Re: [TCML] Why Don't Bleeder Resistors Fail?

[Quarkster <quarkster@xxxxxxx>]

Greg -

I've seen failures of both carbon composition and metal oxide/metal film resistors that were caused by breakdown of the resistor insulation, both arcing across the body (end-to-end) of a single resistor, and arcing from one resistor to an adjacent component

Standard "general purpose" resistors have a relatively low voltage rating, typically 200 volts "working" (constant) and 400 volts "pulse" for 1/4 watt resistors, and 250 volts "working" and 500 volts "pulse" for 1/2 watt resistors.

Vishay-BC Components offer a special series of resistors (VR-series) that have much higher voltage ratings than "standard" resistors, and only cost a few cents more then the standard parts. For instance, the VR-37 series 1/2 watt resistors are rated at 3500 volts DC. I've used these for years without in high voltage applications without any problems. 

Paying a few cents more per resistor to gain around 10X greater voltage margin is (to me) a small price to pay to get resistors that are specifically designed for high voltage applications.

I've seen MMC failures where a single 3 cent resistor failed and took out several expensive capacitors that were nearby.

Of course, good construction practice is probably the single most important factor in preventing HV breakdown:

1. Maintain good clearance between the resistors and any adjacent objects or surfaces (keep at least 1/4" clearance to the capacitor or mounting board);
2. After assembly, carefully remove ALL traces of soldering flux using isopropyl alcohol, etc;
3. Keep all parts of the circuit clean and free from dust buildup and other contamination;
4. Ensure adequate cooling;
5. Ensure adequate in-circuit voltage margin per resistor. 

It's perfectly acceptable to connect two resistors in series across each MMC capacitor, to gain voltage margin and reduce power dissipation per resistor.

See: http://www.vishay.com/docs/28733/28733.pdf  ; for the VR-37 datasheet.

Regards,
Herr Zapp

--- On Fri, 5/15/09, G Hunter <dogbrain_39560@xxxxxxxxx> wrote:
From: G Hunter <dogbrain_39560@xxxxxxxxx>
Subject: [TCML] Why Don't Bleeder Resistors Fail?
To: tesla@xxxxxxxxxx
Date: Friday, May 15, 2009, 5:16 AM

All,

I'm building a new MMC to replace the one that was ruined by a garage flood
in 2005.  The specs are seven series strings of ten, 68nf/1600v film & foil
polypropylene caps, for a total of 47nF @ 16kv peak.  This isn't a very high
voltage rating, but it's for a MOT-based SGTC that will charge it to only
10-12kvdc, so it will do.  The new MMC, like the old one, will employ 10M, 1/2W
carbon film bleeder resistors across each cap.  I believe this is fairly common
practice in the TC hobby community.  Since the resistors are only rated for 350v
or so, why don't they fail?  The whole resistor string is only good for
3.5kv.  I would expect it to flash over within seconds.  Yet I can't recall
a single bleeder resistor failure in my own experience, nor have I heard any
complaints on the TCML.  No big deal really.  I'm just curious.  Any
theories?  Too bad film caps are not so grossly underrated!  Would make MMCs
much cheaper!

Regards,

Greg


      
_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla
_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla