Smoothing capacitors for LM3875 power supply

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I am rebuilding my LM3875 chipamp (new case, adding second transformer and ps board). Currently, the power supply board (based on CarlosFM unregulated PS schematic) has two 10uF 50V smoothing capacitors. Since I am rebuilding the amp, should I upgrade the caps to a higher value?

The setup is run by 21V-0-21V 80VA center tap trafo per channel. Speakers are 8Ohms with sensitivity of about 86db. I also have 6Ohm speakers with sensitivity of 94db but I know that LM3875 does not play well with less than 8Ohm speakers.

Thanks.
 
This is for the older Rev. C of BrianGT's (chipamp) LM3875 kit. Yes, there are only two 10uF caps. There are 1500uF 50V caps too but they are on the amp board. I can't find the schematic, but looking at the PCB it looks like it is simple full wave rectifier with those 10uF smoothing caps. The PS just changes AC to DC.

So if I were to replace those 10uF caps with bigger ones, what value should I go for?

Thanks.
 
The PS just changes AC to DC.

So if I were to replace those 10uF caps with bigger ones, what value should I go for?

Thanks.

The PSU does more than that , a psu has to also filter out the ripple from the AC , It also has to store enough energy to supply enough current for loud transients in the amp , and the capacitors together with the impedance of the power supply form a filter that affect the audio ..... Most poeple claim that the more capacitance in the PSU the better the low frequency responce of the amp ......

:)
 
Thanks for replies.

I will try 4700uF on each rail and see how the amp performs. I think I may have those in my parts bin.

As to the trafo, I have one 80VA for each chip. It should run just fine. Before, I have used only one trafo like that for both chips and never had any problems with it. Not everybody pushes their amps to the limits. :) My main rig is 2x16W Marantz receiver which has plenty of power for my type of listening.
 
With such a small VA transformer, there wouldn't be a lot of point going for too high a value, but I would go for at least 4700uF per rail.
I think the opposite.
When using a transformer that is a bit lower VA rating than normally recommended, I add extra smoothing capacitance to "beef up" the supply.
My intention here is to keep supplying transients and operating current from the smoothing, until the too small transformer has time to charge up the capacitors some time later when demand has dropped a bit.

The effect is that a low impedance of the smoothing is paralleled by the high impedance of the transformer.
Changing one can to some extent be compensated by moving the other in the opposite direction.
 
Last edited:
I can get the 10,000uF and put one per rail.

Now from the scientific point of view, I have two Nichicon caps that I am looking at:

1. VZ series with ripple current of 2850mA, 105degC
2. FW series with ripple current of 4000mA, 85degC

Does the ripple current make a difference here? All other basic specs are the same though the FW series has longer load life which I assume is because of the lower max operating temperature.

Thanks.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.