how to make 100W non inductive resistor
acan...@wwc.com
I need to make a chain of resistors whose total resistance is about
1500 ohms, and where I can pick of in the middle at 40, 500, 900
ohms,etc.. (maybe 10 spots). The chain needs to be able to withstand
5000V peak (AC) at 1MHz, and have as little capacitance and inductance
as possible, to the point where a 6 foot long resistor would be a-okay
if thats what it took. And it needs to hand 250W of power
continuously.
Im thinking if only there was some metal that had a very high bulk
resistivity so I could make a straight line resistor that also had
alot of surface area for power dissipation...anyone do this somehow,
or something like it?
mike
> Hello,
>
> I need to make a chain of resistors whose total resistance is about
> 1500 ohms, and where I can pick of in the middle at 40, 500, 900
> ohms,etc.. (maybe 10 spots). The chain needs to be able to withstand
> 5000V peak (AC) at 1MHz, and have as little capacitance and inductance
> as possible
Not a "specification that can be met" NUMBERS????
You can bifilar wind a wirewound resistor and hook the windings
together at one end so the coupled inductances cancel...sort-of. NUMBERS???
Joerg
NH-series:
http://www.vishay.com/docs/30201/30201.pdf
Also comes in non-inductive:
http://www.seielect.com/Catalog/SEI-kal.pdf
--
Regards, Joerg
http://www.analogconsultants.com/
"gmail" domain blocked because of excessive spam.
Use another domain or send PM.
Jan Panteltje
"acan...@wwc.com" <acan...@wwc.com> wrote in
<126dd916-9310-4a5c...@t17g2000prg.googlegroups.com>:
In the early days we made a light dimmer for stage lights by cutting
of the end of a fluorescent tube, filling it with water and salt,
and hanging a piece of iron in it from a wire, making it a variable resistor.
pimpom
Joerg gave you an elegant solution, but those look a bit pricey.
If you want a less elegant but cheaper DIY alternative, you could
take nichrome heater coils and wind them non-inductively on a
ceramic former.
Tim Williams
news:126dd916-9310-4a5c...@t17g2000prg.googlegroups.com...
> I need to make a chain of resistors whose total resistance is about
> 1500 ohms, and where I can pick of in the middle at 40, 500, 900
> ohms,etc.. (maybe 10 spots). The chain needs to be able to withstand
> 5000V peak (AC) at 1MHz, and have as little capacitance and inductance
> as possible, to the point where a 6 foot long resistor would be a-okay
> if thats what it took. And it needs to hand 250W of power
> continuously.
Well, you can't have more than 100pF across your 1500 ohm resistor, nor more
than 6.4uH in series with your 40 ohm resistor, otherwise it looks more
reactive than resistive. If you need reactive current under 10%, cut these
by 10 and things look more impressive: a 250W vitreous resistor will easily
run 10pF to ambient, more if shielded; and it has to use less than 32" of
wire (assuming 20nH/inch). A noninductive winding will fair somewhat
better.
If your signal is 1.0MHz sine, you can cancel any annoying inductance or
capacitance with some more of the other type. OTOH, if it has harmonics,
you're nearly SOL because they will be affected the most by parasitics.
Tim
--
Deep Friar: a very philosophical monk.
Website: http://webpages.charter.net/dawill/tmoranwms
Bit Farmer
Packed carbon granules in a non conductive tubing.
acan...@wwc.com
> <acann...@wwc.com> wrote in message
Lets say the solution ends up being some kind of linear element (wire,
nichrome, carbon granules in tube etc...) but its a few feet long. Can
I reduce the inductance by sliding the whole assembly inside a close
fitting tube made of mu metal?
Tim Williams
news:86f96034-69d6-47b2...@m27g2000prl.googlegroups.com...
> Lets say the solution ends up being some kind of linear element (wire,
> nichrome, carbon granules in tube etc...) but its a few feet long. Can
> I reduce the inductance by sliding the whole assembly inside a close
> fitting tube made of mu metal?
No, not really. It'll greatly increase the capacitance though (I'm guessing
very roughly 100pF?). Mu metal isn't what you want anyway, it doesn't work
above 10kHz, let alone 1MHz.
You can, however, reduce the inductance by making it physically small. They
make thick film resistors that are expensive as hell but would do nicely at
high frequency. Downside is you need $500 of them to make the series you
want...
What frequencies and waveforms are you using? What accuracy? Can you
cancel reactance with tuning?
whit3rd
> I need to make a chain of resistors whose total resistance is about
> 1500 ohms, and where I can pick of in the middle at 40, 500, 900
> ohms,etc.. (maybe 10 spots). The chain needs to be able to withstand
> 5000V peak (AC) at 1MHz, and have as little capacitance and inductance
> as possible, to the point where a 6 foot long resistor would be a-okay
> if thats what it took. And it needs to hand 250W of power
> continuously.
Odd requirement. You can chain a lot of resistors (say, 100 15 ohm
1W
resistors) and dunk the array into a gallon of dielectric oil, with a
simple
massive-plate heatsink, or just run some cooling water through pipes
in
a 1-gallon paint can.
Inductance can be low, OR you can choose to model the individual taps'
voltage drops as I*R + L* dI/dt, and note that a bit of added mutual
inductance, in series with the tap wiring, can cancel the L-containing
term.
Because power is dissipated in the resistors, there WILL be a B-field
at the resistor surface, and that means there IS always some
inductance.
Zero inductance means no B field, no B field means Poynting's vector
is
zero, and Poynting's vector surface integral around a resistor is the
power dissipation (which we know is NOT zero).