[Leo V]

Okay, just finished some more mods! Older ones in blue, newer ones in red. New mods:

1) Removed & bypassed headphone coupling caps (only 5-6 millivolts DC offset) with silver wire.
2) Removed headphone muting transistors. Still no power on/off transients--the muting relay is definitely doing its job!

3) Removed the second pair of 470-ohm output resistors (R307/309), bypassed with silver wire in teflon jacket, directly from the RY201 switch to the output terminal.

I can't believe it, the sound really improved, presumably from the resistor bypass--it's like a veil was lifted from the speakers! Clearer, definitely smoother sound. I cannot wait till the other resistor pair (R239/240) gets replaced with 50-100ohm Holcos...
Leo

 

[mirlo]

Good to see that you are getting good results. I guess the real fun begins when you sub the op-amps.

But I wonder: how did you find the electronic copy of the service manual? I searched Sony's site and the links to the DVD stuff appear to be down.

[ I'm curious to find out if they have these on line somewhere. It would be interesting to study the difference between the "consumer level" product and the ES series. ]

Just curious,

-- mirlo

 

[Leo V]

Mirlo, I found this website mentioned in Asylum forums, and downloaded the manul from them for some $11. This saved me SO MUCH trouble, I hope Sony doesn't go after these guys for whatever reason! (Like being too helpful, hm...)

In the heat of the moment, I popped out the other two 470-ohm resistors (R239/240) and replaced them with ghetto Radio Shack 47-ohm resistors, direct from opamp outputs and into switch inputs:

It seems to help even more! The voices no longer sound muffled on my speakers, the bass seems tighter... soundstaging is improving, as the singer is now more prominently in the center. I'm still evaluating the effects of the resistor stuff...

As for the opamp mods, I'm just about the order the caps to go with them. My only question is whether to pick AD8620 or OPA2134 for the headphone preamp? It will be fed by the first AD8620 (which will be the line-out preamp).
Thanks,
Leo

 

[Budgie]

Tangent made some notes on his web site regarding the sound of these chips, so you may want to read that. I don't think you will go wrong with either chip.

Very interesting reading about your exsperience with the resistors. Makes me think it is time to crack open this Denon that just doesn't sound anywhere as good as my old Onkyo, and start poking around in it. I know it has resitors all over the place in the output lines.

Your work looks great! You have some soldering exsperience don't you?

 

[Budgie]

(Later that night-) Okay, I opened up the Denon (dcd-3520) and did a bit of reverse engineering (pain in the ass). Although this is a very well made player and near top of the line from when it was made, it has never really sounded that good. The circuit was basically the da's feeding three stages of NE5534AN or similer. I decided to see how it would sound if I took audio off the first opamp in the chain, which I decided to change while I had it apart. I soldered in two OPA228's and connected the output to 47 ohm resistors of the Radio Shack metal film type. I was able to open up a couple of resistors to free up the exsisting output jacks from the remaining circuitry. The other end of the 47 ohm resistors are soldered to the traces near the factory output jacks so the signal path is now only about two inches from the opamp to the jacks.
Then the bummer- This player has mute transistors! What the hell were they thinking? My Onkyo is about the same age, and was a much cheaper unit originally (but built very nice) and it has relay muting in it. At this time I have not removed the mute tranny's, but I am thinking about it. I hate it when a player has pop's and click's and such noise's. I guess I will try to figure out a relay mute to retrofit into this player.
I suppose I should mention the sound. It is much better now. I will give the player a chance in my main system later this evening. The sterile sound and lack of fine detail has pretty much gone away. I don't know if it can beat my beloved tweeked Onkyo, but it is definatly a much better sounding player now.
Thanks Leo, for the inspiration!

 

[Leo V]

After further testing, I'm hearing problems at 47 ohms... I did A/B comparisons using Line-out #1 (47 ohms) versus Line-out #2 (~500 ohms, it still goes through the 47-ohm Radio Shackles plus the 470-ohm resistors).

47 ohms sounds much "edgier" to the point where smoothness is gone and the highs are really harsh, brittle, and hoarse. It may be distortion. At the same time, 500 sounds smoother and more full-bodied, but lacks the "edge" (although not as suffocated as the original 940 ohms).

I'm going to compare again, using 1V output voltage instead of 2V. (Update: ) So far, Sony's 1V setting seems to solve the apparent distortion at 47 ohms... this may be the way to go with such a "hot" source...

If 1V won't fix it, I'll try some 100-ohm Radio Shack resistors later today...

Budgie, I've only done two soldering projects before so I'll take your comments as a compliment

I'm glad your Denon is sounding better. Note what Marc told me, however: the high resistance is *needed* because the muting transistors short out the circuit and may cause your opamp to overheat. So (he says) it's safer to remove the muting transistors *first* and then cut down on resistance. So be careful about your Denon, maybe you'd be safer popping out those transistors.

 

[Leo V]

I just tested 47 ohms with 1V outputs for 45 minutes, and I'm very impressed. The low-impedance distortion is all but gone with 1V output (my speakers allegedly have 800mV sensitivity, anyway...) This is as good as I've yet heard my speakers sound! Even the notoriously loud Aerosmith remastered SACD (worst harsh/brittle highs offender) is sounding decent now.

 

[Budgie]

No problems with the mute circuits as is. (Most preamps don't care if an input is shorted.) I have not decided on taking it out yet, though. I did solder in sockets for the opamps and have spent the morning trying different one. The OPA 627 does sound really good, but I am not certain it is what I want in there. The OPA 228 isn't in the running anymore. Nice chip, but did not do it for me. I am currently listening to older Motorola chips that I had in the junk drawer. MC34081P. Can't find a data sheet on it, but I doubt it is anything special, but it sounds pretty damn good. Tried the NE5534AN's which in alot of players, and yuck! Maybe it can't drive the cables directly. I will play around and see what other chips I have laying around that will drop in. This is kinda fun, trying different chips.

Sounds like you are happy with the results you are getting. It is nice you can select the output on the Sony, that is a nice feature. Do you think you will continue into soem of the other mods to see if they offer any improvement?

 

[Leo V]

Quote:

Originally posted by Budgie Sounds like you are happy with the results you are getting. It is nice you can select the output on the Sony, that is a nice feature. Do you think you will continue into soem of the other mods to see if they offer any improvement?

Of course I'll continue modding!

Now I am finally getting the "open" sound I really enjoy, I am all the more encouraged to do the other mods. The Black Gate caps are on their way, the Holcos resistors should be here anytime. I was thinking of reversing my current configuration by running a wire directly from the opamp to the switching box, and then running two pairs of resistors (of different values) to Line Out #1 and Line Out #2. Say, 30 ohms for #1 and 100 ohms for #2. Then the appropriate Line Out port could be chosen for driving different equipment.

I'd install a socket, except all my opamps are of the solder-on kind. After reading through Tangent's opamp section, it seems that AD8620 should be miles ahead of the OPA2134. While he didn't directly refer to AD8620, my general impression is that OPA2134 would offer another level of smoothness at the expense of detail. However, I prefer the detailed "edge" in sound, provided that it is already smooth.

Budgie, if you find any further interesting results comparing opamps in your setup, I'd like to hear about it

 

[Budgie]

I would describe the opa2134 as having more of a warm, smooth, "tube like" sound. The detail is still good, but a bit more of a relaxed presentation. It does sound like you will probably like a zippy-er chip instead of the 2134.

I have definatly decided on using that old Motorola chip. Sounds very natural and spacious. Excellent bass detail and crystal highs. Voices sound very realistic and the background detail is clear. No fatigue after several cd's. I am happy.

I look forward to hearing how it goes with the Blackgates. I have read that they take some time to burn in,though, so full results are suppose to take awhile.

 

[Leo V]

Budgie, in the meanwhile, could you offer me a suggestion? From my posted schematics you can see that my NC650V has Line-In connectors. However, it doesn't "officially" allow using those inputs unless you're connected to a Sony CD changer.

I am thinking of "enabling" this Line-In, so that I can get sound from my computer. My only concern, of course, is sound quality degradation. I could unplug the computer side of the cable when not in use, which may eliminate the computer-side noise...

The task seems easy... (refer to this diagram) -- the RY201 switch is the one that selects what gets output: the CD player's sound (default) or the externally connected signal (as seen in the diagram).

I'm not sure what actually needs to be done, though. Can I simply connect pins 5 and 6, as well as 7 and 8, on the RY201 switch (and play only one source at a time)? Or do I need something more sophisticated?
--Leo

PS: My built-in headphone opamp (IC207) is labeled either 556 or 5556 or 55564, it's very difficult to tell. Google searches turn up Philips or Motorola.

 

[Leo V]

What I'm thinking now, maybe there is a way to control the input to the RY201 output selection device. Maybe I could mount a switch (like a physical switch) onto the back of the unit, which would send the corresponding signal to RY201 to select the right ouput...

 

[Leo V]

Updates:

1) The headphone opamp is actually either "1556" or "1556A" or "15564".

2) I measured the pos/neg voltage feeding the opamps (for potential future Class A biasing), and found something strange: negative voltage (pin 4) equals -8.95V, but positive voltage (pin 8) equals +10.70V. What to make of this?

Is it reasonable to try biasing first with cheap Radio Shack resistors? (To evaluate whether it helps.)
Thanks,
Leo

 

[Budgie]

You should be able to use that second input by figuring out the switching scheme. I see a second relay that is also involved in that circuit that switches at the same time. I will see if I can make sense of it.

The headphone amp is probably a 4556 or equivelent with a sony part number. A goggle search for NJM4556 will turn up the data sheet. A very common opamp for headphone outputs. (Seems like someone took apart a Grado RA1 headphone amp and found that chip inside, if memory serves me.) I still think a opa2134 will outperform it. The power supply is kinda weird, but the non-symetrical power should not be a problem for an opamp.

I think a single resistor would be worth a try, to see how the class A trick sounds. I should sound better with an improved current source, like ppl's design, but why not try the easy way first? I will try this as soon as time permits, myself. You want to make sure the resistor is not too small and draws enough current to fry the opamp. The Jon Risch web page has good info about doing this, and mentions values for the resistor. http://www.geocities.com/jonrisch/index2.htm
(You do not want to do this to the headphone amp IC.)

 

[Budgie]

http://www.leov.net/~leo/sacd/block_diagram.png

The block diagram shows whats going on. The second relay is used to also send the input to the front channels output for the 5.1, so the eatra input will end up at both the stereo and the 5.1 front channel jacks. It looks like all you have to do is wire the switch to apply the control voltage to the control line.
(edit)- hold on, we need to figure out if it is a apply voltage or an apply ground to the control line.
(edit again)- Looks to me like you apply +5 vdc to the control line to select the input.