Remote PSU advice sought

[john blackburn]

I have an amp with LM1875s that has been screwed to a piece of wood for "testing" for nearly a year now. It uses the values from the Chipamp.com kit http://www.chipamp.com/lm1875.shtml but is done point to point in my case.

The time has come to fit it in an enclosure so I want to try a remote power supply with it. The cable between the amp case and PSU will be at most one and a half metres long.

Should I add some capacitors in the amp case, or will the 1500ufs on the supply be enough for a cable of that length? I had planned to use some 100nfs at the amp connector, but thought I would check your thoughts on whether I need to add any extra capacitance before going any further.

John

[twest820]

Depends on your ripple target and how those relate to your fidelity target, output powers, impedance of the supply cabling, any capacitance multiplication or regulation you might be using, and so on. Off the top of my head my bet would be the 1875's Vcc PSRR is adequate for an unregulated supply without cap mutipliers but Vee will be sketchy on the highs. Toss it into spice and let us know what you find out.

As a data point I'm currently working on a 15VA remote supply. So far it's looking like 10000uFs bypassed with 47uF and 100nF on the supply board and 47uFs and 100nFs on the amp board with cap multipliers. But the output devices are 75dB PSRR and I'm targeting a power supply contribution to the N in THD+N that's 10dB under the 1875's 0.1% nominal after allowing for supply coupling between the high and low channels in the biamp.

[Ted205]

i'm making something similar.

it makes sense to use some capacitance inside the amp, also if you use caps either side of the power umbilical chord you have a CRC filter
you could always use larger rail caps at the chip too.

i would start with out them and then experiment, find out what effect is has on sound and see what its like on the scope etc

[pacificblue]

Originally Posted by john blackburn

The cable between the amp case and PSU will be at most one and a half metres long.

Should I add some capacitors in the amp case,

Yes. Either that or move the capacitors to the amp case. Does the board you copied not come with capacitors for power supply decoupling (47µF + 100nF?).

[john blackburn]

pacificblue

The capacitors are there at the chip pins. I just wondered if it was advisable to add a little if using a long cable. I realise this is a very basic question and there probably isn't a yes/no answer to it.

They make for a tiny amp when done point to point.




Thanks to all for the input.

John

[Kinnja]

This was also briefly discussed in this thread:

External CPU for chipamp

No doubt I'll be trying this myself some day.

[pacificblue]

Originally Posted by john blackburn

I realise this is a very basic question and there probably isn't a yes/no answer to it.

There is a yes answer to it, but that is based on a more classical way to design a power supply.

Since yours is a Gainclone-based design, why not have a look at the original? There the amount of capacitance is similarly small as in your amplifier and all of it is right next to the IC to make up a bit for its limited size.

[john blackburn]

I will leave enough room in the new amp case to add capacitors if the need arrises and go with what I have got for now.

I am well used to the sound in its present form, if anything changes for the worse once it is boxed up I will take another look at it.

Once again, thanks for the advice to all.

John