Bleed current?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I did a search and found references to small bleeder resistors across PSU caps to discharge them at power off. Has anyone experimented with bleeding a much larger current - like 5 to 10 times the idle current of the chips? This would have some negative side effects with relation to peak current available and possibly PSU hum, but I wonder if it may provide some benefit in relation to damping PSU resonances and modulation of rail voltage with the music signal. Any thoughts?
 
Hi,
most bleeders schematics show the resistor connected all the time, not just at power off (no relay).
Your scheme will increase the size of the whole PSU; fuse, transformer, rectifier, capacitor and finally volume & weight.
When the opamps ask for more current the PSU already has a demand and the total current is still going to modulate the PSU output. Your gain will be reduced percentage modulation.
I think this is why over-rated PSU's are reputed to work well (without the bleeder). And I agree with this philosophy.
How about a shunt regulator instead of a resistor?
If you make it really fast the modulation at any audio related frequency will be negligible. Followed by lo inductance caps to reduce vhf hash.
 
Why was this topic moved??? It isn't a general power supply question. My query is specific to chip amps. I want responses from chip amp forum readers who may not read this forum or may not now realise that I'm referring to chip amps because it has been removed from the context of the chip amp forum. If I'd wanted this in the PSU forum, I'd have posted here in the first place. Please move it back to the chip amp forum where it belongs.
 
Ex-Moderator
Joined 2002
Why should psu design for chip amps be any different to psu design for other Solid state amps? I moved it here because I thought you would get a better quality of reply. But if you only want replies from one selection of our community, rather than a wide range of members then I'll move it back.
 
pinkmouse said:
Why should psu design for chip amps be any different to psu design for other Solid state amps?

Because the idle current in a chip amp is very low and most other SS development here revolves around class A. Mainly, I want to hear from people who may have actually tried it with a chip amp.

Thank you for moving it back.
 
I can understand why put some constant loading on a regulator, as most of them have better specs that way (including lower output impedance).
I used 91R between Adj. and Vout on the LM338 PSUs and everything got reassuringly warm:D : the trafo, the regs, the LM3886 chips.

But on an unregulated PSU I see no advantage in going to low value resistors... the (unregulated) PSU voltage will always be moduladed by the current demand of the amp.
I use 2.2k bleeder resistors to bleed the big caps at power off, just that. The caps last longer, and the amp shuts up faster.

Now this thread got me thinking...:scratch:
Hey, a how about for a bleeder a multi-turn trimmer paralleled with a big resistor, and tuned by ear with the amp playin'?:p
 
carlosfm said:
But on an unregulated PSU I see no advantage in going to low value resistors... the (unregulated) PSU voltage will always be moduladed by the current demand of the amp.

Yes, but I do think the percentage of modulation would be reduced. With the oversized transformers that some use the chips present almost no load at idle. I recall seeing a larger minimum magnetising current specification on some of the big trafos.

I think there could be some resonance damping benefits, too.

carlosfm said:
Hey, a how about for a bleeder a multi-turn trimmer paralleled with a big resistor, and tuned by ear with the amp playin'?:p

I've got a couple of 100ohm 50W Dale WW that I can easily try. I'll report back. No trimmers though.
 
jeff mai said:
I've got a couple of 100ohm 50W Dale WW that I can easily try. I'll report back.

OK, here's the data all measured off the negative rail (easiest to measure in my amp) using the min-max-avg function on my multimeter while the amp was playing a pop song at a moderate-loud level:

No bleeder -
Max: -24.68
Min: -25.44
Avg: -25.08

With 100ohm bleeder -
Max: -24.08
Min: -24.60
Avg: -24.31

As you can see, there is an improvement in the modulation of the rail voltage. It's less than I expected, though. Not a very scientific test, I admit, as it was subject to mains line variation during the test.

Listening tests will take longer.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.