From: Chris Carlen
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.0.1) Gecko/20020826
X-Accept-Language: en-us, en
Subject: Re: Hot Op-Amp!
Date: Wed, 23 Oct 2002 03:17:05 GMT
NNTP-Posting-Date: Tue, 22 Oct 2002 20:17:05 PDT
> I made a PCB that has four TI (Burr Brown, actually) OPA547 op amps. One of
> them is getting REALLY hot while the other three are not. They are all fed
> from the same power supply (+12/-12). They are all configured for
> non-inverting operation and the three cool ones are set for a gain of 2 while
> the hot one has a gain of 1. The test I ran was with no inputs on all four op
> The way I am getting this is by using two feedback resistors. R1 is from the
> inverting input to ground and R2 is from the output to the inverting input.
> The gain formula is Vo=Vin*(1+R2/R1). On the hot op amp, I am using a jumper
> wire for R1 and R2 (as was suggested from this newsgroups, or maybe
> sci.electronics.basics??) Maybe I need a resistor at R1 even though I want 0
> ohms for R2??? On the three that have a gain of two, I am using R1=R2=3.01k
> ohms. If I do need an R1 resistor, what value?? I'd like to use the 3.01 k
> just for commonality of parts.
> Also, I know that running an op amp where the output is at the midpoint of the
> power supply is the worst case for heat disipation. My inputs will not go less
> than ground, so I intend to change to a +12/-5 V power supply which should help
> in the heat problem...right??
> Any thoughts would be appreciated.
Hello there. I have used OPA547 quite a bit so maybe I can help a
little bit. You have noticed that three OPamps seem to be not running
too hot. Yes the fourth is cooking.
In cases like these I check first for a wiring error. Chances are
you've got the power connected to the wrong terminal of the hot chip. I
hate when that happens. Chances are high the chip is broken after that,
but sometimes they survive. Trouble is if you correct the wiring error
(if there is one), and the circuit still doesn't work, you don't know
why. So it's best to try a new chip if you are sure the power
connections are right, and it still doesn't work.
Oh, but I see you are trying to build a unity-gain follower with the
fourth amp. If you really have a jumper from the inverting input to
ground, AND a jumper from the inverting amp to the output, then that is
your problem. You have connected the output directly to ground. This
For the unity gain amp, you want the value of R1 to be infinity, which
really means *don't connect anything* from the inverting input to
ground. Then you can connect the output directly to the inverting input
with a jumper, or perhaps use a resistor equal to the value of the
resistor to the non-inverting input, if you have one. (These are
sometimes used, to cancel input bias current induced offsets, but not
really applicable to the unity gain follower). If your circuit is
basic, then you only need a wire from the output to the inverting input.
Once you stop shorting the output to ground, the amp should run at
normal temp. Having the output midway between the power supplies is not
a problem. The amps will get a little warm from their own quiescent
current, but nothing to get alarmed about. But if you short an output,
then you will get very hot.
Also, the OPA547 has a current limit resistor that you can attach to one
of the pins. Have you calculated an appropriate limit current for your
application and attached the correct resistor? The data sheet explains
this, but it takes some patience to figure it out.
Once you get the OPA547 working right, it is a very nice chip. I tried
LM675 once, oh what a pain in the butt. I like the OPA547 and OPA548.
Christopher R. Carlen
Suse 7.3 Linux 2.4.10