|
ifatfat Post subject: What Causes Opamps to Blow? Posted: Tue Oct
07, 2008 5:00 am Lieutenant Joined: Tue Oct 07, 2008
4:56 am Posts: 3 A big hi to everyone... I would like
to know what causes opamps to blow? If I have a non-inverting
amplifier, and I have the output constantly hitting the power rails,
would it cause the opamp to blow? Thanks in advance!
Top darcyrandall2004 Post subject: Posted: Tue Oct 07,
2008 10:50 pm Colonel Joined: Tue Feb 27, 2007 6:16
am Posts: 46 Hello, The opamp will amplify the difference
between its input terminals. The output can not exceed the power rails.
If the op amp amplifies the difference between its input terminals
and the output does reach the power rail voltage, it will not be destroyed
as a result of this. If your opamp is blowing up, I think you
will find that the output reaching the rail voltage is only a sympton
not a cause. Cheers _________________ Regards, Darcy
Randall, Perth, Western Australia Top ifatfat
Post subject: Posted: Tue Oct 07, 2008 11:15 pm Lieutenant
Joined: Tue Oct 07, 2008 4:56 am Posts: 3 Hi... Thanks
for the reply! I'm wondering what might cause the opamps to
blow? I'm not 100% sure... but I guess excessive current at the input
might blow the amps? Are there other stuff I should look out for?
Top IR Post subject: Posted: Tue Oct 07, 2008 11:56
pm Site Admin Joined: Mon Jun 27, 2005 2:02 pm
Posts: 373 Location: Germany An excessive output current can
cause the op-amp to blow, this can happen if the amplifier drives a
load which is too small. Top nubbage Post subject:
Posted: Wed Oct 08, 2008 4:18 am General Joined:
Fri Feb 17, 2006 12:07 pm Posts: 218 Location: London UK Hi
I agree with both diagnoses and add that the differential voltage
at the inputs has a spec that should not be exceeded, and I have had
trouble like this in the past due to excessive voltage spikes on one
of the two inputs that resulted in this spec being exceeded for a few
microseconds. This led to unpredictable consequences within the structure,
basically "punch through" that resulted in thermal run-away and permanent
unrecoverable damage. Put a good scope on all interface pins and take
a look. Top yendori Post subject: Posted: Wed
Oct 08, 2008 10:19 am General Joined: Thu Sep 25,
2003 1:19 am Posts: 50 Location: texarcana What IR said.
Top ifatfat Post subject: Posted: Thu Oct 09, 2008
8:08 pm Lieutenant Joined: Tue Oct 07, 2008 4:56 am
Posts: 3 Thanks for all the replies! The output of the op
amp circuit is going into a 0.01uF cap followed by a 50ohm resistor
in series. How do I determine if this 50ohm is too small a load for
the op amp to drive? In the schematic design, the resistor to
ground at the negative input is absent, so there's only a resistor in
the feeback loop (from output to negative input). Would this design
cause a problem? nubbage, currently there's a resistor to ground
that's parallel to the positive input. Would adding a cap in parallel
to this resistor (to ground too) help out with the sudden voltage spikes?
Top IR Post subject: Posted: Fri Oct 10, 2008 4:31
pm Site Admin Joined: Mon Jun 27, 2005 2:02 pm
Posts: 373 Location: Germany From your description, it looks
like you are using your op-amp as buffer. The absence of the
resistor to ground should not cause a problem. This resistor is
used to set the gain of the op-amp in non-inverting configuration.
You can determine if the load is too low by simply finding out
from the data sheet the maximal output current that the op-amp is capable
to source/sink. By dividing the output voltage to 50-ohm, you will see
if the current exceeds the value mentioned in the data sheet.
Top nubbage Post subject: Posted: Mon Oct 13, 2008
4:58 am General Joined: Fri Feb 17, 2006 12:07 pm
Posts: 218 Location: London UK Hi It is difficult to advise
regarding spike suppression without knowing the exact application. A
spike has a frequency spectrum, parts of which may overlap the bandwidth
of your input signal. In such circumstances it is possible to low-pass
filter the input using a small passive element LC filter, so that the
spike spectrum and the wanted signal are frequency-separated. Many circuits
I have seen also use a degree of capacitive shunt across the gain-controlling
feedback resistor acting as a kind of integrator, an integrator being
a kind of low-pass filter. Putting a good high frequency hi impedance
oscilloscope across the input, the inverting input, the power rail,
and the output should reveal if spikes are the problem. Top
Philip_WNL Post subject: Posted: Fri Oct 31, 2008 1:29 am
Lieutenant Joined: Thu Oct 30, 2008 2:16 am Posts:
2 Hi ifatfat, some few comments: - indeed the
opamp amplifies the difference, but the gain is made so high...that,
in open loop, this statement doesnt translate into useful/measurable
info(your output usually stick to the supply). The Opamp is considered
most of the time in closed loop configuration (as it is the case here
with the non-inverting configuration). -Therefore in closed
loop ONLY, try to find the right common mode voltage by puting a small
AC signal (few millivolts) over a DC voltage -as an input to your non-inverting
buffer- that you shall vary starting from vdd/2 until you see a convenient
output. - an opamp to the contrary to operational transconductance
amplifier (OTA) have an output buffer to drive small load. How small
? this is what you should investigate. try with 1 kOhm first. to be
in the safe side, then bit by bit decrease the resistance and augment
the amplitude of your AC signal to see how much you can get of your
opamp. Also pay attention to the unity gain bandwith that will
tell you the maximum frequency that your overall buffer can sustain.
-finally look carefully into the datasheet to check what type
of load your opamp can accomodate: for instance the output capacitor
may not be suitable for opamp stability reasons.
Posted 11/12/2012
|