Marantz CD63 & CD67 mods list

My first impression is: yes.
It's much solid-state-with-lots-of-feedback vs. less solid-state-with-some-local-feedback, and the latter wins. Maybe the next level will be to compare the sound of different types of transistors? :clown:

The only drawback is that I have to go look for good decoupling caps again, since there's plenty of DC at the output now :mad:. I think i'm going to use a 2u2 Auricap for both, or maybe a BG for the transistor stage? That one can be low voltage :).

Ray.
 
6h5c said:
My first impression is: yes.
It's much solid-state-with-lots-of-feedback vs. less solid-state-with-some-local-feedback, and the latter wins. Maybe the next level will be to compare the sound of different types of transistors? :clown:

The only drawback is that I have to go look for good decoupling caps again, since there's plenty of DC at the output now :mad:. I think i'm going to use a 2u2 Auricap for both, or maybe a BG for the transistor stage? That one can be low voltage :).

Ray.

This sounds very interesting, keep up the good work ;)

Brent
 
6h5c said:
My first impression is: yes.
It's much solid-state-with-lots-of-feedback vs. less solid-state-with-some-local-feedback, and the latter wins. Maybe the next level will be to compare the sound of different types of transistors? :clown:

The only drawback is that I have to go look for good decoupling caps again, since there's plenty of DC at the output now :mad:. I think i'm going to use a 2u2 Auricap for both, or maybe a BG for the transistor stage? That one can be low voltage :).

Ray.

how much dc?

allan
 
The tubestage does between 110 and 145V DC, depending on how much current the triodes are biased. The signal is taken from the anodes, so that's obvious :-D.

The transistor stage has about 6V offset at its output. The voltage on the collectors of the differential input stage is about 7V, and that's buffered by a darlington emitter-follower, so minus 2 * Ube.

Regards,

Ray.
 
wojtek.l said:
Thank You Ray. What do You now prefer - tubes or transistors ?

You're welcome. Actually i've done some listening today. Both stages are very close, and sound very good. I find them both sounding better than the original opamp based filter, that's for sure. But with their own characteristic sound. Of course there are also differences between the components I used that could cause that. At first I used Nichicon Muse (ES) bipolar output caps for the transistor stage, and I was a bit disappointed with the sound. But then I changed the caps to 4,7uF MKC's that I had laying around, and that was a big improvement.

The actual active circuit has become so minimalistic that the other components in there become much more important. So in the end I think that will play a great role in how the sound of either stage will be.

Regards,

Ray.
 
philpoole said:
Hi Ray,

I saw the article on ultranalog.com, but couldn't find the schematic anywhere. Surely you could have done a discrete equivalent of your tube stage?

Cheers,
Phil

Hi Phil,

The tubestage is not taken from Ultranalog, only the transistor stage. The schematic is here on the second page of the SACD Enhancer article. I changed it slightly to get a bit more output headroom, and put the passive filter stage in front. Here's the schematic I used. But both the tubestage and the transistorstage a similar: following the passive filter is a differential input stage with current sink in the tail to combine both DAC outputs, and in case of the transistor stage a buffer for low output impedance.

Regards,

Ray.
 

Attachments

  • cd67_enhancer.pdf
    65 KB · Views: 221
Hi Ray,

Didn't realise there was a second page. D'oh!

I did do a port of your tubestage to transistors a while back.
Its refreshing to see I wasn't too far off.
It was like yours, but without the buffering. My current source used two transistors, one replacing your two diodes (akin to the ultranalog) and the current source resistance I calculated was 300ohms.
I didn't worry about a buffer output in my design, but realise now it might be useful, considering the signal does need to travel down an interconnect.

I think the outout stage could be improved with a series complimentary pair (although, it might be similar already to that darlington-esque arrangement {showing my ignorance there!}) using a current source. I saw great improvements with a headphone amplifier by improving a simple emitter follower this way - but it was driving a very different load.

I'm really distracted with decorating and stuff, but REALLY want to give this a try.

Glad it sounds better than the opamps. 'ear witness' reports to back up a hypothesis are always welcome.

Cheers,
Phil
 
Hi Ray,

Just looking at your schematic, and desperately trying to remember theory...
I think R17 is redundant. Mainly because Q2 and Q3 are trying to follow each other's base voltage(I know that description is wrong!) and R20 and R21 can be removed in some circumstances. My design didn't have them. I can't remember the benefits of keeping R20 and R21 (there are situations when useful, and some when not, perhaps frequency dependent).
I'm going to have to dig out my books tonight to back up my feeble statements with some real facts! However, there is potentially room for some tweaks.

I love transistors (the active circuit here potentially sounds better than a £15 op amp, but costs maybe 75p in parts), but I can never do anything with them without a textbook by my side!

Cheers,
Phil
 
philpoole said:
I think R17 is redundant. Mainly because Q2 and Q3 are trying to follow each other's base voltage(I know that description is wrong!) and R20 and R21 can be removed in some circumstances. My design didn't have them. I can't remember the benefits of keeping R20 and R21 (there are situations when useful, and some when not, perhaps frequency dependent).
Cheers,
Phil

Hi Phil,

Yes, this stage is cheap and simple, and it sounds good too! It's a lot less efford compared to a tubestage.....

R17 is not needed to keep the circuit working, that's true. BUT....the voltage across Q2 will not be the same as Q3, because the voltage drop across R17 is gone. As you may know, the point a transistor is biased in also is influenced by it's Vce, so for symmetry reasons I tend to keep this 'redundant' resistor, just like in the tubestage.

R20 and R21 can also be removed theoretically, but you will find the circuit then amplifies way too much (140x/43dB in my simulation). These resistors act as a local current feedback, and can be used to set the amplification. As you may have noticed, the original SACD Enhancer uses a higher 8k2 and a 5k pot there, and therefore it has less amplification. The passive filter attenuates the signal with some 6.4dB, so I lowered them to 820R to maintain 2Vrms output level.

Regards,

Ray.
 
rowemeister said:
I wonder if some nice descrete component cards could be made up to plug direct into the opamp sockets. :)

Hi Brent,

I suppose it could be done, especially if you use SMD. But if you use the passive filter you can bypass the opamps and use the modified HDAM instead. Whatdoyouknow....maybe HDAM will become useful after all.....

Ray.
 
6h5c said:


Hi Brent,

I suppose it could be done, especially if you use SMD. But if you use the passive filter you can bypass the opamps and use the modified HDAM instead. Whatdoyouknow....maybe HDAM will become useful after all.....

Ray.

Having not spent any time working with the HDAM and just bypassing it. What would it sound like with no opamp and DAC straight to HDAM?

Will the gain and filtering be all wrong

Brent
 
Hi Brent,

You will have some signal, but the first opamp amplifies 2.7x, so it will be a bit low. There will probably be some noise residue because the filter part around the second opamp is missing.

Just yank out the opamps and jumper pins 3 and 7. You'll have to use a coupling cap then, because of the DAC's offet voltage. You should try my passive filter part + modified HDAM :D.

Ray.
 
6h5c said:
Hi Brent,

You will have some signal, but the first opamp amplifies 2.7x, so it will be a bit low. There will probably be some noise residue because the filter part around the second opamp is missing.

Just yank out the opamps and jumper pins 3 and 7. You'll have to use a coupling cap then, because of the DAC's offet voltage. You should try my passive filter part + modified HDAM :D.

Ray.

Just looked at the op amp circuit, the gain of 2.7x is from the 27K and 10K resistors then :D

I'll have to have a go with the passive filter

Brent
 
Hi Ray,

would your discrete output stage work on a cd6000? Think the main difference between the two is that the cd6000 has two dacs not one. If it should be ok to use, should I proceed it with the filter you have designed, or will it not be appropriate for cd6000?
Wher do you get the inductors from, and should they be high precision, or matched components?


Cheers arthur