Strike rail arcs to secondary. -coupling & tuning- / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

Strike rail arcs to secondary. -coupling & tuning-

first previous 4 5 6 7 next last

Move Thread

LAN_403

Physics Junkie

Thu Mar 28 2013, 05:30PM

Registered Member #7267 Joined: Tue Oct 16 2012, 12:16AM
Location: Detroit, Michigan
Posts: 407

Uspring wrote ...

Physics Junkie wrote:

You say you start by tuning just slightly lower. I assume by just a few kHz or so. Are you just guessing really and/or do you ever do that tuning method where you hang a piece of wire from the toroid?

Yes, it is a guess. The amount of detuning depends on the ratio of arc capacitance to toroid capacitance, so it is different for every coil. You might start with 2-3%, which is surely at the border of the range.
Do you know, how big primary current gets? There also might be a problem with the electronics. And don't start using burst length beyond 200us.
I dunno about tuning with a wire hanging from the toroid. Can you explain?

150uS is max PW that I've been testing with. I dont know exactly how big primary current gets but I've wound a CT with some burden resistors so I can verify that. I sure hope there isnt a problem with the electronics, I've spent a great deal of time troubleshooting and finally got a very nice gate waveform etc... Possibly a lot of parasitic inductance on the full bridge due to poor placement of wiring and so forth. Might need some more decoupling caps across the positive and negative of the DC power supply.

Sure, you know how you typically find the natural frequency of the secondary + toroid using a signal generator and hanging an oscilloscope probe about 3 feet away... then you take a piece of wire and cut it to the length of an expected arc and hang it from the toroid. That drop in frequency from adding the wire is supposed to be like a "rough estimate" of where to start tuning the primary, of course there are a few factors that can bias tuning that way, like capacitance of the wire differs from sparks, and human error.

Uspring

Thu Mar 28 2013, 06:59PM

Registered Member #3988 Joined: Thu Jul 07 2011, 03:25PM
Location:
Posts: 711

Ok, wires are a poor approximation to arc load, though. For starting a tuning search, it is probably best to turn your variac low, so that the IGBTs are safe and then look for the biggest sparks. Then increase the variacs voltage and find the the best performance and so on. If you don't have an OCD, you should definitely have a look at the primary current.

Physics Junkie

Thu Mar 28 2013, 07:14PM

Registered Member #7267 Joined: Tue Oct 16 2012, 12:16AM
Location: Detroit, Michigan
Posts: 407

Uspring wrote ...

Ok, wires are a poor approximation to arc load, though. For starting a tuning search, it is probably best to turn your variac low, so that the IGBTs are safe and then look for the biggest sparks. Then increase the variacs voltage and find the the best performance and so on. If you don't have an OCD, you should definitely have a look at the primary current.

Beginning with low voltage from variac is a good idea and thats what I had begun doing. For some reason It just seemed like at low power, natural resonance worked best, which makes a bit of sense because sparks are not considerable in size. I do have OCD but Ill take a look at primary current anyways. Also, would you recommend using a strike target when searching for optimal spark length/tuning, or to just go by largest sparks in free air?
Thanks again

Edit: after carefully inspecting my IGBTs I realized that the mounting screws were acting as conductors between the heatsinks and the metal bottom of the IGBTs
Also here are some current shots, I started tuning at about 75 kHz which is about 3.8% detuned, but still clearly not in tune

1364526500 7267 FT1630 Primary Current 50us

Above Photo: roughly 126V supply measured at the bridge, 100Hz PRF 50uS PW

1364526501 7267 FT1630 Primary Current 100us

Above Photo: 126V in at bridge, 100Hz PRF 100uS PW

1364526501 7267 FT1630 Primary Current 50us 200v

Above Photo: roughly 200V in at bridge, 100Hz PRF 50uS PW. Notice how ch1 shows 73.9 kHz. Say for example that I tune the primary to that frequency now instead of 75, and retest at exactly 200V with same interrupter settings, will that be resonance? It seems like a silly question but im not sure what that lower frequency reading means.
Current transformer is 1:100 with two 10ohm burden for .05 V/A
(In process of winding. 1:1000 CT)
I have all day tomorrow to mess around with the coil so I will upload some better current shots as I tune.

Uspring

Fri Mar 29 2013, 05:29PM

Registered Member #3988 Joined: Thu Jul 07 2011, 03:25PM
Location:
Posts: 711

Physics Junkie wrote:

Also, would you recommend using a strike target when searching for optimal spark length/tuning, or to just go by largest sparks in free air?

The latter. Ground arcs detune the coil and therefore make the primary current ring up.

Above Photo: roughly 200V in at bridge, 100Hz PRF 50uS PW. Notice how ch1 shows 73.9 kHz.

A current with a beating envelope like that is a superposition of 2 frequencies. It is not clear, what the scope value of the frequency is about. If your scope does FFTs, you'll see 2 peaks.
These 2 frequencies are the upper and lower poles. They are not the resonant frequency of the primary and secondary.
"In tune" means that the prim and sec frequencies are the same. The upper and lower pole can never be the same.

Physics Junkie

Fri Mar 29 2013, 06:00PM

Registered Member #7267 Joined: Tue Oct 16 2012, 12:16AM
Location: Detroit, Michigan
Posts: 407

Uspring wrote ...

Physics Junkie wrote:

Also, would you recommend using a strike target when searching for optimal spark length/tuning, or to just go by largest sparks in free air?

The latter. Ground arcs detune the coil and therefore make the primary current ring up.

Above Photo: roughly 200V in at bridge, 100Hz PRF 50uS PW. Notice how ch1 shows 73.9 kHz.

Gotcha. I do have FFT on my scope, ill mess around with it and see if I can get the upper pole more flat. So if I understand this whole frequency splitting thing, that beating frequency is the difference between the upper and lower poles. And the more the primary is detuned to its optimal point, the more the upper pole will become flat (more like a small hump), in turn eliminating that notching which allows the primary energy to increase nicely for the pulse burst. Basically, when the notching is minimal and upper pole is a small hump, then that is where tuning of the primary is good? Or are we simply just looking at the spark and finding where its longest no matter what notching and beating occurs?

Uspring

Fri Mar 29 2013, 07:48PM

Registered Member #3988 Joined: Thu Jul 07 2011, 03:25PM
Location:
Posts: 711

Physics Junkie wrote:

Or are we simply just looking at the spark and finding where its longest no matter what notching and beating occurs?

In my coil the beating goes mostly away once I tuned it for max spark length. I'd be interested in seeing, whether that also happens with your coil if it is tuned to max arc length at max variac.

Physics Junkie

Fri Mar 29 2013, 08:50PM

Registered Member #7267 Joined: Tue Oct 16 2012, 12:16AM
Location: Detroit, Michigan
Posts: 407

I'm making a table of data with frequency of primary/% detuned, Voltage in, PRF & PW, Peak current, rough estimate of spark length. I will post pics and data when complete.

I will say, thus far, at medium power in (200-250V at bridge) spark length is not as big when I tune for less notches. Whereas tuning closer to natural resonance still seem to give me larger sparks but with more beating. But then again its still not max voltage in yet...

I have to re adjust the OCD, its set for 500A and kicks in precisely at 496A, limiting my testing.

Not sure if this is normal or not, but at 500A 250V 100uS PW primary detuned to 69kHz nice looking waveform no beating or notching, I get weak sparks less than a 12" or so.

EDIT: I readjusted the OCD to 1000A so I could really test the limits. Still tuned to 69kHz with a clean looking current waveform, at 390V max power, 50uS PW 100Hz PRF, the OCD trips at 1000A and a half bridge section blows... im so stumped here lol. sparks were barely 12". I mean, its clear that the primary was not 'perfectly' tuned and current was way too high with the OCD set high, but not even 12" sparks at 400V 1000A?

Went out on a limb and tried detuning 20% to 61kHz, All that did was smoke eight 10A fuses in row and several other 15A fuses.

EDIT AGAIN: another two IGBT's exploded violently. im down to my last four

Here are two shots of primary current with improved tuning. Note these are mid voltage.

Above photo: Detuned to 71kHz 200V measured at bridge 100Hz PRF 100uS PW

Above photo: detuned to 69kHz 250V measured at bridge 100Hz PRF 100uS PW the OCD is limiting the current here in this image.

So apparently, going by primary current signal, 69-71 looks healthy, but If I turn the voltage up anymore than 200-250 then the IGBT's blow or fuses blow or OCD kicks in permanently.

spark length never exceeded more than 14ish inches, if that.

In my coil the beating goes mostly away once I tuned it for max spark length. I'd be interested in seeing, whether that also happens with your coil if it is tuned to max arc length at max variac.

Overall, Im having a troubles tuning at max variac for max spark length because I cant even get much more than half variac without something going up in smoke. but you can see in these two new images that there is no notching. perhaps my threshold for tuning is no more than 10-11%. At 20% detuned the OCD kicks in at like 100V at the bridge not even 50uS PW and it get hung up there and I have to shut the whole system off to stop it.

Uspring

Sat Mar 30 2013, 01:27PM

Registered Member #3988 Joined: Thu Jul 07 2011, 03:25PM
Location:
Posts: 711

I'd initially ignore the notching issue.
You should find a noticeable minimum of peak primary current due to arc loading when you try out various primary freqs. The decrease is likely more noticeable for larger arcs, but you should be able to see it with safe bus voltages and currents.
This minimum is an indication, that power transfer to the sec takes place. If you don't see it, possibly the Lsec/Lpri ratio is too high. Typical would be a value of 5000. But I'm not sure which range of values works. What are the params of your coil?

Physics Junkie

Sat Mar 30 2013, 04:34PM

Registered Member #7267 Joined: Tue Oct 16 2012, 12:16AM
Location: Detroit, Michigan
Posts: 407

Uspring wrote ...

I'd initially ignore the notching issue.
You should find a noticeable minimum of peak primary current due to arc loading when you try out various primary freqs. The decrease is likely more noticeable for larger arcs, but you should be able to see it with safe bus voltages and currents.
This minimum is an indication, that power transfer to the sec takes place. If you don't see it, possibly the Lsec/Lpri ratio is too high. Typical would be a value of 5000. But I'm not sure which range of values works. What are the params of your coil?

Didnt know that, thanks for the tip.
Heres data from JavaTC

Above photo: not tuned, this is the full pri and sec parameters. Full Primary is 9turns. Lsec/Lprim=5018.19

Above photo: This would be tapped at 6.5 turns. closest I could get to when I tested at 69kHz. Lsec/Lpri = 7938.9
I used LDC for the secondary, or should I use LES when doing that calculation?
Note that I went by what the coupling is when detuned, not what it says for the full primary of 9 turns

When messing with inductances of the primary at every half a turn, the value of Lsec/Lpri is closest to 5000 at around 8.5-9 turns and that is whether I use Ldc or Les for Lsec. But like you said, knowing a range of values that works would also be beneficial.

Im not sure I saw any minimum peak current at all, it mostly ramped up and slowly down. Like Before, Largets sparks to air that I could pull was about 2 feet long but that was with the primary tuned to 77kHz, practically natural resonance of the secondary (78kHz) - primary tapped to about 5.475 Turns. Lsec/Lpri at this inductance is WAY higher than 5000, a little more than double that, and the current signal waveform would probably look something like the first images I posted with the noticeable notching but higher peak current of course.

Uspring

Sat Mar 30 2013, 07:44PM

Registered Member #3988 Joined: Thu Jul 07 2011, 03:25PM
Location:
Posts: 711

I think your coils params are in order. You still might try increasing primary inductance by 50%, so you'll have to reduce primary capacitance to get similar frequencies as before. Note that you'll get larger cap voltages then for a given current. I don't think you'll need to tune that much low, since your toroids capacitance is pretty big (compared to mine). Arc capacitance then shows less effect.

first previous 4 5 6 7 next last

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob