Capacitor in series - Resistor across ?

[32198]

When capacitors are in placed in series because the proper value cannot be found voltage limitation raise but is there a side effect.

In another thread somebody suggested to put a high value resistor across the capacitor without precisely suggesting a value. I asked about the value of this resistor the membre didn't know and suggested to ask in here... somebody here knows ?

http://www.diyaudio.com/forums/multi-way/162461-little-refresher-please-capacitors-parallel.html

 

[Arnulf]

Depends on your situation but you don't want resistors to affect anything at the frequency of interest. 100K and up into megaohm range is generally sane value when it comes to tube applications.

For starting value calculate reactance at the frequency of interest and use resistor at least 10-100 times that value.

Also keep in mind power dissipation across those resistors !

 

[HollowState]

In the case of electrolytic capacitors used in the power supply for filtering, putting them in series will add their respective voltage ratings. The reason for paralling them with a resistor is to be sure the voltage across each is divided appropriately. (evenly if the voltage ratings are equal) The thinking is to choose a resistor value that is lower then the leakage of the capacitor it parallels. Usually an educated guess is all that is necessary baised on the quality of the capacitor. For new high quality capacitors typical values are from 470K ohms down to 100K ohms. The actual value is not really that critical. And typically ½ to 1 watt is sufficient for voltages up to approx. 500V. For poorer quality capacitors I have seen resistors as low as 50K ohms at 5 watts used, but this is the exception. For non-electrolytic (coupling) capacitors, much higher values should be used if need at all.

 

[pauldune]

When capacitors are in placed in series because the proper value cannot be found voltage limitation raise but is there a side effect.

In another thread somebody suggested to put a high value resistor across the capacitor without precisely suggesting a value. I asked about the value of this resistor the membre didn't know and suggested to ask in here... somebody here knows ?

http://www.diyaudio.com/forums/multi-way/162461-little-refresher-please-capacitors-parallel.html

The story talks about caps in parallel rather then serie?

Could you tell us the application? You have a center tapped transformer, and in what voltage? Solid state or tube rectification?

If its center tapped, and solid state, you might want to use a bridge rectifier, put the C's in series, and connect the center tap to the middle where the caps are connected. This way you ensure the proper voltage division.
250v-0v-250v gives you approx. 700v
I would put some 470kOhm resistors over the caps, just to make sure the c's are properly drained after shutdown.

I was once "shocked"😱 by a 2200uF 450v after 2 days power off....

gr. Paul

ps If you dont use a center tapped transformer; dont be tempted to use the "halfway potential" for something... I once did an experiment: 6.3v CT and bridge rect. and 2x 35v 10000uF caps in series behind them. Without load, it was about 8v, and it was more or less evenly divided across the caps. 4v-4v.
Then I put the filament of some old tube as a load on it. the division stayed appox the same; something like 3.8v-4.2v. Then I put the filament on ground and a the halfway tap: and something weird happened: the loaded c dropped to 1v, and the unloaded raised to 7v. So 1v-7v. If this not were a low voltage experiment, but a 900v (450v-450v) powersupply, the uneven load would give 100v-800v. One of the elco's would surely explode.

 

[Johan Potgieter]

I usually use 330K for voltages (per cap) of up to 450V.

One other thing: It must be noted that resistors also have maximum voltage ratings. With too high a gradient across a physically small resistor, internal breakdowns can occur leading to failure in the end. Thus I would not use physically smaller than 1W resistors, otherwise more than one resistor in series. Specs for good resistors should include maximum voltage ratings.