Electrochemistry: Liquid Power Resistor for 2000VAC

[D from BC]

I'm setting up a test circuit and I need a 100k resistor that can handle
2000VAC for a few seconds.

I was going to run off to the local surplus shop but got the idea of
making a liquid resistor. Saves a trip.

I added some salt to water and got 100kohm. Copper electrodes (all I got
at the moment).

Question is... Will this stay 100kohm +/- 10kohm up to 2000VAC?
My circuit only needs to run for 5 seconds.

iows...Does salty water resistance vary with voltage?

 

[Jon Slaughter]

I'm setting up a test circuit and I need a 100k resistor that can
handle 2000VAC for a few seconds.

I was going to run off to the local surplus shop but got the idea of
making a liquid resistor. Saves a trip.

I added some salt to water and got 100kohm. Copper electrodes (all I
got at the moment).

Question is... Will this stay 100kohm +/- 10kohm up to 2000VAC?
My circuit only needs to run for 5 seconds.

iows...Does salty water resistance vary with voltage?

Why not experiment with it since you already have the setup? Try putting a
100k res in series and measure the voltage. Sweep up to 200VAC and see if
the volage devates from 100VAC at the divider.

I imagine if the liquid gets hot the resistance will change so you might try
an ice bath. You could then try with and without the ice bath and find the
deviation to get some temperature dependence relation.

 

[D from BC]

Trying out another "natural selection" experiment ?

...Jim Thompson

lol

 

[D from BC]

Copper will be electrolyzed into solution and the resistance will
drop. And bubbles will form on the electrodes and resistance will
increase!

Neither happens fast at low currents.

But the dissipation is only 40 watts for 5 seconds. A string of, say,
ten or so 2-watt carbon film resistors would work fine.

John

Yup. I don't have any but odds are better that the surplus store will
have 2W power resistors as opposed to a single power resistor.
It'll probably be better if I run out and get the resistor/s.
groan ..

 

[Tim Williams]

Copper will be electrolyzed into solution and the resistance will
drop. And bubbles will form on the electrodes and resistance will
increase!

Not at AC faster than the diffusion rate of the liquid. Even a pokey 60Hz
modulation is notorious for having terrible electrolysis capacity.

Tim

 

[D from BC]

Why not experiment with it since you already have the setup? Try putting a
100k res in series and measure the voltage. Sweep up to 200VAC and see if
the volage devates from 100VAC at the divider.

I imagine if the liquid gets hot the resistance will change so you might try
an ice bath. You could then try with and without the ice bath and find the
deviation to get some temperature dependence relation.

I might. I do have a isolated variable transformer and a single 2W 100k
resistor.
And I could try graphite electrodes. I found a pencil.

I'm just wobbling between curiosity /laziness vs practicality.
To set up an experiment or... forget about that and run off to the
surplus store and get some real resistors. mmm

 

[Jon Slaughter]

I might. I do have a isolated variable transformer and a single 2W
100k resistor.
And I could try graphite electrodes. I found a pencil.

I'm just wobbling between curiosity /laziness vs practicality.
To set up an experiment or... forget about that and run off to the
surplus store and get some real resistors. mmm

If it's just for an experiment then why not put the resistor in an ice bath?
I'm not sure how well this will help out but it should easily allow a 1/4W
resistor to last 5 seconds? I imagine the resistor may last 5 seconds
without the bath. The main issue is, of course, that the resistance may
change drastically but you could try it and see.

Use the 2W 100k resistor in an ice bath and you should be fine. You could
put 200 liquid containers in series or parallel if you really wanted to have
fun

 

[D from BC]

If it's just for an experiment then why not put the resistor in an ice bath?
I'm not sure how well this will help out but it should easily allow a 1/4W
resistor to last 5 seconds? I imagine the resistor may last 5 seconds
without the bath. The main issue is, of course, that the resistance may
change drastically but you could try it and see.

Use the 2W 100k resistor in an ice bath and you should be fine. You could
put 200 liquid containers in series or parallel if you really wanted to have
fun

I'll do that. Will report later if the 2W 100K with 2000VAC across it
failed in cold bath.

 

[Jon Slaughter]

I'll do that. Will report later if the 2W 100K with 2000VAC across it
failed in cold bath.

If you can, monitor the current or use a voltage divider to record the
voltage and see how long it takes to change... or record the temperature of
the bath.

 

[Jon Slaughter]

I'll do that. Will report later if the 2W 100K with 2000VAC across it
failed in cold bath.

You might want to insulate the wiring too, just to be on the safe side.

 

[D from BC]

I've seen big (as in 50' high, 12 megavolt) Marx generators that used
water-filled clear plastic pipes, about like a garden hose, as the
capacitor charging resistors. They were blue inside, copper sulphate
solution I think.

John

I do have CuSO4.

http://en.wikipedia.org/wiki/Electrolysis_of_water
'The following cations have lower electrode potential than H+ and are
therefore suitable for use as electrolyte cations: Li+, Rb+, K+, Cs+,
Ba2+, Sr2+, Ca2+, Na+, and Mg2+. Sodium and lithium are frequently used,
as they form inexpensive, soluble salts.'

Those are all group 1 and 2 elements..
So I suppose I should use CuSO4 to prevent H2 and O2 bubbles creating
resistance error on the electrodes.

btw.. I barely know electrochemistry.

 

[markp]

I'm setting up a test circuit and I need a 100k resistor that can handle
2000VAC for a few seconds.

I was going to run off to the local surplus shop but got the idea of
making a liquid resistor. Saves a trip.

I added some salt to water and got 100kohm. Copper electrodes (all I got
at the moment).

Question is... Will this stay 100kohm +/- 10kohm up to 2000VAC?
My circuit only needs to run for 5 seconds.

iows...Does salty water resistance vary with voltage?

EBG do a nice range. You could for example put 5 560k SSP52 in parallel:
http://www.ppmpower.co.uk/downloads/products/PPM295.pdf

Mark.

 

[D from BC]

EBG do a nice range. You could for example put 5 560k SSP52 in parallel:
http://www.ppmpower.co.uk/downloads/products/PPM295.pdf

Mark.

I might track down where I can get one.

 

[Tim Williams]

I do have CuSO4.

http://en.wikipedia.org/wiki/Electrolysis_of_water
'The following cations have lower electrode potential than H+ and are
therefore suitable for use as electrolyte cations: Li+, Rb+, K+, Cs+,
Ba2+, Sr2+, Ca2+, Na+, and Mg2+. Sodium and lithium are frequently used,
as they form inexpensive, soluble salts.'

Those are all group 1 and 2 elements..
So I suppose I should use CuSO4 to prevent H2 and O2 bubbles creating
resistance error on the electrodes.

If you use Cu electrodes in CuSO4, you'll get straight electroplating. A
few seconds run time won't make a difference.

Tim

 

[markp]

I might track down where I can get one.

It seems from the spec that they can also be over-rated for 5 seconds at 5
times the power rating, which means you may well get away with just one 100k
SSP52, or possibly two put 220k's in parallel for belt and braces.

Mark.

 

[D from BC]

It seems from the spec that they can also be over-rated for 5 seconds at 5
times the power rating, which means you may well get away with just one 100k
SSP52, or possibly two put 220k's in parallel for belt and braces.

Mark.

I checked Mouser using keyword SSP52. No results.

I'll get resistors when I've build up a parts list worthy of making a
Digikey order.

I might order these .. 4 X $2.00ea..
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=RSD-
25KRCT-ND

OR

$13.00 for a single R.
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=
850F100KE-ND

In the meantime, I'm taking my chances with a liquid cooled 2W resistor
from the junkbox.

 

[markp]

I checked Mouser using keyword SSP52. No results.

I'll get resistors when I've build up a parts list worthy of making a
Digikey order.

I might order these .. 4 X $2.00ea..
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=RSD-
25KRCT-ND

OR

$13.00 for a single R.
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=
850F100KE-ND

In the meantime, I'm taking my chances with a liquid cooled 2W resistor
from the junkbox.

Either option seems to be OK, although in the single resistor version you're
closer to the edge of over-driving it (i.e. there's more slack in the 4
resistor version voltage wise). Although you can think of the liquid as some
kind of ideal heatsink the 2W resistor package is going to have a finite
thermal resistance, so localised heating might damage it before the heat can
flow out of the package into the liquid. Careful too with dielectric
breakdown voltage.

Mark.

 

[markp]

Either option seems to be OK, although in the single resistor version
you're closer to the edge of over-driving it (i.e. there's more slack in
the 4 resistor version voltage wise). Although you can think of the liquid
as some kind of ideal heatsink the 2W resistor package is going to have a
finite thermal resistance, so localised heating might damage it before the
heat can flow out of the package into the liquid. Careful too with
dielectric breakdown voltage.

Mark.

Another thing to watch is the type of resistor package you use. Most of the
moulded types transfer their heat through the legs (unless they have an
integral heatsink or provision to add one), not through the package itself
which has a much higher thermal resistance, so make sure both legs are fully
immersed in the liquid if it is this type.

Mark.

 

[Ecnerwal]

If you use Cu electrodes in CuSO4, you'll get straight electroplating. A
few seconds run time won't make a difference.

Tim

We used copper sulfate with copper electrodes as resistors on Marx
generators. Works. May not be completely stable, but that's not going to
be a drastic change in a few seconds thing, more like over weeks or
months the resistance will change. If the copper sulfate solution is
fully saturated, the resistance will be less likely to change. Something
like 2" diameter tubing, and the whole mess drowned in transformer oil.

Voltage was DC and 10's of KV for charging. IIRC the big one erected to
over a million volts, but none of the resistors would see all that.

 

[Fred Abse]

John Larkin wrote:



That must be Lenin generator.

More likely a Stalin generator - propensity to arbitrarily liquidate those
around it.