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 5 6 7 8 next last

Move Thread

LAN_403

Physics Junkie

Sat Mar 30 2013, 08:22PM

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

Wait, I think I'm misunderstanding you. If I reduce primary capacitance then at 9 turns (full primary) inductance, the resonant frequency is 74 kHz, so to achieve that lower frequency I would have to add turns as well.

Uspring

Sat Mar 30 2013, 09:17PM

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

I meant reducing the pri caps by a factor of 1.5. You should then still be able to tune 15% low with 9 turns. Both the somewhat higher coupling and the lower transformation ratio Lsec/Lpri might help bringing your primary current down.

Physics Junkie

Sat Mar 30 2013, 09:27PM

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

Ahh gotcha. So my pri cap is .3uF/1.5=.2uF so If I reduce pri cap to .2uF by taking 2 strings off my MMC, then the new pri cap =.1999uF or approx. .2uF. At 9T this gives me roughly 64kHz which is detuned 19.75% (all from JavaTC)

Sound about right?

This will bring Ipri down to allow for me to apply more voltage to the bridge without blowing stuff up, I hope

So this is what I'm going to do now:
-Take 2 strings off the MMC
-Measure all the pri frequency's at every half turn just for reference.
-start by tuning a few % low and redo all these tests

So recapping what we're talking about. The hypothesis is that After reducing Cpri, I will be able to apply the same voltage to the bridge as before, with less current. Thus allowing me to apply more voltage to the bridge so that the increase in current is not so high as it was before.

Uspring

Sat Mar 30 2013, 09:33PM

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

Yes, sounds right, but check that you're not overstressing the caps, either in current or voltage.

Physics Junkie

Sat Mar 30 2013, 09:38PM

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

if overstress occurs, I can add 1 cap in series and 1 more string for a 4series x 5parallel MMC which = .1875uF, still close to a 1.5 reduction factor. but i wont do that unless completely necessary. I know those CDE's take abuse

EDIT: I'm glad to report much better results after raising the secondary a little more, it seems that the coupling really needs to be finely adjusted in my system, it still might even need to adjusted a little more. I'm greatful you suggested lowering pri tank C, I can get nice 2' streamers to air at 250VDC 110 uS PW with Ipk of about 620A, no racing sparks at all

I got the coupling in a good spot right now, but I wont be entirely sure until I get out and try full power again.

Since I lowered the MMC capacitance I've made a whole new table of data (since old data is useless now) so I will post that all when its complete, so far I have best performance at 5.5% detuned but i'm still not done mid power testing.

One thing I'm worried about is that I can see a pattern in my data where the Ipk in the primary circuit is increasing pretty rapidly (not constant, ever so slightly exponentially) every 25-50VDC at the bridge (PW is always 100-110uS for this data), and I can 'somewhat accurately' predict that the Ipk will be right on the edge of 1000A (or more) at +350VDC and likely blow the bridge again, especially at 150uS PW. the OCD is set at 740-750A. hopefully a little better tuning will allow for much larger sparks at these same power levels, thus I wont have to worry much, we'll see. The primary signal waveforms haven't changed much which is another indication that better tuning is in order (but no notching though) . Lsec/Lpri is still much over 5000, typically 7000-8000

Uspring

Mon Apr 01 2013, 06:17PM

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

Physics Junkie wrote:

One thing I'm worried about is that I can see a pattern in my data where the Ipk in the primary circuit is increasing pretty rapidly...

It looks like your input impedance is still too low. Why not look for the best "Arclength/Ipri" ratio when tuning instead of just arc length. Primary current seems to be the limit not bus voltage.

I'm surprised, that lowering coupling works for you. From theory it would decrease input impedance, which is not really want you want.

Physics Junkie

Mon Apr 01 2013, 06:37PM

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

Well I am taking the Ipk of pri into consideration when I tune, mostly now because I'm on my last set of IGBTs so I have to be careful not to blow them. What I want to find through my data is where my upper limit of bus voltage/Ipri compared to spark length (just to be more careful to not blow things up)

I dont understand why such low coupling works for me either. It just seems.. oddly too low for exact reason you pointed out. When I compare this build to other systems that are very similar in spec, like S. Ward's drsstc1 and D. McCauley's drsstc1&2, I just assume a K of about .18 would be fine.

I'm tempted to try going back to K=.18 to .17, remove the strike rail, and see If the racing sparks go away or develop from the pri to sec. I definitely need a strike rail though. Its been lowered even more and now sits half inch or less above the pri. where i still had racing sparks at K=.16

Impedance increases as capacitance decreases, right? Maybe slightly lower the MMC a bit more if all else does not work

teravolt

Mon Apr 01 2013, 07:48PM

Registered Member #195 Joined: Fri Feb 17 2006, 08:27PM
Location: Berkeley, ca.
Posts: 1111

if you tuned pri and sec to the same freq, then the low notch should be detuned to if you get a arc. The tuning should not be much below sec resonance if the primary and secondary have the same resonance.

matching your bridge to your tesla is a game because you reduced your tesla primary impedance by removing mmc caps so now you should be able to increase you bridge voltage while keeping your max current.

hope that sound right

zzz_julian_zzz

Tue Apr 02 2013, 02:17AM

Registered Member #3964 Joined: Thu Jun 23 2011, 03:23AM
Location: Valenzuela City
Posts: 332

I agree with teravolt, also i dont know if this can help you guys, but my pri coil is tuned much lower of about 20% compared to Sec coil., the drawback is that i had to lower the coupling to ~0.12, and also the Pri impedance of about ~9 ohms.. I can get decent spark with this set up, in fact, all of my 4 coils(different sizes) are tuned this way., i easily get 3x length of sec. power efficiency is at its best when i lowered my On time to only 60us and max my V input.

Pri current will depend on the ON TIME, and the impedance (ring up) & of course the V in.. in my case, i had to remove the variac from my system because it is too small(500 watts @ 260vac) for my medium sized drsstc, (s wards drsstc1) so given that the mains are to be 'always' the V in of my system, first i tune the pri to 20% lower, (coupling requirement will be low), then I set the impedance & consider On time for my 'target' bridge current. i can see this on a spice simulation.

i tried your approach before, on a high coupling of 0.19K, tuned of about 3% lower(if i go lower tuning, i get pri-sec arcing so i had to tune high), what i noticed is that if i go to higher impedance, the chance of pri-sec arcing is higher, if i go lower impedance (~5ohms) chance of arcing is also decreased(i dont know maybe because Vpri is lower??) but both impedance and high tuning gets me wimpy sparks output of about 2x the sec length.. again, i dont have any documentation of this experiments, and i did it for my own risk, i just want to share my experience when tuning my coils.. Thanks.

EDIT: I forgot to mention, that testing your system on low voltage in, would not exactly replicate the behavior of the system when in Max V in. Increasing inductance / lowering the tuning gets me to explosion of arcs in the order of 85-90% of V in but in 50% i got only small sparks to about 1.5- 2' arcs, but again when i increase the V in, the output DRASTICALLY increases.

Physics Junkie

Tue Apr 02 2013, 05:14AM

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

Julian, that's all great information. Thanks. The only problem that I dont understand is when I try detuning 20%, I get violent explosions of silicon and fuses at low voltage and low PW, and I really dont know why.
Recall:

at 390V max power, 50uS, and a half bridge section blows...

From what you say, that makes me want to try detuning low and raise the secondary a bit more. But this would imply that the coupling is still the issue and needs to be lowered to .12 or less. That just seems odd though.

then I set the impedance & consider On time for my 'target' bridge current

I like that idea, but what range of impedance do you try for?
Julian, what are your primary tank cap specs for your medium sized coil?
_________________________________________________

Above: My data table so far. I will add higher Vin data the next time I get outside for testing

Above photo: bit of improved current signal, best performance so far with 32" spark to air, 36" to grounded target

You can see how performance decreases as I detune. Spark length decreases and racing sparks become more common. So what do you guys think, tune much lower and raise the secondary even more? right now k=.14

first previous 5 6 7 8 next last

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