Strange frequency behavior

[flyingperson23]

I'm trying to tune a small drsstc I built. For the secondary, Fres=235khz, 215khz with simulated streamer. I can't seem to get the primary to resonate at 215khz. As I raise the voltage, it switches resonant frequency. All scope shots are primary current, roughly 80A/div. It starts out like image 1 with very low voltage, then develops a prominent current peak in image 2. For both of these it's resonating at ~190khz.  I'm not sure why it does this, but I've seen it before. Then at a certain voltage in, in image 3, this starts from the end of the pulse and propagates towards the beginning where the current jumps up and the frequency jumps to 230khz. Image 4 shows it propagating further, and image 5 shows what I'd expect; the 290A OCD triggering throughout the pulse at near half input voltage. Changing the primary Fres doesn't seem to change the 230khz it always ends up resonating at. Is this upper pole operation? If not, what is it, why does it only start with high enough input voltage/big enough arc, and what should I do about it?

The coil is a BSM150 half bridge, 0-320V bus, old onetesla secondary and toroid so I don't actually know their dimensions. I built it with long pulse operation for good music in mind. Is there some kind of tuning that would make music sound the best?

[davekni]

Is this upper pole operation?

Yes.  You are starting at lower pole and changing to upper pole.  This is normal with most drivers.  Frequency switches when arc capacitance brings secondary resonant frequency slightly below primary.  If you want to avoid switching poles, reduce primary frequency and increase coupling.  (Presumes you have margin to increase coupling without getting racing sparks up lower part of secondary.)

[flyingperson23]

So in this case primary resonance frequency is close to secondary? Because it seems to be 190khz; I'll test it without the secondary later. I don't have much room to increase coupling. It should already be quite high.

[davekni]

So in this case primary resonance frequency is close to secondary?

Yes, primary slightly below secondary.

Because it seems to be 190khz;

190kHz is lower pole, not primary frequency.

I'll test it without the secondary later.

That will tell you the primary frequency.

I don't have much room to increase coupling. It should already be quite high.

May be plenty high already.  Looks like 35us period for energy to transfer from primary to secondary and back.  That along with 190kHz frequency should be enough to calculate coupling.  I don't recall the formula off-hand and don't have time to derive it at the moment.  Simulation would be one way to see what parameters produce your measured results.

[flyingperson23]

What is the ultimate goal in tuning? Am I trying to get the primary resonance frequency to be the same as the secondary with streamer loading, or am I trying to get the lower pole to be equal to this? In lower pole operation, does it run at the lower pole frequency or the resonance frequency of the primary?

[davekni]

Am I trying to get the primary resonance frequency to be the same as the secondary with streamer loading

Yes, that is typically the goal of tuning.

[flyingperson23]

Ok, the primary by itself resonates at 215khz, same as secondary with 16" wire streamer. The coil together resonates at 200khz, then jumps to 230khz with enough voltage. Am I not trying to get the coil to resonate at the secondary resonance frequency with wire streamer?

Edit: I do have pretty long primary leads and secondary ground wire. Everything was tested with these in place. Should I be measuring the secondary's resonance frequency without the primary there?

[davekni]

Am I not trying to get the coil to resonate at the secondary resonance frequency with wire streamer?

No.  Coupled resonators have an upper pole and a lower pole frequency.  Once coupled, no primary or secondary resonant frequency exists independently.

Edit: I do have pretty long primary leads and secondary ground wire. Everything was tested with these in place. Should I be measuring the secondary's resonance frequency without the primary there?

To measure individual primary and secondary frequencies, they need to be isolated, with the other coil not present.  (For measuring secondary frequency, it is sufficient to disconnect one primary lead rather than removing primary.)  To measure primary frequency, lead wires should be in place.

[flyingperson23]

Ah thanks, I'll try measuring secondary frequency with the primary disconnected.

[flyingperson23]

Disconnecting the primary had no effect. Now it still resonates at ~195 khz before switching to the upper pole which is now at 215 khz?? I didn't change anything really. It takes a lot of input voltage and pulse length before it switches, sounds kinda unstable, and is also really prone to flashing over and doesn't perform super well. I just want it to be a normal drsstc, not switching poles and performing optimally (having the lower pole at 215khz? or primary fres? idk)

[davekni]

My DRSSTC switches poles when arcs get long.  It is a relatively low impedance design (~35k secondary impedance), opposite yours.  Relatively short on times and impulsive arcs.  I play music, but it isn't optimum for that.

[flyingperson23]

Solution(?): If I just run realllly long on times and high voltage all the time it'll be at upper pole pretty much always and sound good enough. Arc length could probably be better though.

https://youtube.com/shorts/gyPdY8dJFyg

[Mads Barnkob]

The long flat lining waveform is a sign up your coil reaching steady state operation. All power fed into the inverter is transferred to the secondary coil and ends up as sparks.

A continuously rising "normal" DRSSTC envelope would be the forever power hungry system, that is only stopped by the interrupter or OCD.

In my experience, I can grow just as long sparks with low voltage steady state long on-time operation, than with higher power, fast ringup, short on-time. It all comes down to the impedance matching through the entire system Which is very different from coil to coil.

[Uspring]

What is the ultimate goal in tuning? Am I trying to get the primary resonance frequency to be the same as the secondary with streamer loading

That is usually the best you can do. Better would it be to have the operating frequency, which is either the upper or the lower pole in order to achieve zero current switching, to be the secondary res with streamer loading. That is impossible to achieve exactly, since the poles never are equal to the primary or secondary res frequency.

or am I trying to get the lower pole to be equal to this?

As stated above, the res frequencies of either the primary or the secondary are never pole frequencies. This is due to the coupling. The larger the coupling, the more the poles move away from the res frequencies.

In lower pole operation, does it run at the lower pole frequency or the resonance frequency of the primary?

It runs at the lower pole.

The possible best tuning is when primary and secondary resonances match. When secondary res frequency drops below primary res frequency due to arc loading, then it is possible, that the coil switches to upper pole. The probably easiest way to avoid this is to tune the primary so low, that secondary res frequency never drops below primary res frequency even at the strongest arc loading.