Hi all,
I've recently acquired a lovely condition Denon DCD-820, as I was keen to see how the old 80s Burr Brown DACs compare to newer examples.
I'm thoroughly impressed with how it sounds - I can usually tell sod all difference between different DACs and players where modern equipment is concerned, but there's a definite difference here. I'm hard-pressed to say whether I actually prefer the sound using the Denon's onboard DAC, or my Audiolab 8200, but it's all good fun and it seems to vary depending on what I listen to.
Now I've read a few threads about people modding older players, and I'm wondering if it's something I should dabble in.
Is there anything I can do to it that will garner a real, noticeable difference?
I've read about things like removing muting transistors, re-capping, changing op-amps etc. The latter is probably beyond my capabilities as an amateur, but I'm interested nonetheless.
Will these changes make any difference, and/or introduce any drawbacks?
Anything else I should consider?
Thanks!
I've recently acquired a lovely condition Denon DCD-820, as I was keen to see how the old 80s Burr Brown DACs compare to newer examples.
I'm thoroughly impressed with how it sounds - I can usually tell sod all difference between different DACs and players where modern equipment is concerned, but there's a definite difference here. I'm hard-pressed to say whether I actually prefer the sound using the Denon's onboard DAC, or my Audiolab 8200, but it's all good fun and it seems to vary depending on what I listen to.
Now I've read a few threads about people modding older players, and I'm wondering if it's something I should dabble in.
Is there anything I can do to it that will garner a real, noticeable difference?
I've read about things like removing muting transistors, re-capping, changing op-amps etc. The latter is probably beyond my capabilities as an amateur, but I'm interested nonetheless.
Will these changes make any difference, and/or introduce any drawbacks?
Anything else I should consider?
Thanks!
As long as you don't expect miracles... you could try to exchange the NE4558's with cheap & "fail-safe" NJM5532 (IC304,305,306) and swap the electrolytic cap on output for a bipolar Elna or Muse or... (C318,319), maybe also C327+328.
Thank you. I have been reading about the opamps (it looks easier than I first thought). It seems that two common "favourites" are the OPA1642 and OPA627. After looking at the datasheets they seem compatible - is there any reason they wouldn't be?
I'll add a DIP socket so I can easily pop in and out different chips.
I'll add a DIP socket so I can easily pop in and out different chips.
The OPA627 is a single rather than dual and so not an option unless you use adapter boards.
I would agree with looking to swap the 4558's though and would perhaps suggest the OPA2134.
One thing to watch (I haven't studied the circuit) is that the 4558 has a low power consumption whereas the NE5532 is quite a bit higher. Although it is all relative and we are only talking milliamps it all adds up and three 5532's pulling 15ma each is quite a bit different to three 4558's pulling say 4 milliamps each.
Have a read at this:
Swapping Op-Amps... you have checked to see it's stable haven't you ?
The OPA2134's should be a straight swap with no issues although they pull around 8 milliamps per package.
I would agree with looking to swap the 4558's though and would perhaps suggest the OPA2134.
One thing to watch (I haven't studied the circuit) is that the 4558 has a low power consumption whereas the NE5532 is quite a bit higher. Although it is all relative and we are only talking milliamps it all adds up and three 5532's pulling 15ma each is quite a bit different to three 4558's pulling say 4 milliamps each.
Have a read at this:
Swapping Op-Amps... you have checked to see it's stable haven't you ?
The OPA2134's should be a straight swap with no issues although they pull around 8 milliamps per package.
That's the kind of info I love this forum for 🙂 I didn't notice that. Thank you! I'll strike it off my list.
I'll have a look at the current requirements as well.
What are your thoughts on using the 1612 / 1642 / LM4562?
Also the 2604? Although I notice that's a lot older.
Last edited:
The 2604 would have been the device to use imo but production was stopped some time ago due to problems when run at its absolute maximum ratings. Although that only affected a tiny minority of applications the device is no longer available and so any you see could well be fakes.
Never ever buy things like semiconductors from anywhere other than authorised distributers.
I haven't used either of the devices you mention and these may not even be available in the big 8 pin DIP packages. So many are SMD only... the data sheet should say what packages are available.
The current issue usually surfaces when the supply is a simple resistor and Zener shunt regulator, something some manufacturers do use.
Never ever buy things like semiconductors from anywhere other than authorised distributers.
I haven't used either of the devices you mention and these may not even be available in the big 8 pin DIP packages. So many are SMD only... the data sheet should say what packages are available.
The current issue usually surfaces when the supply is a simple resistor and Zener shunt regulator, something some manufacturers do use.
The LM4562 is the successor of the NE5532 and is excellent in pretty much all respects however some believe (I'm one of them) that bjt input opamps don't always perform as well as FET based type when fed from 'noisy' DAC's. Some evidence in tests done years ago that the transistor input stage can suffer non linearity issues due to the junctions attempting to rectify the high frequency noise. The FET is immune to that.
More brilliant info, thank you. eBay is littered with 2604s so I guess they're all fakes?
For the others, there are eBay listings that promise they're genuine - should these still not be trusted?
So I'm thinking I should try the 2134 - any others you can recommend? Then maybe I'll try the 1612 and 1642 afterwards in the future. I'm happy to give the LM4562 a try, too. Unless it's likely to cause damage of some sort?
Thanks!
For the others, there are eBay listings that promise they're genuine - should these still not be trusted?
So I'm thinking I should try the 2134 - any others you can recommend? Then maybe I'll try the 1612 and 1642 afterwards in the future. I'm happy to give the LM4562 a try, too. Unless it's likely to cause damage of some sort?
Thanks!
None will cause damage as long as you don't accidently fit a single by mistake which could damage the IC rather than the player.
If you are fitting sockets then try a selection but be sure to give each a fair listen over a few days.
More on the OPA2604 here:
Drop in replacement for NE5532?
And then I posted this some time later:
Drop in replacement for NE5532?
I haven't looked for ages at the status but as far as I know not available...
If you are fitting sockets then try a selection but be sure to give each a fair listen over a few days.
More on the OPA2604 here:
Drop in replacement for NE5532?
And then I posted this some time later:
Drop in replacement for NE5532?
I haven't looked for ages at the status but as far as I know not available...
Farnell currently have 'No Longer Manufactured' against the OPA2604. Pity because they were genuinely brilliant in the right application.
Oh well. Doesn't look like there's any new old stock on eBay either!
I think I'll start with the OPA2134 along with a recap, and removing the muting transistors, and go from there. Are there any other recommended mods I should do at the same time?
I think I'll start with the OPA2134 along with a recap, and removing the muting transistors, and go from there. Are there any other recommended mods I should do at the same time?
You would have to study the circuit in detail to try and see what possible mods there might be but one trick I always like to do is to very neatly solder a small 4.7uF 50v cap directly to the supply pins of the chip.
If you do it neatly and form the cap leads first then you can add the cap to the top part of the pin without solder getting to the part that goes in the socket. Wipe the pins with a cotton bud and iso if you try this to remove any flux that might have run.
If you do it neatly and form the cap leads first then you can add the cap to the top part of the pin without solder getting to the part that goes in the socket. Wipe the pins with a cotton bud and iso if you try this to remove any flux that might have run.
Pins 4 and 8 are the supply 🙂 Pin 8 is positive, pin 4 negative.
Pins 2 and 3 are the two differential inputs for one half the chip (one amplifier of the two).
The caps gives low impedance supply as close to the chip as possible. Just use ordinary electrolytics, nothing exotic needed.
Pins 2 and 3 are the two differential inputs for one half the chip (one amplifier of the two).
The caps gives low impedance supply as close to the chip as possible. Just use ordinary electrolytics, nothing exotic needed.
Sigh, you're right. I looked at it too quickly 😀
Thank you. I will do some modding and post back my results!
Thank you. I will do some modding and post back my results!
One more question actually - I was going to measure the frequency response before and after of the player (using a sweep burnt to a CD and Room EQ Wizard) - do you think differences will be measurable this way?
Sadly I don't have a 'scope or any other way of measuring things like jitter.
Sadly I don't have a 'scope or any other way of measuring things like jitter.
You will see absolutely no difference whatsoever, even if you use a scope. Strange but true. The differences that are audible are way below the resolving power of a scope trace using a sine wave.
The better opamps will sound different to the 4558's though.
Even a squarewave using the 4558 and a 4562 or 2134 will show no visible difference in a CD player... I've tried 😉
Last sentence and picture. If you fit one channel with 741's and one with LM4562's and overlay the traces there will be no visible difference.
The best sounding audio integrated opamps
The better opamps will sound different to the 4558's though.
Even a squarewave using the 4558 and a 4562 or 2134 will show no visible difference in a CD player... I've tried 😉
Last sentence and picture. If you fit one channel with 741's and one with LM4562's and overlay the traces there will be no visible difference.
The best sounding audio integrated opamps