|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
81.246.236.6
Hi all,
I recently recapped an old TDA1541-based Revox B226 CD player.
While the sound has globally improved, I'm annoyed by high frequencies that are grainy, splashy, "dirty". I use horn speakers wich tend to emphasize said faults.
What chances do I have to improve that?
I have LM4562 opamps on the way. I will fit those plus decoupling caps on the 4 and 8 pins to ground, as recommanded. But I doubt it will cure the problem.
Reclock? would that help?
Separate feed for the SAA7220 ( I feel that chip might be the culprit)?
If I reclock, I saw this clock module wich clocks both SAA7220 and separately at half frequency the TDA1541. Would that help?
New caps are still a bit fresh. Probably less than 50 hours of playing. Would more burning-in improve things?
I know it's an old player with only 16bit and 4FS. But it's got good reputation, and so do his Studer A727, A730 cousins sharing the same chips. I have the feeling it can be improved further.
Follow Ups:
Some pics of the recapped player.
I know most of you will shout in awe seeing the output coupling caps standing back to back in a strange way above the PCB. I know it looks very unprofessional. What I did was make some standoffs made of 7mm slices of cable isolation gain, the two legs of the caps going to the PCB trough it, and the caps laying on top of it. It feels solid enough, sounds good, and I don't have anymore this afterthought of having a polar electrochimic on the signal path. It's reversible of course, and I plan to buy a second B226 decoder board on the big auction site or somewhere else in case I wanna sell the player, or in case this board dies from my own hands (wich won't happen I hope). i wish i could take closer pics.
closer
general view
Until now I had replaced all the caps besides the regulators output caps; someone had told me it was best to left the old ones because modern caps would most likely have too low ESR and the regulators might become unstable and oscillate.But today I replaced them all with same values as originals but elna silmic II. From what i've read Silmic II doesn't have particularly low ESR compared to modern low ESR caps; so I took a chance and gave it a try. (I've also seen examples of recapped ReVox players wich had been recapped that way with no problem).
It seems that doing this, plus having simulated a bipolar electro cap by connecting two back to back (as the output coupling cap) instead of the single chimic that was stated in the SM, cleaned the high frequencies quite a lot.
Not sure about the simulated bipolar cap (might be placebo effect) but replacing the ouput caps on the regulators seemed to have a big effect. Probably the originals had badly drifted?
I'm not saying the high frequencies are now up to high end modern players, but at least its not agressing my ears with a splashy, distorted highs, and i think it's as good in that department as my 1996 unmodded Kenwood DP7090 player (obviously not high end, but a very good value "mid-fi" player at the time).
I'm still open to suggestions as to what could be the best improvement without having to destroy the board completely. A separate PSU rail for the SAA7220, or for the TDA1541, or for both, sound like the way to go on those old players, but I need something that doesn't require engineer skills... so i'm not sure.
Thorsten, I will thinker a bit and will send you an email, thanks! :)
Edits: 04/21/16
Elna Silmic II take over 150 hours to burn in and they will become smooth as I recall when I installed them in my PHILIPS 15 years ago. They did not work well so I went back to Elna Cerafine since they were the stock caps. As someone mentioned below, the power supply could be an issue here and I suggest getting good quality filter caps. Perhaps some of the experienced members can suggest some brands to use. Just an advice here, you can be chasing a sound that you cannot achieve with old players.
t.
It is a Band-Aid solution but you could try a tube cathode follower.
Dave
Hi,
The Revox Players are probably the worst examples of TDA1541 Players, short of Philips own lowball budget efforts.
The powersupplies are grossly incompetent (by the standards of the day) and this is were your problem lies.
Solving them means additional independent (new Transformer etc) supplies for the bits that need clean power. New Op-Amp's that are worse than the original NE5534 in this application are not helping.
Get the service manual, analyse the circuit, if afterwards you insist to proceed, e-mail me directly.
Ciao T
At 20 bits, you are on the verge of dynamic range covering fly-farts-at-20-feet to untolerable pain. Really, what more could we need?
The subjective effects you describe are classic bad digital sound, almost certainly due to some implementation choice(s). The potential culprits here are many, not all of which are easily modded after market. You wouldn't have the schematic to post, would you?
_
Ken Newton
I do have the schematics. It's a PDF file, not sure how to post it here?
I doubt the B226 can be described as "badly implemented digital"; it sure dates (1987) but was one of the best at the time...
That said, the SAA7220 chip that provides clock and oversampling for the TDA1541 is said to be very noisy...
I just did some experiment wich seemed to help clean up the treble (but it might be my imagination): seeing as the output coupling cap is a polar electrolytic (silmic II), I added a second, back-to-back capacitor in serie, opposite polarity (-++-) to simulate a bipolar. It seems that high frequencies are now more fluid and cleaner. But it might just be in my head, or placebo effect.
If you have a PC open the pdf and makevit occupy as much of your screen as possible. Then press ctrl alt and print screen at the same time. Open paint and then click edit and paste and there is the pdf. Crop as needed and save as a jpeg.
ET
You can open the PDF file with a PDF application (Apple's Preview in my case) and export it as a JPEG file, which can then be uploaded to AA.
I'm not suggesting that it's badly implemented, merely conventionally implemented. Which, in itself, is sufficient to account for the sound character that you describe. If you can post that PDF it may help.
_
Ken Newton
Edits: 04/19/16
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: