| ralfs |
Good afternoon,
I'm newby here and done some search already, so far didn't
find anywhere on this forum. Please share your experience and
help me.
I have soldered so-called Gainclone, just a dual LM3886
amplifier.
Problem is about 50mV AC saw on each power rail and some
pulses (very short ones, cannot draw, I'm sorry) in each output
regardless of load and power voltage - tried 23 and 33 V.
Saw form makes me think that something draws a lot of power
for a short moment and then the cap starts charging again.
Caps are 50V / 10000 mF, should be fine for this amps I think.
First schema I used can be seen there, the first
figure on the page. PCB is the same.
http://www.elektronet.gower.pl/a021.htm
NFB chain didn't seem right to me and I changed schema to more common one, like this example,
http://hepso.dna.fi/misc/LM3886_amplifier.html
This didn't solve the problem, however.
All bypass caps are in place, even some extra caps directly on
IC pins. Still cannot get it why should IC draw power to make
even small saw in idle and with no load. I not sure how
precise is my oscilloscope but the output pulse seems to be
a 200 Hz, this is 2 x 100 from each polarity, this makes me mad :(
as I cannot get it what's wrong.
I haven't solder a thing for some 15 years, did I become stupid?
Please help
Ralfs |
|
|
| ukram |
| LM3886 takes about 100 mA when idle, so that is the reason for the small ripple on the power line. It is not a problem, because the amp has good power supply rejection ratio. The output may have some high frequency noise, but it cannot be heard. |
|
|
| jackinnj |
| quote: | Originally posted by ukram
LM3886 takes about 100 mA when idle, so that is the reason for the small ripple on the power line. It is not a problem, because the amp has good power supply rejection ratio. The output may have some high frequency noise, but it cannot be heard. |
i only had "noise" and oscillation when I first started using the LM3875 and LM3886's -- the 0.7uH choke seems to have taken care of the problems. |
|
|
| theChris |
| make sure your using a star ground of some form. next, move the signal wire away from the power transformer. i had hum in my small gainclone during testing becuase the signal wire ran close to the transformer. moving the wire fixed the problem. |
|
|
| ralfs |
Hi there and thank you all for responses.
So far didn't have any luck to remove these spikes. I'm thinking
to get rid of PCB. This could be useful to others as well, if
you use AVT designed PCB (some amateur magazine from Poland)
for stereo LM3886 based amp. It had NC pins connected to
some other pins and I didn't pay attention at first
but when I disconnected all NC pins the pattern of output noise
changed. These pins are actually connected
inside the chip, I think it's mute control without a resistor.
I measured a bit more precisely and it appears to be the mains
hum and chips cannot reject something instead it is amplified.
Inputs shorted and this is not signal cable layout fault.
This is not a real photo, just a weak attempt to draw the problem.
I'll post another time if I solve it.
Bye, bye
Ralfs |
|
|
| ralfs |
Good morning here,
as promised I post a solution of the problem.
It all was an input grounding fault, I connected signal ground
directly from input sockets. Signal wires are shielded, indeed,
and go to pots and then to amp. The other side of the shields
was no connected to ground, I'm used to it because in balanced
setups you usually connect only one side of the shield, at
source. This was my big mistake :(
I disconnected extra grounding wire from input sockets and
signal ground is connected now through the signal wire shield.
Amp works perfectly.
Thank you all for helping me. If needed I can post some
visual materials here, schema and amp picture. Let me know
if someone is interested.
Best regards
Ralfs |
|
|
|
