Power supply testing q's...

[GibsonGM]

Good tips, thank you!  There is a place for a LPF at the output I've already found, the output cap is on a daughter board.  Decreasing bass output will probably help a lot!

[GibsonGM]

Ok, the replacement cap really helped.  There is still a "mmmmmm" very bass-end hum present in the noise floor, but it's quite low.  Not as nice as my 'real' amps, but live-able I guess.   

I compared this using your 'basic run of the mill 12VDC power supply' - which was VERY much noisier (same humm, like a 'stuck welding rod'), so my cap is doing what it should.   <I wonder why the humm sounds like a short...hmm>

What i'm thinking now tho is...how to filter this even BETTER?    Is it 'legal' to break the circuit somewhere, add an R and C to clean it up even more?  If so, what values are 'appropriate'?  Would probably need a low R given that the thing has to draw current to drive the speaker...this is now kind of an academic question lol.

[antonis]

Is the speaker directly driven from JFET Drain, Mike..??

[anotherjim]

Power amp and opamp IC's can have pretty decent power hum rejection and you're right, there isn't much "R" you can add before the "C" to improve ripple smoothing. Usually, there's enough series resistance in the PSU secondary - rectifier - cable to help with a large capacitor albeit it's probably <20R on the round trip.

The FET preamp, on the other hand, can have RC power filtering all of its own. You can pick 1/10th of the drain resistor value for the series dropper then there's no significant voltage drop and add a cap big enough to hit the ripple. So if it's a 1k dropper then 10uF is big enough down to 15hz, but 100uF almost hits DC.

Giving a discrete front end its own power filter is Guitar Amplifiers 101. They don't have their own hum rejection and what hum they do get, they will amplify and send on to the power amp.

[GibsonGM]

No Antonis, FET is a preamp for the PC amplifier module, which is a chip amp > speaker.    The humm is the same whether I input to the FET first, or just the amp module, which still has its independent line in.  The hum is present with nothing plugged in, as well.  It's power supply junk.

The FET preamp has a 100u cap to ground where the power comes in, Jim...no R.  It's a Tillman made w/2N5457, so has 10k drain R.  I can put a 1k there, yes, to lower the cutoff...thanks.    We'll see...the "mmmmmm" (albeit very quiet now) may be something else I did by pushing the thing to its extreme, LOL.  I keep guessing that I messed up an on-board filter cap (as I did in the PSU), but can't detect that manually (each seems to have a rising resistance up into megs when tested in circuit).  And it will become not worth it pretty soon to keep mucking around in there.

This may be as good as it gets!   Neat project, it doesn't sound half bad, good for bedroom level stuff.  Worse comes to worse I'll use noise reduction to take out the hum, ha ha.  I've played many amps with worse noise floor.

[Rob Strand]

This may be as good as it gets!   Neat project, it doesn't sound half bad, good for bedroom level stuff.  Worse comes to worse I'll use noise reduction to take out the hum, ha ha.  I've played many amps with worse noise floor.

Most of the hum should be fixed with the filter on the preamp like you have.   You can still get hum from the power amp as those chips have a reasonable amount of gain.  Some hum can be reduced by adding "bypass" caps to specific pins of the power amp (see the LM386 datasheet).   Other times the hum comes from the PCB layout.   One tricky cause is some power amps IC have a + and - input.  If you are using the + input to input audio there is a "good" place to route the signals to the - input which will prevent hum/buzz.   Many designs just connect the - input to the ground near the device but you can find cases which it's not the ideal thing to do.  There was a thread in the last year  which had this issue (LM386 or LM380 can't remember now).

[GibsonGM]

Cool.  Damn chip is covered with a solder on heat sink, so I can't get the ID off it.  At least, right now I can't   LOL  That may change...so no clue what it is right now.

[Rob Strand]

Cool.  Damn chip is covered with a solder on heat sink, so I can't get the ID off it.  At least, right now I can't   LOL  That may change...so no clue what it is right now.

Ha, yeah sometimes we should draw a line and be happy with what we have 

[GibsonGM]

It's not bad; given what the thing is, I should be happy, ha ha. Sounds killer with a Dist + up front, and of course that does away with the 'hmmm'      I'm spoiled, my 18 Watt came out nearly noise-free, I shouldn't expect all of them to be that quiet!