enough to easily see on the thermal imager the outer wires were hotter. To the touch, you could tell they were warmer but after 10 minutes or so they were hot, while the center tap wire was not even warm.
-
That LTSpice result surprised me enough that I tried it myself. With ideal inductors, that does look like a valid result - an oscillation based on the supply inductor L4. Adding 0.2 ohms of series resistance to L4 damps that mode, revealing the intended oscillation.
Inductors are typically farther from ideal than are resistors and capacitors. Accurate simulation often requires adding appropriate series and parallel resistances. Those can be added as explicit resistors on the schematic, or as additional parameters of the inductor. Right-clicking on an inductor brings up a parameter table.
Leakage inductance in your transformer will eventually be important to model too. You will need three separate coupling factors to be precise. The two primary halves are likely a bit more coupled than is primary-to-secondary. For simplicity, you could start by making the one common coupling factor a bit lower, perhaps 0.95. (Coupling and inductance both drop as the core saturates, but that is not easily modeled.)
-
Awesome thank you, I'm glad I wasn't that far off and I actually thought about ideal values but since I noticed the default series resistance for L's were 1 milliohm I assumed it was taken into account and that it was "good enough". Lesson learned, next time I'll play with those values regardless.
I'll use this simulation to do a parametric study on the C1 capacitance. I already noticed that something is happening somewhere between 4.7u and 2.2u. Some other unintended oscillation perhaps.
On a side note, I put together a very crude test with a glass bottle and a couple of fittings. I didn't use argon because I don't have a great seal, but it's good enough to bring the pressure down enough (no gauge right now) to see the real time effect of pressure on the discharge. When I have a gauge hooked up I'll make a quick video as it's pretty interesting.
-
Yes, it is fun to play with pressure, especially with argon. My little one (8kV peak) when pumped down causes glow to extend through the evacuation tube.
-
Oh and I understand what you meant by dealing with the energy store in L4.
I added the most basic interrupter I could think of and here's a with / without side by side comparison.
Edit: Couldn't find the right fitting for the vacuum gauge, but I did capture a video showing power draw as vacuum is pulled.
As opposed to the previous transformers (2250 and 3000+ turns secondary) this one (1600 turns) won't allow discharge to take place at 1atm unless there is a sharp breakout point on the electrode. This video, I only have the smooth brass ball and it will only start discharging once the pressure is low enough.
-
I wouldn't call that progress, but I added a P-channel mosfet to turn off the inductor and the primary through a interrupter signal. I used a N-channel mosfet off the signal generator but a NPN transistor should do as long as it was withstand 30V.
That didn't fix the oscillation still taking place in the transformer though. What's surprising to me is that outside of the first 5ms after the signal generator goes low, there is no damping at all and for the following ~45ms it's still oscillating with no damping.
I checked that the capacitor and L1/L2 all had series resistors to make sure energy had a way out (was mainly looking for an explanation for the lack of damping.
Edit: I'll plot the current coming out of the power source. I think this is likely where the energy is coming from, although the switch open, current can still flow through the 470 ohm resistor looping through the ZVS mosfets which will continue opening/closing regardless of the signal generator. If this is what's going on, perhaps it's time to power the gates with 12V and a lower R, that won't prevent current from still flowing through the primary but at least it should be significantly less. And even then, that oscillation is not really a problem.
edit: Yes, that was it. Feeding the 2x 470 ohm resistors from the P-Channel drain fixed the issue.
-
Move the bottom (anode) of D5 to ground and you will be getting close. Switching the main supply (current through L4) is the best option I know of, even though it requires an additional high-current switching device. (Look at the voltage across L4 in your previous un-switched circuits. Ignoring series resistance, the voltage across any inductor must average 0. Placing a diode across an inductor shorts out one polarity, which will build up high current until the opposite polarity has a correspondingly low voltage-time integral.)
Yes, gate resistors provide enough power to keep oscillation running at low level. That has a key advantage: Makes startup of full-power oscillation faster, minimizing the startup current spike in L4 and following voltage spike on the FETs. So, I suggest not "fixing" that unless you really need it to turn completely off.
Once mostly working, go through the circuit looking at device voltages and currents, comparing them to allowed values in device specifications. LTSpice models usually don't show you that a 0.1A diode is really running at 100A automatically. You need to look. (There are tools that will examine Spice output to check for device parameter violations, but I've seen those only for specific semiconductor processes. I've used them in designing analog integrated circuits, but not for board design.)
-
Thanks Dave, I'll try simulate with the anode to ground.
I'm actually thinking I'll add a switch to use/bypass the interrupter. The interrupter on that circuit for a plasma ball is more of an experiment than anything. I have not seen anyone tried that and it's probably for the good reason that it won't create any nice effect; at best it will probably kill the tendrils prematurely. It won't hurt to try.
On the other end, the adjustable power supply combined with the gates being powered by a 12V Buck through 100 ohm's should give me the power adjustment I am looking for.
Thanks for the tip on checking all voltages and current; I had a hint that it probably wouldn't give give warnings on simulations with errors where the results happily reported thousands of amps in the primary with no warnings, as they say GIGO :)
-
Dave: Would you have any advice in sizing that diode (anode to GND)? Also I was wondering if with the anode to ground it would make sense to use a TVS instead.
I am separating the interrupter circuit with a rocker switch so I can run un-interrupted. The gates are powered by a 12V buck and the main power input is DC 12-60V or so; So technically, when running with the switch on the interrupter position, the voltage supplying the inductor can range from 24 to 60V, and this is giving me a hard time picking the right diode...
-
I happen to know someone in Germany designing a similar circuit at higher voltage (rectified 220V line) for an induction cooking appliance. He is using a FET in series with the supply inductor both for on/off and as a buck regulator to adjust average power. If you notice, that FET, diode to ground, and inductor is exactly buck-converter topology. Just in case you ever want to switch the FET faster (ie. >20kHz) as a buck converter, I suggest a fast diode. Schottky would be best, especially at your relatively-low voltage. A fast-recovery diode (ie. <=75ns reverse recovery time) would also work. If you are interrupting at lower frequency, so the inductor current ramps to zero before the next enable pulse, then diode speed isn't important. Even a TVS would work for such slow operation, but no reason to use a TVS diode.
If interrupting infrequently, then the diode's peak current is what matters. Even a 1A schottky such as DSS110 or NTE642 (2A) would be fine. If interrupting rapidly for buck-converter use, then the diode's average current needs to be over half our max DC supply current, higher to be safe. For 5A parts, perhaps SK520B or MBRD5100 or VSSAF510. There are literally hundreds of options. Obviously you need >60V, so at least 80V to have margin.
Edit: Rather than specific part numbers, here's what I did to find them: For part searching, I find digikey.com most useful. Often not lowest cost, but best parametric search capability I've found. (Would love other suggestions, as "best" is far from ideal.) Select "Discrete" under "Products", then select "Diodes-Rectifiers-Single". In the menus, pick "Schottky". Highlight all currents from 5A and up. Highlight all voltages above 60V. Select "Active" status unless you want obsolete parts included. After "applying", sort by price (or whatever parameter you want). Price sorting is more useful if you also enter quantity. The cheapest ones were surface-mount, so that is what I listed above. You can include "Through Hole" in the menu selections if you want only those package types.
To find the lowest price (and availability) in normal distribution, enter the specific part number (or at least initial part of it) in oemstrade.com.
-
Thanks Dave, 5A should be fine for what I'm trying to do; after all the goal would be to pull as little current as possible.
-
Quick update, I started playing with 3D printed resin (LCD printer) and it looks pretty promising. I found a resin with good dielectric properties and I'm currently fine tuning settings to get my design to print correctly. The cured resin doesn't seem to be less rigid than PTFE I used on the current working transformer. And as far as my design is concerned the angled cutoff joining two adjacent chambers is getting printed without trouble. I am pretty hopeful that this design will allow greater output voltage, based on design parameters and my experience with the PTFE housing, I should be able to reach 3000 turns which almost double the current working secondary.
I'll post an update soon, I have not yet finished making all the parts for the 370mm plasma globe, so I'll likely start testing the SSTC getting converted to DRSSTC before I can test the 370mm plasma globe.
-
Well bad news, I performed a quick test on one of those printed housing wound with about 100 turns per chamber (28 chambers) with 9 turns (2x 4.5) on the primary and I experienced corona through the the final shoulder which is over 4mm thick.
This is not submerged in oil, so the oil would likely help, but this results a bit inconsistent with the 15kV/mm dielectric strength I was given for this resin; that statement could be off and is purely based off experimenting with PTFE which I thought had similar dielectric breakdown.
It appeared at the very end so I will probably try to unwind the last chamber and test again. At that the point the weakest link will be the first to last chamber but this time on the ID of the winding. I'll continue unwinding chambers one at a time to get somewhat of an approximation of the actual dielectric strength. From there I'll be able to tell if that concept is worth pursuing. Perhaps, I'll try to find a better resin too.
Rough estimation from the LTSpice model, I should not be much higher than 15 kV (I was using 12VDC input). So I have a feeling that the resin has a significantly lower dielectric breakdown than my PTFE stock.
It's a bit disappointing because the mechanical design of that secondary housing worked as expected. Last improvements on the groove joining two adjacent chambers helped a lot, the wire sits fully recessed in the groove as I wanted and changing the angle of the groove to make it land tangent to the winding cylinder no longer makes the wire want to come out of the groove; it now stays in here perfectly. It tool me half a dozen tries.
I think Formlabs has higher end resin (which I am not sure are compatible with this LCD printer) but I doubt that they will cure at the wavelength of the LCD printer.
Here's an interesting doc: https://drive.google.com/file/d/1UdcHkNFkwhVlOhqBHQZhIcoYr1_m5VnL/view and I think I am really far from the numbers on slide 17
edit: removing the last chamber helped, but there is still somewhat of a corona forming in the same general area, hard to tell why precisely. The resin is semi-clear and it glows under corona making it difficult to pinpoint its origin(s). I am building some sort of spark gap with 2 large sphere to try to estimate the voltage roughly.
-
Dielectric breakdown through solids is a complex phenomenon. Dielectric strength is commonly stated as voltage/distance (ie. kV/mm). That is valid at only one thickness. Breakdown voltage is closer to a sqrt(distance) function than to a linear function. Thin films can withstand much higher field than can thicker films. That is why HV film capacitors are internally multiple series layers of dielectric and aluminum (rather than a single thicker dielectric layer). Check the thickness used to define 15kV/mm for that resin.
Is the resin specification for DC or AC? AC dielectric strength is generally lower than DC.
Just as with Tesla-coil breakout points, field is higher at edges and corners, such as at the top or bottom of a wound segment. The field at these corners is higher than just dividing total voltage by total insulator thickness.
Finally, does this resin specify strength with some ideal curing conditions? I have little experience with such, but could imagine that dielectric strength changes based on cure light intensity, build layer thickness, humidity, air entrapment, etc.
Yes, I agree that performance will be better in oil. Good luck! Thank you for keeping us informed.
-
Thank you Dave, it took some time before I realized the non linearity, that I would expect it is linear but the dielectric strength be having a Voltage/Distance unit initially led me to believe that it was (linear) when it's obviously not.
I used a x2 multiplier when I put this design together (so if I x2 voltage, I would x4 the thickness).
LTSpice is a good tool for early estimation, but I think it would be great if I could at least estimate voltage just to validate my model. My HV probe is limited to 10kV so that won't work. I did connect the secondary to a spark gap (made of two 1.5" brass balls), and the longest arc I could create off these 2 smooth balls was 13mm, at 20VDC. That was done with the amazon ZVS with an additional capacitor (4.7uF) across the primary. I am not sure this is a good test, I need to do more research.
The information on the resin is limited but it was measured under D149-97a which suggests 60Hz I think.
-
Quick updated, I submerged in oil the secondary as is (all 28 chambers but 1 filled with about ~ 100 turns each).
It's been really stable compared to tests in ambient air.
I also connected to my custom ZVS driver and was able to push the voltage up to 32V (Max of that PSU).
I also performed brass ball separation testing and the maximum arc length was obtain just a little over 22mm.
And then I had the probably stupid idea of using the brass balls as (short" jacob ladder). So I pulled in the two brass balls to about 10 mm. It works wonderfully for about 5seconds before the 1K 1A diodes diodes (1N4007) along with the 2x 15V gate zener's. I repeated this 3 times, by replacing the diodes and experienced the same result every time. There is nothing fundamentally different than the separation test I did earlier order than that was test was just a on/off test, rather than a continuous test (although the failure occurs rapidly around the 3-5s mark). No oscilloscope was hooked up but the current at failure read 8-10A on the PSU.
Before the separation test, I ran the system at full voltage (~32V) for several minutes with no failure whatsoever, but the secondary ground was hooked up to the DC ground as I normally do when I want to get the secondary electrode to shoot plasma (which is the goal of that setup). But I wanted to have a remote idea of voltage so I did that separation test, and I now I'd like to understand the failure. I wish I had a video recording. As opposed to a normal jacob's ladder, this experiment ran a very short period (1 to 3 Hz I would say), so perhaps this created unusual stress on the circuit.
The mosfets being largely oversized did not seem to get damaged on either of the 3 back to back failures.
Any idea what might be happening? Would it be possible that my 1N4007 diodes are too slow?
edit: As it turns out, it was a diode speed issue. I replaced the 1N4007 by BYV26E and that fixed the issue immediately.
That allowed me to redo the separation test and confirmed the 22mm value. The transformer ran for a good 30-45 mn (not continuously) for various tests/experiments and no failure so far. So definitely some progress with this new design despite the poor initial results; removing that last chamber and more importantly submerging it in oil made really shine when compared apple to apple with the previous machined unit. I know I can improve it further, but the highest voltage I'll reach will definitely be on the UY30; Speaking of which, I've received them and with these results I should be able to implement these lessons to the bigger ferrite. If the 22mm value is correct, the output could be in the range of 60 kV, which seems a little high. I don't have a precise figure for the number of turns but I think at least 100 turns per chamber, but it could be upward of 150 turns. On the next one I'll use the counter for sure. But there might be something off, because 60kV and 2700 turns would yield over 22V/turn which seems a bit high.
-
Yes, 1N4007 diodes are much too slow for ZVS oscillators. They are for line-frequency use. There is a fast-recovery variant UF4007, and many other fast diodes that will work fine.
-
I think I edited my post as you were responding; Sorry about that.
Yes, too slow of a diode. That was it!
I'm a bit surprised the 1N4007 worked for a while but it's only that ladder test that made them fail, and really fast too. I wonder what in that test stressed the diode so much more? I mean, the 1N4007 never failed until today when it was running the smaller secondary (1600 turns) at the same input voltage. The only things that changed are the secondary with more turns (~2700+) and the test (before, the secondary ground was tied to the DC ground and the electrode was either discharging through a sharp breakout or through a brass ball under 1atm of argon).
That ladder test was a bit different than traditional ladders I built in the past; this one was short (limited by the radius of each ball) and because of that it would pulse with a much faster period, kind of like a spark gap.
-
The slow diodes may have been operating on the edge of failing before. It doesn't take much change to move from almost to actually failing.
Concerning output voltage, I'd guess that high frequency (relative to line) might jump the sphere gap at slightly lower voltage, but wouldn't expect any large discrepancy. Sphere roughness will allow sparks at lower voltage. It may be that you are getting 50-60kV peak. Running the core into saturation makes non-sinusoidal waveforms with higher peak voltage for a given RMS voltage. You could add an additional winding of a turn or two around the core at the ground side of the HV bobbin. That will allow scoping something closer to accurate volts/turn for the secondary including at least some of the effects of leakage inductance between primary and secondary.
-
That's very interesting - Are you suggesting some kind of feedback winding (i.e. a second secondary of some sort)? I could use some better estimates of voltage on the secondary.
-
I'm suggesting a winding for the sole purpose of connecting to your scope. (The term "feedback" defines how a winding is used - to control oscillation. The geometry of this winding might be identical. If you want to name it, it could be called a "scope" winding or "monitor" winding or "measurement" winding.) The goal is to get this winding as close to the secondary winding as possible without dielectric breakdown issues. Another option would be to wind a turn around the first segment of the secondary using wire insulated for 2kV or whatever you are getting per segment.
BTW, even the labels "primary" and "secondary" refer only to the intended use. Transformers are occasionally used in reverse of their design intent.
-
I appreciate your patience; I didn't realized the physical proximity to the secondary winding mattered; I might be able to get around with a couple of turns on the first chamber. If the goal is to mimic the diameter/shape of the secondary winding this is the way to go. The ferrite itself is not even round on the 2 joining sections; and I have quite a bit of resin thickness around the round section (over 4mm). I am probably close to 2000V per chamber, but I can dial that down easily by reducing the input voltage and allow for measurements.
-
One turn would be enough if that is easier. Would allow wire with thicker insulation to fit in the first section.
Another option is to notch the inside diameter of the bobbin at the low-voltage end of the secondary. That would allow a turn or two around the ferrite with the bobbin fitting over this small monitor winding.
Yes, the closer this monitor winding is to the secondary, the more accurately matched is the field through the secondary and through the monitor winding. That way the voltage/turn of the monitor winding will more accurately match the voltage/turn of the secondary. Magnetic field lines are not all inside the ferrite core. Field lines extend outside the core, especially when the core is saturated or when leakage inductance is significant (coupling factor significantly below 1).
-
Thanks, perhaps on the next print I will use the first chamber specifically for monitoring. These "housings" (no idea what is the right technical term for that "blank" on which the wire is wound around) are symmetrical in my design and they shouldn't be. Well at least not for my plasma ball project since one end is grounded. That end doesn't need as much "flap thickness" as the voltage is not that high (as opposed to the other end/flap that sees the maximum voltage). So I could make that end close to ground much thinner for additional chambers or to host a monitor winding.
The same applies to the thickness of the material around the ferrite cylinder. That thickness could increase with chambers. Not sure if this could be leverage to improve the design much.
-
The terms I see most often are "bobbin" and "coil former".
I've seen asymmetric segmented bobbin designs on this forum. Symmetric designs seem more common. That could be for simplicity or to allow either end to be grounded.