Go Down

Topic: HV supply ideas (Read 2216 times) previous topic - next topic

jackrae

Maybe a flux-capacitor might be worth considering.

wanderson


Yikes that's expensive,
I was looking on ebay and 20$ isn't that bad for it, but I was gonna try and make the frequency as high as possible(probably limited by the mosfets or igbts I get) to allow me to use smaller size capacitors and less ripple
now for that inductor, I guess since its high frequency it wont have to be too large right? Perhaps in the range of something small air cored?


Many of the smaller capacity high voltage capacitors I purchased on ebay failed quickly.  The only ones I have been reasonably happy with are used maxwell caps.  Ran me near $100 on ebay.  They measure 6" x 4" x 2" and are rated to 35kV, but are designed to handle the pulse currents associated with a flash.  You might be okay with the one hung low HV caps, depending upon what you are using them for...  But high voltage parts, even used from the 'bay, are expensive.  I paid nearly $20 for 12' of 50kV rated wire...
New true random number library available at: http://code.google.com/p/avr-hardware-random-number-generation/

Current version 1.0.1

PeterH


I figure I could make a smaller hv transformer at higher frequencies


How easy is it to make a transformer that is insulated to 50KV? Is yer average bit of varnished copper wire up to the job?
I only provide help via the forum - please do not contact me for private consultancy.

winner10920

Just an idea, been wanting to attempt this for a while now so just throwing thoughts around

winner10920

Ok so I spent 28$ on some stuff for my project from digikey , a hip4081a hbridge mosfet driver capable of up to 1Mhz, im gonna aim for 500-800khz area although the arduino max pwm frequency is like 32khz so maybe ill try that first or have the arduino control something faster
4 http://www.rohm.com/products/databook/tr/pdf/r6010anx.pdf for the hv hbridge(I think these will work, maybe without even a heatsink)
3 22uf 450v caps and 1 2a 1000v bridge rectifier
im gonna try an aircoiled inductor to allow me to change the rms into the multipler
from ebay I got some 1n4007 1000v diodes, and some 1000v .01uf caps so ill start with a 5 stage multiplier
hopefully ill be able to get this thing adjustable from 240v to 3.2kv which I guess is a good start,
how many multiplying stages do you think is practical? I think with my high frequency I should be good if I did 15 total to get me to 10kv
all ill need now is some current limiting resistor in the 10-20k range, disconnect switch, rubber tape....
think id need a bleeder resistor from the hv output? Or since the caps are .01uf it'll dissipate quickly?
hopefully ill be able to measure it too, I remember a friend has one of those hv adapters for a multimeter so I could actually get a read on it

wanderson

These folks, http://www.amazing1.com/, sell many useful components and kits for this type of activity.  I would suggest reviewing their site, some of their designs have schematics, etc. online.
New true random number library available at: http://code.google.com/p/avr-hardware-random-number-generation/

Current version 1.0.1

winner10920

That's a cool site indeed, although im gonna try and make it myself for the sake of learning and experience which for me is half the fun  :)

winner10920

Do you think monitoring the current draw before the voltage multiplier would give me a good resolution on the output current? I borrowed my friends HV multimeter probe so ill be able to test the voltage on the output, although when it comes time to make a permananent measurement practice I still gotta figure that out, it'll have to be really high impedance I was thinking a 220M and 22k resistor divider would work? Im also probably gonna buffer that with something before it hits my analog input in case something goes wrong
now in order to change the rms at the input of the voltage divider id have to pwm even faster than the waveform I want to make right? So with the arduinos max pwm of 125khz how fast of a waveform could I emulate to functionally change the multipliers input? are there better external controllable solutions?

wanderson

When constructing voltage divders, it is worth putting a number of smaller resistors in series to build the higher resistance, since that will also increase the safe voltage that can be handled. If you just used one 220Mohm resistor it would have to be rated to the voltage your using.  I constructed a discharge probe using 50 100K 2 watt resistors in series which together have a safe voltage approaching 50 kV.
New true random number library available at: http://code.google.com/p/avr-hardware-random-number-generation/

Current version 1.0.1

winner10920

How much did that cost? One 220M 2w Is like 20$ so I guess that's cheaper?

wanderson

I doubt you could find a 220Mohm resistor for $0.20 that was rated for 50kV.  In my case I believe I spent a few cents apiece for the 100k 2w resistors that I purchased, maybe a little more.  The difference is that each individual resistor was rated to about 1kV, but fifty in series not only increased the total resistance, but also the effective voltage that could be applied across that resistance.  I have a small selection of true high voltage resistors, rated to about 20kV, that I picked up off of ebay for about $5 each, but the selection is very limited.
New true random number library available at: http://code.google.com/p/avr-hardware-random-number-generation/

Current version 1.0.1

winner10920

I saw that resistor on ebay for $20 not 20cents lol, I think I may take that route tho sounds alot cheaper

Go Up