NST driven CW gen keeps burning out capacitors / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

NST driven CW gen keeps burning out capacitors

1 2 next

Move Thread

LAN_403

Robert Francis

Thu Jul 05 2018, 09:08PM

Registered Member #2355 Joined: Thu Sept 10 2009, 09:12PM
Location:
Posts: 71

I have a NST, 15kV 30mA. It feeds a CW gen with 2 stages. According to the calculator that should be about 43kV output.

Problem is the capacitors in the CW gen, particularly the one nearest ground, keeps burning out. At first I thought it was a fluke and something happened to it, bad wiring or something. But when I replaced the capacitor the sparks were normal length again. I then attached a Marx gen, 2 stages, and was getting sparks between the spark gap in the Marx and as an output when I put the grounding wire close to it.

But as the device was spewing out sparks they started getting smaller and smaller until the grounding wire had to be right nearby the marx output to get any spark at all, just like before.

The capacitors are rated to 50kV 2200pF. I don't understand why they would be burning out like they are. Is there something I can do to correct this?

Any ideas?

Thank you.
Bob

Sulaiman

Fri Jul 06 2018, 05:43AM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

Have you got a p/n for, or better a link to the capacitor datasheet ?

Robert Francis

Fri Jul 06 2018, 02:57PM

Registered Member #2355 Joined: Thu Sept 10 2009, 09:12PM
Location:
Posts: 71

No, here is the link to them on ebay

Sulaiman

Fri Jul 06 2018, 06:40PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

The voltage rating is adequate (sqrt(2) x Vrms x 2) so I guess that it has to be current,
it could be rms current (especially in the lower capacitors) heating the esr,
but it may be the VAR heating up the dielectric - which must be a high loss type judging by volume.

try a 2x2 mmc to replace each capacitor ?

As a test, try just the NST with one capacitor across it ?
Caution; at mains frequencies 2.2 nF is getting close to being resonant with the NST output inductance.

radiotech

Sat Jul 07 2018, 03:02AM

Registered Member #2463 Joined: Wed Nov 11 2009, 03:49AM
Location:
Posts: 1546

NSTs can ring unless they terminate in neon tubes. Sign makers
take pains keeping cables away from sign metal.

Amps burn up capacitors. Your CW works on spark gaps ?

radiotech

Sat Jul 07 2018, 03:22AM

Registered Member #2463 Joined: Wed Nov 11 2009, 03:49AM
Location:
Posts: 1546

Missing figure

johnf

Sat Jul 07 2018, 07:42AM

Registered Member #230 Joined: Tue Feb 21 2006, 08:01PM
Location: Gracefield lower Hutt
Posts: 284

Dielectric loss is the problem
associated with current through the cap
the clue is the bottom cap in the multiplier is the first to go where the current is the highest
suggest quadruploing the caps on stage one ie 8.8nF then two caps on stage two 4.4nF

if this does not work source some Murata ones of the same value that have known losses and known voltages

Robert Francis

Mon Jul 09 2018, 12:08AM

Registered Member #2355 Joined: Thu Sept 10 2009, 09:12PM
Location:
Posts: 71

@ radiotech
The 2 stage CW gen feeds an 8 stage Marx gen

@ johnf
Thanks, I will try adding caps in parallel on the CW gen. I'll let you know how that works.

Patrick

Mon Jul 09 2018, 12:15AM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

here is my oscilloscope pic of an unloaded NST.

link to previous work.

.

Sulaiman

Mon Jul 09 2018, 10:48AM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

Just doubling the capacitance in the first stage may not fix the problem as the voltage and dV/dt are unchanged.
That's why I suggested try a 2x2 mmc first, and testing one capacitor directly across the NST.

The practice of using C for the top stage capacitors, 2.C for the next stage down, 3.C for the next .....
does significantly reduce output ripple under load, and use a lot of expensive capacitors.
(also, very messy if just using multiples of one capacitor value (to save cost by 'bulk' buying) or expensive if you want to buy pairs of capacitors in a 1:2:3:4:5...... range of capacitance)

I have used high-loss capcitors of excessive capacitance (e.g. electrolytics if the a.c. is mains, high-k ceramics or ANY film capacitors otherwise) for the string that goes from ground to output.
(As I have quite a few, I used ex-disposable camera flash capacitors, at up to 12s8p x 120uF mmc) (4-stage CW, existing mmc)
The charge-pumping-rate is then determined by the lower capacitance, higher quality capacitors of the other string (or the other two strings if using full-wave)
(voltage ripple ).(cost) is lower, but you need to be careful with the charged capacitors

1 2 next

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