The solid diy cabinet is much stronger, the flimsy old one tordated tremendously. Easy to make with mdf. Sound? Maybe little more relaxed. CDM2 is rather good isolated allready, my cabinet helped some further i can bounce harder before it skips. The resonance frequency of the cdm pickup is lowered with the lead dampening. The springs have new foam dampers.
TDA1541 sounds rather transistor-like (beefy, chunky) PCM56 is little more transparent, like a fet. Solved the hum almost, a ground-loop between dacs did it.
First i had Pedja's AD844 common base I/V but after a while listening treble annoyed. Now with a tube stage and passive I/V it has some more treble roll-off, like the tda in non-os (os filters have often a filter- compensated output to get a lineair output level to 20k) But must do some more extensive listening tests to evaluate sound. A simple OP627 I/V should work also, but i like tubes and wanted to try some.
It is non-os, and discovered i am still not the biggest non-os lover.
Tried a simple filter but my wife came back and then the fun is over, they have so much & many other priorities
TDA1541 sounds rather transistor-like (beefy, chunky) PCM56 is little more transparent, like a fet. Solved the hum almost, a ground-loop between dacs did it.
First i had Pedja's AD844 common base I/V but after a while listening treble annoyed. Now with a tube stage and passive I/V it has some more treble roll-off, like the tda in non-os (os filters have often a filter- compensated output to get a lineair output level to 20k) But must do some more extensive listening tests to evaluate sound. A simple OP627 I/V should work also, but i like tubes and wanted to try some.
It is non-os, and discovered i am still not the biggest non-os lover.
Tried a simple filter but my wife came back and then the fun is over, they have so much & many other priorities
To get an idea of what the PCM63 is, in terms of anything I might've heard I Google'd it and I have heard the Rotel RA-971. It wasn't a great player, but sounded quite warm and detailed in a friend's system. Of course I know one model using a DAC only tells a snippet of truth... but it's interesting to see 
Have you heard the RingDAC? I gather this one was also good (Arcam Alpha 9, FMJ CD23 and some others).
After this CD650 I might try something different again to get a feel for what I really like most.
Simon
Have you heard the RingDAC? I gather this one was also good (Arcam Alpha 9, FMJ CD23 and some others).
After this CD650 I might try something different again to get a feel for what I really like most.
Simon
I agree that the TDA1541A is good.
I've compared mine to a friend's reasonably modded CD63 (reclocked, lots of regulators (albeit off the shelf), improved electrolytics, improved opamps no HDAM etc, deadening, usual stuff), a Micromega Stage2, Martin Clark's Naim CD player (modded with fleas if I recall) and another friend's Consonance Opera (I gather quite poncy, NOS with a valve stage) and except for Martin's Naim, I preferred mine most of the time.
My friend who has the consonance has also acquired a Copland HDCD (and I think PCM63) so I'm hopefully going to have a listen to that soon.
Its just I enjoy investigating different solutions. I've made great progress investigating from NOS to SAA7220 to PMD100 for instance.
I've compared mine to a friend's reasonably modded CD63 (reclocked, lots of regulators (albeit off the shelf), improved electrolytics, improved opamps no HDAM etc, deadening, usual stuff), a Micromega Stage2, Martin Clark's Naim CD player (modded with fleas if I recall) and another friend's Consonance Opera (I gather quite poncy, NOS with a valve stage) and except for Martin's Naim, I preferred mine most of the time.
My friend who has the consonance has also acquired a Copland HDCD (and I think PCM63) so I'm hopefully going to have a listen to that soon.
Its just I enjoy investigating different solutions. I've made great progress investigating from NOS to SAA7220 to PMD100 for instance.
You have some very interesting references then!!
My main "rivals" at the moment are Brent's über CD63KI (with millions of lovely regs, transformers, super passive parts etc.) which is the most detailed thing I've heard by a mile.... and Thomo/Lee's Marantz CD94 (with double-crown DAC, clock, nice regs, caps etc.). That is possible the "nicest" CD player I've heard, in that everything is balanced and it coaxes the music off the CD perfectly. Both of those players have staggering bass. Sadly I've not heard these two players in the same room, but I already know Brent's would smoke the 94 in terms of detail and clarity. Every little noise the mics picked up is heard. I suspect even high-end/stock SACD doesn't come close to that one.
My main "rivals" at the moment are Brent's über CD63KI (with millions of lovely regs, transformers, super passive parts etc.) which is the most detailed thing I've heard by a mile.... and Thomo/Lee's Marantz CD94 (with double-crown DAC, clock, nice regs, caps etc.). That is possible the "nicest" CD player I've heard, in that everything is balanced and it coaxes the music off the CD perfectly. Both of those players have staggering bass. Sadly I've not heard these two players in the same room, but I already know Brent's would smoke the 94 in terms of detail and clarity. Every little noise the mics picked up is heard. I suspect even high-end/stock SACD doesn't come close to that one.
I added a diy (windings from thin copperwire on ferrite body's) analog filter from this scheme: Caps are mkt's orange drops
Its from Triodedicks pages.
Well: THATS more Like it! Now we are gettin' somewhere. The distorsion in mids/colouredness is allmost gone, but the non-os virtues are there (nice stereo effects, reverb sounds are very obvious, details are not hidden behind some sort of curtain anymore.) Only a slight little hum is hearable, will search this out, and swap different I/V resistor values. Coupling caps must be chosen too, now some old mkt's.
(Sorry for offtopic Simon)
An externally hosted image should be here but it was not working when we last tested it.
Its from Triodedicks pages.
Well: THATS more Like it! Now we are gettin' somewhere. The distorsion in mids/colouredness is allmost gone, but the non-os virtues are there (nice stereo effects, reverb sounds are very obvious, details are not hidden behind some sort of curtain anymore.) Only a slight little hum is hearable, will search this out, and swap different I/V resistor values. Coupling caps must be chosen too, now some old mkt's.
An externally hosted image should be here but it was not working when we last tested it.
(Sorry for offtopic Simon)
Yes, I've read a LOT about Brent's beast, as you can imagine. I knew Thomo had a CD94, but don't recall seeing much about it.
Yes, Martin's Naim is a Single Crown I believe. Very nice indeed. Perhaps that's what I should consider next. Not sure. I've heard that the single and double crown are different as opposed to one superior to the other (or maybe I read that from someone selling a single crown on Ebay
).So, trying to get back on thread. Tidying up the power rail noise near the SAA7220 might be a good next step. They are very noisy chips, and can add quite a lot of jitter to the I2S going to the DAC.
When you do get a clock, consider reclocking the I2S (using some d-type flip flops, 74ls174 for instance) might be a good idea.
The reclocking might be a bit tricky, but the power rail decoupling and regulation and stuff is probably right up your street.
I had spectecular sonic results with an old Ortofon STM-72 trannie used as as IV and hardly any low-pass filtering added. I m using a FET input buffer after that as output.
Just waiting for the Lundahl LL1636 I just ordered. See also:
http://www.audiodesignguide.com/cdplayer/dac3.html
Just waiting for the Lundahl LL1636 I just ordered. See also:
http://www.audiodesignguide.com/cdplayer/dac3.html
More small mods
Hi everyone,
I did some more cap swaps last night. I changed a couple of the smoothers for some Panasonic FC I had waiting to be used. I also replaced the post-reg caps (x4) for Rubycon ZA 100/16V.
I then changed the 3 DAC decoupling caps (previously changed to Cerafine, and 2 x Pana FC) to Ruby ZA 220/35V, as I suspect the FC were causing a slight bit of nastiness in the midrange. One idea that springs to mind is the caps were too large (some of the techie types say this introduces noise).
Finally, I changed the decoupling caps next to the RAM & processor to Pana FC 33/50V, and the decouplers next to the digital filter and decoder to Ruby ZA 100/16V.
Unfortunately, I'd upset something and it then wouldn't read the TOC! I cut out my IC decouplers and put the originals back - no change. So I checked all obvious voltage rails - all good. I then removed the main PCB again and retouched every solder joint, and did the same to the processor daughter-board.
It now works. Not had much chance to listen as the repair effort ran late into the night, but it's certainly better! The main thing I notice is the tight, defined bass and a greater sense of ease and naturalness further up.
I realise these are noddy mods to most people but have no fear, I will get onto power supplies, super regulators and clocks in due course. Importantly, with such simple component swaps and improving the mechanical integrity the lowly old cd650 springs to life to make wonderful music!
Simon
Hi everyone,
I did some more cap swaps last night. I changed a couple of the smoothers for some Panasonic FC I had waiting to be used. I also replaced the post-reg caps (x4) for Rubycon ZA 100/16V.
I then changed the 3 DAC decoupling caps (previously changed to Cerafine, and 2 x Pana FC) to Ruby ZA 220/35V, as I suspect the FC were causing a slight bit of nastiness in the midrange. One idea that springs to mind is the caps were too large (some of the techie types say this introduces noise).
Finally, I changed the decoupling caps next to the RAM & processor to Pana FC 33/50V, and the decouplers next to the digital filter and decoder to Ruby ZA 100/16V.
Unfortunately, I'd upset something and it then wouldn't read the TOC! I cut out my IC decouplers and put the originals back - no change. So I checked all obvious voltage rails - all good. I then removed the main PCB again and retouched every solder joint, and did the same to the processor daughter-board.
It now works. Not had much chance to listen as the repair effort ran late into the night, but it's certainly better! The main thing I notice is the tight, defined bass and a greater sense of ease and naturalness further up.
I realise these are noddy mods to most people but have no fear, I will get onto power supplies, super regulators and clocks in due course. Importantly, with such simple component swaps and improving the mechanical integrity the lowly old cd650 springs to life to make wonderful music!
Simon
When poking around I wanted / needed a better grasp of what connects to what in terms of power rails, so I compared my schematic to the board and metered some tracks and parts. This is hard for me because the sch is not actually for the CD650 but a similar machine, and the 650 has NO part numbers on the PCB. In this respect, the machine is not a suitable mods candidate for a total newbie, unless the very lightest mods are enough to please.
I noted that some 5 volt feeds are only at about 4.7V, due to using large decoupling resistors - from an ohm or two up to 10R in places. The decoder IC also uses what appears to be an inductor (shiny green bug) to decouple its 5V supply.
These resistors will be useful locations for injecting fresh volts from super regulators. I may try the digital filter on just a 7805 first if I get a chance this weekend. I might as well mount in a transformer and power supply to feed it. I've not looked to see what other voltages it uses, if any.
The DAC, annoyingly, uses -5 and -15 as well as +5.
Simon
I noted that some 5 volt feeds are only at about 4.7V, due to using large decoupling resistors - from an ohm or two up to 10R in places. The decoder IC also uses what appears to be an inductor (shiny green bug) to decouple its 5V supply.
These resistors will be useful locations for injecting fresh volts from super regulators. I may try the digital filter on just a 7805 first if I get a chance this weekend. I might as well mount in a transformer and power supply to feed it. I've not looked to see what other voltages it uses, if any.
The DAC, annoyingly, uses -5 and -15 as well as +5.
Simon
Sounds like you're making good progress.
The digital filter needs just 5v, and yes the DAC does need 5 -5 and -15v (its worth the effort though )
I wouldn't worry too much about the 4.7v - that's fine for these chips. Admittedly, it would obviously be better to have seperate regulators for each, so an excuse to make them 5v.
A seperate feed (well, regulator at least) for the 7220 is a good idea. It would help reduce the noise injected into the power rails by the chip upsetting other circuits - especially the DAC.
Maybe a topic for discussion elsewhere, but I think a 7805 is probably more than ample for the SAA7220, but I suppose it could be argued that it doesn't block interference going back onto the power rails, which a better regulator would. IOW I don't think a super quiet regulator is crucial for that, more like better power rail decoupling.
For DAC outputs, and analogue stages, obviously things are different.
Anyway, I'm willing to be corrected on that.
The digital filter needs just 5v, and yes the DAC does need 5 -5 and -15v (its worth the effort though
)I wouldn't worry too much about the 4.7v - that's fine for these chips. Admittedly, it would obviously be better to have seperate regulators for each, so an excuse to make them 5v.
A seperate feed (well, regulator at least) for the 7220 is a good idea. It would help reduce the noise injected into the power rails by the chip upsetting other circuits - especially the DAC.
Maybe a topic for discussion elsewhere, but I think a 7805 is probably more than ample for the SAA7220, but I suppose it could be argued that it doesn't block interference going back onto the power rails, which a better regulator would. IOW I don't think a super quiet regulator is crucial for that, more like better power rail decoupling.
For DAC outputs, and analogue stages, obviously things are different.
Anyway, I'm willing to be corrected on that
.
Re: More small mods
I have had the same prroblems. Most of the time the grey sockets have some bad contacts then on the wires in it. I solved that after complet dismantling of plug, (lift careful the two clips of the lightgrey cap) remove all pins, they have a fork wich grips the solid wire, these forks must be tightened up a little with a small griptool. Then reassemble, shorten wire some. The surplus of wiring can be cut off with a flat cutting copper cutting knife or how do you call that. Place cap back on, best is to align all wires with some tape next to eachother. Then press the whole together again carefully with a small bench vice with smooth claw surfaces. Since then i had never those annoying momentairy power-up faults again. And keep the wires sorted as they where
The CD880 has better sockets for PS. CDM has only some grey ones.
SimontY said:
I have had the same prroblems. Most of the time the grey sockets have some bad contacts then on the wires in it. I solved that after complet dismantling of plug, (lift careful the two clips of the lightgrey cap) remove all pins, they have a fork wich grips the solid wire, these forks must be tightened up a little with a small griptool. Then reassemble, shorten wire some. The surplus of wiring can be cut off with a flat cutting copper cutting knife or how do you call that. Place cap back on, best is to align all wires with some tape next to eachother. Then press the whole together again carefully with a small bench vice with smooth claw surfaces. Since then i had never those annoying momentairy power-up faults again. And keep the wires sorted as they where
The CD880 has better sockets for PS. CDM has only some grey ones.
I've been modding again
I took my player to Brent of Audio Upgrades, who is a friend and occasionally enjoys helping me with upgrades! Firstly we compared the lightly modified CD650 to a customer's machine he had been working on - a Marantz CD85 with 8 x S Power regulators, single crown DAC and 4562 op-amps, Black Gate caps etc. I am very fond of this player, and duly it left my effort in the dust - it had far more of everything, most notably a deep, sledgehammer bass and a wonderfully sweet and detailed presentation. The basic character was similar, though, as both use the same chipset.
Now onto mods of my CD650.... Brent had a TDA1541A IC lying around so we swapped it over and had a quick listen. It sounded a lot brighter and I felt the treble was no longer very recessed. There was more detail and space, the sound appeared to be more open. This stayed in.
Next we (ok, he) installed a C1 clock to replace the standard xtal, fed from an Audiocom Super Reg I had. For now, the feed comes from under the big (5v rail smoothing?) cap. Brent cleverly fixed the Super Reg right onto the C1 and fixed these lying down next to the digital filter IC. An ultra-neat install that I shouldn't break too easily. This lifted the sound again, most notably in the character of vocals, dynamics and treble quality. I think most people are aware of the usual gains from installing a great clock.
I took the player home and had a quick listen, which confirmed I had gained a lot in all areas!
Concerned about the extreme sharing of power supplies and regulators in the player (10 seconds of listening is enough to recognise the poor power supplies), I got to work lifting the PSU decoupling resistor that feeds the digital filter. I inserted a short wire that is fed by a 5 volt Audiocom Super Reg (I bought a few of these 2nd hand). For now the reg is fed from the same place as the clock (not ideal but much better than it was). I checked everything looked ok and fired it up. I also replaced another electrolytic cap that must be for decoder decoupling.
ORDER OF MAGNITUDE - this was a whopper. I don't know where to start, but this has transformed the player's abilities. It is no longer mid-fi but can play music with conviction. There's reasonable detail, some sense of space, fantastic bass quality, very sweet and alluring female vocals (these were poor on the stock player), and half decent treble reproduction. A very pleasing musical experience. Piano is a lot more lifelike.
So Phil and Lee - you were right - a separate regulator on the digital filter is vital, more so than a better clock or changing any caps. For academic purposes I should've used a 7805 first but I didn't have any to hand. My experience of other players is that almost any circuit is improved with a good regulator.
Pics to follow.
Simon
I took my player to Brent of Audio Upgrades, who is a friend and occasionally enjoys helping me with upgrades! Firstly we compared the lightly modified CD650 to a customer's machine he had been working on - a Marantz CD85 with 8 x S Power regulators, single crown DAC and 4562 op-amps, Black Gate caps etc. I am very fond of this player, and duly it left my effort in the dust - it had far more of everything, most notably a deep, sledgehammer bass and a wonderfully sweet and detailed presentation. The basic character was similar, though, as both use the same chipset.
Now onto mods of my CD650.... Brent had a TDA1541A IC lying around so we swapped it over and had a quick listen. It sounded a lot brighter and I felt the treble was no longer very recessed. There was more detail and space, the sound appeared to be more open. This stayed in.
Next we (ok, he) installed a C1 clock to replace the standard xtal, fed from an Audiocom Super Reg I had. For now, the feed comes from under the big (5v rail smoothing?) cap. Brent cleverly fixed the Super Reg right onto the C1 and fixed these lying down next to the digital filter IC. An ultra-neat install that I shouldn't break too easily. This lifted the sound again, most notably in the character of vocals, dynamics and treble quality. I think most people are aware of the usual gains from installing a great clock.
I took the player home and had a quick listen, which confirmed I had gained a lot in all areas!
Concerned about the extreme sharing of power supplies and regulators in the player (10 seconds of listening is enough to recognise the poor power supplies), I got to work lifting the PSU decoupling resistor that feeds the digital filter. I inserted a short wire that is fed by a 5 volt Audiocom Super Reg (I bought a few of these 2nd hand). For now the reg is fed from the same place as the clock (not ideal but much better than it was). I checked everything looked ok and fired it up. I also replaced another electrolytic cap that must be for decoder decoupling.
ORDER OF MAGNITUDE - this was a whopper. I don't know where to start, but this has transformed the player's abilities. It is no longer mid-fi but can play music with conviction. There's reasonable detail, some sense of space, fantastic bass quality, very sweet and alluring female vocals (these were poor on the stock player), and half decent treble reproduction. A very pleasing musical experience. Piano is a lot more lifelike.
So Phil and Lee - you were right - a separate regulator on the digital filter is vital, more so than a better clock or changing any caps. For academic purposes I should've used a 7805 first but I didn't have any to hand. My experience of other players is that almost any circuit is improved with a good regulator.
Pics to follow.
Simon
I haven't yet mentioned a repeat of the DISC ERROR problem, but it reared it's ugly head when I finalised my work and closed up the player with all screws in place.
I found that if I pressed my meter's probe down hard on the mech connector it would read! So I thought maybe the connector was duff and followed tubee's helpful recommendation. I unclipped the cover of the connector and pulled each wire further through. I then clamped the connector back together and trimmed the excess... plugged it in and no joy!
I then realised I could make it work by applying pressure around that side of the main board. I took it out and re-soldered some areas and cleaned the PCB with a toothbrush to remove any metallic detritus.... no joy. I gave up in the end, for now, and left it with one screw missing - it plays like this. I must have a cracked track or lifted pad that I simply can't see. I'll investigate later
For now I want to listen more!
Simon
I found that if I pressed my meter's probe down hard on the mech connector it would read! So I thought maybe the connector was duff and followed tubee's helpful recommendation. I unclipped the cover of the connector and pulled each wire further through. I then clamped the connector back together and trimmed the excess... plugged it in and no joy!
I then realised I could make it work by applying pressure around that side of the main board. I took it out and re-soldered some areas and cleaned the PCB with a toothbrush to remove any metallic detritus.... no joy. I gave up in the end, for now, and left it with one screw missing - it plays like this. I must have a cracked track or lifted pad that I simply can't see. I'll investigate later
For now I want to listen more!
Simon