Is this circuit a good idea ?
The reason is that I have 2 pcb's with 6 caps each and I can use 2 of them in parallel to feed a channel each. I do not see any earth loops, especially if I connect BOTH the OUTPUT zero connectors together.
What do you think ?
The reason is that I have 2 pcb's with 6 caps each and I can use 2 of them in parallel to feed a channel each. I do not see any earth loops, especially if I connect BOTH the OUTPUT zero connectors together.
What do you think ?
Connecting capacitors are adding up the total capacitance without any problems.
But do avoid the ground loop on your drawing. There are good configurations for this setup.
But do avoid the ground loop on your drawing. There are good configurations for this setup.
Just use ONE set of V-, V+ and 0 connections to whatever you power, and you have no ground loops.
Real question is will it be beneficial.
You haven't said what you want to accomplish, and what this is for.
Jan
Real question is will it be beneficial.
You haven't said what you want to accomplish, and what this is for.
Jan
Paralleling of caps is right because of two desired effects: you add capacitance and divide ESR and ESL. But it is still better to parallel different kinds and values of capacitors: the cap vale still adds but (exactly as paralleled resistances), the final value is always less than the lower of them.
But why is that better? Why is doubling of capacitance better? Why not halving them, at least that makes the price only half?
Jan
Jan
The negative here is you increase the inrush current, and the time of that inrush current, on power up. The rectifiers and primary fuse should have a peak rating to accommodate this. Also, you should have smaller value caps right at the amplifier. The LM3886 circuits I've seen have 25uF with short traces to the chips power terminals.
That is correct, even though I have a 1000VA toroidal, I also do have a timer/thermistor based inrush limiter already to handle that. I have 3x10 ohm 1000 joule thermistors in series easily capable of absorbing the current spike, which will get shorted out by a timer relay after 3 seconds: https://www.macromatic.com/entity_reference_dialog_formatter/commerce_product/48/quick_view?v=4436
I just wanted to know if anyone sees an issue with any ground loops.
Effectively, with my proposed use, the circuit looks like below. The left channel serving capacitors are one board and the right channel serving capacitors are on the other.
My only issue is whether any mysterious ground loops would end up being created or not.
Thanks for your responses.
Last edited:
You lower ESR and ESL and by using different types and values you potentially make it more linear.
I know but the current is more likely to be drawn from the closest capacitor first, albeit differences will be extremely small. Basically those boards I have available makes this geometric configuration very easy & convenient in my chassis.
All I wanted to know is if a hum introducing ground loop can be observed since the 0 volts is not going to be a perfect star configuration any more.
I suspect, it would be very, very, small difference. There should be very minute resistance in the traces.
If you are mounting the boards back to back, like that, make sure the cap polarity gets changed on one board.
I don't have the experience to comment on the possible hum. Sorry.
If you are mounting the boards back to back, like that, make sure the cap polarity gets changed on one board.
I don't have the experience to comment on the possible hum. Sorry.
Can you elaborate on this 'halving', technically spoken?
(or pm me, in native tongue if prefered)
Halving is the reverse of doubling. What is there to elaborate?
The reason for my original question was to see if people would be willing to actually think about it.
Why is it better to doubling the capacitance? Why is it not better to halve the capacity? At the very least, you'd save half your money.
Jan
The reason for my original question was to see if people would be willing to actually think about it.
Why is it better to doubling the capacitance? Why is it not better to halve the capacity? At the very least, you'd save half your money.
Jan
Last edited:
Ok, how about this configuration.....
I have 2 other boards like the following:
https://jimsaudio.com/super-heavy-duty-crc-clc-power-supply-pcb/
I very much like the fact that the rectifiers are on board.
Due to very limited height that I have, I have to limit the capacitance on the board and therefore to increase the capacitance, I would like to use TWO of those boards, one per channel, but fed from ONE transformer. Here is the typical Nelson Pass two bridge power supply circuit.
Can I use TWO of these circuits but fed from ONE transformer with dual secondary windings ?
The ground lifter thermistor connected point would be the only common output, nothing else would be.
Any problems ?
Thanks
I have 2 other boards like the following:
https://jimsaudio.com/super-heavy-duty-crc-clc-power-supply-pcb/
I very much like the fact that the rectifiers are on board.
Due to very limited height that I have, I have to limit the capacitance on the board and therefore to increase the capacitance, I would like to use TWO of those boards, one per channel, but fed from ONE transformer. Here is the typical Nelson Pass two bridge power supply circuit.
Can I use TWO of these circuits but fed from ONE transformer with dual secondary windings ?
The ground lifter thermistor connected point would be the only common output, nothing else would be.
Any problems ?
Thanks