Built a basic lm3886 amp as per datasheet on veroboard. It has worked well for years and then decided yesterday to remove it for some cleaning. Upon re-installation I applied the lamp testby precaution and everything went fine without speakers. Upon connecting a crap test speaker the lamp lights up continuously and there is loud noise sort of loud hum coming from the speaker. After some seconds the amp switches off so does the lamp also. Can anyone help to solve the problem?
if you have a digitalmm on hand, try to measure the dc-voltage at the output of the amp without a speaker.........
(perhaps you lost a psu-connection -
or shorted some wires)
(perhaps you lost a psu-connection -
or shorted some wires)
Found the culprit. I's me. sorry I tested the amp without Volume control, the reason for loud noise upon switching. Just connected it and everything is ok.
One more question: There is some difference between L and R dc offset V. Left 1.5mV and Right 2.8V. Is it normal and acceptable> will it affect musicality?
One more question: There is some difference between L and R dc offset V. Left 1.5mV and Right 2.8V. Is it normal and acceptable> will it affect musicality?
I hope you mean 1.5mV and 2.8mV which are very low and would be no problem at all
2.8V however is far too high and would indicate a fault somewhere
Brian
2.8V however is far too high and would indicate a fault somewhere
Brian
Sorry it is drawing error after a beer or two. The 22k on the right goes to pin 9 together with one end of R680.
Got another question. My toroidal transformer is a 28 0 28 vac and is delivering 38 o 38 vdc and the amp is driving a4ohms troels nomex 164 speakers. Should I reduce the gain to avoid distorsion and overpowering since the recommendations on the TI datasheet are for using 28 0 28 +- dc for 4 ohms?
No, you should NOT exceed the chip's rated voltage. You should not even come close. Especially at low load Z. Chip-amps really do not like excess voltage. They die. (I have a mini-graveyard.)
The user can always find enough more gain to overcome any "lame gain" game you play.
Also the LM3886 has a lower limit on stable gain which is not much less than we usually run. (You could pad it down but that goes against some people's beliefs.)
The limit acc to datasheet is 84 V with input signal, divided by 2 comes to 42 V +-. The speakers go 4 ohms only between 200-350 hz. Is there a real danger since the limit voltage is not exceeded?
The 22 kΩ to ground provides a DC path to ground for the non-inverting input on the LM3886. The output of the LM3886 provides the DC path for the inverting input. You're good no changes needed.
Assuming you have proper decoupling and a mute cap in place there's no reason the amp should misbehave on start-up. Are you sure that 22 kΩ on the input goes to ground?
As far as I can tell, yes.
Tom
No danger. You just get worse performance. The built-in protection circuits in the LM3886 kick in pretty early with a 4 Ω load when you exceed about ±30-32 V. You'll get the full 65-70 W into 4 Ω at ±30 V with proper heat sinking, but only about 30-35 W (AFAIR) at ±35 V.
Tom