Cheap JFET input op-amp

I just put NJM 072BD and then LF353N in my PCM1798 DAC. They measure very differently. The LF353N (old National parts) performed amazingly well. The TL072 performs poorly in that application. I put them in the I/V and the filter locations. (Using the BB schematic.)

In order the graphs are NJM 072BD, LF353N, LF356M and LM4562NA.

These are single ended measurements at 6kHz (to make it harder than 1kHz) -3 dBFS. Later I will do balanced measurements with my E-MU 1616m.

Wow! Excellent. As far as i know lf353 has low noise, higher bandwidth and better drive capability compared to tl072 but I didn't expect this, the difference is very apparent. Also Lf356 performed really well. He he...:radar: :devily:
 
I was shocked. When LF353 & LF356 were initially mentioned in the thread I was tempted to post. However I am glad that I waited and remembered that I had some from 20 years ago in storage boxes. I dug those out for the measurements and they are surprisingly capable.

They seem perfectly reasonable to use as long as you are happy with the noise levels. The LM4562NA performs really well, as does the OPA1612A. For price to performance ratio some like the LF353 and NJM4580D are bargains. The LM4562NA is really a bargain too considering it has a "fairly" reasonable price and excellent performance. By the way I recommend staying away from NJM4562 as in my measurements it does not compare with LM4562NA. Also I suggest the NJM4580D over the NJM2068D.

I think genuine ones can be bought for reasonable prices from Electronic Components Distributor | JLCPCB & EasyEDA Parts Online store - LCSC.COM (with of course the shipping cost issue.)


Keep in mind that different results can be seen (in terms of performance ranking) depending on the application. The results with my CS4398 DAC output filter (as a test platform) are not the same as with the PCM1798 DAC I/V and output filter (as the test platform). For example, the OPA2134 measures very well with the CS4398 but not with the PCM1798.
 
Last edited:
The LF356 graph was the balanced measurement picked from the wrong folder. The correct (matching single ended) graph is attached.
 

Attachments

  • Old National LF356M 4x Adapters Right -3dBFS.png
    Old National LF356M 4x Adapters Right -3dBFS.png
    107.9 KB · Views: 281
Attached is some tabulated data from my PCM1798 DAC. This is still the single ended data. Later I will work on the balanced data.

It is all at 6kHz which makes it harder than 1kHz.

The first is at -3 dBFS. The second is at -10 dBFS.
 

Attachments

  • PCM1798 Opamp Comparison -3 dBFS 6kHz Sinewave.png
    PCM1798 Opamp Comparison -3 dBFS 6kHz Sinewave.png
    51.8 KB · Views: 406
  • PCM1798 Opamp Comparison -10 dBFS 6kHz Sinewave.png
    PCM1798 Opamp Comparison -10 dBFS 6kHz Sinewave.png
    35.2 KB · Views: 372
I actually don't know much about popcorn noise. I don't have many LM4562 and I don't detect even a hint of any pop-corn popping noise or really any noise when I use the ones I have with my PCM1798. I do like the LM4562NA, perhaps even more than OPA1612A.

How would it affect a typical DAC FFT measurement? Or does one need to look for the sudden up and down step transitions of bi-stable noise with a scope?

What other op-amps (modern, audio) have this issue?
 
Hello everyone, recently i purchased OPA2134PA & going to order few more. Now i've TL072A, TL082A, LF353N & OPA2134 jfet opamps in my possession. My question is do you know why OPA2132 is more expensive than OPA2134 ? According to their datasheet they'r almost identical.

Regards

20221215_205923.jpg