stereo chipamp channel separation

[tonitonitoni]

have built lm3886 based chipamp, on same pcb stereo.

thing is that when i unplug left channel ,signal from right channel input is heard at left channel output(and other way around),few db less than on right.

schematic is regular from datasheet ,100milion times checked and working in few other amps i make, only difference is it is on same pcb this time.

i am baffled and feel dumb ,but certain that i overlooked something very simple.
please help,ty

 

[OlegSh]

Hi Toni,

If you want to get help attach your PCB design pictures...

Regards,
Oleg

 

[tonitonitoni]

it is very bad pcb i made few years back and has placement errors but is correct schematic ,would it be better to put schematic?

 

[OlegSh]

So far the PCB design which may be faulty is unavailable for checking by the audience here...

 

[mrWagner]

Show something... How do you think anyone could help without knowing anything from your build?

 

[tonitonitoni]

here is all,please focus on this stereo problem ,not rest of errors. thank you

 

[OlegSh]

I may be mistaken but your R12 connects to both channels outputs on the PCB...

 

[tonitonitoni]

damn i copy when making schematic other half, and forgot to change same name of feedback node... i am sorry for being this dumb, thank you very much on help

 

[OlegSh]

one trace cut should solve it...

 

[tonitonitoni]

it did solved problem, thank you

 

[OlegSh]

I'm happy it helped!

Regards,
Oleg

 

[zanden30]

Is this scheme really working? The end transistors are only driven by 6k8 and 100uF. IMO this will only work with high frequencies.

 

[tonitonitoni]

for now left channel only works without problems, right has still unindentified problem with hum, but both produce normal,fullrange sound, why not, maybe later i post measurments of working channel if i have time.

i have few new pcbs and slightly altered schematics.new are mono and lot better than this. probably gonna sell few of those if tests/measures well.

 

[zanden30]

At max output, the C = 100uF must not loose more than circa 1V. If the Hfe of the 2SC5200 = 50, the Ib must be 1/50 of the Ie = I speaker. With max speaker current = 7 A, the Ib must be circa 140 ma. The C of 100 uF will be unloaded within less than 1 ms with 1 V. This implies that this amp will only work at full output power till 1.000 Hz.

Did you really try this and did it work with 30 Hz at full output power? I cannot believe so.

 

[tonitonitoni]

what is full output power? that is very relative speaking.

isnt hfe 50 more than enough for 100watts in this config? or i missing something ,if yes please give me some source where i can read/learn more.

will try to post at least some measurments tomorrow.

 

[zanden30]

Hfe ° 50 means: Ic = 50 x Ib. If Ic = Ie = I speaker = 7A max, then Ib should at least be 140 mA. And it is impossible to impose this current more than 1 ms through a 100 uF capacitor, unless diminishing its Q with the corresponding 1 V ==> This implies that you cannot output in full power low frequencies.

 

[tonitonitoni]

but if i want only 100w isnt that 5A max out, so i need only 100mA at ib? and thats "full power" i need . if rollof is happening will measure it later today,but i cant hear it with my speakers.

 

[tonitonitoni]

btw i use fairchild encapsulation which has hfe stated at 65.

 

[zanden30]

5A and Hfe 65 (with low signals!!!) does not matter so much. It means somewhere at 500 Hz cutoff.....
I am curious that it will work. You d better choose the Dr. Frost solution. Easy, stable and proven low distortion.

 

[tonitonitoni]

dr frost? what? ????
this works for real, on my bench speakers,speakers not good much below 50hz but they play as on other amplifier,what baffles me now how it works when shouldnt???? i am not that deaf i think i should hear 500hz cutoff.

on working channel noise is very low,cant hear it, also distortion i cant hear on my test song(hells bells). now i need to make measurment tool out of my soundcard to post measurments here when come home from work.