LM3886 Bypassing Revisited (applies to other chip amps as well)

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I saw very little benefit going from 470 uF to 1000 uF. I wouldn't bother with more than that on the PCB. It depends on the inductance and resistance of the wiring from the main supply cap to the board, however.

The main supply reservoir caps need to be sized for the biggest acceptable ripple voltage under worst-case load conditions.

~Tom
 
If you plan to always use a regulated PSU. The cap choice is not critical. A regulated PSU acts more like a battery with the caps only supplying MF and HF. So far all the tests have been done with a regulated PSU. Because it is all so very critical and you have the tools, a good design including all the PSU components would give you added value.
 
The main caps also need to be spec'ed for low ESR as that translates directly to output impedance of the supply. The caps need to be able to handle the ripple current. Aside from that, it's not rocket science...

I think the most difficult part is to figure out how much ripple you're willing to accept at the output.

For tube amps, I've generally been pretty happy with less than 1 mV (and preferably less than 500 uV) of ripple/hum on the output. That's with no signal. For full signal, I'd prefer to have all ripple induced components below the THD floor, so, say 60~80 dB down. Now you can work your way back to the supply requirements given the PSRR specified in the data sheet.

~Tom
 
In my solid state amplifiers I expect hum to read less than 0.1mV in it's prototype, no chassis, stage.

When fully installed in it's chassis with all the paraphernalia that goes in to make it complete, I try to get as close to 0.1mV, occasionally for stereo I have to settle for 0.3mV of hum and noise.
 
That's in line with my expectations for solid state amps as well. For tube amps, I'm willing to cut them a little slack... :)

With my 87 dB (1 W, 1 m) efficient speakers, I find that 1 mV of hum is just audible from the listening position on a quiet night. Below 500 uV, I really have to stick my head right up against the speaker to hear the hum. That's where my <1 mV came from.

~Tom
 
Last edited:
Because of many other factors, I do a lot of critical analogue layout and the correct positioning of components, feedback loops etc. are critical, as well as decoupling caps (think of the power delivery system as a whole, its not just the main supply). For small signal surface mount devices are the way to go, small loop areas, shorter signal paths. Whether level of analogue you are laying out, the same attention to detail should be observed.
This is a good read:
https://www.analog.com/media/en/training-seminars/tutorials/MT-101.pdf

Other manufacturers also have relevant information.
 
Because of many other factors, I do a lot of critical analogue layout and the correct positioning of components, feedback loops etc. are critical, as well as decoupling caps (think of the power delivery system as a whole, its not just the main supply). For small signal surface mount devices are the way to go, small loop areas, shorter signal paths. Whether level of analogue you are laying out, the same attention to detail should be observed.
This is a good read:
https://www.analog.com/media/en/training-seminars/tutorials/MT-101.pdf

Other manufacturers also have relevant information.

In this article seem like the main use of HF bypass capacitor in audio amplifier is to short HF noise to ground but in many articles that i have read, people mention that those capacitor used to supply fast transition current for HF so i'm really confused.
 

PRR

Member
Joined 2003
Paid Member
You think it is an AUDIO amplifier but the amplifier does not know that. In fact all "audio" amps have gain far-far above the audio band.

If you take an "amplifier", with some capacitance in it, and run a wire out and back, it may become a radio frequency oscillator. The inch or feet around the power supply loop CAN act as a tuned circuit and encourage the amplifier to oscillate ultrasonically.

A *short* path around the supply circuit puts the natural frequency of the tuned circuit above the useful frequency of the amplifier. (Which is why we can be a little careless with fat tubes and have to be just-right for UHF tubes and high-strung GHz chips.)
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.