I don't think it matters...
If you want to ensure they get equal always-positive signals, put diodes across them to prevent reverse voltage. Note that when the capacitors are uncharged, one diode switches on and apparent C is just one; on the next cycle, the caps are charged and will look like two caps in series, i.e. C = C/2.
Electrolytics aren't good for continuous AC use, hence motor start caps are smaller than motor run caps, which are film in oil.
Tim
If you want to ensure they get equal always-positive signals, put diodes across them to prevent reverse voltage. Note that when the capacitors are uncharged, one diode switches on and apparent C is just one; on the next cycle, the caps are charged and will look like two caps in series, i.e. C = C/2.
Electrolytics aren't good for continuous AC use, hence motor start caps are smaller than motor run caps, which are film in oil.
Tim
Which way do I do it for mains DC blocking (with the antiparallel diodes across them)?
BTW, I only found 6A diodes locally, whereas I need 10A. When I parallel them, do I need to put resistances in front? Somewhere I read that should be done, to drop 0.3-0.5V, otherwise as the diodes aren't a perfect match, one will conduct most of the current. But that means 0.05Ohm diodes, and I can't find anything like that.
BTW, I only found 6A diodes locally, whereas I need 10A. When I parallel them, do I need to put resistances in front? Somewhere I read that should be done, to drop 0.3-0.5V, otherwise as the diodes aren't a perfect match, one will conduct most of the current. But that means 0.05Ohm diodes, and I can't find anything like that.
Yes, a bit of resistance will level out current sharing. The only thing it costs you is dissipation and voltage drop, whether that will be a problem is up to you.
With or without diodes, direction doesn't matter, as long as the (cap+diode)'s both point in opposite directions.
Wazza mains DC blocking about?
Tim
With or without diodes, direction doesn't matter, as long as the (cap+diode)'s both point in opposite directions.
Wazza mains DC blocking about?
Tim
Erm, the 1N5404 diodes turn on when DC > 0.6V.
Also, you only need one cap.
I don't see DC getting on the line anyway, unless you have a **LOT** of half-wave-rectified equipment running on your circuit. Consider that line voltage comes from a transformer (pole pig) which delivers about 240VCT at 200A or more... you think any DC that matters to your equipment is going to get past that short circuit of a winding?
Tim
Also, you only need one cap.
I don't see DC getting on the line anyway, unless you have a **LOT** of half-wave-rectified equipment running on your circuit. Consider that line voltage comes from a transformer (pole pig) which delivers about 240VCT at 200A or more... you think any DC that matters to your equipment is going to get past that short circuit of a winding?
Tim
Tim, if you unsymmetrical sine wave you HAVE DC, by definition. DC in the mV range is enough to get toroids to hum. Tim, the only reason for using a DC trap is to get rid of annoying hum from toroids especially, nothing else!
My solution look like this
Lars Clausen has also a DC trap but I can't find it. Lars, where do you have it? You have it in the pricelist but the link goes to something else.
My solution look like this
Lars Clausen has also a DC trap but I can't find it. Lars, where do you have it? You have it in the pricelist but the link goes to something else.
Why? That cap could get a sustained up to 0.6 V (in my case 1.2 V as I used two diodes to block more DC) in the wrong direction, which over time will still electrolyze the aluminum oxide.
What I'm wondering is how two capacitors a per Peranders' design and the one I posted are supposed to protect against this. I need to know whether to go out and buy two double sized ones to replace the single polarized one I'm using.
Er, one pair of caps. 
Anyone who's dealt with distortion knows that a sine wave can be asymmetrical without DC offset... how else does 2nd harmonic distortion get through an output transformer?
Which begs how it starts in the first place - the power company produces a quality 50/60Hz sine wave of very good frequency stability and generally good regulation (in here it stays between 120 and 125V, YMMV up to 5% I think). Noise pickup, particularly from the many DC power supplies running off it, bast@rdize the waveform but only with odd harmonics (symmetrical distortion). It would take a half wave rectifier of large capacity (or many, many small ones) on your local circuit (which includes your house and most likely a number of neighbors) to produce a significant offset.
Hm. If a 15A circuit can draw 15A with negligible drop (maybe 1VAC), it would take 9A rms DC on that circuit to produce a 0.6V offset. On a seperate circuit, you have far less, and lower resistance, wire in common with said circuit, I would expect maybe 30mV from same load on a different circuit. And none if on the opposite leg of your house circuit (US residents get 120-0-120V (240VCT) at maybe 100A, with half of the circuits running on each leg, with any 240V loads (water heater, dryer, etc.) going between legs).
If you truely have 0.5V DC offset, and it's not a measurement glitch... you honestly should have your wiring or appliances looked at.
Tim
Anyone who's dealt with distortion knows that a sine wave can be asymmetrical without DC offset... how else does 2nd harmonic distortion get through an output transformer?
Which begs how it starts in the first place - the power company produces a quality 50/60Hz sine wave of very good frequency stability and generally good regulation (in here it stays between 120 and 125V, YMMV up to 5% I think). Noise pickup, particularly from the many DC power supplies running off it, bast@rdize the waveform but only with odd harmonics (symmetrical distortion). It would take a half wave rectifier of large capacity (or many, many small ones) on your local circuit (which includes your house and most likely a number of neighbors) to produce a significant offset.
Hm. If a 15A circuit can draw 15A with negligible drop (maybe 1VAC), it would take 9A rms DC on that circuit to produce a 0.6V offset. On a seperate circuit, you have far less, and lower resistance, wire in common with said circuit, I would expect maybe 30mV from same load on a different circuit. And none if on the opposite leg of your house circuit (US residents get 120-0-120V (240VCT) at maybe 100A, with half of the circuits running on each leg, with any 240V loads (water heater, dryer, etc.) going between legs).
If you truely have 0.5V DC offset, and it's not a measurement glitch... you honestly should have your wiring or appliances looked at.
Tim
I should add--
What I mean by glitch is this: I just measured 120.2 to 120.4VAC (fluctuating a bit) at the computer here. The DMM measured DC as 260mV to 280mV (quickly alternating), however it read the same "offset" in the reverse direction. It should read -270mV if it were true, but instead it's all meter error. Even so, it's the same AC creating the error and the same meter, so half the difference in these measurements should show actual DC - but the numbers are exactly the same. So as I expected, I have no DC offset on this circuit. YMMV, but I doubt by much.
Tim
What I mean by glitch is this: I just measured 120.2 to 120.4VAC (fluctuating a bit) at the computer here. The DMM measured DC as 260mV to 280mV (quickly alternating), however it read the same "offset" in the reverse direction. It should read -270mV if it were true, but instead it's all meter error. Even so, it's the same AC creating the error and the same meter, so half the difference in these measurements should show actual DC - but the numbers are exactly the same. So as I expected, I have no DC offset on this circuit. YMMV, but I doubt by much.
Tim
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.