I had a close look at the tray closed switch and took some photos. It has no option for adjustment other than bending the leaf switch tabs. It is triggered by a lug on the side of the tray which again isn't adjustable in any way.
Attachments
Last edited:
I couldn't agree more! I have the unit in front of me with schematics and service manuals but he has come up with some unique ideas.
I didn't think to try the CD93/73 option pin. The tray didn't oscillate but it seemed to be ignoring the leaf switches and waiting for the optical sensors, which obviously mine doesn't have fitted. Pulling the two control pins for those high and low didn't have the desired effect.
One combination actually engaged the clamp motor with the tray open. I had to shut it down pretty quick as it sounded like it was going to strip the plastic teeth off the mech!
<<This project would become vastly easier if we had a functional CD-73 to compare with>>
These models are a rare find now and even rarer in a working state. Of the 3 recently sold on eBay, 2 were faulty. The only one for sale at the moment is also faulty.
The only contact I've had from other owners is to ask if I've managed to fix mine.
These models are a rare find now and even rarer in a working state. Of the 3 recently sold on eBay, 2 were faulty. The only one for sale at the moment is also faulty.
The only contact I've had from other owners is to ask if I've managed to fix mine.
I suggest temporarily connecting jumper wires from the “leaf spring” tray switch MCU inputs over to the “optical” tray sensor inputs on the MCU. In other words put the same signals on both sets of MCU inputs.
This will permit testing it in both the CD-73 and CD-93 modes.
Hopefully the signal polarity is the same (logic low for end-of-travel). If not you will need an inverter or two.
It’s curious that the CD-93 retains the leaf switch for “tray open” in addition to having an optical sensor for “tray open.” Akai may have done that to permit the user to push the tray in a tiny bit to initiate a “close” cycle. Most disc players support pushing the tray in to cause it to close.
If the tray moves too fast in CD-93 mode you could keep the collector of the speed control transistor grounded to make everything move at a more leisurely pace. It probably wouldn’t cause any problems if the clamper moved more slowly. My guess for why Akai made the tray motor spin more rapidly on the CD-93 is because it has an extra stage of motor RPM reduction with 3 belts rather than 2.
-EB
AAARGH! Sorry about that.
It would be a good idea to temporarily disconnect the clamper motor while experimenting with the tray drive system.
-EB
First to answer some earlier questions...
For how many seconds does the motor run after pressing open/close button with the tray remaining in the fully open position?
(Don’t touch the tray for this test. Just let it stay in the fully open position)
The motor runs for 4 seconds and then stops.
Repeat this test by manually moving the tray in part way while the motor is still running. What happens?
The motor runs for 4 seconds in one direction, then 4 seconds in the other direction. This repeats continuously all the time the tray is half way.
Repeat this test by waiting for the motor to stop before moving the tray part way in by hand. What happens?
Again, the motor runs for 4 seconds in one direction, then 4 seconds in the other direction. This repeats continuously all the time the tray is half way.
Does the clamper oscillate or does it operate normally at this time?
(After the 5 seconds of tray motor rotation in opening direction)
The clamper doesn't oscillate or move during these tests.
For how many seconds does the motor run after pressing open/close button with the tray remaining in the fully open position?
(Don’t touch the tray for this test. Just let it stay in the fully open position)
The motor runs for 4 seconds and then stops.
Repeat this test by manually moving the tray in part way while the motor is still running. What happens?
The motor runs for 4 seconds in one direction, then 4 seconds in the other direction. This repeats continuously all the time the tray is half way.
Repeat this test by waiting for the motor to stop before moving the tray part way in by hand. What happens?
Again, the motor runs for 4 seconds in one direction, then 4 seconds in the other direction. This repeats continuously all the time the tray is half way.
Does the clamper oscillate or does it operate normally at this time?
(After the 5 seconds of tray motor rotation in opening direction)
The clamper doesn't oscillate or move during these tests.
Last edited:
I tried linking up the open and closed switches to pins 33 and 34 with pin 57 pulled to 0V for CD93 mode. The schematic for the CD93 shows 33 and 34 pulled down to 0V via 2K7 resistors and the photo transistors switching them to 5V.
I put 5V on one wire of each switch and the other wire of each to pins 33 and 34. I also grounded TR5 collector to slow the tray. The tray will close when the panel button is pressed but the tray motor continues to run, slipping the belt, before reversing and opening again.
The clamping phase doesn't start and the MCU appears to be ignoring the levels on pins 33 and 34. The oscillation doesn't happen but I think this is because the MCU isn't seeing a tray closed signal.
I put 5V on one wire of each switch and the other wire of each to pins 33 and 34. I also grounded TR5 collector to slow the tray. The tray will close when the panel button is pressed but the tray motor continues to run, slipping the belt, before reversing and opening again.
The clamping phase doesn't start and the MCU appears to be ignoring the levels on pins 33 and 34. The oscillation doesn't happen but I think this is because the MCU isn't seeing a tray closed signal.
This project begs to have several units lined up on one workbench. Hopefully not all of them with the same identical tray oscillation issue!
One is plenty enough for me!
It’s too bad they are so rare.
How good does it sound when it’s working?
It sounds nice when working but the challenge is getting it to accept the disc. Cycling the power switch repeatedly does eventually get it to load fully.
I've also looked at the test point for the PLL . I don't have a frequency counter at home, nor the test disc mentioned in the manual, so I just used my scope to see what was visible on the test point.
It is a nice clean square wave at 8695Hz. There is a hand written note on my service manual that says 7350Hz. I haven't tried adjusting VR1.
It is a nice clean square wave at 8695Hz. There is a hand written note on my service manual that says 7350Hz. I haven't tried adjusting VR1.
That LED/optical sensor used on the CD-93 model has two photosensors which are only separated from each other by about 10mm. I don't see how that could be used for "tray open" detection. I think one of the CD-93 photodetectors will generate a signal when the tray is "almost closed" and then the other one (or both of them) will generate signals when the tray is all the way closed. Without actually testing a working CD-93 it is impossible to predict what the timing of these signals would be. Also the CD-93 has a leaf switch for "tray open" detection.
I haven't read any complaints of "tray not working" for the CD-93 model yet. It seems the CD-73 is the only model with this tray oscillation.
This remains one of the oddest faults I've seen yet. And I've been working with audio gear for 50+ years.
-EB
Let's make a truth table which relates "motor activity" to the conditions of the tray switches. For this test the switches could either be operated by manually moving the tray or by having the tray part way in (both switches open) and then using clipleads or temporarily wired up switches to simulate the input conditions.
Switch state:
Tray-open | Tray-closed | Motor activity
1) closed | open | rotates 4 sec and stops
2) open | open | reverses at 4 sec intervals
3) open | closed | ?
4) closed | closed | Is this even possible?
Notes:
Condition 1 seems correct for the case of a broken belt. If the tray doesn't move at all then the motor will stop after 4 seconds.
Condition 2 represents what might happen if the disc is crooked in the tray and jamming it. The motor keeps reversing to try to release the jam.
Condition 3: This is where oscillation starts up when the tray closes. But if condition 3 is met statically (with "power-off") then switching on the power won't start the oscillation. It will be informative to get the unit into this condition (statically) with the clamper still open and then switch on the power. If I read your posts correctly i think the clamper will close at power-on without oscillation and then the disc will initialize and can be played normally.
Condition 4: Perhaps this simply cannot occur in real life? There would have to be some sort of lever that pushes on the "tray open" switch at the point where the tray gets to be fully closed.
There's just nothing I can grasp that would explain why this oscillation is occurring. It doesn't make sense.
-EB
Page 5 of this thread has a comment from another member (nullzie) who asked:
Hey Andy, did you ever resolve the tray issue? I have the same one on my CD93, but only on when it's closing and only with a disc inserted. Such a weird fault...
This got me looking at it again after a break of a couple of months.
I think one of the CD-93 photodetectors will generate a signal when the tray is "almost closed" and then the other one (or both of them) will generate signals when the tray is all the way closed.
This backs up your theory that the CD73 is looking for more than a 1-0 state when closing but I can't see how the single switch and single lug on the tray cound generate a 1-0-1 signal.
I've worked in electronics for nearly 40 years and my current job is test and repair to component level in aerospace. Some of the older military aircraft parts I work on have very weird and complex designs which, coupled with being almost as old as me, can be a real challenge to fault find.
They pale into insignificance compared to the Akai CD73 though, which defies all attempts, logical or otherwise to resolve it!
I know its bad when I take a break from working on it to replace a 176 pin QFP DSP audio chip... which is therapeutic by comparison
Hey Andy, did you ever resolve the tray issue? I have the same one on my CD93, but only on when it's closing and only with a disc inserted. Such a weird fault...
This got me looking at it again after a break of a couple of months.
I think one of the CD-93 photodetectors will generate a signal when the tray is "almost closed" and then the other one (or both of them) will generate signals when the tray is all the way closed.
This backs up your theory that the CD73 is looking for more than a 1-0 state when closing but I can't see how the single switch and single lug on the tray cound generate a 1-0-1 signal.
I've worked in electronics for nearly 40 years and my current job is test and repair to component level in aerospace. Some of the older military aircraft parts I work on have very weird and complex designs which, coupled with being almost as old as me, can be a real challenge to fault find.
They pale into insignificance compared to the Akai CD73 though, which defies all attempts, logical or otherwise to resolve it!
I know its bad when I take a break from working on it to replace a 176 pin QFP DSP audio chip... which is therapeutic by comparison
Coincidence?
I have one of those big bad DSP chips to install also. A Yamaha HTR-7065 A/V receiver is waiting patiently for it. This receiver had an intermittent TI D80YK113-family DSP chip. Evidently these were plagued by faulty metallization which caused intermittent faults at 3-5 years of age. Some of the big consumer electronics manufacturers put pressure on TI and got free replacements. TI referred to this fault as “premature aging.”
Over time I accumulated a stack of A/V receivers for almost nothing. Some had DSP chips which were still good but came with other problems which were easier to fix. I plan to use the HTR-7065 in one of my own systems where I don’t have enough shelf space for an A/V receiver of “truck-sized dimensions.” The HTR-7065 has features, adequate power output, & great sound in a convenient sized package.
-EB
I’ve encountered at least one disc player which had an optical detector to identify whether the tray had a disc in it or was empty. I can’t provide a good reason why that would be needed. But audio gear of “Byzantine complexity” does exist.
-EB
There are a total of 4 leaf switches. Could there be a hidden interlock which connects the tray section to the clamper section in some bizarre manner? I’m thinking about a tiny plastic tab somewhere which might crack or fall off.
Your photo of the end-of-travel switch for the tray leads me to think this switch doesn’t close until the tray nearly hits the backstop and then it should stay closed. I didn’t see any bumps or cams which would cause this switch to do anything other than a single “open to close” transition.
I really don’t have a clue about the clamper system. I’ve never dealt with a disc player with more than one motor to operate both the tray and the clamper.
Another complicated portion of the CD-73 CD-93 is the linear motor which moves the optical pickup carriage. It’s controlled by the same MCU as the tray and clamper. But I don’t think it has any leaf switches associated with it.
-EB
I too entered the world of premature aging TI DSP chips earlier this year somewhat inadvertantly... A routine blood test last December discovered I had developed a rare blood condition called Essential Thrombocythaemia and a subsequent MRI showed I'd also had a stroke. This left me shielding at home from Covid for 3 months while recovering.
During that time I 'treated' myslf to a faulty Pioneer VSX-2021 to pass the time and discovered the problem with TI DSP chips.This model uses a D810K013BZKB4, which is a 256BGA device. I had a go a replacing it but it still remained silent. I subsequently acquired an almost identical VSX-LX55, which developed the UE22 error after only a few days. Then I discovered that the faulty DSP can cause the DSP flash memory to become corrupted. Copying this from the 55 to the 2021 got it working, so my BGA install was good after all.
Since then I've been buying faulty Pioneer and Onkyo AVRs from eBay and repairing them as I have a good source of replacment parts in China. This week was a couple of VSX-922s that use the D808K013DPTP4, which thankfully isn't a BGA. Again it needed a working flash dump for the DSP but I managed to source this online. If nothing else it gives me a break from the CD73 and saves a few expensive AVRs from the scrap heap.
I haven’t jumped into the world of BGA removal/installation yet.
I’m gradually getting competent at removal/replacement of larger QFP packages. I started by practicing with defunct network cards. I do have a PC board preheater and a low-end SMD hot air rework gun of the “858” style. And my trusty Hakko 808 desoldering gun which is about a million years old but still works great. Sometimes I use that “chip-kwik” low melting temperature alloy for desoldering. My goal is always to get the old part off the PC board without damaging any foils or pads.
For BGA rework, there’s a big community of young Asian tech people where I live (a major state University with massive engineering & computer science departments). I’m thinking I’ll go to one of them when I need a large BGA package replaced. These guys fix Android, iPhone, & laptops all day long at the component level. I don’t know how they do it, but they do.
-EB
Last edited: