Dividing resistors & Capacitor voltage in series.

JamieMcC · 1113

0 Members and 1 Guest are viewing this topic.

Offline JamieMcC

  • Hero Member
  • *****
    • Posts: 1167
on: March 14, 2014, 04:47:21 AM
I came across a old post about connecting capacitors in series, here it is below. And have a question I hope someone might be able to help.

Two short notes:

1) putting caps in series requires the dividing resistors; otherwise one cap will inevitably hog the voltage until it blows out. However, that means the combination will have leakage current. You can't do this with an output cap such as the 100uF in Crack - you'll get DC current through your headphones.

2) The technical problem with bypassing caps is the possible resonance between the small cap and the parasitic inductance of the large one. If that resonance is damped by the parasitic resistance, you are OK - unfortunately those values are hardly ever specified. If the resonance is well above audible frequencies, you may also be OK - it's been argued both ways. Usually a small resistance (0.1 ohm to 1.0 ohm) in series with the larger cap, with the small cap bypassing the combination, will damp the resonance pretty thoroughly. At least in theory it will...  :^)


I think this is what is being described above in no 2 but am not sure if I am following correctly.


My question relates to the dividing resistors.  Is it possible to substitute a capacitor for the resistor and thus block the dc current leakage to the headphones. Its not a Crack specific question just looking to see if I might be able to utilise some of the lower voltage film caps I have been trying in the Crack in perhaps the Sex or Stereomour later on.

I have a little collection of caps now which are rated in the 300-400 VDC range and wondered if they could be connected in series to reach 600V higher ratings I see used in the Sex or Stereomour in conjunction with a paralleled 600v bypass capacitor in place of the dividing resistor to prevent the dc leakage to the headphones in the Sex amp or even if this effect is still relevant for the Sex amp or Stereomour.

Apologies if I am way off in my thinking, with no electronics background I  find such topics hard to grasp.


« Last Edit: March 14, 2014, 04:52:50 AM by JamieMcC »

Shoot for the moon if you miss you will still be amongst the stars!

Offline Paul Birkeland

  • Global Moderator
  • Hero Member
  • *****
    • Posts: 19507
Reply #1 on: March 14, 2014, 05:09:48 AM
Hello Jamie,

This advice primarily concerns power supply caps.  If we have two power supply electrolytic caps in series, with the overall supply voltage being higher than the individual voltage ratings (say 400V B+ with two 250V caps in series), then uneven leakage current between the two capacitors can shift the ideal balance of having 200V on each cap to being very skewed.

Consequently, when we put a resistor across each capacitor, there is standing current across those resistors that is generally much larger than capacitor leakage, and that swamps out any differences that would appear. 

When you move to interstage and parafeed caps, these are DC blocking caps, so putting resistors across them kind of defeats the purpose of having them in the first place, as some of the DC voltage that the caps are trying to block will appear at the outputs.  On the other hand, the leakage current of polypropylene capacitors is very low, but not low enough to ignore.  What you will sometimes see is a 10-100M resistor across each capacitor to maintain voltage balance across them and to minimize the voltage appearing at the output. 

To say the least, implementing this can be somewhat trivial, or it can completely change the design of the circuit entirely.  Buying higher voltage rated components generally looks a lot better after working through something like this.


Paul "PB" Birkeland

Bottlehead Grunt & The Repro Man

Offline JamieMcC

  • Hero Member
  • *****
    • Posts: 1167
Reply #2 on: March 14, 2014, 06:09:55 AM
PB I am following you just, and was hoping you might reply. I did see some math that should give the relative load each capacitor sees when in series to given load which I might make an attempt at. Even though it is sure to lead to some serious brain ache on my part.

Even though the intended application/use is a little way off on my part it does offers something to consider in a future build.

Thanks for the reply.

Shoot for the moon if you miss you will still be amongst the stars!