Good opamp for that1646

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

warpie

Well-known member
Joined
Feb 7, 2009
Messages
1,566
Can you recommend some decent op-amps to drive a that1646 chip that it's reasonable priced?. What op-amp would be consider good nowadays?

Also, does the buffer before thet 1646 offers any benefits other than the improved CMRR?
 
There is no 'special' requirement in terms of driving the 1646.
So any of the 'usual suspects' is fine. Depends on your signal source impedance whether you go with BiPolar / FET/ BiFET...
 
Well I guess what I wanted to ask is what are the good usual suspects nowadays... ☺

Depends on your signal source impedance whether you go with BiPolar / FET/ BiFET...
Can you elaborate a bit please?
 
So, what is the signal source that you want to balance and drive with the 1646 ?
It's just an idea really but I was thinking to experiment with a 1646 instead of an output trafo on fripholm's TG1 limiter. Here's the output section.
 

Attachments

  • out.jpeg
    out.jpeg
    88.7 KB · Views: 95
It's just an idea really but I was thinking to experiment with a 1646 instead of an output trafo on fripholm's TG1 limiter. Here's the output section.

I see. I'd say just go with a good bipolar opamp with suitable noise/distortion specifics.
It's kind of boring to say I know - but NE5532 / NE5534 are just as good as they've always been (you can see what people reckon on the different manufacturers if you like) although they do take a significant current if that is a consideration.
LM4562 is an obvious contender. Quite a few others - just look at TI and AD data sheets and application notes.
Bear in mind that many newer devices are available in SMT only.
Others may have more esoteric suggestions but I'm not big on 'OpAmp Rolling' or discussing the "sonic imprint of a TL072 in the upper mid vocal range" :rolleyes:
I'd also say stay away from very 'fast' high GBW opamps unless they particularly interest you. The slew rate isn't needed and they can become a bit 'twitchy' wrt layout / decoupling etc. Particularly if using 2-layer pcb (depending on the detail layout) or stripboard / wire wrap techniques.
 
I see. I'd say just go with a good bipolar opamp with suitable noise/distortion specifics.
It's kind of boring to say I know - but NE5532 / NE5534 are just as good as they've always been (you can see what people reckon on the different manufacturers if you like) although they do take a significant current if that is a consideration.
LM4562 is an obvious contender. Quite a few others - just look at TI and AD data sheets and application notes.
Bear in mind that many newer devices are available in SMT only.
Others may have more esoteric suggestions but I'm not big on 'OpAmp Rolling' or discussing the "sonic imprint of a TL072 in the upper mid vocal range" :rolleyes:
I'd also say stay away from very 'fast' high GBW opamps unless they particularly interest you. The slew rate isn't needed and they can become a bit 'twitchy' wrt layout / decoupling etc. Particularly if using 2-layer pcb (depending on the detail layout) or stripboard / wire wrap techniques.
Many thanks! I'll give it a try. So, omitting the op-amp and connecting straight the 1464 isn't the best idea right?
 
Many thanks! I'll give it a try. So, omitting the op-amp and connecting straight the 1464 isn't the best idea right?

Well what's the output impedance of your circuit ? Without looking at the detail I'll guess a few tens of Ohms (happy to be corrected -). OpAmp[ buffer output will be a fraction of an Ohm so it wins out in optimising CMRR in the system.
 
Well what's the output impedance of your circuit ? Without looking at the detail I'll guess a few tens of Ohms (happy to be corrected -). OpAmp[ buffer output will be a fraction of an Ohm so it wins out in optimising CMRR in the system.
Yeah I also believe so. I just thought I'd ask in case I'm missing something.

I'd like to compare the CMRR with and without the buffer but can't find a way to measure CMRR. At least something straight forward.
Anyway, many thanks for your help!
 
Yeah I also believe so. I just thought I'd ask in case I'm missing something.

I'd like to compare the CMRR with and without the buffer but can't find a way to measure CMRR. At least something straight forward.
Anyway, many thanks for your help!

Well it's not straightforward to quantify experimentally without at least some at least some semi-serious kit.
I guess you might take a look through "The Chamber" here for ideas short of a Prism or AP setup.
tbh I'd just put the buffer in there seeing as how you're using the THAT device. Why go to the effort of using it then not use it optimally.
If you really want to have the option to omit a buffer stage then just build it with the opamp and bypass it to test if you experience any subjective difference.
 
Well it's not straightforward to quantify experimentally without at least some at least some semi-serious kit.
I guess you might take a look through "The Chamber" here for ideas short of a Prism or AP setup.
tbh I'd just put the buffer in there seeing as how you're using the THAT device. Why go to the effort of using it then not use it optimally.
If you really want to have the option to omit a buffer stage then just build it with the opamp and bypass it to test if you experience any subjective difference.
I do have some test gear, (soundcard, software, a tek scope, function generators, PSUs, etc..) so I'll see what I can come up with. :)

The reason I'd like to test both setups is for educational purposes mostly.
 
sniffing out CMRR is relatively easy, but everything is relative.

IIRC the 1646 wants to be driven from a low impedance to preserve differential feedback networks. If the + and - inputs are driven with different source impedance the two paths won't match and deliver good cancellation.( I never used a 1646 but a lot of people here and about have.)

Presumably we are talking about CMRR wrt the 1646 output. CM like an AC potential difference between sender and receiver 0V (which aren't ever really 0V). You can artificially introduce a ground potential error by floating the safety ground on the DUT (device under test) then inject a small voltage there. I would keep it small, and not float the safety ground completely maybe start with something like 100 Ohm. Then inject some CM noise , a proper test involve a frequency sweep.

If course be careful when messing with safety grounds.

JR
 
sniffing out CMRR is relatively easy, but everything is relative.

IIRC the 1646 wants to be driven from a low impedance to preserve differential feedback networks. If the + and - inputs are driven with different source impedance the two paths won't match and deliver good cancellation.( I never used a 1646 but a lot of people here and about have.)

Presumably we are talking about CMRR wrt the 1646 output. CM like an AC potential difference between sender and receiver 0V (which aren't ever really 0V). You can artificially introduce a ground potential error by floating the safety ground on the DUT (device under test) then inject a small voltage there. I would keep it small, and not float the safety ground completely maybe start with something like 100 Ohm. Then inject some CM noise , a proper test involve a frequency sweep.

If course be careful when messing with safety grounds.

JR
Thanks John. How small is small?

I've built one of these hummer jigs (from jensen). Can this be of any use?

https://www.jensen-transformers.com/wp-content/uploads/2014/08/as032.pdf
 
Yeah I also believe so. I just thought I'd ask in case I'm missing something.

I'd like to compare the CMRR with and without the buffer but can't find a way to measure CMRR. At least something straight forward.
Anyway, many thanks for your help!

We have a gadget that would help you. However, although it also does non 500 series equipment, it is powered from a 500 rack. So, if you do not have one then you will need to power it externally.
https://groupdiy.com/threads/500-rack-eq-one-micpre-one-mid-side-one-by-total-audio-control.52462/
Download manual
http://www.robotica.co.uk/tac/CMRR_TRIMMING_JIG_MANUAL_V3.pdf
The manual has good amount of information including the schematics.
 
sniffing out CMRR is relatively easy, but everything is relative.

Quite ! In the distant past I've spent working weeks looking at stuff with an AP System 2 or, on a less refined level a Neutrik A2. These days in my home setup I'm usually 'reduced' to looking at some LED bar meters on a A&H or Soundcraft desk - and, of course, my ears !
 
How about the OP97(fpz)? Do you think it's a good contender? I believe it's lower power than the 5534 and also I remember reading that the 5534 isn't very stable with gain less than 3 but this was a long time ago so I might be wrong.
 
How about the OP97(fpz)? Do you think it's a good contender? I believe it's lower power than the 5534 and also I remember reading that the 5534 isn't very stable with gain less than 3 but this was a long time ago so I might be wrong.

IIRC NE5534 is stable for minimum gain of five ? Don't have time to look through the datasheet right now. But you can easily compensate it to be unity gain stable.
I recall that it's noise spec' is slightly lower than the equivalent unity gain stable dual NE5532. But you can parallel the two parts of a NE5532 to lower the noise.

I'm not clear why you've landed on the OP97 for an audio buffer. It's not particularly quiet (voltage noise) and has no real audio specs' in the data sheet and is basically described as a slow, precision amplifier. Compare the specs' against an 'everyday standard audio OpAmp' that gets used everywhere eg TL071 or (better) OPA134.
 
Last edited:
OP97 seems to be unity gain stable. Check the page 12 of the data sheet. 10k resistor is adviced on the feedback path for small signal application.

https://www.analog.com/media/en/technical-documentation/data-sheets/OP97.pdf

Thank you guys. Actually I'm reading the data sheef as we speak. I think it looks good but also a bit more sensitive when it comes to PCB design . Another question I have is if I can run it at +/-18V. Absolute maximum is +/-20V but +/-15V is suggested.
 
Last edited:

Latest posts

Back
Top