LM358 Test circuit

Thread Starter

Ford Prefect

Joined Jun 14, 2010
245
So far you've got three suggestions for a test circuit; all of them will do the job, and I can't think of anything simpler that will do a meaningful test. Are these three circuits inadequate in some way? And if so, what's wrong with them?
Nothing that I can see but I'm just asking if anyone has any other good ideas it would just be nice and informative to hear them and I'm very grateful to all who contribute.
 

Audioguru again

Joined Oct 21, 2019
6,673
The lousy old LM358 has many problems such as crossover distortion, noise and poor high frequency response. Maybe these problems should also be tested?
Instead of throwing away the worst ones, why not select a modern opamp that does not have these problems?

I designed an equalizer module for a new speaker. The equalization had many peaks and nulls that needed accurate levels and frequencies. Low noise and distortion were important. I designed and built a tester for all these things and had thousands of equalizers built with ON Semi TL074 quad opamps.
Only one opamp failed because it was mounted backwards. One module failed because an electrolytic capacitor was shorted.
None of the thousands of modules were returned. So the quad opamps were very reliable.
 
If the op amp isn't used as a comparitor, then measuring the differencef the + and - inputs should be close to 0.

The osc test sounds good because it can also test drive.

Gain and offset tests are two other ones. Basically the same circuit. Set one for a gain of 10. Use an LM10 as source of voltage (200 mV min) and 0V. You would measure offset voltage x10.
 

Thread Starter

Ford Prefect

Joined Jun 14, 2010
245
I want to thank everyone who contributed and helped me with a LM358 test circuit, particularly OBW0549
I built the circuit the other day that he presented in Post #3 above and despite having to use slightly different value components (because I didn't have the ones in his circuit), the circuit works very well.

The schematic is here:
Capture3.JPG

I made the PCB on a 64x56mm board with the traces only on the bottom as below:
Capture4.JPG

Board without showing the traces:
Capture1.JPG

Board with only the traces:
Capture2.JPG

The finished board (without an LM358):
IMG_20200802_100915.jpg

On testing I checked an LM358 which I knew worked without fault and the LED's flashed at first at the same time but I think because of very slight variations of component values the LED's flashed alternately at a rate of about once every second...
IMG_20200802_101735.jpg

I also had 2 LM358's that I suspected to be faulty and put them to one side and later tested.
This LM358 did not flash the LED's and was getting extremely hot!

IMG_20200802_101421.jpg

I had tried the other faulty LM358 and discovered that ONLY the right half of the op-amp was working, (the green LED was the only LED flashing).
IMG_20200802_101434.jpg

PS. I wired a 0.1uF ceramic decoupling capacitor at pins 4 and 8 of the IC socket on the underside of the board.
:)
 
Last edited:

Audioguru again

Joined Oct 21, 2019
6,673
I designed an audio equalizer product and had over 20,000 built, thoroughly tested and sold. Not a single one failed but one had a defective electrolytic capacitor and another had its TL074 quad opamp installed backwards.
Guess what, the opamps were not bought on ebay. They were genuine and were made and bought locally.
 

DarthVolta

Joined Jan 27, 2015
521
Well I pitty the 1st guys trying to calculate the outcome of an op-amp, without already knowing that you could set the ouput in your equations and work backwards for some of it too. But I suppose they could still say that they wanted a certain output, and work backwards......so I'd firgure it out too after a while.

What I certainly can't do yet is tackle a complex audio amp for instance. I could cheat, and skip ahead to that chapter and just use the equations if I had the values, but that's not my goal. I've had a glimpse of the k-factor or something, that determines if a feedback loop is more + or neg FB, I should find and learn that now.
 
Last edited:

Lo_volt

Joined Apr 3, 2014
316
One suggestion to any further test circuit builders: Avoid using the stamped sheet metal contacts. They tend to wear out after a few insert and extract cycles giving intermittent contact. This may interfere with proper testing.
1632922382974.png

I'd suggest using the machined pin style sockets or better yet, a zero insertion force socket that has a lever to clamp on the part under test.
1632922454665.png
Zif socket:
1632922544220.png
 

Thread Starter

Ford Prefect

Joined Jun 14, 2010
245
I'd suggest using the machined pin style sockets or better yet, a zero insertion force socket that has a lever to clamp on the part under test.
View attachment 249074
Yes of course, thanks for the suggestion this is a good idea
I did actually change the IC socket to a machined socket a couple of weeks after the build. It was only because I didn't have any machined IC sockets at the time of the build.
 
Top