Quarter shrinker charging resistors / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Quarter shrinker charging resistors

1 2 next

Move Thread

LAN_403

MRMILSTAR

Sun Jun 17 2018, 03:08PM

Registered Member #62119 Joined: Sun Feb 04 2018, 04:59AM
Location: Cedar Rapids, Iowa
Posts: 136

I am in the process of building a quarter shrinker. Output could be as high as 24 KJ but will probably be used at something more like 5 KJ. My question concerns the use of charging resistors between the transformer and the energy discharge capacitor. It appears that the purpose of the charging resistors is to limit the current that can be supplied by the transformer to the capacitor. Since I am using a 15 KV NST with a maximum output of 60 ma as the charging transformer, do I even need the charging resistors?

Hazmatt_(The Underdog)

Sun Jun 17 2018, 04:54PM

Registered Member #135 Joined: Sat Feb 11 2006, 12:06AM
Location: Anywhere is fine
Posts: 1735

The NST's current limit into neon tubes at 60mA, but the tubes are not a dead short like a discharged capacitor, and your estimated 220uF @ 15KV is certainly going to 'look' like a dead short to that small transformer.

Your transformer's impedance is 250,000, do you need more external resistors to 'protect' the transformer from killing itself, is what you are asking here.

Theoretically you don't need more resistors because the transformer will current limit, but from experience, these transformers don't survive short circuit operation, the 36 - 32 AWG wire just doesn't stand up to the abuse.

Each coil is probably about 6000 ohms of wire if you measure it, and 7500V un-loaded terminal voltage, which will drop to ~5500V loaded into a tube in CC @ 60mA.

So if we guestimate that your capacitor is going to be a short circuit instantaneously and you are supplying 5500V per terminal with an internal resistance of 6000 ohms, your inrush current is approximately 900mA, so that is quite a bit of stress on the transformer.

Course, these are all approximations, neglecting a lot of impedances and actual AC calculation.

If you are dead set on using an NST as the source though, I would get a few in parallel though.

Also, you don't have to use resistors for charging ballast. You can use capacitors in AC systems for charging and as resistors because of their reactance. A 2nF cap 'looks' like a 1.3M ohm resistor at 60 Hz, so you could always use a couple of doorknob caps to current limit your charging system to the big capacitor.

If we guestimate this as 2.6M ohms and 220uF, it would take approximately 50 minutes to charge, but you would never have to worry about blowing up your transformer.

klugesmith

Sun Jun 17 2018, 07:37PM

Registered Member #2099 Joined: Wed Apr 29 2009, 12:22AM
Location: Los Altos, California
Posts: 1716

In general, a series resistor can protect the cap-charging supply from overcurrent when the voltage on big capacitor reverses.
It also protects the charging supply against being blown up in the event of an internal failure.
Protection must be traded off against the increase in charging time.

IMHO, it would also be good to disconnect the charger with a switch before firing the discharge event.
Switch needs to be able to hold off high voltage. Could be interlocked for safety.

As Hazmatt said, a NST by itself takes care of limiting the charging current. Not 'cause of the winding resistance but 'cause of the leakage inductance (an integrated inductive ballast).

Nameplate current value, 60 mA, is the AC short circuit current -- not much different from current in a lighted neon tube. I think the transformer can run for decades with output shorted, just as it can light a neon sign for decades. But must respect the experience which Hazmatt cites.

Short-circuit DC current from a full wave rectifier would be of similar magnitude, if the Secondary Ground Fault Protection circuit doesn't turn it off right away.
60 mA would get 220 uF up to 7 kV in less than 26 seconds, as you know. Actual current will dwindle as the capacitor voltage builds up, toward the steady state (NST's no-load peak voltage, on the order of 10 kV w/r/t center tap).

Klaatu barada nikto.

MRMILSTAR

Sun Jun 17 2018, 07:40PM

Registered Member #62119 Joined: Sun Feb 04 2018, 04:59AM
Location: Cedar Rapids, Iowa
Posts: 136

Here is some further information about my capacitor:

Manufacturer: Aerovox
Capacitance: 102 uF
Maximum Voltage: 22 KV

Hazmatt_(The Underdog)

Mon Jun 18 2018, 04:31AM

Registered Member #135 Joined: Sat Feb 11 2006, 12:06AM
Location: Anywhere is fine
Posts: 1735

Why not use a flyback?

I know I'm being silly, but why not charge the cap with a flyback. You could have it cranked up so it arcs across a spark gap to charge the cap, you can get a kw out of a flyback or more pushing it hard with everyone's favorite ZVS circuit, and if you kill a flyback or two so what.
I'd rather kill a flyback than a $200 NST.

In theory the NST's could run short-circuited, but when a capacitor bank shorts it usually kills the NST too. I've lost a couple, and I have a friend who lost a PT due to his mica caps shorting.

So a luminous tube is a short circuit, but it still has a much higher resistance than a solid copper wire, that and the peak currents are momentary (its flashing at 60 Hz).

klugesmith

Mon Jun 18 2018, 08:38PM

Registered Member #2099 Joined: Wed Apr 29 2009, 12:22AM
Location: Los Altos, California
Posts: 1716

We know better than to argue about little stuff. One of these years, one of us can measure and report NST secondary currents with various loads & rectifier topologies.

Mil, I think your responders agree that some sort of charging resistor is in order, if just to protect NST and rectifier diodes and kV-meter against a 5 kJ learning experience.
My Aerovox 22 kV capacitor is only 52 uF. (Looks like 22 kV is or was a popular design point for some physics contracts.) There are a couple of big tubular ceramic resistors between the capacitor and the charging/switching/metering box. Design as pictured is not intended for use at 22 kV.

(pic from 2006).

Another useful accessory is a resistive load for discharging the cap rapidly but nonviolently, should the need arise. (Your hard-wired bleeder resistors could take care of that, if you're prepared to wait a while before the apparatus is safe to touch.) Charging resistor could also do that job, if the charger end is switched to ground. I've used a "water resistor" -- a plastic bottle of copper sulfate solution with two copper plates, designed for about 100 ohms at room temperature. Connecting it to a 52 uF capacitor at 5 kV is undramatic. kV meter reading changes to zero, water temp goes up a degree or two, and one of the copper plates turns black IIRC. In many years on the shelf, it hasn't dried up or grown algae inside.

MRMILSTAR

Mon Jun 18 2018, 09:54PM

Registered Member #62119 Joined: Sun Feb 04 2018, 04:59AM
Location: Cedar Rapids, Iowa
Posts: 136

The preliminary design that I am thinking of after a lot of thought and investigation is this:

* Aerovox 106 uF, 22 KV capacitor (very convenient voltage rating for a 15 KV AC NST)
* 15 KV, 60 ma NST with full wave bridge rectifier as charger
* 10K ohm charging resistor with appropriate power rating (225 watt or as large as is practical)
* 50K ohm discharge resistor with appropriate power rating
* appropriate protection devices, circuits, lockouts, and blast containment

Thoughts?

klugesmith

Mon Jun 18 2018, 10:12PM

Registered Member #2099 Joined: Wed Apr 29 2009, 12:22AM
Location: Los Altos, California
Posts: 1716

Full wave bridge.
So you are going to float the capacitor case at some undefined high voltage level,
instead of keeping it grounded? Or does your cap have two HV isolated terminals?

One alternative, which I'm actively working on for a different project,
is to float the NST so a full wave bridge rectifier can have one DC output grounded.
Will put a separate HV isolating transformer between wall plug and NST primary.
Need some resistor arrangement to fix the NST case potential (midpoint of high voltage AC)
at the mid voltage of rectified DC output. Bridge rectifier doesn't do that automatically, during phases when no diodes are conducting (as in your cap-charging application, I think).

A much simpler & safer approach is to start with grounded capacitor case, grounded NST, and two-diode full wave rectifier. Could charge to almost 11 kV, maybe a bit more if NST primary voltage is boosted with a variac. Inherent safety against overvolting your capacitor.

Signification

Tue Jun 19 2018, 06:41AM

Registered Member #54278 Joined: Sat Jan 17 2015, 04:42AM
Location: Amite, La.
Posts: 367

Well, Here is how I did it, no charging resistors!...I used TWO 70uF @ 12kV pulse caps in parallel--I never charged then past 10kV. At this max setting (3.5kJ x 2 = 7kJ) the quarter was VERY "over-shrunken" It's diameter was almost equal to its thickness! It was "ugly" So I stepped the energy down until I got the quarter a bit smaller than a dime--This looked AWESOME! I used an NST 15kV @ 60mA controlled by a variac whose output was half-wave rectified by high voltage diodes. You can get small diodes (20kV @ 200mA) for $5.99 on ebay, one of which is, theoretically, sufficient; but I have a strange 'bad' gut feeling about using just one! Currently I have a dozen 6kV @ 200mA diodes in series (dirt cheap from all Electronics, Don't put then in a 'tight' zig-zag formation! To be safe maybe I should have used four of the 20kV diodes in full wave configuration, but everything was fine, so I kept the dozen. I also made certain that I had a 10-20kV DC voltmeter on the cap line. I think your 102uF / 22kV cap (24.5kJ) will totally mutilate the coin at full energy. You probably won't need half that voltage!

At first, I would disconnect the power before blasting the coin to protect the NST, which started humming after about ten shots if set off with everything connected. I put two high energy (standard would probably work) automotive ignition secondary coils in series with the two HV output lines to block damaging pulses with high inductance during firing, and everything worked perfect! You probably only need one, but I used two at each HV output line just for a look of symmetry. You could fire now without disconnecting the power--but turn it down or off first. Don't forget that most variacs step up 120VAC to about 140VAC--that's ~25kV peak DC on your caps!!!!! Watch this carefully! I never used the NST at full power--I just watched the voltmeter. The 60mA NST can charge the 140uF caps in less than 30 seconds!...so no hurry there--don't turn the variac up high--charge slow. Before putting power to ANYTHING, make sure you are bulletproof and have good eye and ear protection. I think the voltage rating printed on the NST is just safe for a short time, as to strike the neon light, which then pulls the voltage down significantly also, don't forget, the caps put a nice pull-down load on the NST also. However, I have never had any problems running them at full voltage. Also, you need earth ground connected as the insulation of an NST isn't rated for the labeled voltage---the case is the ground center-tap.

Since I am rushed right now, I don't have time to proof-read...probably left out some small details--will check.

Klaatu barada nikto.

MRMILSTAR

Tue Jun 19 2018, 02:12PM

Registered Member #62119 Joined: Sun Feb 04 2018, 04:59AM
Location: Cedar Rapids, Iowa
Posts: 136

This is to answer the questions posed. My capacitor has 2 HV terminals. I plan to float the NST. Nothing will be grounded to the mains ground. I also plan to disconnect the charging system from the capacitor before firing so that the only components connected to the firing circuit will be the capacitor and the work coil. I don't plan to ever fully charge this capacitor. One reason is to keep the capacitor within its 20% voltage reversal spec. I plan to charge it to no more than about 7 KJ. My initial idea was to have no current limiting charging resistors and to just rely on the inherent current limiting of the NST but I then decided to to use 10K ohms just for some current limiting.

1 2 next

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