Author Topic: Purpose of a discharge resistor across a smoothing capacitor after a rectifier?  (Read 2697 times)

0 Members and 1 Guest are viewing this topic.

Offline BlastFXTopic starter

  • Newbie
  • Posts: 4
  • Country: cz
I like to open things up and while looking at LED light bulbs using a simple dropper (capacitive or even resistive) to drive the LEDs, I've noticed most have a discharge resistor connected across the smoothing capacitor after the bridge rectifier (well, the ones that do have a smoothing capacitor in the first place).
The way I understand it, the diodes should block any current from going from the smoothing capacitor back out the contacts. If one or more diodes fail open circuit, it's still safe. If a diode fails short circuit, you've got bigger problems than possibly getting a zing off the contacts (the kind of problems that cause other things to go open circuit… and also flames).
So what's the point? Am I missing something?
 

Offline exe

  • Supporter
  • ****
  • Posts: 2647
  • Country: nl
  • self-educated hobbyist
What's the voltage across the cap? May be it's lethal, that's why they do it.

I used to be a big fan of bleeding resistors. Not anymore as, apparently, electrolytic caps don't like to be discharged, it's bad for oxide layer inside. Somebody said it's good that PN-junction has a voltage drop of ~0.6, this way ICs cannot  discharge caps to zero. So, I'd put a diode in series with the resistor so voltage doesn't drop much below 0.6V.

I myself didn't verify this, but what I did notice that after charging caps to their rated voltage and waiting for a few minutes the leakage current starts to drop. So, I can believe that maintaining cap charged helps it.
 

Offline CatalinaWOW

  • Super Contributor
  • ***
  • Posts: 5571
  • Country: us
Depending on the speed of the bleed it could be just to help get the lights off quickly.
 

Offline BlastFXTopic starter

  • Newbie
  • Posts: 4
  • Country: cz
What's the voltage across the cap? May be it's lethal, that's why they do it.
It is potentially lethal, it's full peak mains voltage in some cases (minus two diode drops), that's 310V where I'm from, but, as I said in the first post, there's no way I can conceive of for you to get to that voltage unless you disassemble the light bulb (at which point anything that happens is on you).

Depending on the speed of the bleed it could be just to help get the lights off quickly.
That kinda makes sense. The caps I usually see are in the tens of micro farads range, the voltage across the LEDs is usually from 20 to 80 volts (and you only need to nudge it down by a couple volts to make the LEDs go out) and the discharge resistors are usually half a megohm, so that works out to tens to hundreds of milliseconds, which would be a fairly sharp turnoff, I suppose.
But — specialist applications aside — I don't see why a light bulb trailing off for a couple seconds would be a concern. The glass in halogen bulbs also keeps glowing for a couple seconds after they're turned off and that doesn't seem to bother anyone.
 

Offline CatalinaWOW

  • Super Contributor
  • ***
  • Posts: 5571
  • Country: us
Just think of CFDs and how minor differences in turn on/turn off from incandescents put off their acceptance.  While it really made no difference, perception is everything.  If the LEDs would stay on for more than a couple hundred milliseconds in the absence of the bleed it could be the reason for them.
 

Offline BlastFXTopic starter

  • Newbie
  • Posts: 4
  • Country: cz
OK, guys, after a lot of googling, I think I figured it out.
The resistor isn't as much across the capacitor as it is across the LEDs.
Apparently, if your light switch is shitty enough, a few microamps can get capacitively coupled through it and that's enough to make the LEDs glow dimly even when the light is “off.” The purpose of the resistor is to eat up that current and keep the light bulb from lighting up when it shouldn't.
 

Offline mikerj

  • Super Contributor
  • ***
  • Posts: 3382
  • Country: gb
I used to be a big fan of bleeding resistors. Not anymore as, apparently, electrolytic caps don't like to be discharged, it's bad for oxide layer inside. Somebody said it's good that PN-junction has a voltage drop of ~0.6, this way ICs cannot  discharge caps to zero. So, I'd put a diode in series with the resistor so voltage doesn't drop much below 0.6V.

I myself didn't verify this, but what I did notice that after charging caps to their rated voltage and waiting for a few minutes the leakage current starts to drop. So, I can believe that maintaining cap charged helps it.

This is nonsense.  Whilst electrolytic capacitors do eventually degrade with age if not used (the reason that very caps on old valve radios etc. need to be "reformed") it's a slow process and it's just not possible to indefinitely maintain a voltage across a cap if the circuit is not powered.   Electrolytic capacitors have built in leakage, as do diodes, so the voltage will always fall to zero at some point.
 

Offline exe

  • Supporter
  • ****
  • Posts: 2647
  • Country: nl
  • self-educated hobbyist
This is nonsense.  Whilst electrolytic capacitors do eventually degrade with age if not used (the reason that very caps on old valve radios etc. need to be "reformed") it's a slow process and it's just not possible to indefinitely maintain a voltage across a cap if the circuit is not powered.   Electrolytic capacitors have built in leakage, as do diodes, so the voltage will always fall to zero at some point.

I don't think this is nonsense. 1) leakage current is not that big, I have caps maintaining charge at least for weeks. 2) I read that oxide layer, while it is dissolving, creates potential, which in turn slows down degradation. I don't know if the source I read is reliable, or not. But please don't jump to conclusion without data.
 

Offline mikerj

  • Super Contributor
  • ***
  • Posts: 3382
  • Country: gb
This is nonsense.  Whilst electrolytic capacitors do eventually degrade with age if not used (the reason that very caps on old valve radios etc. need to be "reformed") it's a slow process and it's just not possible to indefinitely maintain a voltage across a cap if the circuit is not powered.   Electrolytic capacitors have built in leakage, as do diodes, so the voltage will always fall to zero at some point.

I don't think this is nonsense. 1) leakage current is not that big, I have caps maintaining charge at least for weeks. 2) I read that oxide layer, while it is dissolving, creates potential, which in turn slows down degradation. I don't know if the source I read is reliable, or not. But please don't jump to conclusion without data.

Sorry but it is nonsense, you've read something somewhere and jumped to a daft conclusion.

You certainly can have a cap maintaining charge for fairly long time if it's connected to nothing, but usually not when it's in a circuit.  The voltage across the vast majority of capacitors will be zero within seconds or minutes of switching off a circuit, and yet designs using quality capacitors within their voltage and temperature limits will last a very long time even with intermittent use.
 

Offline Psi

  • Super Contributor
  • ***
  • Posts: 10385
  • Country: nz
I have caps maintaining charge at least for weeks.

My 2600 farad maxwell supercaps have sat on the shelf for a year and only self discharged from 2.5V to ~1.7V.
Which is pretty amazing.

They drop pretty quick from 2.5V down to 2.2V but after that it's very slow
Greek letter 'Psi' (not Pounds per Square Inch)
 
The following users thanked this post: exe

Offline dcbrown73

  • Regular Contributor
  • *
  • Posts: 114
  • Country: us

My 2600 farad maxwell supercaps have sat on the shelf for a year and only self discharged from 2.5V to ~1.7V.
Which is pretty amazing.

They drop pretty quick from 2.5V down to 2.2V but after that it's very slow

By on a shelf, you mean outside of a circuit?
Why exactly do people feel I should have read their post before I responded?  As if that was necessary for me to get my point across.
 

Offline John B

  • Frequent Contributor
  • **
  • Posts: 865
  • Country: au
Polarised electrolytic caps are also used in audio coupling where there is either (almost) no voltage across the cap, or even a slight AC voltage going into reverse polarity. They can last decades being used like that.
 

Online ejeffrey

  • Super Contributor
  • ***
  • Posts: 4034
  • Country: us
The 0.6 V for a pn junction is at high current.  A diode will still conduct at any nonzero voltage.  An ideal didoe will have forward conduction about equal to it's reverse leakage at 25 mV.  This isn't much but will discharge ordinary size capacitors in a few minutes.
 
The following users thanked this post: exe

Offline ArthurDent

  • Super Contributor
  • ***
  • Posts: 1193
  • Country: us
It depends a lot on the capacitor type and voltage rating. Where I worked we did some very high voltage testing and used large high voltage oil filled capacitors. When these were being stored out of circuit these capacitors always had shorting wires across the terminals to prevent charge build-up that could be dangerous. I generally store some of my not-so-big HV caps the same way. Here's a post explaining what happens:

For large, high voltage capacitors, storage with shorted terminals ensures that no lethal voltage is present across the terminals, and therefore no unfortunate accidents happen. For well constructed capacitors in dry climates the voltage will not decay adequately on its own even after months of storage. In fact, a capacitor can be discharged to zero volts, but without the wire the voltage will slowly build up again in a phenomenon called dielectric absorption. It turns out that a fraction of the capacitor charge soaks deep into the dielectric and can slowly raise the voltage back to potentially hazardous levels over days or weeks . The continuous presence of a shorting wire during storage is a wise safety measure, and often required by law.""

 
The following users thanked this post: dcbrown73


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf