Forum Replies Created
-
I’ve spent the last few days pulling my hair out trying to figure out all of the mechanical switches and levers on this thing. After several “one step forward, two steps back” moments (I had to disassemble the carousel twice because the elevator kept trying to shove CDs into already occupied spots) I finally got to the point where it could reliably once again cycle through a tray of CDs – loading and unloading, but failing to play anything – even failing to play with my prior hack of “upright position”. So with nothing playing, I was consigned to spending a few hours looking at ribbon cables, cleaning any mechanical switch I could find, and scouring the Sony service manual for clues as to why it wont accept a CD as “Loaded”. I finally zeroed in on this board, which handles the elevator mechanism.
Perhaps a bad seating or broken trace was the issue. Among other things, the board has 2 photo sensors – one to determine if a CD is in the player, and one to determine whether the CD is fully seated in the player tray.
I concluded that one or the other of these sensors was likely intermittent, causing the controller to determine that there was a CD jammed in the system. The readings for the “CD inserted” was indeed reading a medium short. And yet, it appeared to always be working for a specific dark-gold-colored CD, no matter where it was placed in the carousel. Odd!
… and then it hit me. These photosensors are designed to operate in total blackness. And here I was – for days – trying to poke and prod and shine my 3000 lumen desk light inside the internals to try and discern why the darn thing was not working, all the while causing the “CD Inserted” photoreceptor to constantly report a CD jam.
Ugh – I’m an idiot! I switched off my desk light, and everything began to function properly. For now – I have a hunch there are still a few gremlins left to be discovered. And they don’t like bright lights either.
Here’s a cursory look at the 5103 internals.
It’s a fairly straightforward design, with a plastic light “Power” rod that I managed to snap the end off by not noticing it as I was disassembling the panel. Much of the logic is buried behind shielding, with a smattering of capacitors, including one huge 2200uF capacitor. For those individuals in 240V communities, the transformer does appear to be a dual-voltage transformer, part #8013435, which would make a voltage conversion simple. It also appears to share the same mounting points as several Beogram transformers I’ve been working on as of late, so a simple switch to an already-wired one would also be a possible solution. The transformer supplies +25V though a rectifier to two IC regulators, one producing 5v for internal logic, and the other producing 12V for the CD carousel.
Although there are shields on both sides of the board, the chipset is only on one side, with the shield on the other side covering the pin outs. Hiding inside the shield is a ROM chip encoded as 8341743 S/W1.0 and an MHS microcontroller – basically, the 1980’s version of an Arduino!
All-in-all, it looks like a straightforward solution from B&O to quickly address an apparent hole in the product lineup, and also reasonably well made. The capacitors all tested quite excellent, so if there’s an issue with corrupted status messages, it’s not a simple capacitor swap that will right the ship. If I had more time and patience (and interest, frankly), I’d look into updating the ROM with more capabilities. The easier solution will instead be to intercept status signals coming from the carousel, and interject proper status messages onto the Datalink bus, which the Beomaster will happily interpret as being valid. This would also be the spot to activate pin 6 with Datalink ’86, and code for DL ’86 interoperability. All doable, but admittedly well down in my list of priorities.
So, what can this unit do? For starters, it does work with a subset of basic Datalink controls. ‘Play’ plays, ‘Stop’ pauses, ‘< Step’ and ‘Step >’ work as they would with a standard Beogram CD player. ‘<<‘ and ‘>>’ have been commandeered to advance and return from one CD to the next, instead of used as ‘FF’ and ‘REW’ through a song (which may be commands that the A20 doesn’t support anyways). Although I have not tried it, ‘>>’ + number will skip to the CD in that position in the carousel. When one CD finishes, the next one is automatically loaded and played in sequence. On the plus side, the system is quite responsive, on par with or slightly more quick than the standard CD5500-7000. Unfortunately, on my unit, no current status information is transferred via Datalink, such as ‘Track number’, or even ‘transport status’, either via the Beolink 7000-specific protocols, a MCP 5500/6500, or a Penta status update. So in that sense, it’s a bit of a one-way system that responds to commands, but offers no listener feedback.
So herein lies my disappointment. the Sony CDX-A20 is designed to be used with a head panel on a vehicle dashboard, and has a dedicated pin for transmitting CD data to a display. But I get zero status on any B&O controller or Powerlink speaker- a wasted opportunity for a reasonable integration solution, which would have taken only a minimal effort to program.So there are a few places yet to explore for some definitive answers as to where any fault in the status reporting may lie. 1) Check the Sony data-out line for a signal denoting the sending of data. If present, then either 2) the B&O firmware is not responding as expected, or 3) hasn’t been programmed to pass on the Sony display status data via Datalink. My unit has firmware version 1.0 – I have seen a picture of a 5103 controller with firmware version 1.1, but without more knowledge, I cant speak to what changes are included in the later revisions of the controller.
Given the CMM-10’s limited integration, I can’t imagine that these sold all that well. For starters, the use of Datalink ’80 as a control protocol means that the CD changer has to reside reasonably close to the Beomaster, with it’s own exclusive 7-pin connection tied directly to the Beomaster in order to function, instead of in a cabinet somewhere, perhaps in another room – IMHO a clear limitation that lies in the chosen implementation. To me, any perceived market need by B&O was an opportunity to put a DL’86-enabled unit (or series of units – up to 7 total!) on the Aux MCL, and have it addressed via the Source “Unit” designation that the DL’86 protocol was designed for.Perhaps they were simply intended as a stop-gap solution for those customers who asked for multi-CD changers for background music, or commercial customers who simply wanted an un-managed 10-hour playlist. If anyone has more information on these strange-and-likely-rare-for-a-reason units to help me better understand the what/why/when/where of their genesis, feel free to add to the discussion.After some gentle prodding of the main elevator gear, I managed to free the elevator from the grease-turned-glue that had bound it in place, and the unit began responding to ‘Tape Play’ commands. By ‘responding’, I mean it would pull out a CD and position it correctly, then fail to detect that the step had completed properly, and promptly return the CD to it’s holder. It would next raise the elevator to the next CD in sequence, and do exactly the same thing again, until it had gone through all ten slots, without playing a single CD. At least there was some progress.
After verifying the function of the light detection sensors, I began looking for some kind of mis-calibrated spring-loaded switch, or some other sensor that I had not yet seen. A bit more fiddling led me to the discovery that when I positioned the player in the upright position (remember from earlier that it can be mounted in 2 main positions), the sensors correctly detected the CD and promptly lowered it into position over the laser and began to play it – the unit was now working!
… ‘Ish’. Clearly I’m going to have to examine the capacitors and adjust the laser strength/focus, as some of the CDs were prone to skipping and/or getting stuck on sections of the CD. I’ll have to do some research on the laser adjustments for these Sony lasers, and find out of they are prone to the same issues that plagued the old Philips servo boards. My initial examination of the capacitors revealed a few familiar Sanyo units, which I tend to replace in the Beolink 7000 units.I also came up with a better version of the motherboard upgrade for early 5005 and TX-2 boards to prepare them for Datalink use, by placing the 10K resistor in the same location that it appears on the 5500 and newer boards.
I first drilled a hole for the resistor.
Next, I placed the resistor and transistor in their proper locations, taking care to leave the legs long, so I could connect them later under the board. I didn’t make any changes to the transistor placement from the last iteration.
Last, I connected the legs of the resistor and the transistor base, followed by a protective heat shrink cover to prevent shorting.
This go-around, I thought I’d take a stab at creating a plug for the RIAA right on PCB2, like the original. After checking all the caps and diodes for function, I drilled 2 small holes with a 1.0mm bit using a hand drill. It’s a tight squeeze on a US-Spec power supply, due to the long fuse which occupies the space normally reserved for the plug. Thankfully theres an adequate space over the PCB label, where there are no traces to damage.
I added a dab of crazy glue to the backside of the plug, and placed it in the hole. So far, so good!
The rear side required 2 wires. I decided to stick to my convention of keeping the red wire at 12V, and the black wire at 25V.
After checking all connections for connectivity and shorts, I was ready to insert the power supply back into the chassis. I replaced the regulator IC for good measure, as the installed one was reading 11.7 volts – it probably would have worked fine, but the new one was bang-on 12V. Although I have a few NOS barrel capacitors, I elected to keep the original in place, as it was reading a healthy 1100uF / 0.47 ISR, with no signs of leaking. Another step complete.
Another day, another upgrade. This time I’m taking a Beogram 3000 and upgrading it to Beogram 4500 specification. I got this unit because I thought it would fit in stylistically better with the black 7000 series that I have. My first thought was to simply replace the 5500 shell with the 3000 shell, but I decided to leave well enough alone – the 5500 had the latest motherboard with the discrete standby function, which the 3000/4500 could not make use of, because it lacks a front face.
I started with the chassis. Some new feet and touch-up paint to cover up the deep scratches.
I’m using graphite-infused black paint, in order to keep all areas of the base conductive. It’s not a perfect match, but IMHO better than exposed bare metal. My hope was to eventually use this conductive paint as a new semi-black coating for the platter, but that will be a bit later when I’ve experimented a bit with it.
I’m restoring another 5005, and I elected to try another technique for retrofitting a TR11 BC547 transistor into the datalink path of PCB1.
I started with a new BC547 transistor I bought from Mouser. When looking at the flat face, the pins from left to right arre (C) Collector (B) Base and (E) Emitter.
I unsoldered R42 on the side nearest the I/O cable, and drilled a small hole 2-3cm from it, taking care not to pierce any existing traces, or the adjacent R45 resistor.
Next, I took the BC547 and placed the (C) pin through the vacant R42 hole, and the (B) pin through the newly drilled hole. I bent the (E) pin towards the R42 resistor.
Turning the board over, I unsoldered the jumper adjacent to R30, and squeezed a 10K resistor through the hole. Theere was just enough room to have both the resistor and jumper in the same hole, which I resoldered in order to make a permanent connection.
I covered the resistor lead with heat shrink, and soldered the end to the (E) pin of TR11. This heat shrink is just a security blanket to prevent future shorts on the board.
I finished with attaching the TR11 (E) pin to the R42 resistor, and covered the end with a piece of heat shrink. IMHO, this looks much nicer than the other solution.More “remote” work – today it’s a Master Control Panel 7700, which may have been living in a moldy attic for a few decades. Seems to mostly function, with a few non-working buttons. I thought I’d try my hand at ribbon replacement, given the crusty and brittle original ribbon cables. I did the easy one first, just to see how it would go.
On to the more complex cable controlling the LCDs … big mistake. One dead LCD later, I concluded that there’s no obvious way that I can see to replace the connection to the LCD screens except to splice into the existing ribbon cable. The LCDs seem to be connected to the ribbon cable using some sort of random membrane that I cant make sense of. Oh well – thankfully this was just a poor-condition unit I acquired to verify the MCL output, which still works.
A Beolab Terminal that I bought for $50 had lost a battle with a glass of some sort of sticky cocktail. Half the buttons were non-responsive, so I took it apart and bathed the components in Iso. Luckily all the traces were undamaged and tested fine.
Everything looked and worked great after re-assembly.
I still need to re-tape the button pad. The original is very much a one-time-use material. Does anyone have a “favorite tape-medium” they use for this purpose? I was just thinking clear or perhaps thin black packing tape would work fine, and provide a bit of protection.Received a complimentary Form1 with another purchase, it the usual condition – without foam. After reading the archives on the issues everyone was having with the original foam, I bought a cheap set of Logitech 150 foam replacements for $10-for-2-sets, and affixed them with trim adhesive. Not ideal nor exact replacements, but serviceable given the time and effort I was willing to put into it. If someone ever has a better solution, I can always shave them off. Unit seems to work fine without the upper cushions, but then, I’m not overly concerned with comfort, fit and finish of this particular headset. Hopefully someone else can find this information useful to repair their sets without foam.
