Upgrading from 4560s in an SX202 mic preamp in 2023

[Mammals]

Hello, hopefully this is an ok forum to post this query in.

Monte McGuire's detailed SX202 mods checklist still lives, as does my stock Symetrix SX202 dual mic preamp (rev G), and I am about to tackle (as a fairly inexperienced person with a helpful neighbor) swapping out the compensation caps for the two SSM 2015 chips and getting sockets in to replace and upgrade from the three 4560s.

Since time marches on, new ICs have appeared, and others seem to be on their way out, in the dozen years since I've found people last discussing upgrades for the 4560s online.

So, I'm wondering if there are pin-compatible upgrades anyone here would recommend, more recent than and above the LT1469, LM6172, OPA2134—or, if one of those is still your best recommendation, that's great too, just tell me your thoughts.

(I had naively hoped I could just replace them with THAT 1512 but that appears to be wrong.)

Thanks in advance to all of you with decades more experience than I've got, for your input. I yield my time!

 

[Mammals]

Per the suggestion of a forum member, I'm ordering OPA2137. Results tbd

 

[k brown]

Could do more harm than good - personally I'd leave the 4560s in - perfectly good op amp, and that's what the circuit was designed for; which is more important than the specific chip used.

By the way, the THAT 1512 is not an op amp - neither are 1510, INA163, INA217, SSM2015, SSM2017, SSM2019 - they are all instrumentation amps specifically suited for use as microphone preamps.

 

[Mammals]

Thank you for the followup—I appreciate the explanations.

I do like the unit pretty well and have been using it for several years, but if it can be improved by doing a few small things, I'm interested. Based on the online recommendations of several people, I'm still keen to give it a shot. I can always put 4560s into the sockets if I find it has been degraded and I want to undo it, so it seems low-risk to me.

Here's a short excerpt on the 4560s from the modifications info mentioned above.
"At best, these amplifiers are somewhat slow, and this adds excess distortion above 3KHz or thereabouts. The Rohm devices seem to be marginally unstable in addition to being too slow to pass audio cleanly, and this causes the HD spectrum to be littered with many high order harmonics. The end result is an overly bright and hard sound, solely due to the 4560."

 

[k brown]

Poppycock.

If I were to bother replacing the 4560s, I'd just do the readily available JRC (NJM) 5532 (as long as they aren't fake - which has been happening). Folks get all ramped up about slew rate of newer chips, but the (now considered tortoise-slow) 5532 has been proven time and time again to be plenty fast enough for audio, and that faster chips only measure better with no audible improvement; and if the circuit isn't designed for (or competently modded for) them, they can oscillate and/or cause other problems.

McGuire's cap (especially power supply) and grounding mods are more worthwhile. Replacing the gain pots with switched fixed resistors is also good, but not for the faint of heart. I did it with my M-Audio DMP3s, but simplified it to to just four steps of gain (14, 20, 26, and 32 dB, since that's all the range I require).

 

[JohnRoberts]

I suspect some people are listening with their eyes... The spec sheet for the TI 4560 shows 15 mHz gain bandwidth and 5.5V/uSec slew rate, which should be respectable for most signal handling.

I am not familiar with the sx202 but expect it's performance will be dominated by topology and other design considerations.

Did your advice claiming excess distortion above 3kHz show distortion plots?

JR

 

[k brown]

I suspect some people are listening with their eyes... The spec sheet for the TI 4560 shows 15 mHz gain bandwidth and 5.5V/uSec slew rate, which should be respectable for most signal handling.

I am not familiar with the sx202 but expect it's performance will be dominated by topology and other design considerations.

Did your advice claiming excess distortion above 3kHz show distortion plots?

JR

Especially since, in the 202 the 4560 are operating at unity gain.

No such luck on data/plots - here is the mod article Symetrix SX202 Mods, and the 202 schem.

Also, the 202's phantom voltage is a bit low - if you have mics that have to have full 48v, do the mod described in post #43 on this thread: https://gearspace.com/board/diy-ele...hoto-diaries/182097-symetrix-sx202-mod-2.html

 

[JohnRoberts]

Since its a stereo unit you can mod one channel and compare the results with a null test.

JR

 

[k brown]

Since its a stereo unit you can mod one channel and compare the results with a null test.

JR

And even that would only tell you the sound has changed, not necessarily improved ; - )

 

[JohnRoberts]

And even that would only tell you the sound has changed, not necessarily improved ; - )

but sometimes with already decent performance designs the "improvements" from throwing more expensive parts at it does not result in significant changes. The null test can quantify how much difference there is while it requires judgement to analyze.

Null errors will be dominated by frequency response, then phase shift. The audibility of these differences can be subtle, but determining that there is any difference at all, is an important first step.

JR

 

[Ricardus]

Save your money and put it toward some mic pres with a completely different topology. Maybe something API or Neve flavored with transformers which will color the sound.

There are endless threads on Gearspace and such about modding these, and I am sure the people who did the mods believe they can hear a difference, but I suspect there wouldn't be any difference in a null test.

 

[Newmarket]

but sometimes with already decent performance designs the "improvements" from throwing more expensive parts at it does not result in significant changes. The null test can quantify how much difference there is while it requires judgement to analyze.

Null errors will be dominated by frequency response, then phase shift. The audibility of these differences can be subtle, but determining that there is any difference at all, is an important first step.

JR

Suggest to null test the unit unmodded first to check how well the channels match to begin with.

 

[Nen'O]

Hi,
Here is my tip regarding mods and upgrades.
The biggest bonus while executing the mod is that you actually learn about circuits and components and how they affect the sound. That alone makes you better audio engineer/ music producer.
Down side is that the time you spent messing with gear is time lost in making music

My trusty console A&H Zed R 16 circuit is based around 4560 opamp. To my ears the biggest improvement was upgrading power supply. If you want to make your preamp special you can try one of discrete opamps available on market.
This one is direct replacement to 4560 and Andrew is great guy with a lot of knowledge to share . Don’t hesitate to ask him for advice.

https://sparkoslabs.com/discrete-op-amps/
Keep us updated

 

[Mammals]

Update:
1. It turns out that while this unit is Rev. G, and the board is also labeled as Rev. G, it seems to use the numbering system for the Rev. C boards, at least for the compensation caps around the SSM 2015 chips.
2. We started as planned, swapping out the compensation caps, but in trying to be gentle and not ruin the pads, we found the process much more time-consuming than either of us had expected, so solder-sucking and replacing at the next 24 points to install three 8-pin sockets was not in the cards yesterday. My neighbor normally builds guitar pedals, and I think he didn't want to feel responsible for damaging this preamp, so he went slowly and treated it gingerly.
3. To the very point of this query: for all the grousing in this thread about not wasting money on newfangled speed-focused expensive chips, I'd like to point out that upthread I had already posted that I had (following up on the Los Angeles-area mentor thread and asking advice of a forum member who clearly knows what he's doing) chosen three OPA2137 for replacement. Total cost: approximately $11, I believe, and the datasheet says 1997 on it, so these comments seem simultaneously strange (misdirected) and familiar (classic internet forum). As a person fairly new to this forum, I found that disappointing. In any case, they're inexpensive enough that, although I'd love to see if the mod makes an improvement, if it doesn't get done, the cost I'm eating is insignificant.
3b. Also, what a strange assertion that I should save this money (a very small amount) and put it toward preamps with different topology (costing a large amount) when that suggestion for more color is neither what I'm looking for from this preamp nor shows any awareness of other preamps I already do have. Again, classic internet.
4. Finally, while chewing up the time of the person helping me out made null testing unlikely yesterday, I could certainly do it now (post compensation cap upgrades) in the event the sockets work does get done. Will using an app that does a frequency sweep to send stereo signal through these channels be rigorous enough testing to show anything, or is there some other sort of gold standard of putting them through their paces?
Thank you