clock upgrade for Rotel 965BX works !!!
Mar 7, 2007 at 9:25 AM Thread Starter Post #1 of 9

budx3385

500+ Head-Fier
Joined
Nov 7, 2005
Posts
547
Likes
34
WOW. I just have to share this experience. I upgraded the clock in my vintage Rotel CD player, and it's now a winner source !!!

I bought the Rotel 965BX in 1995. It was my first "audiophile" component, along with an Exposure XV integrated and Epos 11 speakers.

The Rotel had good resolution, but it was hard-edged and had digital "glare". When I cleaned out the garage at Christmas, I listened to it, thought about selling it for $100-200, but wondered whether it could do better. I scanned a lot of threads, and I decided to upgrade it myself.

I bought the Rockclock 4 from net-audio.com. There are other DIY clocks - kwak, LC, trident, etc. All the websites describe the effect of removing jitter by stabilizing the clock pulse. It makes a lot of sense. The Rotel had the very simplest RC-crystal clock circuit, and I was convinced that they were all telling the truth. I decided they would all make a difference, and perhaps one clock might be better, but that the differences between the various clocks would be WAY smaller than the primary effect of upgrading the clock.

I chose the net-audio clock because it came with a metal shield, because it had itself been upgraded at least three times, because it came as a kit that included two new caps and opamp sockets to replace the original NE5554s, and because the website gave very clear, illustrated instructions for installing it in the very same CD player, the Rotel 965BX. Also, it was far less expensive than most other clocks. I called and spoke with David, too, and I was convinced I could have his help if I needed further guidance. Along with the Rockclock, I also ordered the clock power supply upgrade, silver solder and the gold RCA connectors offered on the same site.

I did take a deep breath before tackling the mod, because I had to remove the CD drawer in order to unbutton the PCB to get to its backside. But it all went smoothly. Adding the RCA connectors made an audible difference. I swapped the opamps (OPA604s, OPA627s, AD797s) and finally settled on AD8065s.

The results are phenomenal. After letting it burn in now for a week (very important), the SQ is VASTLY improved. The "glare" is completely gone. It is now smooth, detailed, dimensional, extended, clear, and balanced. It has a perfect soundstage with stable focus and air. It has good frequency extension and excellent timbral depth. Whereas before it had been emotionally off-putting, it is now emotionally engaging - I sit and listen to the whole CDs, not just the usual test tracks. It lacks only that last oomph of palpability, the heart-wrenching presence that I have with, say, my Audio Aero Capitole 24/192. But DUH, that costs ~ 50x as much ! The Rotel now drives the SP PPX3 SLAM > Equinox > HD600s at my listening / reading chair just fine, thank you!

The next step is installation of the VDC power supply, which is supposed to make another audible difference by providing clean and isolated DC power to the clock. And I've already ordered the Burson discrete opamps, which should just plug in to the 8-pin sockets. Gee, this is FUN !!!

I hope this will encourage other DIYers. Most of the DIY threads I found on clock upgrades were old. This is worthwhile, believe me. For example, I remember what the Arcam and Musical Fidelity players sounded like when I auditioned them in my system just a couple of years ago, and this is WAY better, for a mere pittance, and besides, I did it myself.

600smile.gif
600smile.gif
600smile.gif


IMHO, of course.
 
Mar 7, 2007 at 2:51 PM Post #2 of 9
WTG. Makes a pretty big difference, doesn't it?

I just upgraded the clock in my Rotel RCD855 a couple of weeks ago. I used a Tentlabs XO module (the actual oscillator, not the complete XO2/3 board) and built a PinkFishMedia.net Flea board (low noise PSU) that contains and powers the XO. I power the Flea with one of Tangent's Tread boards with its own transformer. I also am clocking the TDA1541A DAC chip directly (/2 circuit), which takes the SAA7220 digital filter out of the clock distribution circuit. The 7220 is a known noisy chip and probably the reason why people hear so much difference when they switch to NOS as a clean clock going into the 7220 will come out noisy and is then distributed to the DAC and receiver chips). Total cost: ~$37 for the Tent XO, ~$20 in parts to build the Flea, $5 for the Tread board and the cost of the transformer.

Next mod: DIY discrete output stage. I have already rolled several opamps thru it, but am convinced that discrete is the way to go.

Chris
 
Mar 7, 2007 at 3:59 PM Post #3 of 9
Quote:

Originally Posted by Pars /img/forum/go_quote.gif
I also am clocking the TDA1541A DAC chip directly (/2 circuit), which takes the SAA7220 digital filter out of the clock distribution circuit.
Chris



Neat, Chris. That's much more advanced, and really cool. Thanks for the link on the PS. I'll be reading up on pinkfishmedia now. Do you have a link explaining the digital filter bypass that clocks the DAC directly (/2 circuit)?

Like you, I was also convinced by opamp rolling that a discrete circuit will make a huge difference. I've been wondering about the caps on the PCB just in front of the opamp sockets that filter its power input, too --- David at net-audio made a comment about replacing those. I hope to receive the Burson amps by this weekend. I'll let you know.

Bud
 
Mar 7, 2007 at 4:24 PM Post #4 of 9
Quote:

Originally Posted by budx3385 /img/forum/go_quote.gif
... Do you have a link explaining the digital filter bypass that clocks the DAC directly (/2 circuit)?


Bud,

It basically just involves cutting the trace between the 7220 (which is what the stock clock feeds) and the DAC, then using a D flip flop (74HC74) configured for /2 to feed the DAC directly (100 ohm resistor, although a 47 ohm probably would be fine). The player is still running OS, BTW, the clock doesn't get mucked up in the filter chip anymore. In my case, it didn't immediately work (nice digital noise
eek.gif
), and I needed to insert a bit more delay by running it thru an inverter. The frequency that the DAC is expecting will also vary by player/chip set/architecture.

The caps around the analog stage in mine are BG standard. I replaced the output caps with BG NX and removed the muting transistors, but I don't see a big need to replace the others (yet
biggrin.gif
).

Chris
 
Mar 7, 2007 at 9:58 PM Post #5 of 9
Gratz! Makes me want to invest in a good CD player. I really should but I am content with my iPod right now.
 
Mar 8, 2007 at 6:45 AM Post #6 of 9
Quote:

Originally Posted by Pars /img/forum/go_quote.gif
Bud,

In my case, it didn't immediately work (nice digital noise
eek.gif
), and I needed to insert a bit more delay by running it thru an inverter. The frequency that the DAC is expecting will also vary by player/chip set/architecture.




Yup, Chris, that sounds more advanced than me. I don't have a 'scope here at home, so no chance that I'd pick that up. I see that net-audio sells a NOS replacement for the 7220, though, so maybe I should call David and see whether that might help.

Right now I'm sad, because I installed the "VDC" clock PS this evening, and the sound became worse. It has a slight sibilance now, which sure smells like a timing problem. Can it be that the DC voltage is slightly different (I didn't measure and compare) and now the new clock needs to "break in" again? I'm letting it sit "on" so that maybe it'll settle down. Else I'll take the PS back out and wait patiently for the Bursons.
 
Mar 8, 2007 at 7:46 AM Post #7 of 9
Hmm...you'd probably need a scope to figure out where the problem lies. Maybe the PSU is injecting noise into the circuit and you're getting ill effects from that
frown.gif
 
Mar 8, 2007 at 2:46 PM Post #8 of 9
Quote:

Originally Posted by budx3385 /img/forum/go_quote.gif
Yup, Chris, that sounds more advanced than me. I don't have a 'scope here at home, so no chance that I'd pick that up. I see that net-audio sells a NOS replacement for the 7220, though, so maybe I should call David and see whether that might help.

Right now I'm sad, because I installed the "VDC" clock PS this evening, and the sound became worse. It has a slight sibilance now, which sure smells like a timing problem. Can it be that the DC voltage is slightly different (I didn't measure and compare) and now the new clock needs to "break in" again? I'm letting it sit "on" so that maybe it'll settle down. Else I'll take the PS back out and wait patiently for the Bursons.



Bud,

The part that didn't work immediately was reclocking the DAC. The master clock worked right away (which is what you are doing). Lowered jitter should make the highs smoother and more "analog" like, as well as tighten up the bass. From my experiences clocking the DAC directly, it pretty much works or it doesn't... the digital noise I got at first was just that... static, no music (well, very faintly I could hear some music in the left channel). I also don't really believe in break in
eek.gif


Unfortunately, the equipment required to measure jitter is really expensive (not even sure what they use for the plots I have seen... spectrum analyzer?), so there is no good way to verify that the clock you just paid for is working up to spec or not, other than what you hear. I do know that noise (even PSU) and clocks are not compatible. If you read the thread on pinkfishmedia, you probably ran across the board redesign that I initially did putting the 74' series logic on the flea board (PSU), and was told not to do that (I did it anyhow for now, but that isn't the point). That is one reason why many clock upgrades recommend a completely separate PSU for the clock. Some even go to the trouble of powering the DAC chip itself with yet another supply. The Pedja Rogic DAC I am building has 7 or 8 supplies on it, each fed by sep. transformer windings. Not sure what else to tell you regarding your situation...

Chris
 
Mar 9, 2007 at 7:45 AM Post #9 of 9
Well, Filburt and Chris, I'm really grateful for your suggestions and discussion. I agree, the PSU must have been injecting noise somwhow, and so, after a telephone conversation with a headphone buddy, I had another go at the clock PSU tonight, and I found a ground connection in between the AC line inputs. It's supposed to be a "virtual disconnect" PS, which I guess means multi-stage separation from the source, ie floating, so I grounded the connector to the main chassis ground pin

and VOILA

it now makes GORGEOUS music. The Rotel is now plain and simple the best source I have for the SP headphone amp. I'd say that the PSU did indeed make another improvement in the lineup of mods -- Rockclock 3, opamp sockets rolled to AD8065s, gold RCA connectors, and now the floating PSU. I also added a new 6922 Amperex PQ driver tube in the PPX3 SLAM that is just exquisitely clear and tuneful, and now my system really makes me smile !!!

600smile.gif
600smile.gif
600smile.gif


Chris, I will try to read the pinkfishmedia thread this weekend. I'd very much like to see now what else can be done. I would easily believe that cleaning up the DAC clock would also be very audible.

all the best
Bud
 

Users who are viewing this thread

Back
Top