Sound troubles

[bamera]

First post here, I hope someone can help me with my project.

I did a search but couldn´t find an answer to my problem.

I´ve put together 3 different LM3886 application circuits from the datasheet.. same problem with all.

I put together the split supply schematic using a +/-18V 3A power supply.

All three circuits work perfectly... but present the same problem. After a few seconds of perfect and cristal clear playing, the audio peaks begin to distort and the audio output sort of stutters.

I find it hard to believe that this is a heatsinking issue because I´ve tried several and have settled on a massive block I´ve cooked up joining (with screws and thermal paste) three large heatsinks. I also used artic silver thermal paste just to be on the safe side. The heatsinking barely gets warm now.

I´m testing using a 200W 8ohm 12" speaker and a 60W 4ohm 8", both present the same problem. I substituted the C2 and C1 caps from ceramic to polyester and no change, the problem persists.

I´ve tried with another split 18V PSU (1A this time) and still the same problem.

The sound will not distort at low volumes but even at medium bedroom levels, the damn things start distorting. Although I know at 18V split I won´t get maximum power, even at 5W volume (roughly) The distortion starts.

I followed the schematic to the letter, 0.5mm indutor wire at the 10ohm resistor and all. All the circuits were hand soldered each in a singular way so I´m out of ideas.

I appolgize for the long post but I´d really appreciate some help here.

Thanks for your help, Daniel

[AndrewT]

Hi,
the delay could be temperature related and may also be oscillation causing that temperature increase.
5W average power from a 30W amplifier is a lot.
Most of the peaks will be clipped if you are using normal dynamic range music sourceS.
Highly compressed CD & MP3 may not sound as badly clipped.
It is normal to use an average power level of 1% of maximum to avoid most of the extreme peaks getting clipped.

[bamera]

Thanks for the feedback AndrewT... I´m dissapointed because I get more power out of the TDA2050 than the 3886 with the same power supply. I used MP3´s to test the output by the way.

Any tips on how to handle possible oscillations ?

Thanks again, Daniel

[h3ad]

18v +- is the lowest voltage that chip supports (according to the datasheet)

up the voltage and you should be right, because as most people know, an amp running below spec. voltage is going to distort very easily

haydn

[markiemrboo]

Originally posted by h3ad
18v +- is the lowest voltage that chip supports (according to the datasheet)

A +/- 9v split rail supply would be the typical minimum according to the datasheet, which is 18v in total, from rail to rail.

Not to be confused with a +/- 18v split rail supply, which would be 36v from rail to rail.



OP: Do you have a scope, or can you borrow one from friend / family? That would be a good way to check for oscillation.

Otherwise you could try the light bulb trick (A simple 50W light bulb in series with the mains live, before the transformer) to see if the amplifier is consuming more power than is considered normal whilst idling. When switching on the light bulb should flash brightly for a couple of seconds and then fade dim or possibly not lit at all. If it stays lit up brightly it could be oscillation.

[h3ad]

ah i didnt realise that... then it probably is oscilation then

regards haydn

[bamera]

Thanks for the replys guys!

The problem is solved, After giving up on the 3886, I tried the 1875 where I encountered very similar troubles. There´s a long post at dyistompboxes where the problem was solved....

turns out my transformer ground not hooked up to the circuit.

The TDA2050 didn´t present this problem and R.G explained as to why.

[markiemrboo]

Originally posted by bamera
Thanks for the replys guys!

The problem is solved, After giving up on the 3886, I tried the 1875 where I encountered very similar troubles. There´s a long post at dyistompboxes where the problem was solved....

turns out my transformer ground not hooked up to the circuit.

The TDA2050 didn´t present this problem and R.G explained as to why.

Glad you got it sorted, and thanks for following up with a solution

[h3ad]

interesting problem... oh well glad you fixed it

regards haydn