News:

SMF for DIYStompboxes.com!

Main Menu

LM358 dual op amp

Started by amz-fx, February 28, 2021, 11:08:42 PM

Previous topic - Next topic

amz-fx

I posted a brief article about the LM358 chip in my Lab Notebook:

http://www.muzique.com/lab/main.htm

Enjoy!

Best regards, Jack

PRR

Quote from: amz-fx on February 28, 2021, 11:08:42 PM
I posted a brief article about the LM358 chip in my Lab Notebook:.....

If the show starts in an hour, the 324/358 can be salvaged with a heavy pull-down. IIRC, I used 680 Ohms. This ate the battery in 4 hours but it was a 3 hour gig so we got through.

If the load is "light" then the pulldown does not have to be so brutal. And we do see these chips in audio with no pulldown (above 100k it is moot). But I was driving the last of the true 600-Ohm equalizers.
  • SUPPORTER

Rob Strand

#2
There were some official notes about the pull-down resistor.   There will be some minimum bias current which helps and after that it depends on the load - since you are operating the output as buffer transistor + emitter resistor.

Here's one, but it looks like a later note.  The one I was thinking of was old style,
https://www.ti.com/lit/an/sloa277/sloa277.pdf
section 3 and section 4, page 9
Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

anotherjim

The 3rd gotcha is the asymmetric output swing. It runs to negative supply but about 1.5v short of the positive supply. Having the "helper" resistor sourcing from the +Vcc lets the positive peaks reach the max before clipping and still smooth the crossover. Also, set Vref halfway in the output swing, which is less than 1/2Vcc if you want max before clipping.

The single supply chip used in the original MIJ Boss DS-1 is a lot like an LM328.


amz-fx

The noise spec on the LM358 would lead you to believe that it is more noisy than the tests indicate. This is probably because semiconductor manufacturing techniques have greatly improved in the last 45 years, but the datasheet has not been updated.

Even so, the 5532 has around 20db lower noise floor than the LM358 in the circuit that I was using.

I tested a lot of dual op amps while I had the test jig set up, and most of them are much more similar to each other than they are different.

Best regards, Jack

iainpunk

so, in my background project ''the worst overdrive ever'' i should use the LM358! i have a lot of them laying around.

cheers
friendly reminder: all holes are positive and have negative weight, despite not being there.

cheers

Rob Strand

QuoteI tested a lot of dual op amps while I had the test jig set up, and most of them are much more similar to each other than they are different.
Interesting result.

You would think the (actual) input bias current would be an indication of the noise voltage.   We have to assume the impedances are low enough that we can ignore noise currents.   The NE5532 gets it's low noise from a high bias current for sure.   One caveat is some arrangements of the transistor can double the noise voltage for a given bias current.    (There's also a caveat for modern opamps since some have input bias circuits which make the input bias current from outside the chip lower than the base current of the actual transistors.)
Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

PRR

> much more similar to each other

Transistor is transistor.

They may be different hFE and different Ic but over wide range this makes small difference.

In 1968 they may have been made on a "dirty" line, not good enough for military or computer chips, and specs written to cover a bad day in the factory.

PNP used to be very low hFE, and '741 (and LM358!) stack two at low current on each input. '358 sucks 30nA each input pin. Input hiss resistance dozens of kOhms. Even so this is not "huge" compared to many of our uses.
  • SUPPORTER