which TI OPA opamp is best

Status
Not open for further replies.
Jump to Latest
#22
MaccAu said:
Thanks Derfnofred, so you are steering towards the 1642, and you find the 1602 and 4562 about the same in terms of sound? Caps on the dac supply pins
Click to expand...

I'm hardly a fan of Abrax, but, yes, the 1642 is the smartest opamp for that location, all things considered. JCX explains it far, far better than I do.

Frankly, I wouldn't mess too much with the DAC itself, as it should be sufficiently bypassed and you're more likely to do harm with those tiny pins than help. The schematic doesn't have any local bypass caps labelled on it, but I do wonder if they're on the board itself. Any photos perchance?
 
#23
abraxalito said:
What I was meaning was installing EFs (emitter followers) as buffers between the MASH DAC and the subsequent analog opamp stages. This would allow the internals of the DAC to operate at their best. After that, attention would turn to improving the subsequent opamp stages. I'd raise the working impedances (change to JFET based opamps as jcx is recommending) and bias the outputs into classA with external CCSs.
Click to expand...

Thanks Abraxalito 🙂 The impedences would be the capacitors the bps 22uf 16v ones. Emitters would go along the signal path? or on the dac pins

JMFahey said:
A non solution to a non problem.
Electronic products are good or bad because of design, not because they useºexpensive parts.
If it were so, it would br EASY to design good stuff, just buy most expensive available, forget studying Engineering,reading all datasheets, having a well equipped Lab, eperimenting a lot, etc. , just replace a $1 part with a functionally same $15 one and presto, our amp sounds 15 times better !!!
Clearly it does not work that way.
Even IF the new Op Amp were 15 times better than the old one, ... the circuit is not made to take advantage of that, sound/performance will improve only IF the old part was used beyond maximum performance (say, a 741 asked to deliver clean 20kHz at a significant level) and were the only limiting factor.
Doubt a competent designer would do that.
Click to expand...

Hi Fahey 🙂 Thanks i think you are saying the MASH is not worth spending much on, and to use cheaper parts, a lowest cost than the 1642.

Derfnofred said:
I'm hardly a fan of Abrax, but, yes, the 1642 is the smartest opamp for that location, all things considered. JCX explains it far, far better than I do.

Frankly, I wouldn't mess too much with the DAC itself, as it should be sufficiently bypassed and you're more likely to do harm with those tiny pins than help. The schematic doesn't have any local bypass caps labelled on it, but I do wonder if they're on the board itself. Any photos perchance?
Click to expand...
Hi Derfnofred 🙂 Thanks, the 1642 for IC802 and maybe also IC892. Appears to be some reddish films near IC802 as in the photo
An externally hosted image should be here but it was not working when we last tested it.
 
#25
MaccAu--those red caps are between the transistors and the opamps in question, as you can see their numbering and where they sit on the schematic. Not bypass but for seemingly the deemphasis circuit.

Also, what does the backside of the board look like?

Cheers,
D
 
#27
Yes, they're both SOIC8. Not seeing much in the way of local bypass on them, neither top nor bottom, which is of some concern wrt stability. 100 nF is recommended bypass.

http://www.ti.com/lit/ds/symlink/opa1642.pdf

See page 28. Although I really recommend the a good read throughout the datasheet, as our own DIYA member Johnc124 rewrote much of it (and wrote it in the first place) and added a lot of very good information.
 
#28
Hi D 🙂
Yes Johnc124 helped me with my Technics SU-V8 phonostage opamp - it was the 1642. I havent had a detailed listen to it, though I noticed it had less bass than the LME, but very happy with it.

Actually near IC802, 2 caps BUT they have resistors in their place instead, ie C805 and C806 - wonder why Technics used resistors in their place. IC891 has the bipolar caps 22uf 16v, which I could increase capacitance on them.

TI talks of bypass on page 28, notice much to bypass, though that C805 and C806, maybe that was for the model up from the PG440A

OPA1642 seems the favourite to use for IC802 and IC891.
 
#29
#30
Hi Abraxalito 🙂
Thanks for the link of the diagram. The MASH output has 4 which is the corners of the dac with larger amount of solder.

C805 - C808 on service manual = 120pf 50v, in their place is resistors **
 
Last edited:
#31
Well, unless the passives were changed, it's hard to see how the transfer function (especially when there's a huge amount of loop gain at low frequencies) would have changed materially between the different opamps. LF noise, certainly.

Those two caps would be for low-passing (to rid of HF DAC spurs), and you can see them in the schematic in front of IC802.
 
#33
found DAC datasheet - outs are labeled PWM output - inplying no op amps, no analog filtering internally

ef buffering would very likely be a disaster for the performance # due to rectification, asymmetric slew

OUTL(–) Left channel inverted phase PWM output pin.
Click to expand...

I would repurpose the output op amp as an additional low pass filter if you have MHz+ noise shaped PWM 0-5 V DAC out
 
#37
jcx said:
found DAC datasheet - outs are labeled PWM output - inplying no op amps, no analog filtering internally

ef buffering would very likely be a disaster for the performance # due to rectification, asymmetric slew



I would repurpose the output op amp as an additional low pass filter if you have MHz+ noise shaped PWM 0-5 V DAC out
Click to expand...

Hi Jcx 🙂 Thanks. So it would be a disaster to change IC802, the output opamp would be IC891. I can see the 4 output of the dacs the chip is labelled


abraxalito said:
The DS I have shows 47k/47pF LPF on the DAC's 4 output pins then leading to unity gain opamp followers, I'd keep those RCs in place prior to the EF buffer.
Click to expand...

Thanks Abraxalito. I am not sure what to do 😱 as i usually just change caps, and opamps.
phase said:
Wow, that opa164X would appear to be a great part, made for audio, and fet also.

Does it respond to biasing?
Click to expand...
Hi Phase, not sure but being fet, i think it would be a good replacement for my P990 opamps someday, lol it has so many not sure which to change. OPA1642 is good in the SU-V8s phonostage, less issues with high frequencies I have with the LME4462


mash dac - surface photo
An externally hosted image should be here but it was not working when we last tested it.


P990 opamps :crazy: heaps of opamps compared to MASH player
An externally hosted image should be here but it was not working when we last tested it.
 
#39
abraxalito said:
So many cooks, hard not to spoil the broth😛

If you're dead certain you only want to change opamps, then just do that. Find JFET input ones that suit, upgrade the power supplies to taste once you have them installed.



Save​
Click to expand...

😛 lol Abraxalito, i am confused about the broth

I would love to find opamps to suit my Technics SL-P990 someday, i actually have two of them. TI sugguested the fet opa1642, though i am not sure which to change. Going by a Japanese website they said it was Op-amp "M5238FP" Mitsubishi made in I / V conversion. Its Jfet



IC805,806,807, 808 are IC Buffer AMP M5238FP

IC 809, 810 are IC Amplifer NJM5532 ( near adjustment screws)

IC 811,812 are IC Filter NJM5532

IC 813, 814 are IC Amplifer M5238FP

The MASH, if shown what to do i can try to do it, so both opamps can be changed ? the OPA1642 would it suit? or another jfet. I would increase capacitor - bp on the output x3 times or more 22uf maybe use 100uf
 
Last edited:
#40
Stick with a JFET in this location. OPA1642 is excellent for this application.

Only extra recommendation is to find a good ground point that you can attach a pair of 100 nF cap (+ve to ground, -ve to ground x2, that's 4 in total if you replace 2 opamps). Lazily, that could be the bottom of c80x
 
Status
Not open for further replies.
Top Bottom