DC at output again

[Mattyo5]

Ok, i'm doing a 6 chip lm3886 BPA-200. I've got 3 chips in parallel +3 chips parallel. On 3 of the chips, i get 1.9V at output. On the other 3, i get 25V at output and i'm running 27V rails. I used the same exact components as the BPA-200 suggests, except .15ohm 1% output resistors. I have not tied the output resistors together so essentially i've still got independent channels. On page 12 of the BPA-200 schematic, it shows the connections, except PS, and mute pin...i tied the mute pin to -V rail w/ a 33k resistor, pin 4-8, pin 3 tied to 9 w/ feedback, no obvious shorts from rails to output. I have pins 1 and 5 tied together V+ and pin 10 is + and pin 9 is - .

Oscillation? Since all 3 chips act the same way, i must have wired something wrong, but all the connections look good, and i doubt i'd have 3 identical bad solder joints. At least i'm consistent! It still happens with the input grounded. hmmmm

-Matthew K. Olson

http://www.national.com/appinfo/audi...ation_Note.pdf

http://www.national.com/ds/LM/LM3886.pdf

yes i did search

[officeboy]

Did you use the basic, or the detailed plans (servo or no servo)?

Also you might want to review http://www.national.com/an/AN/AN-1192.pdf as it is the updated bpa200 document.
*edit*
reading a bit closer I see you used page 12, the basic plan.
It's hard to say what could be wrong, maybe you could show us a few pictures of what you have put together? something where we could follow the traces to see if we can spot a problem.

[Mattyo5]

officeboy, i emailed you ...don't know if you received it. I shorted the inputs ...which i could have sworn i already did. the dc stopped...now i've got <5mV dc on each of the chips.... so now the question is, do I just keep going with the project? when i hook everything up, i don't want oscillation obviously? so does that mean i have to use shielded cable for each of the 6 inputs? dang.

-Matthew K. Olson

[officeboy]

I never received your email.

As to shorting the inputs, You need to make sure that your amp has a reference to ground. That will be taken care of when you install the Rin resistor.

I would keep going, 5mv is certainly fine.

I don't see a reason to use shielded cable. It's always nice, esp with longer runs but these will probably be less then a few inches.

Btw, you are using U5, or some type of preamp right?

[Mattyo5]

email me

mattyo5@aol.com

thanks


uh, i was thinking about that reference to gnd as well. I need to take a 20k ? resistor and put it to gnd from input and see what happens ...see if the dc offset is still the same. i will be using u5 yes, but firs i watned to check that all individual chips were working hehe. please email me.. sounds like you've got some good ideas... don't know why the diyaudio email never got to ya... hmm

-Matthew K. Olson

[Mattyo5]

Ok, here's an update. If I short inputs to gnd everything is fine. So, I soldered 100k resistors from each chip input to ground. The 3 chips that were giving me 25v on the outputs are now fine...with 7mV on output :-D the other 3 chips however are still giving me 1.4v on output. Now, i noticed that if I added a 10k from input to gnd in addition to the 100k's....i get less dc on the output. down to 600mV or so. If I keep with the 100k's, and leave off the 10k, and just hook it up to the input buffer...everything should be fine right?

-Matthew K. Olson

[officeboy]

You really should just stick to the schematic.
Build it like is shown and have faith that if you build it carefully it will work.

[Mattyo5]

I am sticking to the schematic, the only change is adding the 100k's to gnd as a reference for the inputs, but without the inputs shorted i'm still getting dc, the chips are essentially individualized still, except for the inputs being tied together... so anyway, when i reach critical mass for parts, i'll throw it all together, add the buffered input and have faith that no dc will come from the chips until then, i strive to get it all correct on the bench...thanks for the help

-Matthew K. Olson

[Mattyo5]

one other update, the side of the amp that is still giving me trouble is the side where the input coupling capacitors are actually attached to input, the lower 2 chips on the bpa-200 schematic page 12. ... i removed the input coupling caps, and got less dc, from 1.7v down to 1.3v, but still dc at output, next step, toss it all together w/ input buffer and see if dc is still there

-Matthew K. Olson

http://www.national.com/appinfo/audi...ation_Note.pdf