CW multiplier

[Rocket Man]

i've got an old x8 cockroft walton voltage multiplier in a parallel config, i'm looking to get it running again after it fried my oscillator on a number of occasions.

what type of oscillator would i be able to run it off? considering it tends to drop it's guts through the oscillator lead as well as the output.

it's limited to 100v due to the caps i used

[lavenatti]

If it's frying your oscillator you most likely need to limit the current it can draw. Try an appropriate resistor, or a heavier duty driver circuit for the oscillator.

[YT2095]

Inductor couple it.

 

most CWs are inductor fed anyway, and at those voltages you`ll get away with a small audio transformer in reverse, the 8 Ohm out and 1k Ohm in type.

 

drive the 8 Ohm with your Osc, and pull the HT AC from the 1K side.

 

simple eh!

[Rocket Man]

i ended up scrapping the original oscillator, it was too big anyway.

it's a 100v output limit so i'm really only safe with a 12V input. could i use a 1k:1k ct isolation transformer and a blocking oscillator?

 

i've got parts for that at least. i'm just concerned that i'll be dropping the 100v through the coil somewhere on an output short and that could get nasty for the oscillator and the transformer (it's what kept cooking the original.. mmm, burnt silicon).

 

i built it with fast-ish discharge in mind so a current limiter isn't a big option.

to save the oscillator, could i put two opposing zener diodes in series across the transformer output? i have a few 12V ones lying around, i'm pretty sure that'd stop low voltage but shunt the output.

[YT2095]

inductor coupling, isolates the free running osc anyway.

 

and where did you get the idea that because the caps are rated at 100v the output will be also???

 

an AA 1.5v is a fixed value also, now put 100 of them in series and see what voltage you get!

[Rocket Man]

it's a parrallel config so the last cap is charged to the output voltage.

when i made the thing, i should have built it in series but i didn't know much about it when i built it.

i'll put a schematic up when i figure out how.

[YT2095]

if yours is an 8 stage then you have 16 caps (or should) 2 lots of 8 in series, one lot will be your charge caps the others will be the decoupling caps, linked with Diodes in a Zig-Zag pattern.

That`s a CW, I have No idea what you had though, but it doesn`t sound like a CW.

[Rocket Man]

right, i've been scouring google and i couldn't find the schematic i started with.

i am using 16 caps.

i'm not sure it's a true cw, but it's quite similar

the diodes are all in series and the caps sit one lead between each with two rails either side of the diode string. alternating caps link to these in parallel with each other.

(two sides, caps in parallel, outer edges are uninterrupted conductors, inner line is a string of diodes)

 

the two rails are across the inductor and charge one cap through another cascading down the chain. each sucessive cap charges to the input potential plus the cap behind it connected to the other rail.

[YT2095]

here, the last picture right at the bottom: http://mxp.physics.umn.edu/s03/Projects/S03Microwave/apparatu.htm

 

and a bit more detail showing the Inductor coupling here: http://home.earthlink.net/~jimlux/hv/cw1.htm

[Rocket Man]

right, i found one.

http://www.osaka-gu.ac.jp/php/nakagawa/TRIZ/eTRIZ/epapers/eMorihisa010217/Image139.gif

that's the basic arrangement.

edit: in fact, that's pretty much the exact arrangement, the voltage source is ac.

can you suggest a website to post my schematic on?

i just had another thought, all capacitors in the ladder are actively boosting the voltage, that "osaka" company is making digital watches so the number of parts is critical to them.

[YT2095]

I`ve no idea unless you have your own webspace, anyway, I believe your question is Now answered

 

use the cct in the last diag I provided and your OSC won`t blow-up

[Rocket Man]

i'll still have one rail charged to 100v across the inductor, could i use opposing zener diodes to shunt the discharge from one rail to the other?

[YT2095]

no, you should build it Exactly like the cct diag.

 

I`ve made Dozens of these in the past, I don`t see what the problem is?

[Rocket Man]

well, i don't want to buy new caps. i have the ladder already, i just needed a more surdy oscillator.

i don't doubt the series config is better, that's how i'd build the next one. i simply have one that i want to get working.

the parallel config can dish out higher continuous current so i tapped the string of diodes for variable output from input to 16x input.

i'm quite limited as far as power supplies go so i want to capitalise on cheaper transformers. (preferably without buying new equipment)

i really don't like the idea of re-building the oscillator after every misadventure so i need a fail safe.

i was unsure whether i could use an audio tansformer, thanks a bundle!

would i be able to shunt with zeners?

[YT2095]

small audio transformers like the one in pocket AM radios are good up to a couple 100 volts (low current), and of course have the advantage of having an 8 ohm load for your input.

you can use a 2 transistor bistable osc or even a 555 to drive a transistor with the 8ohm load emitter fed.

the only shunt you MAY need is a small reverse polarized silicon glass diode across the 8 ohm load to prevent any back emf, but if your transistors rated for the job, this is optional.

you don`t need Any zeners anywhere at all! and certainly Not after the output of the inductor, although it Is a good idea to make your 1`st 2 diodes in CW chain rated at a higher PIV than the others, I`ve found through experience that when a CW blows and it`s not the caps, it will always be one of the 1`st 2 diodes.

[Rocket Man]

i've settled on a blocking oscillator at 6V both sides of a 1k-1kct isolation transformer (if i find a suitable ac output plug pack it'd be ideal)

 

it is a different config, so the rules are slightly different.

i can see why the first diode would go in a cw, you're dropping an entire side through them. the collective tension must be huge!

 

this one however, has roughly even load through all the diodes. the worst case would be an over current in the last diode.

 

with this config, the input is the weakest link;

when you dump one side of the ladder, the cathodes of the entire other side need to discharge through the transformer coil to get to the output.

 

that's why i'm thinking about a breakdown shunt bypassing the transformer, i want this to last with minimal maintenence.