diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Digital Source (http://www.diyaudio.com/forums/digital-source/)
-   -   Marantz SA8001 Modification to Eliminate Distortion (http://www.diyaudio.com/forums/digital-source/170560-marantz-sa8001-modification-eliminate-distortion.html)

whaleman 21st July 2010 02:55 AM

Marantz SA8001 Modification to Eliminate Distortion
 
3 Attachment(s)
Here is a mod that solved quite a bit of distortion coming from the analogue side of my Marantz SA8001. The gain amplifier on this machine is quite hot leading to a really nice output, but in many newer CDs, this sends the equalizing amp over the top and it results in distortion inside the unit and in my older amps due to the too much output during loud passages, especially voices.

The mod involves replacing the 4 1.2K ohm resistors (R651 R652 R601 R602) that interface the DAC and the equalizing amp. Ideally, you can probably use 2.2K, 4.7K or 6.8K ohm which is what I used.

All distortion due to too much gain is now gone and it will not over drive my amps inputs as the signal is attenuated by about 25% or so.

Proceed at your own risk, I would not touch one of these units unless you have the skill and a problem already.

SoNic_real_one 21st July 2010 02:01 PM

1 Attachment(s)
You basically reduced the input signal. Because the noise of the HDAM stage will remain the same, you decreased the Signal/Noise ratio. It might be still OK, but it is a shame in my view.

I would rather lower the negative feedback resistor to decrease the gain or add a divider somewhere close to the output.
I don't have your schematic, but some similar one - red mark the places where I would make the above changes. It is also easier since you keep balanced stage... balanced (you have to match only 2 resistors, not four).

whaleman 22nd July 2010 04:35 PM

1 Attachment(s)
Quote:

Originally Posted by SoNic_real_one (Post 2249629)
You basically reduced the input signal. Because the noise of the HDAM stage will remain the same, you decreased the Signal/Noise ratio. It might be still OK, but it is a shame in my view.

I would rather lower the negative feedback resistor to decrease the gain or add a divider somewhere close to the output.
I don't have your schematic, but some similar one - red mark the places where I would make the above changes. It is also easier since you keep balanced stage... balanced (you have to match only 2 resistors, not four).

Your idea about the feedback sounds good, the output from the DAC was way overcoming one of the amp stages downstream, rather than to try to single it out, I just went for the "top kill" and it worked. I could have probably used lesser resitors, but had 4 matched metal films in the drawer.

The sound before was unbearable on about 75% of my CDs. A remarkable improvement now, but as you state, I did sacrifice STNR level. The good thing is that new CDs are maxed out in signal and do distort in some instances on my other players, but not on this one after the mod.

I have attached a schematic, would love to see what you recommend for the negative feedback resistor replacements, I may try it down the road. You'll see the DAC on the left side and RCA outputs for the 8001 on the dog legs all the way to the right. Signal paths are highlighted.

Only question I have about your suggestion is if the distortion occurs early in the differential amp stage, does changing the HDAM feedback resistor have any effect there? Thanks.

SoNic_real_one 22nd July 2010 07:55 PM

I would try first to attenuate the signal at the end - a resistor in place where C618,668 empty pads are (NC means not connected). Attenuation there is in a place with the highest level of signal so won't be that bad for SNR.

Other place would be feedback resistors R641, 691. I doubt that the whole thing is distorting due to high levels, rather your amplifier is to sensitive. That's why I won't change nothing drastically like the feedback or the differential stage.
If you change the R615,655 in the diff HDAM, then the filter freq change too, so that is a more involved change.

whaleman 22nd July 2010 09:03 PM

Quote:

Originally Posted by SoNic_real_one (Post 2250947)
I would try first to attenuate the signal at the end - a resistor in place where C618,668 empty pads are (NC means not connected). Attenuation there is in a place with the highest level of signal so won't be that bad for SNR.

That was my first thought until I heard the distortion in the headphone tap that goes right before this point.

Quote:

Originally Posted by SoNic_real_one (Post 2250947)
Other place would be feedback resistors R641, 691. I doubt that the whole thing is distorting due to high levels, rather your amplifier is to sensitive. That's why I won't change nothing drastically like the feedback or the differential stage.
If you change the R615,655 in the diff HDAM, then the filter freq change too, so that is a more involved change.

I agree with you on making the least involved change. The distortion was for sure inside the unit as I attenuated the end signal right at the RCA outputs to eliminate the amp, and no change. That left the other end as least involved and it works beautifully now. Come to think about it, the STNR was quite low before this mod.

What is a puzzle is why this mod was needed to make the unit listenable.

Mooly 24th July 2010 06:40 AM

Seems very odd. CD has absolute limits on max output levels and it seems inconceivable that this is a design error. I would look for other causes, even something as simple as a supply issue to the analogue stages (low rails). A scope check would quickly reveal the problem using a test CD at 0db level. No music signal can ever exeed that level.
It could also be an issue with the muting circuit just quickly looking at the circuit. Easy to remove the four mute transistors to check.
Whatever the issue is... I think you are just fixing the symptoms and not the underlying cause.

SoNic_real_one 24th July 2010 02:22 PM

I agree, the CD player should NOT distort on the headphone output (with any CD's).
I was under the impression that your amp has the input level to low...

stvnharr 24th July 2010 07:08 PM

Hm,
I've been reading this thread and just wondering............................
The Marantz 8001 has the exact same analogue stage, after the dac, as the 7001, 8260 and 8400 players, right down to the spelling of AUDIO SUPLLY on the board. There are quite a number of these players floating about. Never heard of an issue as described here. Usually the issue with these is a transport/laser issue. It all sounds to me like something downstream from the player.

Replacing the resistor on the dac output and reducing the voltage output from 2.78V to about 1.5V doesn't seem a good idea as the filter circuit is based on the incoming 2.78V. Changing feedback resistor seems okay but maybe not do too much.

If you want better sound, always good to bypass the muting transistor circuit by just removing the wire link and running wire, or 75R resistor, from output cap to rca. You get a "click" sound when disc is ready to play, but the sound is easily heard to be better.
For the best sound out of these players, best to replace the whole analog stage after the dac. But that's not really the issue in the OP.

stephensank 26th July 2010 03:45 PM

Looking at the OS schematic, there is NO WAY that circuit should be able to clip at 2.8V RMS. It should be able to swing damn near 7VRMS or so without clipping. I
So, I quite firmly believe that the problem is actually the muting transistors. I know for a fact that 2SC2240, an otherwise nice transistor, does not like being connected as an 'upside down' muting transistor, with grounded collector like this. When thus configured, they will ALWAYS eventually starting leaking current to ground.
I very strongly suggest that you either disconnect them, or replace them with 2SK117 or 2SK170 jfet, provided the mute line goes negative when unmuted, which it darn well should. I do this on all Nakamichi Dragons I rebuild & they work brilliantly; totally silent operation with no current leakage at all, and no discernible effect on sound. This is to replace the 2SC2878's, which do hold up reasonably well in that funky grounded collector muting config, where I used to install 2SC2240's in grounded emitter config, but the jfet's work better & never go bad.
Other than the stupidly chosen muting xstor config, looks like a nice output stage to me, if rather conventional.

SoNic_real_one 26th July 2010 04:52 PM

BTW, do you have any experience with the KTC2875B (lesser version of 2SC2878A/B?)for the muting transistors?

I think you are wrong is some aspect - the usual configuration of muting transistors seems to be "backwards", because of the normal negative base voltage bias - which in fact polarize them "normally", with the collector being more "positive" compared with the base.
The only non-conventional voltage is the emitter one, that fluctuates around zero, with no base-emitter current.
On the other hand, I never saw JFET's in that position - I would assume the are worse (Schottky noise?)


All times are GMT. The time now is 08:15 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2