Hey all,
I’m starting a new thread on this as I can’t seem to find a definitive answer on it anywhere so apologies if this is common knowledge!
For the two opamps on the output stage on a CDP (specifically a Marantz CD63), would a FET or bipolar input be more suitable? I can see both types being readily recommended in the CD63 mods thread specifically but with seemingly no preference either way.
Both the original NJM2114D’s and the LME49720NAs I’m using now are bipolar so if there’s no good reason a FET might be an “upgrade” I’ll probably stick to the original.
I’m starting a new thread on this as I can’t seem to find a definitive answer on it anywhere so apologies if this is common knowledge!
For the two opamps on the output stage on a CDP (specifically a Marantz CD63), would a FET or bipolar input be more suitable? I can see both types being readily recommended in the CD63 mods thread specifically but with seemingly no preference either way.
Both the original NJM2114D’s and the LME49720NAs I’m using now are bipolar so if there’s no good reason a FET might be an “upgrade” I’ll probably stick to the original.
An FET input op-amp won't be better merely because it utilizes FETs. Setting aside the valid fact that most all modern audio op-amps feature supposedly inaudible performance metrics, should you want to 'roll' op-amps, I think that you're going to have to try different devices and see which your ears prefer.
Cheers for the reply. I understand it won’t necessarily be “better” due to that fact but I’m not sure which type would be better suited to the application if I was to choose one or the other.
I’m not interested in “rolling” really but looking at the Muses opamps... the 01 and 02 seem to be sold as the same apart from one is FET and the other is bipolar. Lot of money to roll the dice on so thought I’d ask
Then, your question pivots on the design of the surrounding circuitry. Typically, a surrounding circuit which is designed for a bipolar input op-amp will work at least "okay" with a FET input op-amp. However, the reverse may be more problematic, as a circuit designed for a FET input may not work okay with a bipolar input. Bipolar input op-amps usually require an external bias current path, which FETs do not require. In addition, FET input op-amps are often selected when the driving source impedance is high, which could result in high current noise if used with an bipolar input device.
Thank you. That clears things up - explains why people might be getting good results from FET types but I’ll probably just stick with bipolar types then with that in mind!
FET input stages are reputed to be better behaved in the presence of HF noise and hash and this is one reason that they can be favoured for this duty in CPD audio stages.
I've certainly had excellent results in the past with changing devices such as NE5534's to FET types. Try it would be my advice.
I've certainly had excellent results in the past with changing devices such as NE5534's to FET types. Try it would be my advice.
Interesting thank you... people do seem to have had good results using FET types in this particular CDP. If the results in all other aspects are equally good it sounds like a FET type might possibly be preferable here.
Not at the moment - when I originally modified the player years ago I soldered the LME49720s directly thinking I’d never need to change them (lol). I do plan on fitting sockets before changing them though.
Fet opamps would sound darker and softer, the sound would be low frequency dominant. Sounstage would be bigger and the sound would be less aggressive and harsh (with some recordings) than with bipolar opamps. But there would be a huge loss of resolution, fine detail, ambience and excitement. I tried fet opamps (OP275 and OPA2134) but I could not stand to use it more than a few hours and reverted to bipolars. NJM2114 is spectacular opamp and does not need any "upgrade".
From testing drop in trplacement
I had performed several board surgeries replacing opamps. Then used single tone tests to see what results from this. Never swapped a bipolar for a fet input opamp, only a fet for a bipolar. And swapper bipolar for bipolar. Paid attention to speed and GBW to minimize oscillating.
Results were consistent. Slightly higher noise with the fet input opamp. Also the highest distortion changes. With bipolar input the 3rd harmonic is the highest. With fet the 2nd order is the highest.
The trend changes per application, but odd order is usually higher with bipolar.
The fet input opamps generally have a spectrum that decays with frequency and less high peaks and missing fundamentals. Just a gradual decay. Prettier.
These differences should not be audible. These peaks are way down in level.
I decided power supply changes are more apparent than opamp swaps. But if stock is done well to begin with these do not help either.
I had performed several board surgeries replacing opamps. Then used single tone tests to see what results from this. Never swapped a bipolar for a fet input opamp, only a fet for a bipolar. And swapper bipolar for bipolar. Paid attention to speed and GBW to minimize oscillating.
Results were consistent. Slightly higher noise with the fet input opamp. Also the highest distortion changes. With bipolar input the 3rd harmonic is the highest. With fet the 2nd order is the highest.
The trend changes per application, but odd order is usually higher with bipolar.
The fet input opamps generally have a spectrum that decays with frequency and less high peaks and missing fundamentals. Just a gradual decay. Prettier.
These differences should not be audible. These peaks are way down in level.
I decided power supply changes are more apparent than opamp swaps. But if stock is done well to begin with these do not help either.
Interesting observations thank you.. the effect you describe with the FET units certainly isn't what I'm going for. It's a nice lively sounding player as it is and I don't really want to diminish that.
I might get a few ordered and see how I go on 🙂 I was looking to try the Muses01 or 02 out of curiosity but I really wanted to just pick a type and stick with it rather than "rolling" them as my budget doesn't stretch to 2 of each type lol!
Maybe I will try out some cheaper FET types and see how they compare to the LME4970NAs before making the decision.
Again interesting thank you... the player is heavily modified in terms of the power supplies so is hopefully I've got a good basis to start improving further.
If you don't try them (devices like the OPA2134 etc) then you will never know and will always wonder. Fit sockets, and whatever devices you choose be sure to listen for a few days to form a true opinion. Trust what you hear.
Back in the real world, none of this exists, its all in your imagination. Noone can
discern the difference between different quality opamps at line-level from blind listening tests as the distortion levels are well below 0.01% and frequency response is flat as an extremely flat thing, and noise is inaudible.
If there is a difference, you have a faulty/counterfeit opamp or no decoupling or you've overloaded it.
I would beg to differ Mark. The output stages of CD players are one area where different opamp devices and types do make a big... big being a relative term of course... difference.
Opamps used in these locations are being used under conditions that differ markedly from simple 'gain blocks' in normal amplification stages where the wanted signals are 'clean' and free from HF noise and hash.
Opamps used in these locations are being used under conditions that differ markedly from simple 'gain blocks' in normal amplification stages where the wanted signals are 'clean' and free from HF noise and hash.
This thread is one of the reasons I find the hobby so fascinating 😀
It seems fairly well established in the CD63 thread that opamps can make good improvements right up to the Burson discrete types.
I ordered some of the Muses opamps and while I feel like I must be absolutely insane to have spent that sort of money on 2 opamps, I'll trust my ears when they do arrive. The player has had almost everything done to it at this point so I figure if I'll ever notice a difference in opamps, it'll be in this CD player. I ordered two of the Muses01 which is of the FET input variety after your post Mooly and some of the reports in the CD63 thread - including one alluding to the FET type working slightly better with the HDAM bypassed (which mine is).
It seems fairly well established in the CD63 thread that opamps can make good improvements right up to the Burson discrete types.
I ordered some of the Muses opamps and while I feel like I must be absolutely insane to have spent that sort of money on 2 opamps, I'll trust my ears when they do arrive. The player has had almost everything done to it at this point so I figure if I'll ever notice a difference in opamps, it'll be in this CD player. I ordered two of the Muses01 which is of the FET input variety after your post Mooly and some of the reports in the CD63 thread - including one alluding to the FET type working slightly better with the HDAM bypassed (which mine is).
I've just looked... I knew they were dear but how much 😀 😀 😀
There is always the TLE2072 Excaliber, you could have ten of those for one Muse 😉
There is always the TLE2072 Excaliber, you could have ten of those for one Muse 😉
I know, momentary lapse in sanity I think but only one way to find out whether they’re worth it or not 😀 I’m sure they’ll be easy enough to move on if I’m not impressed!
Will let you know how I get on with them - at least they’re cheaper than Bursons these days 😀