Shunt capacitance in SMPS to chip amp power connections

[Bruce Morgen]

A couple of questions regarding the topic:

1. There seems to be a very wide range in the values of shunt filter/reservoir capacitors found in TPA3255 chip amps -- e.g. 1000 mF in the Aiyima A07, 1500 mF on the TI eval. board, and 2200 mF in the Aiyima A07 Max. Does more capacitance have any particular advantage in that application considering the higher ripple frequency of an SMPS vs. a traditional linear power supply?

2. Is there any advantage to having this capacitance on the amp PCB vs. at the other end of the cable -- IOW, across the DC output of the connected SMPS at the barrier strip to avoid opening up the amp and either replacing the factory part or adding a second capacitor in parallel?

 

[Doodski]

1. There seems to be a very wide range in the values of shunt filter/reservoir capacitors found in TPA3255 chip amps -- e.g. 1000 mF in the Aiyima A07, 1500 mF on the TI eval. board, and 2200 mF in the Aiyima A07 Max. Does more capacitance have any particular advantage in that application considering the higher ripple frequency of an SMPS vs. a traditional linear power supply?

The frequency of the switching in the PWM power supply can vary from tens of kHz to several 100's of kHz+ so the capacitance for smoothing can vary too.

 

[DVDdoug]

1. There seems to be a very wide range in the values of shunt filter/reservoir capacitors found in TPA3255 chip amps -- e.g. 1000 mF in the Aiyima A07, 1500 mF on the TI eval. board, and 2200 mF in the Aiyima A07 Max. Does more capacitance have any particular advantage in that application considering the higher ripple frequency of an SMPS vs. a traditional linear power supply?

It's less critical at higher frequencies. Higher frequencies are easier to filter-out, and its usually above the audio range so you don't hear it if it happens to leak into the audio. But a larger value could help if the power supply has limited current capability... The capacitor can supply a temporary surge of current.

2. Is there any advantage to having this capacitance on the amp PCB vs. at the other end of the cable -- IOW, across the DC output of the connected SMPS at the barrier strip to avoid opening up the amp and either replacing the factory part or adding a second capacitor in parallel?

It doesn't matter where it is physically. The circuit/schematic is the same. Usually there are smaller "bypass" capacitors on the PCB to prevent oscillations and those need to be close to the active circuit. Those are usually not electrolytic because electrolytic capacitors don't behave like capacitors at very-high frequencies.

 

[Bruce Morgen]

The frequency of the switching in the PWM power supply can vary from tens of kHz to several 100's of kHz+ so the capacitance for smoothing can vary too.

The amp designers don't know the internals of the connected SMPS -- the amps are sold with no power requirement other than voltage and current sourcing capabilities, so that doesn't really explain the big differences in capacitor values.

 

[Doodski]

The amp designers don't know the internals of the connected SMPS -- the amps are sold with no power requirement other than voltage and current sourcing capabilities, so that doesn't really explain the big differences in capacitor values.

So you are saying they don't know the internals of the switch mode power supply that have no smoothing caps and then place smoothing caps that are non-existent in the power supply at the amplifier PCB... Highly unlikely.

EDIT: I reread your comment several times and get what you meant now. Then there is no explanation as per why the additional caps are there and what is selected other than as additional smoothing and storage caps. No reason for the variations in capacitance other than storage for transients is all I can think of.

 

[Bruce Morgen]

So you are saying they don't know the internals of the switch mode power supply that have no smoothing caps and then place smoothing caps that are non-existent in the power supply at the amplifier PCB... Highly unlikely.

EDIT: I reread your comment several times and get what you meant now. Then there is no explanation as per why the additional caps are there and what is selected other than as additional smoothing and storage caps. No reason for the variations in capacitance other than storage for transients is all I can think of.

Thanks -- when I got the e-mail notification I had quite a "WTF?" moment.

 

[Doodski]

Thanks -- when I got the e-mail notification I had quite a "WTF?" moment.

Yes, the first time or two that I read your comment I was in a repeating loop of thought and didn't see what you meant but it's clear as day now. Thanks.