Feedback current transformer doesn't work

[nick]

Hi,

I've built a coil using Loneoceans' SSTC 2 https://www.loneoceans.com/labs/sstc2/sstc2schematicv10.jpg. My circuit is the same for all intents and purposes. The only difference - and source of my problem - is the feedback current transformer. In short, nothing works when using ~50 turns around a core through which the bottom of the secondary passes (as described in the linked schematic). What's interesting is that when I use an antenna instead (which works GREAT) and scope the output of the CT, I see exactly what I expect: the coil's resonant frequency at more than sufficient voltage. Any ideas what the problem could be?


To hijack my own post, I had another question - where can one find a nice toroidal top-load? I'm currently using 2 smooth salad bowls glued together (lol) - which works OK - but would love to have something nice like in Loneoceans' coil: https://www.loneoceans.com/labs/sstc2/toroid.jpg

Thanks!

[Weston]

Have you tried reversing the phasing of the current transformer? That would be my first guess.

The three places I know where you can get toploads from are ebay, information unlimited (amazing1), and eastern voltage research. They are somewhat pricey, after years of making tesla coils I recently bought my first real spun aluminium topload.

[nick]

Thanks for the reply! I'm sure I've tried reversing the polarity but will do so again tonight. Unfortunately I'm pretty sure I've already done this and it didn't work

I'm kind of getting tired/bored of my current build (it always feels like the journey is so much more fun than the destination...) so I think this weekend I will upgrade to a full H bridge. I now have 9 or 10 60n60 IGBTs - I might as well put them to work!

Shame about the (lack of) availability of nice toroids... My next step might be to try aluminium ducting that I see others use but I'm surprised that the relatively rough surface works well.

[davekni]

I was just looking at that SSTC schematic with current feedback to answer another post.  It appears to me to have an error.  The two 1N4148 clamp diodes on the current transformer secondary should be on the HC74 input (pin 1) rather than directly on the transformer secondary.  Even better would be to add a second 1k resistor in series with the HC74 input, then place the clamp diodes between the two resistors:
 [ Invalid Attachment ]

Does anyone else have experience with the "http://www.loneoceans.com/labs/sstc2/sstc2schematicv10.jpg" circuit with CT feedback?  Anyone have a reason that it makes sense to have a diode directly across the CT, making it carry a net DC current?

[Mads Barnkob]

It could just be a copy/paste error from the Steve Ward SSTC5 / mini SSTC that he used as inspiration. The antenna single-wire feedback was suddenly replaced by a two-wire CT and it just got referenced to ground.



The diodes are there for clamping the induced high voltages on the antenna, since corona on the antenna is pretty normal, maybe he is just missing a DC blocking capacitor like Steve Ward uses with the CT feedback here:

[nick]

Thanks both! Will try this tonight.

[nick]

Hello again, apologies for another reply. And also apologies for the noobish question but could you perhaps explain why it would make a difference whether you put the DC blocking cap before or after the diode clamp? Wouldn't it be OK that the cap removed the DC bias regardless, as long as it's in front of the 74HC? What am I missing?

Also @davekni - it would be much appreciated if I could get your feedback as to why adding another 1k resistor will help? I am even wondering why a resistor is needed at all give the CT feedback should produce a low current (albeit at high voltage, but that's what the clamp is for)? Further, why is it beneficial to have another 1k resistor in series - what should this accomplish?

Thank you!

[davekni]

That's a good question.  Transformer windings (and inductors in general) have zero DC voltage across them (ideally - ignoring wire resistance).  In other words, the average voltage is zero.  That's because the current through an inductor is the integral of the voltage across it (scaled by 1/inductance).  If the average voltage is positive, current will ramp up until it pulls the average back to zero.

With the diodes clamping the CT output directly, the negative voltage is clamped to ~-0.7V (one diode forward drop).  The positive voltage is clamped to +5.7V.  This ramps CT current up until the duty cycle is lopsided enough to make the negative clamp time ~8 times as long as the positive part to get zero average.

BTW, here's the attachment that failed previously, basically the same as what Mads posted, except the cap and resistor are reversed, which makes no difference.  Hopefully it shows up this time.


Note R2 in the second schematic Mads posted.  That helps kick-start oscillation.  Putting the resistor (R2) between pins 1 and 2 of the HC14 should work instead.  That has the advantage of repeated kick-starting if the first one fails.

I'd also suggest an additional resistor between HC14 pin 1 and the clamp diodes D1/D2, ~1k ohms.  That further reduces current to the HC14 input, forcing the clamp diodes to handle most of the current when voltage exceeds 0-5V.

[nick]

It works! Thanks guys! I ended up with one end of the CT in series with a 100n cap which is followed (in series) by a 1k resistor, before hitting the diode clamp. The other end of the CT is obviously grounded. I did all this before seeing your latest reply Dave, so will try the additional R tonight.

By the way, I still have another 100n DC blocking cap AFTER the clamp - any point keeping it there?

To be entirely honest, I'm struggling to understand all the theory in your post Dave from a first read, but that makes me happy since I can research myself / do simulations, and hopefully learn something new

[davekni]

I'd suggest removing the second blocking cap - it's the perfect place for a second 1k resistor.  BTW, the second resistor should not affect performance, just reduce HC14 input current in high-power situations, improving reliability.  (As it stands, the HC14 pin 1 is not at a defined DC level because of the second cap.  The HC14's internal input clamp diodes conduct once oscillation starts, making that second coupling cap voltage go to roughly zero.)

The startup issue is this:  When the enable pulse comes, the gate drive transformer is driven to one state or the other (one pair of H-Bridge FETs turned on).  Depending on what charge is left on the H-Bridge output DC blocking capacitor, that initial state may or may not generate much current (or voltage, so this applies to antenna feedback too).  Mads suggests adding a ~5-10k power resistor across the H-Bridge output, at least for DRSSTCs.  (For SSTCs, across the DC blocking capacitor works too.)  This makes the initial H-Bridge output state centered, so the initial enable edge generates a half-voltage edge.  That single half-voltage edge must generate enough current signal to change the HC14 state, which then makes another H-Bridge output transition, making another current half-cycle, etc.  (Oscillation starts.)

R2 in the second circuit Mads shared does two things.  It defines the DC level of the HC14 input between enable pulses.  It then changes the level on the HC14 input after the enable edge.  That makes a full-voltage H-Bridge output transition to initiate oscillation.  If R2 were placed across the HC14 (pins 1 to 2), then the HC14 will oscillate even without enable, so will create H-Bridge edges after enable goes true, until oscillation takes over.

[TiagoBS]

Hey guys, how are you doing?
I'm having the same problem on my RSSTC. Once again I'm following one of Zach's from Labcoatz tutorials.
Link: https://www.instructables.com/Building-the-Worlds-Easiest-QCW-Tesla-Coil-Staccat/

My circuit is working and I already got good results using the antenna, however I am having “bad” output with unstable BPS and pulse width.
-



Because of this I would like to try using a feedback transformer.
On the PCB designed by him there is already a connection for the antenna and for the CT. The scheme is as follows:
-


I tried using two different CTs (Also reversing the polarity). One of them I took from another SSTC (About 25 turns). The other one I wound around a ferrite core I took from an old computer power supply (about 55 turns).
-


My coil is currently in this setting
-

-

-
I'm using the CT at about that position.
-


When I'm using just the antenna I can hear the ucc clicking at the frequency of the 555 interrupter, when using the CT I can't hear anything.
I found this topic and decided to ask here because it is something related. I don't have much experience so I might be missing something very obvious.

[davekni]

Have you tried adding the 10k startup resistor from Steve's DRSSTC-.5 schematic Mads posted above?  Also helps to have bleed resistor(s) across primary coupling cap(s) to make sure half-bridge output is at center supply (both FETs have equal Vds before burst starts).

The other one I wound around a ferrite core I took from an old computer power supply (about 55 turns).

How do you know this core is ferrite?  Most supply toroid cores are powdered iron (or iron alloy).  The input line common-mode choke core will be ferrite, if that is what you pulled from supply.  Other cores are likely not ferrite and not suitable for CTs.

Any picture with CT actually connected?

[TiagoBS]

I made some tests and changes.

The other one I wound around a ferrite core I took from an old computer power supply (about 55 turns).

How do you know this core is ferrite?  Most supply toroid cores are powdered iron (or iron alloy).  The input line common-mode choke core will be ferrite, if that is what you pulled from supply.  Other cores are likely not ferrite and not suitable for CTs.
Any picture with CT actually connected?

Really I wasn't sure if this core was ferrite, so I used the core that came on a Chinese SSTC board. Originally this CT had 25 turns, but I redid it and now it has over 50.

With the 10k resistor between 74HC14 pin 13 and before the clamp diodes coil works as if it were in CW mode (using the new CT).

In other tests I kept the resistor but used the antenna instead of the CT. With this configuration the coil worked, but with a very weak output.

Without the resistor and with the new CT the coil didn't work, but I noticed that when turning on the main voltage sometimes I got a small discharge on the output, but the coil didn't keep the oscillations.



-

And here it works with just the antenna.

Any other ideas of what could be causing these issues?

[NyaaX_X]

I guess.. I don't know if this will work.

Make a Faraday cage to cover the driver. And connect the cage to Ground.

[davekni]

Make a Faraday cage to cover the driver. And connect the cage to Ground.

I'd suggest the same.  For testing, a cardboard box covered in aluminum foil works well.  I'm not thinking of any other possible causes for your symptoms than electrical interference from coil to driver board, so grounded shielding should help.  If problems persist, at least the shielding will also make future scope captures cleaner.

The suggestion Zak posts below is great for startup issues, especially for coils like this starting when line voltage is low.

coil works as if it were in CW mode

This is the issue that seems likely to be caused by interference from secondary voltage.  Almost any grounded shielding around driver board helps.  I use a 5-sided box (no top cover) with driver ~6cm below top.  Allows scope probing from top and provides sufficient shielding.
Other less-likely possibility that comes to mind is some bad connection within driver/interrupter circuitry, bad solder joint or poor socket connection if you use any sockets.

[ZakW]

Hello TiagoBS,

Sorry to hear you are experiencing this problem, you are not alone though! I had this exact same issue recently (several times throughout different version too).

For me the problem with the output and feedback happened after I killed the IGBTs after that I could not get the coil to output consistently and it would have the light snapping arc sound. Bringing my hand near the CT would cause the coil to run sometimes. Disconnecting it completely would result in some output too. I replaced EVERYTHING on my board to no avail. What fixed it for me was adding a few components to the 74HC14 feedback section. This allowed the 74HC to self oscillate and trigger the coil to turn on. Here is the example from my thread:


I ended up making C3 220pf and R13 11kohm to output around 400kHz. If you want to try this you can set C3 to a low value like 220pf (depending on target frequency) and then use a potentiometer for R13. Check what the oscillation frequency is and adjust the pot so it is slightly below what your coil runs at. Don't forget R10 as well, the additional burden resistor really helped to improve the signal from the CT.

Here is a link to the post showing some scope captures of the coil failing to oscillate https://highvoltageforum.net/index.php?topic=2338.msg17153#msg17153. If you go to the third page you can see after I added the self oscillation the coil runs great now.

Good luck, the issue is really frustrating. I hope this helps.

[TiagoBS]

Hey guys, thanks for the help!

I added the self oscillation circuit and was able to measure and vary this frequency (320khz). I added the other components as suggested too.
I also made a Faraday cage.

C3 = 200pf
R13 = 22K - (Potentiometer)


-



I measure the frequency via pin 4 of the 74HC14, but in my last tests I was unable to read the signal (Likewise, I can't measure the signal from the interrupter), in addition to being very noisy signal.
-


Measuring GS of the two IGBTs with just the logic part powered, these were the readings:
I don't understand why they are so different.
-

-



When connecting the coil in this configuration it works as in CW. With only the logic part turned on I can hear the UCC clicking on the interrupter frequency. By changing the value on the frequency potentiometer I can notice the increase or decrease in the clicking sound.

If I disconnect the self oscillation circuit the coil doesn't work and I can't hear the UCC clicking. However, when I put my finger on R1 it works (like in CW). With just the logic part connected, by touching R13 I can hear the UCC clicking very quickly, like at 60Hz.

[ZakW]

Sounds like you are making progress! I have complied A TON of notes on all the things I have learned and issues I have overcome with this type of coil. I plan on sharing all of it once I finish boxing my up my latest RSSTC coil. Mine is only a SSTC not dual resonant. Still though, the issues you described below are something I had as well. I will have to check my notes when I am home later.

I measure the frequency via pin 4 of the 74HC14, but in my last tests I was unable to read the signal (Likewise, I can't measure the signal from the interrupter), in addition to being very noisy signal.

Why were you not able to measure the interrupter output?

When connecting the coil in this configuration it works as in CW. With only the logic part turned on I can hear the UCC clicking on the interrupter frequency. By changing the value on the frequency potentiometer I can notice the increase or decrease in the clicking sound.

If I disconnect the self oscillation circuit the coil doesn't work and I can't hear the UCC clicking. However, when I put my finger on R1 it works (like in CW). With just the logic part connected, by touching R13 I can hear the UCC clicking very quickly, like at 60Hz.

Is the feedback signal overpowering the interrupter signal? I am not really sure how this could happen if the UCC is also connected to the interrupter. It would be helpful to see what the output of the interrupter looks like when it is exhibiting this behavior.

The only other thing I can think of at the moment is related to the ZCD. On several versions of my PCBs, my coils would start out working fine then I would have a failure and it would start exhibiting strange BPS/feedback issues similar to what is happening to you. In my builds I have used several different methods for detecting zero crossing. If the ZCD is not working or triggering properly the interrupter will not output and the coil shouldn't run. It may be that you have a second issue related to the ZCD/555 timers.

[davekni]

If I'm understanding your pictures correctly, one issue shows up there:  Shield for driver should be connected to ground including driver ECB ground.  Bottom of secondary should not connect to shield, not until after passing through CT.

I measure the frequency via pin 4 of the 74HC14, but in my last tests I was unable to read the signal (Likewise, I can't measure the signal from the interrupter), in addition to being very noisy signal.

Not sure what you mean by "unable to read the signal".
Noise is unexpected.  Is the H-bridge powered for this measurement of 74HC14-4?  Or is this with only logic powered?

Measuring GS of the two IGBTs with just the logic part powered, these were the readings:
I don't understand why they are so different.

That is more different than I'd expect, but perhaps reasonable.  74HC14 input threshold voltage is usually a bit below 2.5V (below half of supply voltage).  Your chips must be quite a bit below 2.5V.  Are you certain your chip isn't 74HCT14?
This duty cycle error (not 50%) can be fixed by adding another resistor, either across D2 or from 74HC14-1 to ground.  That lowers the input signal level to be centered around 74HC14 input logic threshold.  Resistor value needs to be adjusted to get 50% duty cycle.

[TiagoBS]

I did some more tests.

Why were you not able to measure the interrupter output?

Well, it looks like there's a problem with both 555s (It's not the first time this has happened). I swapped the two and was able to read these waves:
On one of the pins of the 100k potentiometer (interrupter frequency).
-


These here are from the interrupter output:
-

-

-


Low voltage transformer (Disconnected from PCB)
-


Connected to PCB:
-


I believe I need to resolve this noise issue before proceeding. It's disturbing all my readings.

The only other thing I can think of at the moment is related to the ZCD

Maybe I need to change some components on the board, maybe the 2N2222?

If I'm understanding your pictures correctly, one issue shows up there:  Shield for driver should be connected to ground including driver ECB ground.  Bottom of secondary should not connect to shield, not until after passing through CT.

I will make this change.

Noise is unexpected.  Is the H-bridge powered for this measurement of 74HC14-4?  Or is this with only logic powered?

Only logic powered.

That is more different than I'd expect, but perhaps reasonable.  74HC14 input threshold voltage is usually a bit below 2.5V (below half of supply voltage).  Your chips must be quite a bit below 2.5V.  Are you certain your chip isn't 74HCT14?

Wow, once again you got it right!
My chip is actually the HCT14. But I have some HC14 too. I'll make the change and test the result.

About the difference between the two, I found it briefly in a search that:
The basic difference is that HCT14 are designed to work at lower inputs (5v) and HC are rated for higher inputs (15v).

How does this affect the feedback operation?

This duty cycle error (not 50%) can be fixed by adding another resistor, either across D2 or from 74HC14-1 to ground.  That lowers the input signal level to be centered around 74HC14 input logic threshold.  Resistor value needs to be adjusted to get 50% duty cycle.

Did you identify this from the image with the blue wave?