Yamaha CX-1 Thread

You are starting to convince me that I should try a OPA1612 in my sound card instead of the 2107. It's a Asus Xonar Essence STX. I really don't know anything about it internally other than the op-amps in it were socketed from the start. I used the 2107 and 2x 49720 based on someone(?)'s recommendation and like the sound a lot better than before. Looks as I might need to order some more of those 1612's.

Thanks for the discussion, it's been useful and informative.

Cheers,
James
I am familiar with the card, and some of the suggestions. There have been some changes in what we have to choose from. The original LM4562 are actually a fairly decent performer, but I'm not a fan of the JRC. The real issue is what is available in DIP format. The LM4562 (original) and the LME49720 (your replacement) are actually the same die. If you do choose to replace the OPA-2107 with the OPA-1612, you really do need it to be as close to the board as possible. So if you mount it on an adapter board and settle on it, I would remove the socket and place the adapter directly into the board to get it as close to the ground plane as possible.
 
I've swapped op-amps in my Revox B226 CD player and in two different active crossovers. I really like LM4562 but you have to play around with local decoupling until it sounds good. It can sound really wrong with the wrong decoupling... plus it sounds sharp and thin for a good 100 hours before mellowing down. I'm sure some of us dismiss it before giving it time to settle. Another one I like and is less finicky to use is OPA2132, sounds better than OPA2134 and is as easy to use. I've never tried OPA1612 or 2107, OPA1612 sure looks like the holy grail of ICs looking at specs! Pity it's not DIP-8...
 
...with the OPA-1612, you really do need it to be as close to the board as possible. So if you mount it on an adapter board and settle on it, I would remove the socket and place the adapter directly into the board to get it as close to the ground plane as possible.

Am I correct in guessing that you would recommend doing this with an amp or cd player also? (Getting it as close to the ground plane as possible.)

Cheers,
James
 
... plus it sounds sharp and thin for a good 100 hours before mellowing down. I'm sure some of us dismiss it before giving it time to settle....

I didn't really believe this until I actually swapped some op-amps and found to my surprise that the really do start to sound better after 60+ hours of use. And to address the "that's just your brain and ears getting used to the change" argument that often gets stated, I've taken 2 otherwise identical cd players and the one with brand new op-amps sounded more harsh than the one with 2 months of use. So this is me agreeing with you. :)

Cheers,
James
 
I've swapped op-amps in my Revox B226 CD player and in two different active crossovers. I really like LM4562 but you have to play around with local decoupling until it sounds good. It can sound really wrong with the wrong decoupling... plus it sounds sharp and thin for a good 100 hours before mellowing down. I'm sure some of us dismiss it before giving it time to settle. Another one I like and is less finicky to use is OPA2132, sounds better than OPA2134 and is as easy to use. I've never tried OPA1612 or 2107, OPA1612 sure looks like the holy grail of ICs looking at specs! Pity it's not DIP-8...
Just keep in mind, one of the shortcomings of DIP, and one of the reasons SMT technologies must come into the picture, is decoupling very close to the device, and with very short leads and close to the ground plane. This is how something like the OPA-1612 can achieve the specs it does. With riser boards, there really is no way to effectively decouple it within very close proximity (within a couple of mm) and no ground plane.

I can say, in my own experience, my Avantage goes from so-so SQ with the original NE5532's to reference SQ with the OPA-1612. ESS will tell you strait up, even in their lowest performing DAC, the NE5532 is the lowest of the scale in acceptable performance, and in their better DAC's it isn't even on the choice list. It just doesn't have the performance required. I really do wish that Yamaha would finally figure out, we as consumers would be willing to pay an additional 10% (to cover the cost of the part price for the op-amp) to obtain the better SQ, and they could certainly gain more market share with such, rather than trying to sell that a 5th plastic foot (The Anti-Resonance Technology Wedge) under the center of the chassis is some miracle SQ device.
 
Last edited:
Thank you very much if you meant me. If not thanks anyway just for being here and for the heck of it. :)

Was meant for you! I "unfortunately" bought some more vintage Yamaha gear and as I research them, I come across many of your post and they are very helpful to a newbie like me.
 
Was meant for you! I "unfortunately" bought some more vintage Yamaha gear and as I research them, I come across many of your post and they are very helpful to a newbie like me.

That's great! That is mainly why I make these threads, to spread information and let others benefit from my mistakes.

Cheers,
James
 
Btw, everybody can laugh at how difficult it was for an old guy to do this, but I finally figured out how to make a pdf file of my parts list and attached it to post # 182. It's perhaps not perfect, but usable I think.

Cheers,
James
 
Just keep in mind, one of the shortcomings of DIP, and one of the reasons SMT technologies must come into the picture, is decoupling very close to the device, and with very short leads and close to the ground plane. This is how something like the OPA-1612 can achieve the specs it does. With riser boards, there really is no way to effectively decouple it within very close proximity (within a couple of mm) and no ground plane.

I can say, in my own experience, my Avantage goes from so-so SQ with the original NE5532's to reference SQ with the OPA-1612. ESS will tell you strait up, even in their lowest performing DAC, the NE5532 is the lowest of the scale in acceptable performance, and in their better DAC's it isn't even on the choice list. It just doesn't have the performance required. I really do wish that Yamaha would finally figure out, we as consumers would be willing to pay an additional 10% (to cover the cost of the part price for the op-amp) to obtain the better SQ, and they could certainly gain more market share with such, rather than trying to sell that a 5th plastic foot (The Anti-Resonance Technology Wedge) under the center of the chassis is some miracle SQ device.

Hi DB, a couple of questions/thoughts.

If I bought an SO8-Dip-8 adapter without stand-off type leads (something like the one below), attached the op-amp, and then laid it directly on the pcb I would be able to just solder some short leads directly through the board/pcb. Would this get me to where you are suggesting going with this? (this is pre-supposing of course that there is room and nothing under the position that can short)

sop8 dip8 adapter20.jpg
In "decoupling" do I understand correctly that you are meaning a small (0.1µF?) cap from the supply lines to ground?

Can you show us a photo or 2 of how you added the 1612 to your amp(s)?

Also, one thing that I've never seen in any Yamaha amp etc is a part that wasn't of Japanese origin. And the OPA's are from Texas Instruments. I think that we can wait a long time before they use an op-amp that wasn't sourced from Japan.

Cheers,
James
 
Hi DB, a couple of questions/thoughts.

If I bought an SO8-Dip-8 adapter without stand-off type leads (something like the one below), attached the op-amp, and then laid it directly on the pcb I would be able to just solder some short leads directly through the board/pcb. Would this get me to where you are suggesting going with this? (this is pre-supposing of course that there is room and nothing under the position that can short)

View attachment 1303862
In "decoupling" do I understand correctly that you are meaning a small (0.1µF?) cap from the supply lines to ground?

Can you show us a photo or 2 of how you added the 1612 to your amp(s)?

Also, one thing that I've never seen in any Yamaha amp etc is a part that wasn't of Japanese origin. And the OPA's are from Texas Instruments. I think that we can wait a long time before they use an op-amp that wasn't sourced from Japan.

Cheers,
James
James, the adapter shown is not for SOIC, but VSSOP (micro) and the OPA-1612 is not available in that package type, however, the OPA-1602 is. That is an incredibly small package, and very tough to deal with if you don't have a x10 scope and proper tools. But from a concept perspective, yes, the closer you get the SOIC to the board and the shorter the leads, the better off you are, but through hole technology doesn't utilize multi-layer boards with ground planes.

Concerning decoupling, yes, I was speaking to properly decoupling the supply. My Avantage is already SMT based and no mods to the boards, or adapters were required. Simply removing the NE5532 (SOIC) and replacing it.

Most of what Yamaha is selling this past decade from an AVR perspective has a number of TI parts, including the DSP's. The original NE5532's are TI chips. The real issue here is that NE5532 are $0.25 and OPA-1612 are $4.00. Profit margin dictates this decision.
 
Last edited:
You are starting to convince me that I should try a OPA1612 in my sound card instead of the 2107.
If you want to elevate its performance further, use discrete FET modules from companies like Burson. Yes, they are a bit pricier but worth the difference in improved transparency and dynamics IMHO. The only gotcha is they need a bit of height. I replaced OPA2134s/OPA2604 in a Music Fidelity DAC 25.3 with V5s.

Since the op amps were socket mounted, installation took just minutes. They are plug compatible with a large list of single and dual op amps - see at bottom of link.

burson2.jpg
 
Last edited:
A small update.

I've been using this control amp almost every day and really love it. Just recently I found myself deciding to keep a second CD player in use and wanted to better optimize the Tape-3 inputs. Which boiled down to replacing the op-amps in the input buffer.

The supply voltage for these op-amps is 20v so that limits things a little bit. But I have a few LME49860's that I bought just before they went EOL and they can use +/-22v. So I mounted a couple of them to some Brown Dog adapters and mounted them. Heh heh, it will be a few days before I can report on how it sounds with them. The NJM5532S-D's that they replaced have 9 pins but the adapters can be mounted using pins 2-9.

DSC06493.jpg DSC06494.jpg DSC06495.jpg

Cheers,
James
 
Last edited:
While I had the unit open I retro mounted a 0.022µF and 0.01µF Wima film cap in addition to the previous bypass caps in parallel with all of the signal path coupling caps and the CD & Tape-3 inputs.

DSC06489.jpg DSC06490.jpg DSC06491.jpg DSC06492.jpg

Sure it looks kind of ridiculous but it sure is sounding delicious. :)

Cheers,
James
 
Please tell us how it sounds! :)

Sorry for taking so long to reply. I've started off the new year by getting a very nasty cold and am a bit slow at the moment.

In the end, after a lot of listening I've reversed those last things that I changed.

Adding the 0.022µF and 0.01µF film caps brought out an unpleasant hard edge to the high end and I removed them.

Concerning the op-amps. Heh.
When I first tried out the tape-3 inputs I only got a garbled bad sound. After looking things over closely the answer to that was so obvious that it made me feel pretty stupid.

The 9-pin Sip NJM5532 has duplicated V+ supply on pins 1 and 9. The 8-pin SIP adapter that I used to mount the LME's has to be mounted in the PCB positions 2-9 to be pin compatible and gets it's V+ from pcb pin-9. Here is where the stupid part plays out because when one actually looks at the pcb with both eyes open and the brain engaged it's pretty obvious that pin-9 on the pcb is a dummy pin isolated from everything, so, no V+ and the op-amp wasn't getting half of the power it needed.

I could have run a wire from the de-populated pin-1 over to pin-9 and solved that issue but didn't really want to do that and instead re-installed the NJM5532's.

So, back to where I was before that last round of changes. The pre-amp sounds great, and I think that it's time now to keep my fingers off of it except to turn it on and use it. :music:

Cheers,
James
 
Another small update.

After I bought my CD-S2000 cd player and started to study how it was built I noticed that the bridge rectifier in the analog section was a Schottky diode rectifier. P/N D4SBN20. The specs compare very nicely to the D2SBA20 used in the CX-1. Without a heatsink it's rated at 2.2A. It has a similar form as the D2SBA20, just a little larger. And Mouser carries them.

https://www.mouser.de/ProductDetail...&utm_term=D4SBN20-7000&utm_content=Shindengen

So I bought a few of them with my last Mouser order and yesterday I installed one.

The legs need to be bent a bit to make it fit but after that it is a drop in fit.

DSC06906.JPG DSC06908.JPG DSC06912.JPG DSC06913.JPG

Just thought that I would pass that on in case anyone is interested.

Cheers,
James
 
And now I'll mention something so dumb and embarrassing that I don't really want to talk about it but will anyway just in case it helps someone.

When I reworked the CX-1 I did buy replacement relays for protection and muting. But I got dumb and didn't actually replace the old ones. I just made the unfounded assumption that what with being sealed, inert gas filled and having gold plated contacts that they wouldn't really need to be replaced.

Fast forward to yesterday. I have been struggling to come to terms with an irritating bit of noise/distortion that has showed up in the CD input. It has been gradual and using a different input sounds better.

Yesterday when I had the unit open to install that new schottky bridge rectifier I thought about those relays. There is one in the protection section, and one each in the CD and tape-3 input buffers. So I finally replaced all 3 of them yesterday.

And that noise issue is cleared up now. :) My guess is that the tape-3 input hasn't been used very much and that's why that input buffer was sounding better than the CD input.

So if you are working on one of these by all means go ahead and replace those relays. They are less than €3 each.

DSC06909.JPG DSC06910.JPG DSC06911.JPG

Cheers,
James
 
Recently I have had the urge to change a few things in the control amp. I wanted to revise what I had done concerning the bypass caps on the signal path coupling caps, and in 1 case revise the coupling cap entirely. I also wanted to try some low-noise foil resistors in the very first position of the CD buffer.

Specifically, I wanted to remove the cascaded bypass caps and replace them with a simpler set of 0.1µF and 0.01µF polypropylene with copper leads. In addition I wanted to parallel a 1µF PP with the 3x 47µF in each channel. And lastly I wanted to replace the Wima 10µF MKS2 polyester cap in the CD buffer with a PP cap.

It took me a while to find what I wanted for the 10µF, 1µF and 0.01µF, mainly because I wanted to go full OCD and have only caps with tinned copper leads, as well as the size problem when using a 10µF PP cap.

I eventually landed on Panasonic ECWF(A) and ECWH for the smaller caps and Wima MKP4 for the big ones.

https://www.mouser.de/ProductDetail/Panasonic/ECW-H8103RHV/?qs=GoZnPdyRfjZLBdoBmljywg==
https://www.mouser.de/ProductDetail/Panasonic/ECW-F2104JAQ/?qs=/ha2pyFadujK%2BdvILcjJFJLrFVbsOG0AanJ4eg4pEcrBpNEzZAVRTQ==
https://www.mouser.de/ProductDetail/Panasonic/ECW-F2105JA?qs=GoZnPdyRfjZScBQE8XcRvw==
https://www.mouser.de/ProductDetail/WIMA/MKP4D051007E00JSSD?qs=K1lkoHpmGTrGA8Eiln8cbQ==

The CD input has a 220 ohm resistor as the very first component in the signal path, originally a 1/3w carbon film and replaced with a 500mW Vishay RN metal film. I bought some Charcroft Z-Foil resistors for these positions. These things are pretty expensive (€14.46) so only these 2 positions will be changed.
https://www.hificollective.co.uk/catalog/car170-220r-charcroft-zfoil-resistor-p-5046.html

The goal with all of this was to see if it would offer any clarity or detail improvements.

Cheers,
James
 
Last edited:
I started with the 2x 47µF caps mounted on the volume control board. There was barely enough room to piggyback the 1µF directly on the 47µF and mount the 0.1µF and 0.01µF to the trace side.

I bought 100 of each of the 1µF, 0.1µF and 0.01µF and measured them all and then sorted out the best matching pairs, with the focus on matched L&R pairs.

DSC00987.JPG DSC00989.JPG DSC00990.JPG DSC00994.JPG

Cheers,
James
 
  • Like
Reactions: mbz
Back
Top Bottom