Jon,
Thank you. That worked and the whole service manual is now one nice pdf file on my desktop. Now to do some reading of it.
Joe
Looking at the Arcam Alpha schematic that Simon so kindly supplied, I see that it is broadly similar to my Rotel RCD965BX and shares most of the same ICs – except of course, the DAC and analog out, which is TDA151 in Arcam and SAA7323 in Rotel. The other major difference is that the system clock on the Arcam is inserted into the SAA7220 and then shared out. The Rotel has the clock on the 7323 DAC and then shared out.
The ICs around the CDM4 drive are similar, as are most of the components values. The only real difference is that the turntable motor drive is a difference TCA0372 on Rotel and a NJM4560 on Arcam. On the Marantz they are both TCA0372.
Both driver chips are fed by 10V, which, as far as I can see on the schematics, is unregulated on both CDPs (and Marantz). The + and – 10V is taken off after the rectifiers and before the 5V regulators. I can't see where the Arcam has any regulators on the 10V lines, or that any other higher voltage lines are regulated to 10V.
However, both Arcam and Rotel have TDA8808/0 processors for each driver, which seem to be lacking on the Marantz.
So I scratch my head and try to learn as much as I can from all this and make up my mind whether it is worthwhile to virtually dismantle the Rotel on the off chance that I might get an improvement from installing the regs.
Cheers,
Joe
The ICs around the CDM4 drive are similar, as are most of the components values. The only real difference is that the turntable motor drive is a difference TCA0372 on Rotel and a NJM4560 on Arcam. On the Marantz they are both TCA0372.
Both driver chips are fed by 10V, which, as far as I can see on the schematics, is unregulated on both CDPs (and Marantz). The + and – 10V is taken off after the rectifiers and before the 5V regulators. I can't see where the Arcam has any regulators on the 10V lines, or that any other higher voltage lines are regulated to 10V.
However, both Arcam and Rotel have TDA8808/0 processors for each driver, which seem to be lacking on the Marantz.
So I scratch my head and try to learn as much as I can from all this and make up my mind whether it is worthwhile to virtually dismantle the Rotel on the off chance that I might get an improvement from installing the regs.
Cheers,
Joe
Joe,
I'm glad the service manual went back together ok! If you read Martin's last post carefully you'll see he doesn't mention those chips running off regulated supplies, but rather the other chips in the servo section. The driver op-amp chips (Martin calls them an output stage I believe) are run off unregulated supplies. I think Martin was alluding to the fact that almost every power rail in the CD63 runs off that same 10V rail (most in turn through a single 5V reg). That 10V rail feeds those servo op-amps.
There may be a factor beyond what Martin has noted, which is that some have found further gains by increasing the stiffness of the supplies feeding the servo op-amps in the CD63. This suggests those chips are important and warrant decent supplies, not just separation from the others (for the sake of avoiding contaminating the others).
Really that is why I was so surprised to hear no change when regulating the rails and feeding those chips from beefy external supplies.
You guys have pointed out, however, that the arrangement is more different than just power supplies, so some of that must be a factor in determining what should or can be done to extract more from the player.
Simon
I'm glad the service manual went back together ok! If you read Martin's last post carefully you'll see he doesn't mention those chips running off regulated supplies, but rather the other chips in the servo section. The driver op-amp chips (Martin calls them an output stage I believe) are run off unregulated supplies. I think Martin was alluding to the fact that almost every power rail in the CD63 runs off that same 10V rail (most in turn through a single 5V reg). That 10V rail feeds those servo op-amps.
There may be a factor beyond what Martin has noted, which is that some have found further gains by increasing the stiffness of the supplies feeding the servo op-amps in the CD63. This suggests those chips are important and warrant decent supplies, not just separation from the others (for the sake of avoiding contaminating the others).
Really that is why I was so surprised to hear no change when regulating the rails and feeding those chips from beefy external supplies.
You guys have pointed out, however, that the arrangement is more different than just power supplies, so some of that must be a factor in determining what should or can be done to extract more from the player.
Simon
Hello Simon,
Yes, I take your point about Martin's comments on the regulated / unregulated chips. So, if I am going to do the job properly, I should regulate the 10V rails and 5V regulators on the other chips in the servo section. I can do all that from the 10V rails I already have. Ah well, there goes a few more late nights.
Thanks for reminding me about that. I had read on the Marantz thread about reclocking the microprocessor but could not see it on my schematic. So I thought it was an issue specific to the Marantz. However, I went back and had a closer look at the schematic and there it is, hiding in plain site all along – a 4 mHz crystal on the 68HC05C8 chip, pins 38 (OSC2) and 39 (OSC1). This is the same chip, pinout and crystal frequency on both Arcam and Rotel. I think the marantz uses an 8mHz crystal.
I might as well do the reclocking along with the driver / servo regs at the same time.
Cheers,
Joe
Hello Simon,
Yes, I take your point about Martin's comments on the regulated / unregulated chips. So, if I am going to do the job properly, I should regulate the 10V rails and 5V regulators on the other chips in the servo section. I can do all that from the 10V rails I already have. Ah well, there goes a few more late nights.
Thanks for reminding me about that. I had read on the Marantz thread about reclocking the microprocessor but could not see it on my schematic. So I thought it was an issue specific to the Marantz. However, I went back and had a closer look at the schematic and there it is, hiding in plain site all along – a 4 mHz crystal on the 68HC05C8 chip, pins 38 (OSC2) and 39 (OSC1). This is the same chip, pinout and crystal frequency on both Arcam and Rotel. I think the marantz uses an 8mHz crystal.
I might as well do the reclocking along with the driver / servo regs at the same time.
By the way, does Lee have a suggestion for the best sort of reclocking. I use a Tentlabs clock and PSU for the system clock on my Rotel.
Cheers,
Joe
Last edited:
Yes , that's it exactly Simon. I'd only to clarify that some of the Arcams servos do indeed have separate discrete ouput buffer stages run off separate supplies - eg look at the turntable motor circuit, the servo controls the motor through a push-pull pair of transistors run off separate -/+9v supplies.
PS I don't mean that the 63 cannot sound good or great -obviously it can- but look at the effort required, and how much of it spent just sorting out the shared-power mess.
PS I don't mean that the 63 cannot sound good or great -obviously it can- but look at the effort required, and how much of it spent just sorting out the shared-power mess.
Last edited:
Hello Simon,
I have found the 4mHz crystal on the cpu of the Rotel and it is actually a ceramic oscillator, not a canned crsytal. On the Arcam schematic it appears to be part number 1503.
Tent labs does not sell 4mHz XOs, and there is nothing much on ebay, but I have found a 4 pin XO at RS. Its part number is 226-1976 (AU$7.25). I will get one of these and mount it in a socket with its own clean and quiet supply. If it makes a noticable difference I can then hunt down a really low jitter XO and swap it in the socket.
I would be interested to hear if anybody else has done this mod or is contemplating it?
I am also gathering the parts for the servo / driver regulator upgrades. It may take a while but I will report back when finished.
Cheers,
Joe
I have found the 4mHz crystal on the cpu of the Rotel and it is actually a ceramic oscillator, not a canned crsytal. On the Arcam schematic it appears to be part number 1503.
Tent labs does not sell 4mHz XOs, and there is nothing much on ebay, but I have found a 4 pin XO at RS. Its part number is 226-1976 (AU$7.25). I will get one of these and mount it in a socket with its own clean and quiet supply. If it makes a noticable difference I can then hunt down a really low jitter XO and swap it in the socket.
I would be interested to hear if anybody else has done this mod or is contemplating it?
I am also gathering the parts for the servo / driver regulator upgrades. It may take a while but I will report back when finished.
Cheers,
Joe
I've been contemplating it for a while, it's just my player's very busy inside and getting hard to work on. I'd only do it if I was out of other ideas or if someone else did it first.
Right now I need to work on getting it to read more reliably. I want to do the "check of laser current and focus offset" as detailed on page 12/13 of the service manual. I'm just not sure what "test disc 5" would have had on it and if any old music or test tone cd would do.
Right now I need to work on getting it to read more reliably. I want to do the "check of laser current and focus offset" as detailed on page 12/13 of the service manual. I'm just not sure what "test disc 5" would have had on it and if any old music or test tone cd would do.
Ok, reading around it seems probably any original CD should do for setting laser current, though setting focus offset ought to be done with the proper test CD (no idea where to procure one). I also read that the latter adjustment is not critical and if current adjustment doesn't cure a reading issue the laser is probably toast.
I've just had a check of my laser current and focus offset voltages. The laser one read about 53mV so is within spec. I fettled the trimpot and couldn't get it closer than about 52mV but I don't think this is an issue...
The focus offset measured over 7V! It's meant to measure 400mV. I adjusted it down to 6.xxVDC and then ran out of adjustment. I think the 22K trimpot has gone faulty. With the player powered down I am seeing a low ohmic value between two pins and basically open (Megaohms) to the third. To me it does seem pretty broken. Please check my thinking someone: a pot should measure short->22K between any 2 pins, right?? There should be no meagohm values surely.
Once I've repaired all the point to point wiring from removing the audio board I'll have a listen and see if it's any better with respect to cutting out and skipping and being extremely sensitive to vibration.
Has anyone heard of these trimpots going bad?
The focus offset measured over 7V! It's meant to measure 400mV. I adjusted it down to 6.xxVDC and then ran out of adjustment. I think the 22K trimpot has gone faulty. With the player powered down I am seeing a low ohmic value between two pins and basically open (Megaohms) to the third. To me it does seem pretty broken. Please check my thinking someone: a pot should measure short->22K between any 2 pins, right?? There should be no meagohm values surely.
Once I've repaired all the point to point wiring from removing the audio board I'll have a listen and see if it's any better with respect to cutting out and skipping and being extremely sensitive to vibration.
Has anyone heard of these trimpots going bad?
Last edited:
Well I've done some listening to a few CDs and it's sounding great and not skipping or dropping out, and doesn't overreact to physical knocks either. The adjustment seems to have worked well enough, for now.
In the picture the wire link with the probe on it shows where to test for focus offset voltage. The other probe goes to ground. You can also see the 4.7K (laser current adjustment) and 22K (focus offset adjustment) trimpots.
Simon
In the picture the wire link with the probe on it shows where to test for focus offset voltage. The other probe goes to ground. You can also see the 4.7K (laser current adjustment) and 22K (focus offset adjustment) trimpots.
Simon
Attachments
Hi Simon,
The Rotel manual specifies a "Sony type 4 (YEDS 18)" test disc to use when checking laser and focus voltages. I haven't got one but used a normal music CD to reset my the Rotel''s voltages a while ago. The procedure is the same and test voltages are the same. The trimpots on Rotel are tin can type and adjust up through a hole in the board I have drilled holes in the bottom of the case to make this job easier (if I ever have to do it again). The pots look like 500mw type to me.
Did you change the trimpot on the focus offset? There shouldn't be any megaohm values there. For long term reliability I would change the pot, even though its a pain to get at.
Just a recent thought about doing mods under the transport without dismantling the player. I am going to try just removing the drawer from the mechanism and then taking out the whole swing arm and motor assembly and all its trays, just leaving the shell of the mechanism in place.
If the PCB of the Arcam is anything like the Rotel, it is a through hole plated type and the components, or their legs, can be desoldered and resoldered from the top of the board. I haven't tried it yet but will do so for the servo / driver regs.
My Rotel currently has 13 separate regulators with 4 transformers / PSU's, a new clock and big heatsinks, so I know what you mean about removing the PCB.
Joe
The Rotel manual specifies a "Sony type 4 (YEDS 18)" test disc to use when checking laser and focus voltages. I haven't got one but used a normal music CD to reset my the Rotel''s voltages a while ago. The procedure is the same and test voltages are the same. The trimpots on Rotel are tin can type and adjust up through a hole in the board I have drilled holes in the bottom of the case to make this job easier (if I ever have to do it again). The pots look like 500mw type to me.
Did you change the trimpot on the focus offset? There shouldn't be any megaohm values there. For long term reliability I would change the pot, even though its a pain to get at.
Just a recent thought about doing mods under the transport without dismantling the player. I am going to try just removing the drawer from the mechanism and then taking out the whole swing arm and motor assembly and all its trays, just leaving the shell of the mechanism in place.
If the PCB of the Arcam is anything like the Rotel, it is a through hole plated type and the components, or their legs, can be desoldered and resoldered from the top of the board. I haven't tried it yet but will do so for the servo / driver regs.
My Rotel currently has 13 separate regulators with 4 transformers / PSU's, a new clock and big heatsinks, so I know what you mean about removing the PCB.
Joe
Thanks for the response Joe, I will probably try to find a 500mW trim-pot when replacing the 22K one. I agree I should change it for long-term reliability.
I'd not fancy soldering things from the top personally, but maybe I'm just too clumsy. It's bad enough having about 6 wires soldered to the TDA's pins.
Simon
I'd not fancy soldering things from the top personally, but maybe I'm just too clumsy. It's bad enough having about 6 wires soldered to the TDA's pins.
Simon
I installed a 7909 and 7809 regulator on each of the servo chips (TCA0372). It may be only wishful thinking, but it seems that the base has tightened up a little more and the presentation is just that little bit more dynamic. It also seems that instruments and voices are more clearly defined, or perhaps more 'separated' is a better description. The improvement is subtle, but my wife noticed it and told me it sounded a bit clearer than before.
The Rotel transport section is so similar to the Arcam and so it seems to me that the Arcam should show similar improvements.
I also added +ve and –ve 5V regs to the TDA8808 Photo diode signal processor (Arcam IC 6501) and the TDA8809 Radial error signal processor (Arcam IC 6503). However, I can't claim any perceptible difference in performance with this mod. It didn't hurt but it didn't help.
The only things left to do on my Rotel is to reclock the 68HC05C8 processor (Arcam IC 6530) and give it a separated +5V reg. Then I will rebuild the SPDIF out.
All of these mods are common between Arcam and Rotel.
I have included a photo of the Rotel, which has a total of 5 transformers, and now 21 regulators. The analog out is a FET stage based on a Nelson Pass paper and sounds better than any IC I have tried. The output caps are Mundorf bypassed with Russian teflons. Very nice.
By the way, here is a link to an interesting discrete I-V with onboard discrete regulators designed by Greg Ball here in Australia. It looks very good and has the benefit of having DC nulling so it is DC coupled and does away with capacitors. It's worth a read. I have used Greg's low-noise 'minregs' on my Rotel DAC and analog stage.
SKA sponsored Audio Forum - Discrete Stereo I-V with Onboard Discrete Regs
I'll let you know about the reclocking when I can get to it. Here is a link to AW DIY site with a nice little "miniclock" that should do the trick. AW DIY - miniCLOCK Version 2
The Rotel transport section is so similar to the Arcam and so it seems to me that the Arcam should show similar improvements.
I also added +ve and –ve 5V regs to the TDA8808 Photo diode signal processor (Arcam IC 6501) and the TDA8809 Radial error signal processor (Arcam IC 6503). However, I can't claim any perceptible difference in performance with this mod. It didn't hurt but it didn't help.
The only things left to do on my Rotel is to reclock the 68HC05C8 processor (Arcam IC 6530) and give it a separated +5V reg. Then I will rebuild the SPDIF out.
All of these mods are common between Arcam and Rotel.
I have included a photo of the Rotel, which has a total of 5 transformers, and now 21 regulators. The analog out is a FET stage based on a Nelson Pass paper and sounds better than any IC I have tried. The output caps are Mundorf bypassed with Russian teflons. Very nice.
By the way, here is a link to an interesting discrete I-V with onboard discrete regulators designed by Greg Ball here in Australia. It looks very good and has the benefit of having DC nulling so it is DC coupled and does away with capacitors. It's worth a read. I have used Greg's low-noise 'minregs' on my Rotel DAC and analog stage.
SKA sponsored Audio Forum - Discrete Stereo I-V with Onboard Discrete Regs
I'll let you know about the reclocking when I can get to it. Here is a link to AW DIY site with a nice little "miniclock" that should do the trick. AW DIY - miniCLOCK Version 2
Attachments
That's an interesting post Joe, your player must sound very nice now.
I've had a little time and inclination to look at my recently-stopped-functioning Alpha today. It stopped playing a few weeks ago and I thought it was due to the draw not closing right or something silly like that. Sadly it's not so simple.
It will occasionally read the TOC and once on the "test bench" it started to play a CD. This rules out broken PSU wires or badly broken signal connectors or tracks.
I removed the front panel to access test mode and got Error 2: No Track Loss signal at start-up. Not massively helpful.
I thought the laser was probably dying and I then remembered I had a CD52 and that had a CDM4. Sure enough it's even a CDM4/19 like in the Alpha, albeit visually subtly different in places. Anyway, the laser comes out with the sled and can be swapped so I did that... only to find the same fault. I then put my Alpha's laser in the CD52 and it works faultlessly! So my fault isn't the laser.
I'm now stripping down all my mods etc that are in the way of the "mono board", and I will remove that 22K trimpot that seems to be faulty. Maybe that is the cause of the failure, given that focus test point reading of 7 volts....
I just need to find a suitable trimpot to order now.
Simon
I've had a little time and inclination to look at my recently-stopped-functioning Alpha today. It stopped playing a few weeks ago and I thought it was due to the draw not closing right or something silly like that. Sadly it's not so simple.
It will occasionally read the TOC and once on the "test bench" it started to play a CD. This rules out broken PSU wires or badly broken signal connectors or tracks.
I removed the front panel to access test mode and got Error 2: No Track Loss signal at start-up. Not massively helpful.
I thought the laser was probably dying and I then remembered I had a CD52 and that had a CDM4. Sure enough it's even a CDM4/19 like in the Alpha, albeit visually subtly different in places. Anyway, the laser comes out with the sled and can be swapped so I did that... only to find the same fault. I then put my Alpha's laser in the CD52 and it works faultlessly! So my fault isn't the laser.
I'm now stripping down all my mods etc that are in the way of the "mono board", and I will remove that 22K trimpot that seems to be faulty. Maybe that is the cause of the failure, given that focus test point reading of 7 volts....
I just need to find a suitable trimpot to order now.
Simon
Do I need 0.5W or will 0.15W do?
Looking at the service manual schematic I can't see why a good 10K pot won't do. It's just +5V at one end, -5V at the other and circuit feeds from the output somewhere between the two. Surely a good multi-turn pot will do an even better job, e.g. http://www.maplin.co.uk/square-22-turn-cermet-preset-potentiometers-2202
It may or may not have enough leg length to bend it over and lie it flattish if needs be.
Ordered a 10K multi-turn and a 5K to do the laser current pot if I fancy it.
Looking at the service manual schematic I can't see why a good 10K pot won't do. It's just +5V at one end, -5V at the other and circuit feeds from the output somewhere between the two. Surely a good multi-turn pot will do an even better job, e.g. http://www.maplin.co.uk/square-22-turn-cermet-preset-potentiometers-2202
It may or may not have enough leg length to bend it over and lie it flattish if needs be.
Ordered a 10K multi-turn and a 5K to do the laser current pot if I fancy it.
Last edited:
Hello Simon,
I think I would go for 0.5W as minimum and a single turn is sufficient. However, a multi turn will let you set the voltage more precisely.
A quick look at the RS catalog brings up part no 186-189, which is a 22k.
1 turn,linear, Cermet trimpot,16mm and 1Watt rated. It should fit staright in to your board, but you would need to check the pin spacing.
Neither RS nor Farnell seem to have the Bourns or Vishay .5W single turn types.
Anyway, see how you go with what you have ordered.
The other thing to keep in mid is how easy it will be to get at the pot with a screwdriver when you want to adjust it.
Joe
I think I would go for 0.5W as minimum and a single turn is sufficient. However, a multi turn will let you set the voltage more precisely.
A quick look at the RS catalog brings up part no 186-189, which is a 22k.
1 turn,linear, Cermet trimpot,16mm and 1Watt rated. It should fit staright in to your board, but you would need to check the pin spacing.
Neither RS nor Farnell seem to have the Bourns or Vishay .5W single turn types.
Anyway, see how you go with what you have ordered.
The other thing to keep in mid is how easy it will be to get at the pot with a screwdriver when you want to adjust it.
Joe
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.