I was browsing through a book of old Electronics Magazine articles this weekend, and I found one where it suggests replacing the resistor in an R-C power supply filter with a diode. You get additional filtering, without any voltage drop or heat dissipation. A 1N4007 should be fine for the typical supply.
It's a good idea, and I just thought I'd pass it on.
Joel
It's a good idea, and I just thought I'd pass it on.
Joel
Yes Joel, I think so. The problem is that the 2nd cap gets charged by the peak of the ripple. If a resistor was present, it would charge for longer though not to such a high value.
It's probably a valid thing for a boom-box, but not for HiFi.
BTW it'll only work on PP Class B properly, and half properly on class AB.
Cheeers
It's probably a valid thing for a boom-box, but not for HiFi.
BTW it'll only work on PP Class B properly, and half properly on class AB.
Cheeers
DIODE
Hi guys,
A bit off topic but Joels' diode triggered something else:
First of all John-san I agree with what you say about it and I had a bit of a deja vu as well.
However,as I notice people using all sorts of schemes hoping to prolong tubelife along with it (and often overdoing it in the process causing damage) I thought I'd mention this little trick:
A diode like the 1N4007 can be put to good use to prolong tubelife.
Just insert it after the last cap in the PSU in series with the B+ cathode pointing forward of course.
It won't do nothing but conduct when B+ is applied BUT when you shut down the amp it will isolate the HT rail from the circuit.
(Blocking off)
This prevents residual B+ on the tubes while the PSU slowly discharges,preventing cathode stripping in the process.
Kampai,
Hi guys,
A bit off topic but Joels' diode triggered something else:
First of all John-san I agree with what you say about it and I had a bit of a deja vu as well.
However,as I notice people using all sorts of schemes hoping to prolong tubelife along with it (and often overdoing it in the process causing damage) I thought I'd mention this little trick:
A diode like the 1N4007 can be put to good use to prolong tubelife.
Just insert it after the last cap in the PSU in series with the B+ cathode pointing forward of course.
It won't do nothing but conduct when B+ is applied BUT when you shut down the amp it will isolate the HT rail from the circuit.
(Blocking off)
This prevents residual B+ on the tubes while the PSU slowly discharges,preventing cathode stripping in the process.
Kampai,
KIDDIN' ME??
Hi John,
Unfortunately I haven't got them here.
It is however quite simple:
Just after the last reservoir cap wire a diode in series with B+.
When B+ present it conducts,when not it blocks.
The net result is that the tubes don't see the presence of B+ anymore as they would while the powersupply would be discharging.
At your service,
Hi John,
Unfortunately I haven't got them here.
It is however quite simple:
Just after the last reservoir cap wire a diode in series with B+.
When B+ present it conducts,when not it blocks.
The net result is that the tubes don't see the presence of B+ anymore as they would while the powersupply would be discharging.
At your service,
DIODE
Hello,
Immediately,yes.
A few seconds aren't going to hurt the tubes,it's going to take a bit more than that.
Surely the discharge path can't exist when the tubes are not conducting?
The point is not to bombard the cathode with repelled electrons by sudden loss of heaterpower and B+ still present.
I measured more then was and it works,no B+ present with the diode in place.
Naturally any charges left in coupling caps are going to take a bit longer to discharge but noticeable less than when these diodes aren't in place.
Cheers,
Hello,
Immediately,yes.
A few seconds aren't going to hurt the tubes,it's going to take a bit more than that.
Surely the discharge path can't exist when the tubes are not conducting?
The point is not to bombard the cathode with repelled electrons by sudden loss of heaterpower and B+ still present.
I measured more then was and it works,no B+ present with the diode in place.
Naturally any charges left in coupling caps are going to take a bit longer to discharge but noticeable less than when these diodes aren't in place.
Cheers,
John,
Well, I checked with my favorite electrical engineer, and he says a pi filter with a diode instead of a resistor will make a good filter. So we have cap-->diode-->cap
The voltage drop in a silicon diode is .6V, so the 2nd half of the filter will be able regulate 600mV worth of ripple. Now, I know that with any reasonable sized capacitor in the 1st slot, you would never have anywhere near that much ripple left over. Also, the time constant of such a filter would be extremely low.
I think the basic point is that if the supply is beefy enough, it doesn't matter how you charge those caps.
Seems it wasn't done much because, a) diodes used to be expensive, and b) usually the preamp stages weren't run at the same voltages as the output stage.
Well, I checked with my favorite electrical engineer, and he says a pi filter with a diode instead of a resistor will make a good filter. So we have cap-->diode-->cap
The voltage drop in a silicon diode is .6V, so the 2nd half of the filter will be able regulate 600mV worth of ripple. Now, I know that with any reasonable sized capacitor in the 1st slot, you would never have anywhere near that much ripple left over. Also, the time constant of such a filter would be extremely low.
I think the basic point is that if the supply is beefy enough, it doesn't matter how you charge those caps.
Seems it wasn't done much because, a) diodes used to be expensive, and b) usually the preamp stages weren't run at the same voltages as the output stage.
ANOTHER DIODE
Joel,
It may be worth a try.
Keep in mind though that engineers rarely look at things in an audio way.
To them it works or doesn't.
If you tell them you're using Polypropylenes or PIO caps in your PSU they already think your a mental case.
So be careful what you ask for and whom you ask,opinions will vary.
Regards,
Joel,
It may be worth a try.
Keep in mind though that engineers rarely look at things in an audio way.
To them it works or doesn't.
If you tell them you're using Polypropylenes or PIO caps in your PSU they already think your a mental case.
So be careful what you ask for and whom you ask,opinions will vary.
Regards,
Joel,
I will reiterate:
With C-D-C, the ripple will be really low at low current demand, but will be worse with high current demand. [The loud bits].
This is because at that time the diode will be "on", and the filter will just be C+C.
You'll only see a real "improvement" with class B.
Personally I don't like class B.
- - - -
Frank,
I don't believe your diode idea will work either.
I can see I'm making friends by the minute here!
Cheers,
I will reiterate:
With C-D-C, the ripple will be really low at low current demand, but will be worse with high current demand. [The loud bits].
This is because at that time the diode will be "on", and the filter will just be C+C.
You'll only see a real "improvement" with class B.
Personally I don't like class B.
- - - -
Frank,
I don't believe your diode idea will work either.
I can see I'm making friends by the minute here!
Cheers,
Guys, I have to go along with dhaen on this one.
Joel, I ran a sim on the circuit as described and saw no improvement to ripple with a diode unless the current is VERY low.
Frank, I don't see how what you describe will do anything other than cause a .6 volt drop. Where does the reverse voltage to turn the diode off come from?
Later
Bruce
Joel, I ran a sim on the circuit as described and saw no improvement to ripple with a diode unless the current is VERY low.
Frank, I don't see how what you describe will do anything other than cause a .6 volt drop. Where does the reverse voltage to turn the diode off come from?
Later
Bruce
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.