That's interesting, because in my little opamps rolling experiment it did not do very well. I listened to OPA2134, LM4562, and NE5532, and OPA2134 sounded the best. Load was 1k but I'm planning on changing that to something easier like 2k and retesting NE5532 just for fun.
I'm still waiting for OPA1656 in the mail. But once it comes, I'm excited to try to listen to that.
I wonder how the sound of CMOS would compare to the JFET? Since jfets are said to sound warm and tube-like...
Last edited:
Got OPA1656 delivered few days ago. Today had a chance to swap it into my amp and give it a listen.
In my opinion it sounds the best at close to max volume.
Sound is cleaner comparing to OPA2134. But the warmth in mid frequency (1k-4k) is gone.
Great for piano/violin or metal / electric guitar. I did not like it too much when listening to edm / electronic music.
With this IC I can instantly tell the difference between good mastering/mixing tracks vs not so good and muddy.
In my opinion it sounds the best at close to max volume.
Sound is cleaner comparing to OPA2134. But the warmth in mid frequency (1k-4k) is gone.
Great for piano/violin or metal / electric guitar. I did not like it too much when listening to edm / electronic music.
With this IC I can instantly tell the difference between good mastering/mixing tracks vs not so good and muddy.
Does anybody know the max differential input voltage of the OPA1656? The common mode input can go to the supplies, but funny enough, the data sheet doesn't mention the max diff input voltage.
Being a FET input I think it would be the supply voltage, but the fact that they don´t mention it makes me suspicious.
Jan
Being a FET input I think it would be the supply voltage, but the fact that they don´t mention it makes me suspicious.
Jan
Take a look at the functional block diagram in section 7.2 of the datasheet and figure 43. There seems to be ESD protection diodes connected between the input terminals, so the differential input should be close to 0.
Unlike e.g. OPA2191, where the differential input can be anything within the supply rails.
Unlike e.g. OPA2191, where the differential input can be anything within the supply rails.
Yes, but those protection diodes are also limited in current (iirc 10mA) so you still have to take some precautions in the circuit. In this particular case I am upgrading an existing PCB so have to live with what it is.
The OPA1642 for instance has a diff input voltage max to the supplies.
I'll check the OPA2191, thanks for the heads up.
Jan
Thanks @Mooly ! I think this post belongs in the split thread: https://www.diyaudio.com/community/...cmos-audio-op-amp.335416/page-36#post-7082464
Its low gain bandwidth makes OPA1642 a nice option for retrofits, it will work fine on an oldskool PCB. Ideal when taking trash like RC4558 out of vintage amps... 50MHz opamps like OPA1656 (to stay on topic) are risky on single sided boards.
I generally agree with you and think this is a prudent move. On the other hand, I modified a one sided (poor layout) Carver C-2 Preamp to 1656s and had no stability issues, albeit I carefully soldered in .1uF caps as close as I could to the new opamps. One sided boards have really terrible grounding schemes (if you can even call it that).