Bert,
Thanks for the detailed reply! There is a lot of information to go over
here.
I have good and bad news and good news to report...
I removed the PFC's, re-implemented the safety gap, removed some C from the
tank cap (I probably shouldn't have). I ran the coil amp draw dropped to an
acceptable 20a range, spark output also noticeably dropped as well. Then
spark output dropped completely! Arg. I think I fried am NST or two. I
haven't had time to inspect yet. HOWEVER!
I found a killer deal ($200) on a 2KVA 13,200 potential transformer (with
fuses).
So now I'll need to build a ballast and still need to build a phase
controller for the SRSG.
Thanks,
~Dan
On Thu, Nov 22, 2018 at 4:03 PM Bert Hickman via Tesla <tesla@xxxxxxxxxx>
wrote:
> Hi Dan,
>
> If the safety gaps are set properly, they should seldom fire in a stable
> LTR system. If your safety gaps are firing frequently, partial
> cancellation of the NST current limiting function and main resonance are
> likely occurring.
>
> There are two types of LTR approaches: standard LTR and inductive kick
> LTR. A good discussion of standard versus inductive kick LTR systems
> (versus classical mains resonant systems) can be found in the archives:
> https://www.pupman.com/listarchives/2000/August/msg01085.html
>
> A regular LTR system just barely charges the tank cap to the maximum
> rated output voltage of the NST twice on each mains cycle. Either a
> static or rotary gap works well with this setup. The tank cap is ideally
> sized to 1.57 times the mains resonant value, so for your 15/120 system,
> this would be about 33.3 nF.
>
> Because of the heavier capacitive load on the NST, the size of the PFC
> capacitor bank can be significantly reduced or PFC can be eliminated
> entirely. A standard LTR setup with a much smaller PFC bank is the
> configuration I would recommend. The PFC bank may, in fact, be causing
> part of the problems you're seeing.
>
> If you want to use an inductive kick LTR setup, the tank cap size can be
> estimated using the following formula:
>
> C = 0.83*I/(BPS x V)
>
> Where:
> C = LTR Tank cap (Farads)
> I = NST bank short-circuit output current (Irms, in amperes)
> V = NST output voltage (Vrms)
> BPS = Break rate (2X mains Hz or 120 BPS for your system)
>
> For your 15/120 NST bank the ideal LTR size works out to be about 55 nF
> or about 2.6 times the resonant cap size.
>
> However, there are a couple of significant disadvantages using an
> inductive kick LTR system. It requires a properly phased SRSG where the
> gap fires 3-4 milliseconds after zero crossings. More importantly, the
> SRSG can be difficult to "start". The tank cap will only charge up to
> about 30% of the NST open-circuit peak voltage if the gap doesn't
> initially fire. However, if the gap does start firing, then the tank cap
> voltage will significantly increase to near the level of a standard LTR
> setup. This occurs as energy (stored in the magnetic field in the NST's
> secondary inductance) provides an inductive "kick" that boosts tank
> capacitor voltage. But this can only occur IF the gap starts firing at
> the right time.
>
> If your rotary gap is configured to have a total of four gaps in series
> (two sets of gaps located 180 degrees apart on the disk), the total SRSG
> breakdown voltage may be too great to "start" your LTR system. This
> sounds like the problem you encountered when trying to drive your
> Maxwell 60 nF cap. To alleviate this, you may want to use only ONE set
> of stationary electrodes (i.e., only 2 total gaps). This will allow you
> to maintain adequate mechanical clearance while also allowing the gaps
> to initially fire at the reduced starting voltage in your LTR system.
>
> I don't recommend an inductive kick LTR approach since its benefits may
> not justify the added setup complexity and gap firing difficulties.
>
> Hope this helps and good luck!
>
> Bert
>
>
> Daniel Kunkel wrote:
> > Thank you all for the input so far!
> >
> > 1) Bert:
> > In your opinion, with a SRSG/NST setup, how often should a safety gap
> pop?
> > 10% of the time, or less? Based on what you are saying, I'd say the
> > resonant rise symptoms match my situation. I can see the safety gap
> > preventing things from 'rise' escalating, but that doesn't seem like a
> > permanent fix? I like your previous suggestion of removing the PFC caps
> and
> > starting over...but I am guessing the BPS and capacitance are not
> > happy...so ultimately do I need to re-size the MMC? What do you suggest
> on
> > sizing?
> >
> > 2) Gary:
> > I am glad you brought this up. I had initially removed the MMC and
> swapped
> > in my Maxwell .06 uF cap, but no amount of fiddling could get the gap to
> > reliably fire, let alone produce streamers. I have seen situations like
> > this before, and it was always due to the tank cap being too large. I
> have
> > no idea how your 15/60 is charging and firing a .04 uF cap! For my NST
> > setup with 120 BPS, JavaTC suggests a .0553 uF cap. For you, it suggests
> a
> > mere .0277 uF.
> >
> > To answer your question, 'I suspect that
> > you were unable to stop the safety gap from firing so you opened it up to
> > where it stopped?', the answer is YES, you are correct.
> >
> > My cap options are MMC (100 caps of .068 @ 2,000 VDC) or Maxwell pulse
> cap,
> > .06 uF @ 50,000 volts)...what do you suggest I try?
> >
> > 3) Phil:
> > Thanks for digging that out of the archives! I always love reading
> material
> > from Richie. I can't wait get a real transformer one day (it WILL happen)
> > and then I can leave the NST's behind. I think this all confirms what
> > everyone is saying...my tank cap needs to be altered.
> >
> > 4) To anyone else still reading...
> >
> > Here is a pic of coil running...I did manage to take a quick video and
> here
> > is a single frame. By scaling on the photo (secondary is 26.5" inches of
> > winding), the lower streamer alone is 62+" from the breakout point to the
> > end of the frame. It is 72" from the break out straight down to the
> ground,
> > and this streamer ALMOST hit the ground at an angle, so I'd say it was
> > right around the the 6 foot mark. The amp meter was showing 40 amps...4.8
> > kVA YIKES!
> > https://imgur.com/a/xQKOpbx
> >
> > ~Dan
> > Kansas City area
> <snip>
> _______________________________________________
> Tesla mailing list
> Tesla@xxxxxxxxxxxxxxxxxx
> https://www.pupman.com/mailman/listinfo/tesla
>
_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxxxxxxxxxx
https://www.pupman.com/mailman/listinfo/tesla