mic amp type thing problems.

[kittenyberk]

Right,

I have a Rolls GCI404, which is a fairly handy toy, apart from the fact the microphone bit of it (which I use most) makes a bloody horrible noise when it's set loud enough to be useful.

I've poked about grounding it super well with no improvement, and swapped the opamps around (all BA4560 448 H00) now I'm out of ideas.

Schematics for the thing are in Here (warning, pdf)


Anyone have any ideas on how to make this thing not completely bloody useless?

Cheers
Alex.

[robertkd]

Do you use phantom power? (condenser mic)?

does the clip indicator light as well when it's noisy?

Robert

[kittenyberk]

Nope, no phantom power, probably worth mentioning I get the noise whether a mic is plugged in or not.

No clip indicator light i'm aware of.

[robertkd]

mmm Ok I miss understood the peak detector, and it's function is to modulate the ducking function, is it me or does that sound wrong "ducking function"

OK what doo you get if

no mick plugged in

gain turned up

sounds like what?

[kittenyberk]

sounds like

www.hairy-arse.com/rolls.wav


edit: just had a moment of clarity, both in mind, and amp. may have averted my own stupidity

[robertkd]

Oh it's the aliens, there trying to communicate with us

really sounds like the mic amp is oscillating, but why?

Is it possible to tag a 100uf or even a 220uf across the supply pins of U1A/U1B I think this is a 4580??

Anyway what I want to achieve is to decouple the supply for this opamp

[kittenyberk]

yay!

Seems I was kind of half a step ahead of you, just bodged together a wall wart to feed it, rather than computer PSU (Yeah, I know, switchmode is evil)

Still, sounds fine now

[robertkd]

mmm ok still I love supply decoupling, it makes thing kinda mm quite.

great to see it's working

Robert

[rickcr42]

Jan 18, 2009 13:12:40 GMT robertkd said:

mmm ok still I love supply decoupling, it makes thing kinda mm quite.

great to see it's working

Robert

when it is a computer power supply especially.Increasing the size of the electrolytics is usualy moot because they are more for ripple killers than noise reduction so unless you have low frequency abberations they are best left alone or you may in fact be going the wrong way in SQ (for instance most three terminal voltage regulators HATE large caps on the output ).

If you are powering an audio device using a computer power supply it is far batter to go with an inline ferrite element in combination with shunt caps for a dead-block noise reducer,an example of which can be seen here:

www.hagtech.com/images/hagusbschem.gif

I pretty much use this scheme on ALL things digital power feed but with difference of I add another leg of capacitance across the 100uf/1uf combo of 0.01 (or even a few stacked in parallel,must be low inductance type and also must be right at the power feed pin or don't bother since it won't be doing squat)

That it does or does not make a difference in all cases I have no idea since it is an automatic addition if there is any digital audio/computer powered audio involved in the design being the single most audible "Quick "N Dirty" mod/upgrade one can make (the other is BNC Jacks and S/PDIF transformers on the input of any device with a S/PDIF input),the rest being more of a "rip it out and replace it" type of upgrade such as analog stage replacements,power supply replacements and/or clock upgrades.

you can have great basic circuits in a device but it is in the details that make or break real world everyday use and that means proper in/out matchups and elimination of ALL external noise sources which being expensive in comparison to the basic chips/R/C means where a bugjet device will cut corners and luckily where it is easiest to make the needed changes

[robertkd]

Hi Rick, are you suggesting I use series inductors to help get rid of the HF krap usually found in digital gear?

I agree also ceramics well silver mica are also good for the typical hash noise found on digital PSU rails. in this case I guess the noise was from USB supply.

cheers
Robert