From: Chuck Simmons
Organization: You jest.
X-Mailer: Mozilla 4.61 [en] (X11; U; Linux 2.0.33 i586)
Subject: Re: Problem with LM1877 audio amp I.C.?
Date: Mon, 23 Sep 2002 15:30:35 GMT
NNTP-Posting-Date: Mon, 23 Sep 2002 08:30:35 PDT
> I have been using this device (DIP version, LM1877N-9A) for years and
> occasionally have a problem which bugs me. I have just used it in a new design
> and the problem has become more prevalent.
> I'm using a single supply configuration, with a 47uF from the bias pin to
> ground and 1M resistors to the inputs, exactly as detailed in the National app.
> note. Here is the problem - every now and then, when the device is powered up,
> the output will go straight to Vcc and obviously no audio is forthcoming. A
> power reset will correc the fault. Some chips are worse than others - some
> never faulter, while some do it almost always and are throwaways.
> In my various designs, the supply has varied from about 5V to 12V. It's a good
> clean supply with an electrolytic and ceramic bypass. Sometimes regulated,
> sometimes not, but seems not to make a difference as far as I can recall. I
> vary the gain from about 10 to 20. Apart from that, the topolgy is the same,
> but some designs have never been problematic and others, very much so.
> A web search has brough up nothing and I'd be surprised if nobody else has had
> similar problems. I know it has been used in arcade games, so maybe somebody
> here has some experience?
I use these parts as servo power drivers both in voltage and current
feedback modes. Two things. I don't use high resistance values on the
inputs (mainly because the LM1877 has pretty high input bias current). I
consider 20K about the limit in DC applications of the part. The part
really isn't stable below a voltage gain of 10. I have not seen the part
latch at lower gains. It usually sings at about 220KHz or something.
Still, you should make sure the part is used in a voltage gain
configuration above 10. I use a single part as an H driver with one
section tricked into a gain of -1 while it "sees" itself to be at a gain
of -10. If you must use high resistance values on inputs, make sure you
have leakage paths under control. I consider these parts to be pretty
solid. I do find that some loads will require snubbers. The clue,
however, is the part oscillates. I have yet to see one latch.
... The times have been,
That, when the brains were out,
the man would die. ... Macbeth
Chuck Simmons firstname.lastname@example.org