Inductive Bleeder Resistors

[TedN]

I'm finishing up a small plate supply and want to order the bleeders for it.
The supply currently has, if I remember correctly, 1M carbon comp for bleeders.
It's an old filter cap bank that I pulled from the junk box to start testing. I want to order new bleeders in the range of 30K per each new 450 WVDC cap.  The only parts I can find are wire wound power resistors. I can buy non-inductive but they are not in stock at my supplier. I remember reading somewhere not to use inductive type resistors as bleeders. Parallel LC network. I did some research and can't find any reference to this statement. Any fact or just snake oil?

Thanks For the advice.

73
Ted KC9LKE

[WA1GFZ]

Wire wound will be fine and I would use 50K 10 watt

[TedN]

Thanks!
I believe I read this in QST many years ago. Just didn't want to shell out the dough
on parts that cause headaches. I'll try the 50K/10W.

Regards
Ted/ KC9LKE

[KB2WIG]

Some peeps use a few 1meg bleeders accross the wire wound types as  secondary protection. The theory is if the big resistors go, the 1 megs will bleed down the charge. Being that they are so much higher R value than the actual bleeders, they dont enter into things, so thay loaf along. Something like 1meg/1Watt per 400 volts of protection. The breakdown voltage of the 1M/ 1W is around 500.

 -- If u bleed a 1000 Volt supply, you want MORE than 2, 1 Watt resistors accross the thing. 3 is the bare minimun.  This is inclusive of your 30K per new 450 WVDC cap.   (GFZ's 50K is a good idea too)

Look here.....

http://www.ohmite.com/catalog/pdf/little_demon.pdf

  A open bleeder resistor is a verry, verry dangerious thang ...........  klc

[k4kyv]

No need to worry about non-inductivity for bleeder  resistors.  For rf applications yes, for 60~ power supply applications, no.

The main concern is power rating, and getting the heat away from the resistors (without cooking nearby components) as easily as possible.

I always mount wirewound resistors that are hollow inside in a vertical position, to increase convection cooling via the chimney effect.

For a choke input filter, use no more than 1000 ohms of  bleeder for every henry of choke inductance, to maintain critical inductance (so that it doesn't turn into a capacitive input filter, with soaring output voltage when the load is removed).  For example, a 5-25 henry swinging choke should have no more than 25k of bleeder, if you plan to completely remove the load as in the case of an AM modulator or SSB leenyar power supply, or class-C rig used on CW.  If you plan to use on AM only, and run a common power supply, the class-C final load makes up part of the bleeder.  In that case the purpose of the bleeder is strictly to bleed off the filter caps when the power is turned off.  But I would use filter caps with voltage rating at least 50% higher than the rated plate voltage, in case you lose the PA load, so that the voltage doesn't soar high enough to blow the filter caps.

In my HF-300 rig, the voltmeters have 10 ma movements, so the multiplying resistors act as back-up bleeders and will eventually drain the caps.  With a voltmeter permanently installed at the output of the power supply, you can observe the output voltage bleed off.  But don't trust your life with it, just in case the voltmeter craps out at exactly the same time the main bleeder opens.  Murphy just loves to pull tricks  like that.

[WA1GFZ]

failure mode of a carbon resistor is first it goes to a very low value and it isn't rated for 450 volts

[KB2WIG]

Ohmite specks their 1meg/1w carbon comps at 500V.

http://www.ohmite.com/catalog/pdf/little_demon.pdf

R go boom

klc