Jump to content
Sony Insider Forums

Sony MDS JB980 stuck on standby mode

Rate this topic


Recommended Posts

  • Replies 143
  • Created
  • Last Reply

Top Posters In This Topic

Top Posters In This Topic

Posted Images

On 5/15/2020 at 12:40 PM, kgallen said:

Good news! I think there are 3 switches - WP, REFL, LIMIT and OUT. All of these should be probe-able on the ribbon cable at the main PCB end for convenience. All are resistively pulled up to 3.3V and when the contacts are closed are shorted to 0V.

WP and REFL are tiny blue/white pins that detect holes/depressions in the MD disc case.

LIMIT and OUT are part of the drive mechanism, but are harder to get at as they are burried between the drive PCB and the mechanics. However given that in SM you can insert a disc, the OUT should be able to toggle. If you can move the sled in SM then LIMIT should be able to toggle (disc outer edge limit I think).

Frustrating we still can't nail this.

Did you do any visual inspection?

I mesured the different switches with the disc inside.

 Limit I managed to get to togle between 0 and 3.3V, by turning the gear so the sled went out /in. so that is ok. 

Out and REFL was 3,3V , WP and Protect was 0V .

What worries me is WP and OUT.. are they correct? 

I'm puting up a sheet in exel to mesure all the pins on both cn102 and 103 to get an bether overwiew. maby it will get clear to me then.

Link to post
Share on other sites
19 minutes ago, Petter Halonen said:

 Limit I managed to get to togle between 0 and 3.3V, by turning the gear so the sled went out /in. so that is ok. 

Good.

19 minutes ago, Petter Halonen said:

Out and REFL was 3,3V , WP and Protect was 0V .

With a disc in, WP should show the state of the movable flag that is used to indicate a write protected disc - the little white slider in the corner of the disc case. When writeable (slider covering the hole) then I'd expect WP=0V, when write protected (hole is open) then 3.3V.

REFL is the disc reflectivity. With a recordable MO disc then the REFL hole is open so I'd expect 3.3V. If you have a pressed MD-CD disc then the hole is blocked so expect 0V. The REFL hole is next to the WP hole in the MD disc case.

OUT switch should show (probably) 3.3V with a disc out and 0V with a disc in. (The state should be different with a disc in compared to a disc out.). This is possibly the most interesting one given our issue. The switch is burried so it's hard to see. But having this one change state between disc out and disc inserted is obviously important!

 

 

Link to post
Share on other sites

OUT switch should show (probably) 3.3V with a disc out and 0V with a disc in. (The state should be different with a disc in compared to a disc out.). This is possibly the most interesting one given our issue. The switch is burried so it's hard to see. But having this one change state between disc out and disc inserted is obviously important!

I ejected the disc manualy an mesured again. out sw is now 0V and the same is refl.

 

Link to post
Share on other sites

Update: After geting some help from a friend, I think we are on to something.  The CLKA signal from pin 5 on IC803 has an amplitude of only 1 V p-p, so it is only 50% of what it shuld be. Input clock on pin 1 has an amplitude of 2,5 V p-p, so that is correct. the VCC pin 2 and FS1 pin 7 is 3,3 V and pin 4 and pin6 is GND.

So there must be somthing wrong whit IC803 or IC201 making the amplitude go down.

 

MDS-JB980_ic803scema.jpg

MDS-JB980_ic803.jpg

MDS-JB980_ic803p5.jpg

MDS-JB980_ic803p1.jpg

Link to post
Share on other sites

Don't necessarily agree. That chip supports a directly connected crystal across the XIN/XOUT pins. The crystal is excited by a lower amplitude sine wave than the usual digital levels. In this application they haven't used a directly-connected crystal, they are sourcing the clock from elsewhere. The clock they have sourced is probably a full-swing digital signal, and they don't want this for XIN so they've ac coupled it with capacitor C899.

-> I suspect this is as intended.

Hmmm. There are a few different datasheets about but none of them are particularly comprehensive.

From the CLKSEL and FS[1:0] settings looks like the input clock should be 11.2896MHz and the output 90.3168MHz.

It's not stated but it would seem reasonable that the CLKA output swings to the rails of the chip supply, which is stated as 3.3V typical. However there are some voltages annotated around that chip and it doesn't look like 3.3V on CLKA, more like 1.5V. Can you clarify on your document? OK just found the SM, it says 1.7V at IC803 pin5, even though the chip supply is 3.3V. From your top oscilloscope trace, it looks ok-ish to me (at 1V/div) - it's maybe 1.5V pk-pk on your scope but you are approaching the limits of that scope - 90.3168MHz on a 100MHz BW scope. Also the probe cap is probably 20pF on top of the input pin cap and that chip has a CLOAD(max) of 15pF.

Also I note the input the clock is driving, IC201 pin 20 OSCO. This chip also seems to support an external crystal on pins 19 and 20. It's likely it's ok with a low-swing clock signal on pin 20 as there will likely be a Schmitt-trigger stage internally to square-up and level shift the crystal resonance to the correct digital levels of 3.3V.

-> I would still conclude IC803 is ok.

But I do concede the SM seems a bit inconsistent between the schematic, the waveform page and possibly the chip datasheet.

Link to post
Share on other sites
18 minutes ago, kgallen said:

Don't necessarily agree. That chip supports a directly connected crystal across the XIN/XOUT pins. The crystal is excited by a lower amplitude sine wave than the usual digital levels. In this application they haven't used a directly-connected crystal, they are sourcing the clock from elsewhere. The clock they have sourced is probably a full-swing digital signal, and they don't want this for XIN so they've ac coupled it with capacitor C899.

-> I suspect this is as intended.

acording to service manual, the clock signal from ic803 shuld be 2.5Vp-p, mine is 1Vp-p.MDS-JB980_signal.thumb.jpg.fe36be1afdc6ff76e2dc38d7414dd5ec.jpg

Link to post
Share on other sites

From the data sheet for the chip I’d agree the amplitude should be somewhat nearer 3V. But at the level shown in your scope I’d be surprised if this is a fault given the way the circuit is configured. If you’re measuring at IC201 then the clock does have the series termination 100R and ferrite bead in the way. 

Link to post
Share on other sites

I mesured whit another probe on my scope and now the signal is correct, so you were right. My bad... again..

That is why I always have this mistrust to what I mesure and tripel or octcheck what I mesure if the signal look suspitious.

Back to start ( almost).

man this is frustrating...

Link to post
Share on other sites
20 minutes ago, Petter Halonen said:

man this is frustrating...

It is I agree.

On the negative side it's annoying having a great machine out of service, particularly if it's your only one.

On the plus side, look how much you're learning about your machine!

Still keep looking for those 'obvious' / 'simple' explanations though. One rubber belt change can't kill a fully working machine!

If there's anything we've learned on this forum in recent times, these machines are well designed with quality components and rigorous manufacturing and test processes. It's tended to be something daft in the end...

Link to post
Share on other sites

Wondering if it is something caused by ESD. This is like a gohst kidding whit me.

The P down signal is 3,1 V... strange that it is only halfe in standby. mabye I'll have a look at it.

  I do have one md walkman, but this was my fawourite player of all time. 

I had just got it talking whit my win10 computer.

Link to post
Share on other sites
6 minutes ago, Petter Halonen said:

Wondering if it is something caused by ESD. This is like a gohst kidding whit me.

With the components in-circuit you'd have to be very careless with e.g. bare ends of ribbon cables or deliberately rubbing your hands across the PCB after scrubbing your nylon clothes with a cotton duster. Out of circuit it's much easier to damage CMOS devices.

In the industry we use the ESD Human Body Model as one means to characterise the device in this respect.

Link to post
Share on other sites
46 minutes ago, kgallen said:

Explain to me again where we are with the Standby condition...

The display is switched of and the standby LED is litt. but the power is not in standby. The Q911 is on so the rele is swithced on

.P. Down pin19 on IC1 is 3,3V.

I mesured whit osciloscope and then ther is a puls signal there, almost like it is trying to switch of or on but can't.

And that is how far I have got with it.

I mesured the P.Down signal whit the scope and sudenly the rele klikked and now it is dead. All I have is the 3.3V system power. 

Probebly easyer to find the fault now.. I hope.

Link to post
Share on other sites

I thought I'd share my experience regarding "stuck in STANDBY".
I saw this situation maybe 8...10 times. Some cases I could repair, some I could not. A few examples, that are worth to mention:

- an S38 came for repair, the OWH was broken off. After I replaced the OWH, the whole device stopped working. Popped the drive into a good 510, it worked well. Same when swapping the front panels (needed a little trick with the ribbon cable). Then with unit powered down, using my ESR meter, I checked the main CPU IC, and I found this: two pins were down to ground. Both were dead shorts, and were not supposed to be. Something burnt inside the silicon. Reason? Probably ESD related. End of the road.

- a JB930QS needed a new OP, as the old one run more than 3000 hours, and began skipping a lot. With the new OP in, the deck would not boot at all. It went into service mode, but was not able to spin the disc or move the sled back and forth. I swapped the switch board, the motors, yet another laser, no success. Then I swapped a BD board from a 530, and it all started to work again, with the original parts plus the new laser. Something on the BD board broke down. Reason? Again, probably ESD related, but might as well be some aged components on the BD board, like caps. I rescued the 930 by sacrificing a 530 for the BD board of it.

- I got a JE640, that would not turn on, but I could hear the eject motor whining. It seemed the mechanism was forced when someone tried to pry out the disc. I manually repositioned everything as they should have been, but the deck still would not accept a disc, not even pulling it in. After loading a disc "manually", when powering on the deck would just eject it and immediately go to STANDBY. I measured all laser emissions in static service mode, they were spot on, so virtually all OK. Then I replaced the OP by chance, and the device came alive. The OP was broken in a way, that the logic could not properly talk to it, and prohibited the whole unit from working.

- weirdest example: it was a 980 with C13 error. Turned out, that some plastic toy coins were stuffed into the drive slot, that the drive could not digest. I took the whole deck apart, cleaned, checked and tested all parts individually. After putting everything back on, the 980 did not come out of STANDBY. I went thru all parts, over again. No luck. Last I swapped the OP, when the device came alive. Then I put back the original laser, and the deck was still working. Never knew, what was the issue, and how it "cured itself".

Looking back on those, and after reading the whole thread, I would give a chance to swap the OP, a least for a quick sanity check, even without going deep into precise laser settings.

Quoting some of your thoughts above that would also suggest this:

"Is there any protection modes were the CPU shuts down if something is wrong?"
"I think we’re looking at a simple (ha ha ha) drive issue. USB is probably off track"
"are surely indicators that the CPU is alive. Just not talking to the BD"
"They look ok to me, only the disk does not spinn up"

 

Link to post
Share on other sites

Interesting that you comment on it being especially sensitive to ESD and possibly capacitor discharge.
I have an otherwise working 980 which needs its eject belt changing and am soon to attempt it.

Considering the warnings above I think it wise to do the following in advance of attempting the repair.

Ensure the unit is powered off and disconnected from mains for 24-48 hours before disassembling.

Press all buttons on the front in a further futile effort to make sure everything is discharged.

Ground yourself with an ESD wristband.

Ground the chassis with a croc clip to ensure it has no potential to discharge any static to you.

Normally I wouldn’t take quite such care but it seems these are really sensitive.

It might also be profitable to put a pen mark on each of the ribbon cables for orientation and alignment. I’m wondering if there is possibly an issue where people are getting the ribbon slightly out of line. If the pins get misaligned/shorted it could be causing it to not come on or worse, die altogether.

With enough caution, short of physically damaging something or extreme bad luck, it should be possible to repair with no issues.

Edited by mikeyp
Corrected some autocorrect corrections which were not correct.
Link to post
Share on other sites

Sorry for the third post. Possibly a dumb question, but have you checked continuity of the ribbon cable and connectors. The conductors can break and lose continuity. Sounds like that’s possible. It might be worth trying fresh ribbon cables if you can find any.

Link to post
Share on other sites

there are posts on here about availability of ribbon cables somewhere, i think it was by kevin.

when you change your belt, be careful with that clip on the far end of drive unit, if you break it your stuffed.

Link to post
Share on other sites
8 hours ago, M1JWR said:

there are posts on here about availability of ribbon cables somewhere, i think it was by kevin.

when you change your belt, be careful with that clip on the far end of drive unit, if you break it your stuffed.

Indeed. Here it is. I think 980 is the same PCB set as 480, referenced below. Check the dimensions on your own machine though. These ribbons seem quite robust, so unless they've been really abused I'd expect them to be ok. Oxidation of the exposed fingers is what we've generally found.

 

Link to post
Share on other sites

I am  curious to hear about Petter, whether or not he checked his OP (the laser head). I still give it a non-zero level of probability, that the OP is working erratic in some way. Unless it is proven being 100% OK (the quickest way would be testing it in another deck and vice cersa) thus out of question, further guessing might be misleading.

I also would recommend a quick check on the main cpu as I described above (check all pins against ground and Vcc rails, in a safe way, like with an ESR meter). Again, to exclude any silicon damage from possible causes.

Last, but not least, I am somehow missing some waveform checks on the power rails. I read the DC voltages were measured, but what about any AC component (noise) on the DC rails? Are we convinced, that all filter caps are in their good shape? (The 980 came to the market just in the middle of the bad caps disease of the industry.)

I understand this particular "stuck in standby" situation appeared somehow right after replacing the belt, and I certainly can understand if these "theories" considered weird or irrelevant. My point here is if you have no clue (I mean, "no more clues left"), then get back to the start line, and begin excluding all things that can be a cause anyhow, even those that cannot be not directly connected to such an easy disassembling/reassembling process.

Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...

×
×
  • Create New...