last updated 08/31/04 --- I bought this CD player in late 2002 on ebay for $250 - it came with good reviews, and at over 60% off the retail price, I figured I'd get a pretty decent player for the money. Other than having a pretty flimsy remote and not reading CD-R's that are recorded at higher than 8X, I really couldn't find much to complain about.
I stuffed it on my rack, sitting directly on top of the tape deck, and couldn't tell any difference in sound to the 10 year old cheap Denon it replaced. Hmmm. Maybe I was doing something wrong. So I went online and researched the subject. There was a lot to be learned. The biggest thing I didn't realize was how important it was to isolate the CD mechanism from the rack and to dampen the CD player itself.
Some experiments with rubber balls and other materials led me to an isolation platform consisting of several materials. The impact on the player's performance was quite stunning. See this link to learn more about the platform the CD player rests on (it has changed since then - see the details on the 7075 aluminum plate and machined bearing dishes).
Apart from a few bypass capacitors, I had left the electronics inside of the player untouched until late in the project process (August 2004). I was quite happy with the player as stand-alone. In June 2004 I added the DDDAC1543, which at that point in time reduced the Cambridge to a transport. I liked what I heard, so for the time being, I was "done" with the project.
This "done" status changed in July 2004, when I got curious and built me a Toshiba SD-3950 based transport. I figured in the worst case scenario, I'd end up with a tweaked DVD player. However, the Toshiba transport clearly outperformed the Cambridge. Suddenly, there was no room on the rack for the player, but I felt I had to give it one last chance, and this time I would have to go and open the box and replace parts, similar to the popular Toshiba SD-3950 mods. That's how the most serious upgrade of this player came together - it was a do or die approach. See the progress at the bottom of this page (click here to jump there). I'm not done with this part of the project, so the jury on the benefit of these mods is still out.
What has been done other than documented below includes a custom Chris VenHaus design power cord, a dedicated OneAC isolation transformer and Risch-type DIY AC filter outlet strip. The RCA plugs have been treated with Pro Gold and the interconnects are pure silver/air cotton DIY cables with Eichmann plugs. Those IC's are my own design and outperform previous ICs I had built by a large margin. Actually, the last IC upgrade belongs in the top 3 of all my system tweaks.
|
The main "tweak" to get the CD player to sound good was to dampen the case. The stock D500 is a pretty lightweight design - thin metal case with very little structural rigidity. I tackled the rigidity at a later date with the aluminum base plate (see below). The very first thing I did was to spray a layer of "Noize Killer" on the lid that covers the player case. Black gunk, smelled awful for a week, especially when the player drawer was opened. That made little to no difference. Then I learned about "rope caulk" window insulation material. 90 linear yards for $4.00 at the home improvement store. I started to cover anything that could cause resonance inside the player. While doing that I kept looking for the infamous clock chip, which is the part that needs damping the most. I found it months later when I decided to remove the RF shielding plastic/metal-coated cover on the rectangular circuit board next to the transport. A big fat blob of rope caulk on it, plus on the capacitors around it made a significant difference in player performance. Where to find it if you have a D500 is marked on the image to the left. Later even (see Stage 2 below), I found a second clock chip on the board near the outputs, hidden below yet another copper shield. |
|
I figured that the Noize Killer spray was good, but some additional caulk on the lid would be better. However, when I mounted the cover to the player chassis, I realized I had overlooked the lack of clearance above the transport. The caulk on the player cover stopped the CD spindle. The good news about the caulk is that it can be peeled off, although things aren't very pretty afterwards. |
|
I figured what worked well on the timing chip and other electronic components could also be applied to the components on the D/A board in the far right rear of the player. The gold colored covers of what appears like the DAC chips got a full treatment, and all the capacitors around the area are well covered as well. There may be ways to upgrade these components, but I won't bother with that until I decide what to do about an external DAC. |
|
The most important area to dampen with caulk is the transport itself. I was quite
liberal in the application of caulk wherever there may be vibrations from the spinning CD and where the transport is
mounted to the base plate. I filled the plastic screw towers with caulk, and squeezed it into every possible crack that would
not impede proper operation of the mechanism. I did not apply any caulk to the part of the transport that holds
the laser, since added weight on that part may impair operation, however, I may try to apply some caulk with a small knife to
surfaces of the body of it. Since this part of the mechanism is moved to track the CD, damping may actually be very
beneficial in this area. Invisible on this image is the caulk underneath the transport (it had to be removed for that), as
well as the caulk in some of the cavities below the CD drawer. It now opens and closes with a very quiet and solid
sound, and it should rest on the mechanism with a lesser tendency to vibrate. Note the Grundgebuster mat on the
black CD-R in the drive (all media I use is copied from masters to black CD-R media using EAC on a dedicated external 1X speed
CD burner). The player clearly works better of well done copies of my CDs than the originals. what can't be seen on the photos is the "aperture" I painted on the laser itself, using flat black paint. Search the web for tweak info on that. I am not sure if it did anything for the performance of the player, and I guess the risk of messing up your laser is too high to recommend this tweak any further. Far in the back of the image you can see caulk on the fuse near the IEC connector, and covering the power supply transformer. I never opened that cover yet. |
|
Soon after I completed my DIY roller bearings, I realized that the sub-chassis of the D500 was far too flimsy to adequately support the player on the three bearing balls. The metal of the base plate is simply too thin to hold even a 5lbs player rigidly on just three tiny contact surfaces. To compensate, I decided to place the entire player on a 0.125" T6 6061 aluminum plate sourced from onlinemetals.com. Expensive, but they cut to size and ship to your home. What I had to figure out yet was how to bond the plate to the player, especially since the base of the player is not flat and smooth. Feet were removed, but there were some bulges in the base that had to be either cut out, or a spacer had to go between base of the player and the new plate. First I polished the contact areas of the aluminum with some metal polish to a shine (see the shiny areas i the photo below power button, far in back behind that and on the opposite side in the middle of the chassis). The result was so promising that I "upgraded" the plate to a 7075 0.250" plate a few months later. That plate came from a source on ebay (much cheaper!), but needed quite a lot of sanding and polishing to get to the smoothness I desired. |
|
And what better to use to bond the player to this plate than - you guessed it - a full package of rope caulk. Actually, it was more like 1 1/2 packages that was needed to build up the caulk to a level height clearing all the dents and bumps in the base of the player. The hardest thing was to squeeze the aluminum plate against the caulk, as the pressure is distributed across a 17x11" area. I am told that a blow dryer will soften the caulk enough to make this work better. When I bonded the thicker 0.250" plate I used a 2000 watt shop light to soften the caulk. I also flattened it with a rolling pin ("daddy, are you making cookies?") after covering it with some plastic wrap. That made the surface very even and increased the contact area with the plate. The player weight had also increased by about 100% since I started adding caulk. The thick plate bonded to the chassis also adds significant rigidity to the player it didn't posses originally. It's not very "pretty", but it does the job. Comparing this mechanism on the bearings to a friends's Wadia 301 (which due to it's size and weight was sitting on a plain MDF shelf) was a pretty close match. Definitely much closer than the 1500% price difference would have suggested. |
|
And here it is on the bearing balls, in the very latest version as of December 2003. The aluminum plate alone was a very significant upgrade, matching in impact all the other tweaks combined, probably due to a) removing flex from the chassis, and b) giving the bearing balls below a very solid and smooth surface to operate against. |
|
Here the complete setup with some machined 7075 aluminum dishes replacing
the cheap basting spoons, supporting my 3/4" bearing balls. I anticipated just a minor change in performance, but the
machined dishes are a huge improvement, similar to the impact of the plate under the player. Compared to the early
versions of my DIY bearings, this setup is now perform at at about 500% the effectiveness of the early DIY basting spoon setup without
any plate under the CD player. It also looks better.
This image also shows the entire isolation "stack" below the player, starting with the really bad steel rack, followed by (from bottom up) 1/2" MDF board, an underinflated bicycle inner tube, another MDF board, two large 8x16x2" paving bricks, ball bearings (bonded to bricks below with three really thin dots of rope caulk) with 3/4" chrome bearing balls, 0.250" 7075 aluminum plate, a layer of rope caulk (cord weatherstripping) and then the player itself, tweaked inside as shown above. I may try some 0.5" bearing balls to see if smaller bearing balls perform better. Note the mirror finish on the bottom of the aluminum plate. |
|
Here is the next target for tweaking: the power supply caps and voltage regulator of the DAC board in the player. I may switch this part to battery power, but since I am already building a separate outboard DAC, all I need to worry about is the quality of the digital output coming from this board. I am sure that it won't hurt to upgrade a few of these capacitors, though. |
|
In April 2004, I learned about a tweak that allegedly could make some cheap DVD and CD players sound much better: putting two 0.47UF 600V Auricaps across the hot/neutral wires in the power input of the player is supposed to filter all sorts of crud out of the power. Auricaps are pretty expensive, so I tried it first with similar A/C power rated caps, but that didn't do anything noticeable. Since I was building a DAC that needed 0.47uF caps in the output stage, I decided to order some and if the tweak with the AC didn't work out, I would have use for the caps. Well, so I put them in (behind the fuse, so they don't fail and catch fire while the unit is unattended), powered up and sat back to listen. It took a while before I was able to lift my jaw off the floor... AMAZING detail, sound stage definition up a full level of magnitude. Everything in the transport and built-in DAC must be working much better with clean power, as the amount of detail now coming from that player is simply amazing. One of the most effective tweaks, and very simple to install if you know what you're doing. |
|
I figured if two 0.47UF Auricaps can do so much good, why not bypass them with two 47pF 1000V Vishays? Why two? dunno, I just felt that I should stick with the proven model. I put them in parallel with the Auricaps, covered everything with heat shrink (after the photo was taken), and tested it. No significant change, if any. I'll probably remove them the next time I open the player to see if the undoing of the tweak sounds better. |
|
Now that I had started messing with the electronics inside the player, it was time to use some of the 100 Wima caps I had recently bought on ebay. 0.022uF 50V Wimas - perfect for bypassing the low voltage electronics inside the player. And since I am still using the player as DAC as well, the output board and all the large caps on that circuit were my first target. Turns out that there were quite a lot of large capacitors. I think I bypassed anything larger than 4.7UF. Not having any schematics, I figured I better leave alone the small ceramics that were on the board, but I bet I could get better results if I knew what caps I could actually remove and replace with Wimas. Still, the change added some warmth and reduced some of the high frequency harshness in the player. Well worth the $0.50 worth of capacitors the picture shows. The first caps I bypassed were four 2200uF caps that are rather small for their rating. Those have to be the main power supply caps for the board. The rest of them are spread all over the place and I can only guess as to what they are doing. |
|
The largest capacitor in the player is located on the PCB next to the transport, right next to a regulator on a fat large heat sink. I decided to double bypass that one with a 0.022uF Wima and a 47pF Vishay. There was a second slightly smaller cap in that area that also got a Vishay added. the same board also contains the clock chip. I didn't hear much of a difference after these mods, so I may go back and actually replace some of the caps with high grade parts. I'm holding back on the output board, since I am building an external DAC. |
| Player Modification Stage 2 | Now that I have a Toshiba SD-3950 that sounds better, there's nothing left for this Cambridge but to be modded to the max to sound better (if it can). So instead of stuffing it into a closet, I got busy: |
|
Following all the upgrades listed above, I went back to the basics. The player still wasn't sounding quite right compared to my Toshiba transport, so while listening to the Toshiba, I took out the soldering iron and the desolder braid, a piece of paper to take notes, and got busy removing parts from the PCBs in the unit. It took only 10 minutes to strip the PCBs from the player - something I had never dared to do before. |
|
While doing the disassembly I actually found a second clock (see the square tin can piece on
the photo) hidden under a some small piece of copper shielding material I had previously never removed (d'uh!). After 2 hours of
removing parts and taking notes, another 2 hours of browsing the net for adequate replacement parts, I had a list of components
ready to order. I'll post the list in case my changes are successful. The total cost for the capacitors and regulators, schottky diodes and
a few regulators as well as a new opamp for the analog outputs was about $40. This may sound crazy given that the cheap Toshiba bested
this player as a pure transport, but I simply wanted to find out if this is related to power supply parts quality or
other factors. This will help in future projects to allocate mod funds to the parts that matter the most. |
|
Meanwhile I added some additional damping to the laser mechanism to combat vibrations right at the point where the laser tracks the CD. This is not a place I'd mess with caulk, but the sticky Stillpoints material is very easy to cut to the perfect size to be tacked right on the unit. I'll be adding some more of this material to the sub frame of the laser assembly and directly to the motor that's driving it. |
|
There isn't a lot of stuff left on the board that usually is covered by the large square copper-colored plastic cover next to the drive. The clock is one of the few chips left on the top side of it. It is that spam can shaped metal thing next to the spot where the two pole wire connects to the PCB.This area is usually heavily covered in caulk. I removed it to get to some caps nearby. |
| Stage 2 Update | It's been over two weeks now that I have the unit back together with all the upgrade
parts installed. So you probably want to know "is it better than the Toshiba?" - hard to say. It definitely took a huge step forward and
sounds extremely pleasing now (as transport, have not listened to the analog out with built-in DAC yet), with a sound stage that seems
much more relaxed and deeper than the Toshiba's. It really sounds pretty fantastic now, except there are some CDs and some musical material
that it just doesn't like. The problem is a rather noticable high frequency "crackle" that sounds like interferance - it happens mostly with
dynamic piano music, but it also can occur in guitar or similar higher frequency content. It's not repeatable (i.e. if I skip back to hear it
again, it won't be there). This is a problem I have had before this final upgrade, but now it really has become the limiting factor of this
player. The crackle comes through analog outs as well as digital, so it has to be transport related. It may be an issue with my use of CD-R media
instead of my master CDs - this player does not read CD-Rs of any kind written faster than 4X for some reason (any D500 owners out there with
that same problem?). All my CDs are 1X written these days and most of them work just fine (black media, EAC, external burner, etc - no issue on the
creation end of the media - they all work fine in the Toshiba). In a recent test the problem went away when I went back to the master CD
of the problematic CD-R. Could it be a laser issue? Are there spares available for this player? it would be a shame to have to ditch it after all
these mods or to have to listen to the generally worse sounding master CDs rather than the black CD-R copies.
More photos of the final look of the player in its current state once I get the Toshiba back in the system. It's currently "on the bench" for its very own "stage 2" mod. update 09/07/04 - before I wrap this up this quick summary: I've put the Toshiba back in service and am a bit uncertain about which player I like best. The Toshiba clearly reveals more detail, but it sounds a little congested in the midrange. The Cambridge is very smooth and relaxed souding now, yet it has this issue with CD-Rs that makes it pretty difficult to like, given I have duped a lot of my music to black media. Since I've pretty much run out of ideas as to what to do about the Cambridge to make it work, I am probably retiring it. Meanwhile I am not really happy with the Toshiba, but there's more room for improvement yet. Watch for the Toshiba pages coming up soon. |