Optimum Secondary Winding

[alan sailer]

I am making progress with my mini DRSSTC. One step forward, one step back.

An early mistake I made winding my secondary has left me with a too high resonant frequency. I have had the thing running
but it needs too much top-load for my tastes.

So I have wound a new secondary with many more turns. Maybe too many. Since I plan to varnish the coil heavily it will be more
difficult to remove turns after this is done and preserve the smooth finish.

My question is the maximum number turns that is optimal for a DRSSTC secondary. Right now I have about 2200 turns on a
75mm lexan form, a total of 445mm length. Fres is 260kHz/147kHz without and with the topload. I'd like to keep the operating
frequency lower to make it easier on the IGBJTs but I am pretty sure that the devices can handle this frequency or higher.

Is there any rough information about the maximum turns on a coil?

Mads has a table on secondary construction and this secondary is slightly out of the bounds he presents.

Cheers.

[davekni]

The key is running JavaTC on your design to see how it is likely to perform.  My guess is that JavaTC will show secondary resonant impedance is on the high side (~50Kohms is considered about optimum) and coupling factor on the low side of optimum.  (JavaTC's recommended coupling factor is for spark-gap coils.  A bit higher is generally ideal for DRSSTC.)

[alan sailer]

Thanks for the information.

I did as you suggested and found that the fully wound secondary has an reactance of 60kohms.
About 10K higher than the magic 50K (which I suspect is more of a rule of thumb ie make the coil like
other coils that have worked well). The coupling is 0.136 with the amount of overlap I specified in
 the simulation. Removing 4cm of winding only lowers the reactance to 57K which doesn't seem like that much.

I'm still debating what to do. Winding this coil with 32AWG was by far the hardest I have done,
my usual method of tightening up loose winding no longer worked well. So I worry about removing
turns which might be useful in retrospect.

Cheers.

[davekni]

60k should be fine.  If it had been 100k, then I'd have suggested changing.  0.136 coupling should be workable too.  It will be hard to get much more from a tall thin secondary without creating other issues.  Just make sure your primary impedance is high enough to allow long-enough enable times to get energy transferred from primary to secondary.

[alan sailer]

Thanks. I'm moving ahead with the secondary as is. I can remove turns later.

As an educational thing is there any source for learning about the primary impedance and it's relationship to the secondary?

My understanding of this stuff is very rudimentary. I do know that there is primary to secondary coupling that is a function
of the secondary/primary geometry.

I also understand that there is a needed match between the primary resonant frequency and the secondary frequency. In
general, because the spark adds capacitance that lowers the secondary frequency, a well tuned coil generally has a primary
tuned to a lower frequency. As the primary transfers energy to the secondary, sparks grow, lowering the
secondary frequency so that at some point the two are much closer in frequency, allowing for best energy transfer and longer
sparks.

But that is about it and I know that there is much more to a coil and it's tuning than this.

Cheers.

[Mads Barnkob]

Is there any rough information about the maximum turns on a coil?

Mads has a table on secondary construction and this secondary is slightly out of the bounds he presents.

As Davekni points out, you have to be grossly out of bounds for performance to be really hindered, but even here you can still get a decent spark length. Perhaps with higher losses and such, but it would still produce an output.

I tried to make it clear in my guides, that those are proven design limits, that within those it will always work GOOD. Not perfect, not bad. But it will be a sound design. Squishing those last few tens of percent of performance out here and there takes a much larger effort, that is simply of out scope for most people to endure. We want to make sparks, not study for an engineering degree

The table on the website is just a selection of the most common suggested designs, for a bigger picture you should take a look at the linked spreadsheet on the secondary coil page, but here is a direct link as well: https://docs.google.com/spreadsheets/d/1TkpvH7e5geHCIhB_Gr3egwL_THJYqjZIU5INoo54nD4/edit?usp=sharing

I put in your 2200 windings for the given coil form and wire size and it results in 52K. We can properly argue all day that I should not have set limits on the background coloring so tight... its only +/-10K for the small coils and then leaning towards lower impedance for larger coils. But its a guideline to build something that works, I also do write that experimentation is encouraged!

[alan sailer]

Mads,

Thanks very much for the input. I have gone to your DRSSTC pages many times and they have been very helpful.

Your input was useful. I had suspected that much of the information about tesla coil optimal reactance is empirical, which I really like. I'm poor at theory so I tend to trust experiment.

I also will not argue with anyone about tesla coils. I don't have enough knowledge.

I am also pretty sure that my design is close enough because a) I copied it form OneKone's SKM100 coil and 2) I was getting two foot sparks off it before I decided to rewind a new secondary. The first was too high in frequency. I had ordered the proper 32AWG wire, wound the coil and then found out that the darn vendor sent me 30AWG. I decided to see what happened anyhow and came to the conclusion that winding a second coil would be better than trying to force the old secondary to work.

I am assuming the spreadsheet is the one on your website. I was planning to play with it but at work it is behind a firewall. I hadn't got around to trying again at home.

Cheers.

ADDENDUM: I just finished installing and starting up the new secondary. I may be deluding myself but it appears much more like a coil should be. The unit barely buzzed at low voltage/low pulse width (100V/10usec) and now it starts to get noisy quickly. I am also seeing much cleaner looking collector voltage wave forms. The coil is now operating about 20kHz lower and I think the SKM100 devices will be much happier.

Thanks for the help.

[alan sailer]

An update on the secondary. I have been doing some fiddling with the coil and it is behaving like a real DRSSTC.
Yesterday I was getting 23 inch strikes to a grounded rod at about 150 Vbuss. Raising the voltage to 200 volts was
a non-starter as the unit was steady OCD at 500 amps.

After some thought I decided to raise the coupling. I have a nice method of doing this, the whole secondary is
on a plastic screw thread made from a plumbing drain pipe. I just disconnect the secondary ground and spin
the secondary up or down to change the coupling. The range is about 5 cm which should be more than enough.

Lowering the secondary by about a centimeter turned the coil into a new creature. The OCD is staying off (until
there are ground strikes on my garage door). The strikes are at 30 inches (76 cm) and the whole thing is just
nasty (in a good way).

I'm getting some sort of low frequency ringing in the current waveform that I don't understand. I'll do some searching
but may have to ask for a more experienced eye to explain what is happening. The collector voltage wave forms are
wonderful, very little overshoot and I am miles away from the 1200 volt limit of the IGBJTs.

Cheers.