This is more of an info thread for G90 owners, I figure the more info out there the better.

I've posted my set in the f/s section as is, but I'm taking a last attempt at repairing it. THe tech in me can get pretty stubborn.

As per the f/s post, the set fires up fine, projects menu and an image, but none of the functions work, ie you can ramp any command (brightness, focus, convergence, etc) from 0 to 254, and while the onscreen numbers change, the command has no effect on the set. After about 5 minutes the set will shut down with a couple of error codes, 89 and 8A which point to the convergence boards.

I have found that one keypad command does work, the ABG control works as it's supposed to. THe gain and bias controls do nothing.

Also, right before the set shuts down with an error, the image will change in brightness, flicker slightly and on occasion the image will skew and bend slightly, as if the geometry controls change drastically. Then the set shuts down.

I've found that by disconnecting the convergence board input connectors, the set will stay on longer.

I've sent the RGB input board and the main CPU board to Terry for testing, and they check OK.

My theory so far is as follows:

1) The convergence boards are fine. I figure that the rapid changes in signal going to the convergence boards cause current spikes giving a false shutdown.

2) The waveform generator seems to do nothing, yet it gives no error codes, thus at least the power supplies feeding the waveform boards are fine.

I'm now trying to trace how the brightness and contrast adjustments are controlled. I figure the brightness and contrast controls do not go through the wave boards. Can anyone assist in the signal path here?

I think it needs a new muffler bearing :D

Seriously, I wish I could help but I couldnt resist. I stayed at a Holiday Inn Express last night ;)

Disclaimer: I know absolutely nothing specifically about the Sony VPH-G90U, but since I already had the service manual open for another thread I thought I'd take a crack at this issue too.

If the YA and BA Boards were already found to work correctly in another chassis, doesn't that more or less leave only the MA Board?... Which almost narrows it down to basically a bad connector, solder joint, or trace.

The obvious suggestion would be to check around connector CN602 and CN603 (that the YA Board plugs in to) on the MA Board. Especially pins 9-10 and 12-16 on CN603, as those are the clock and data lines for the I2C Bus and Registration DACs. Somehow, given who's asking, I'm guessing it's not going to be quite that simple though, is it? :)

Do you get a ~5MHz (per the Theory of Operation Manual) clock waveform at CN421, Pin C12 on the YA Board?

Do you also see that same clock waveform at CN341, Pin A13 on the BA Board?

And again, let me reiterate, I may have no idea what I'm talking about... Just throwing out a few suggestions that sounded good to me at the time.

-- Ryan

Edit: A couple other things I realized...

I imagine you've already tried this, but if you input a composite NTSC signal on the video input, do the COLOR, HUE, and SHARPNESS controls work? The IC responsible for those adjustments (IC1600) is on a different bus than the DACs for BIAS, BRIGHTNESS, CONTRAST, etc. (IC609 and IC611).

Also, if the D-SCL (clock) waveform I mentioned above is present on both YA and BA boards, are you seeing periodic assertions of D-LD0 at CN421, Pin B13 (YA Board) and CN341, Pin B14 (BA Board) as well?

Well where the heck were you when I posted back a year ago?

:)

It _appears_ that there is no clock pulse coming out of the SCL connector. I get the 20Mhz clock going into the DAC on the CPU board, but very little is coming out of the DAC to the buffer chip. I've checked for shorts on the SCL line, and even tried feeding an external clock to the buss, but no dice.

The only thing that doesn't jive here is that I sent the CPU board to Terry and it checked fine there. This problem doesn't come and go, so it doesn't make sense that the board worked at his place and not here before it left or when it came back.

So I dunno. I've now put about 10 hours into the set, and have learned a LOT more about the G90s than I knew before I started this.

I've got several interested buyers in this set, so I'll take them in order. With luck the problem is elsewhere, and someone that already has a G90 can simply do some board swapping and then get this set running.

As a tech I hate walking away from the unit, but with no spare parts here or out there, I'll cut my losses now.

Thanks for the help on this, it was fun, if not frustrating. And my soldering station died in the middle of it, dammit!

I realize you're probably going to sell this set AS-IS, but here are a few last thoughts...

To ensure we're on the same page, when you say "DAC on the CPU board" you're referring to IC421? And you do get 20MHz at Pin 60, but nothing at TP428 (the D-SLC output)?

Anyway, a little background on my understanding (so no guarantee it's correct) of how this works:

DACs IC609 and IC611 (that control brightness, contrast, etc.) on the BA Board are loaded with data from DAC Controller IC421 on the YA Board via lines D-SLC (clock), D-SDA0 (data), and D-LD0 (load) during a 180µs period syncronized with the start of vertical retrace. The timing relationship is shown in section 17-5 of the Theory of Operation Manual. Hence, I don't think you can just inject a 5MHz square wave (if that's what you did) and expect anything (good, at least ;)) to happen. That clock needs to be syncronized with some other signals.

While D-SLC and D-LD0 are low, D-SDA0 is set high or low depending on the data you want loaded. D-SLC then asserts for ~100ns, loading the bit on D-SDA0 into the shift register of the DAC. That continues until all bits have been shifted in, at which point D-SDA0 and D-SLC again remain low. D-LD0 then asserts for ~200ns, which copies the data from the shift register to the selected latch register on the DAC and subsequently changes the analog output.

So, if you are seeing various pulses (you mentioned "very little") on D-SLC, D-SDA0, and D-LD0, my guess would be it's working. Even if it weren't, I'm not sure how you'd verify the data and timing are accurate without at least a two-channel storage oscilloscope.

One even simpler thing I did miss, though... Is D-RESET at CN421, Pin B12 on the YA Board high? This is connected to every DACs' Preset NOT Input which, if low, would cause them all to sit at mid-level and ignore serial input.

-- Ryan

Hello,

I've bought this Curt's G90 defective unit some time ago and did not found time to investigate it until now...

Yesterday, spent some couple of hours to swapp boards between my first 'working' G90 unit and this defective one.

I've also soldered some probing wires on the YA I²C communication links to probe all I²C lines linked to DD/DE/DF and BA boards (D-SCL, [DLD0-DLD3], [DSDA0-DSDA2] and [M1-M5] with D-Reset signal). I found that, when a failure occurs, the D-SCL signal is fading toward 0V without any reason.... Looks like a I²C bus loading problem because of a defective receiver OR a defective local power on receiver...

To investigate if it's a I²C bus loading issue, I tried to swap all DD/DE/DF boards with the one from my first unit to see if it solve the issue or not : Nothing has changed on the unit behaviour. Then, It's not these boards so not the buses loadings conditions...

Since Curt said that both YA and BA boards were checked OK by someone else, I didn't swapped those two boards yet but after checking all the PSU sections and found nothing wrong, I was wondering if is was not worth trying to swap those YA/BA boards as well, just to re-check :o

Then I made it yesterday and then the unit started to work fine :eek:

Found also that the green tube bad G1/G2 bias (ABL ?) I had was due to the BA board (this failure disappeared with my working BA board)

- Then, I must conclude that the I²C buses failure is relevant to YA board ONLY.
- The Green tube overdrive issue is only relevant to BA board :) .

I will try to change IC13 (74HCT541 SOIC20 chip) I²C bus buffer on YA board and probe the Green channel after IC404 (LM1283) device since the green overdrive is present with internal tests patterns on BA board.

Curt, did you noticed that green tube was overdriven when you've investigate the unit ?

Later

John

Is there any "protection zeners" parallel in those sda/scl lines..I remember time ago(maybe 10 years) when sony car radios did get "nuts". sometimes those prot.zeners was leaky and they maked little shorts and prosessor think that was data... Some macines turned on and pumped volume up :D etc...
And solution was to chance all prot zeners on those lines ( bad component batch)

John, glad you're making progress! No, I never saw the green tube being overdriven, when I did get a pix out of it, it seemed relatively normal for the few minutes that it did run..

I'll check my notes, but I think I sent Terry Ferentinos the Ya board along with others to test, and he said all of the boards I sent him tested OK. But if it's intermittent for you, then I guess that's what happened.

If it is intermittent though, I'd suspect a bad solder joint on an SMT chip. Unless you find otherwise, chips themselves ususally don't go intermittent. Barcos are really bad for bad connections on the SMT leads of chips.

As mentioned earlier, there's a leaky diode (esd) on one of the data lines. That leaky diode will hold down data transmission causing some functions to not work at all, and will also cause intermittent operation.

It can be hell to track down, but it's a very common data problem on a lot of Sony broadcast gear, as well some other stuff.

As mentioned earlier, there's a leaky diode (esd) on one of the data lines. That leaky diode will hold down data transmission causing some functions to not work at all, and will also cause intermittent operation.

It can be hell to track down, but it's a very common data problem on a lot of Sony broadcast gear, as well some other stuff.


Suuuuuure, NOW you tell me!

:)

(then again, I never asked)

Thanks for the points guys,

In fact the issues are not intermittent at all. Then I don't think this can be a solder on boards. I rather suspect something wrong with the YA board I²C lines drivers chips (HCT541) but I do have difficulties to find this part with 20 pins narrow body shape (only wide) :(

About diode, yes that can be a reason if there is such diodes on the I²C lines, but as far as I know (I will check again ;) ), Sony did not use ESD protection diodes on these communications lines ...

Later then for the work progress :D

John

I know this is simplistic, but I discovered that one of the major problems with my G90 was board sag. In particular, the YA and power boards in the back of the set were sagging and shorting out. I propped them up with a small piece of plastic, and solved most of my problems.
Just a note for posterity. E.

Sony did not use ESD protection diodes on these communications lines ...

John

You're right the don't use "ESD" diodes, however they do use regular switching diodes for ESD diodes.

Hello Mike,

Spent some time on the G90 Service manual and didn't see any diodes on the SPI/I²C lines between YA and BA/DD/DE/DF boards. I only seen 100R serial resistors in front of each receivers ICs SDA/SCL inputs on BA/DD/DE/DF schematics.

Ed, Thanks for the tip but checked this when I did cleanup all the G90 unit boards.

John

Hello,

Here is my update on this G90 servicing :

Found BA board Green Bias issue !

Spent some time on probing BA (4/12 schematic) RGB BKG CONTROL section to compare analog signals between Green (defective) and Red/Blue (OK). Found that BKG-G output signal was too low (1.32V) instead of ~3V.

- Tried to change IC601 (MUX) and IC604 (Current to voltage converter) : no fix :confused: .
- Spent one hour with removed BA board on impedance probing around IC601 vs IC600 and IC602 : found that IC601 pin#1 impedance vs +12V rail was 10K higher than the two others chips :rolleyes:
- I then usoldered R603 (100K 0.5% SMD0805) and check it with ohmmeter : found it opened :eek: :eek:
- Tried a new 100K SMD0805 resistor in replacement, plugged back BA board on G90 and fire : works fine now :D :)

BA board failure solved :)

I never saw such open resistor like this on board in my all EE life :eek:
Let's repair the YA then !

John

ARe you sure it was open or just cracked at one end? I've had lots of them crack, but you're right, I've never had one OPEN!

No Curt,

The resistor was not cracked at one end... its glass body and even its terminations looks beautifull under my 20X/40X binoccular magnifier !

It was simply found OPEN : awsome !

John

Wow so the G90 now works? does the unit respond to input on the menus now?

Not yet Don,

In fact I've fixed the first failure of the two BA/YA boards: the BA board Green bias failure :rolleyes: . I still have to fix YA now (and I really hope it's not the CXD305-127R DAC controler chip :o !)

John

Please keep us posted, I think this is one of the first if not the first G90 repair threads.

Pictures would be interesting too.

Frank

John-
Do you mean open as in "no resistance"? If so - how the heck does that happen? E.

No, as in infinite resistance..:)

It must be a hairline break in the substrate..

WIth no offense to John, I'm sort of glad it's a difficult problem. THe world might have been down one CRT tech had John posted that it was something simple like an EPROM reset..:)

I spent about 20 hours on that G90 before I sold it (with as much info as I knew about the problem) to John.

John seems far more adept at low level digital signal stuff than I am.

No, as in infinite resistance..:)

It must be a hairline break in the substrate..

WIth no offense to John, I'm sort of glad it's a difficult problem. THe world might have been down one CRT tech had John posted that it was something simple like an EPROM reset..:)

I spent about 20 hours on that G90 before I sold it (with as much info as I knew about the problem) to John.

John seems far more adept at low level digital signal stuff than I am.

1 projector to fix might be easier to devote time too than having a workshop full of sets waiting for repair :rolleyes:

Yes, but sorry newbieDAN, I also have two others G70 to repair as well !
See my workshop state (some pics for Frank ;) ):
http://johnhwman.chez-alice.fr/G90/Atelier1.jpg
http://johnhwman.chez-alice.fr/G90/Atelier2.jpg

Cur't original G90 under servicing and my brother's G70 unit (that have D board failure), then another friend's G70 that do have a C block incompatibility with unit B board :eek:

So, I should say that my workshop is almost full of sets waiting for repair :D

John

1 projector to fix might be easier to devote time too than having a workshop full of sets waiting for repair :rolleyes:

Not when you have ADD. If I get stuck on one, I can move to another...;) :D

John,

Glad to see you have the beer fridge close at hand in your shop!!!

Or perhaps, given your location, it is the wine cooler.

In any case - High Five

:)

John, have you tested those "tv-serviceman´s trouble finding routines" hairdryer & coldspray??and are you taked good look at those smd elyt.caps, many times leaky caps..Forget! you know this stuff allready :D

No Curt,

The resistor was not cracked at one end... its glass body and even its terminations looks beautifull under my 20X/40X binoccular magnifier !

It was simply found OPEN : awsome !

John

This is a first.. :confused:

I see in this thread that ES protection is built into the data chips. And the more I think about that resistor being bad, the more I'm thinking this set has been exposed to shock (static, etc).

According to the HT people I bought it off of, it simply stopped working one day at a customer's house. Something like 8K hours on the chassis.

Congrats John!!!!!! Keep it going so we can absorb your EXTREMELY VALUABLE knowledge. :^)

Cliff

Thanks you Cliff ;)

Okay guys, after BA board repaired, next was YA.

As said earlier, I was expecting that the I²C bus problem that YA board exibit was maybe due to the IC13 (74HCT541AF - narrow body) buffer located between IC421 (DAC controler) and all the rest of unit boards (DD/DE/DF and BA) :rolleyes: .

So I bought a 74HCT541 chip to replace on-board. But despite hard searching, I was unable to find the same narrow body chip as Sony used. Then I was forced to order one wide body version from Farnell some days ago.

Just received the chip today and then update the defective YA board. Unsodering of original IC13 was very difficult because Sony did put some Red glue large points (4) under the chip for the wave soldering process. But I finally succeded to remove it :D .

So I soldered the new 74HCT541W (wide version) on the narrow IC13 footprint by bending all the pins (thus rizing the IC over the PCB). Here is the pictures of the result (hey ! it's for Franck ;) )
http://johnhwman.chez-alice.fr/G90/YA/IC13_1.jpg
http://johnhwman.chez-alice.fr/G90/YA/IC13_2.jpg
Pretty good indeed :D

OK, then I put back the updated YA board within the unit and fire-up again.

You know what :confused: the unit power up gently and display a perfectly stable image :eek: without been choppy as before. The only thing is that the projected image was too dimm.

I let the projector for a while to see if the random 8A/89 errors comes again... Nothing happen and the image was perfectly stable. So I guessed that the IC13 was the source cause of the I²C communication problem and now is gone :D

Next post to follow-on

Okay, so I wonder why the display was so dim :confused: I tried to rize the Contrast/brightness and this done nothing on the display :confused:

So I was wondering if this was maybe something wrong with the default user memory setup (tubes Bias or Gain) and decided to try to RESET all the user memory with factory Data :

- I pressed [RESET] on the remote for 5 seconds and then answer that would like to update all memories.
- Then the unit ask if I want to reset them with Service or Factory Data, I answered "Factory Data",
- And immediatly the unit switched off itself and display a Red 'd5' code at its back display (Factory NVM write error).
- I then tried to power-up again the unit with no success. I switch-off the main for 20 sec. and switch on back the main :
---> after a short time red '88', unit displays green error code '39' and hang.

Tried to power-up again the unit with no success :confused: unit displays green error code '39' and hang all the time.

What was happening ?

The Green error code '39' isn't described in the SM :eek: and the green color of this error SHOULD let me power-up the unit with remote : it's not the case :(

I tried to create Factory Data using the SM Page 5-10 procedure with no success (nothing happened and the display still shows '39').

Is the 'Factory NVM' IC334 chip ? If yes, this is a standard flash memory device and it's on socket ;) .
http://johnhwman.chez-alice.fr/G90/YA/IC334.jpg

In the Service manual they called the 'Factory domain NVM' ('d5' code)

I'm puzzled :confused:

Could somebody tell me what I did wrong with factory memories recall (all reset) ??

I just did this on my brother's G70 two days ago and it was woking nice :o

Now, I can't switch on my G90 anymore and Service manual can't say why, indeed :(

Any idea ?

John

Excellent pictures!

Really sorry to read about the subsequent problem.

IF you blanked the system you will have to recreate the factory memorys or you get an error. I can't remember but I saw this before you have to hold one of the two reset buttons down by the factory dip switches while you power it on. Then use the write system and factory information as explained in either the service or theory of operation manual. After that the re-write will be fine.

Thanks for the point Don, but I think I made it without success (see above : "I tried to create Factory Data using the SM Page 5-10 procedure with no success (nothing happened and the display still shows '39')).

As far I understand, according to SM page 5-10, you have to put S201 dip switch#1 to OFF while the unit is OFF from main then power-up the unit from main and put S201 dip switch#1 to ON again. This should re-create factory memorys but no success for me :( ...

John

Dam Wonder if reloading the Firmware will get you anywhere?

I tried Glassact method to initialize the User SRAM :

- When the YA displays '39' at main start, I pressed S202 for 5 second,
- Then YA display a big red 'd5' error again :eek:
- Tried again : same result.

OK, so I wonder if this is something wrong with Dallas DS1245 SRAM device since this device seem's to be the cause of this 'd5' error.

- I found one blank Dallas DS1245-100 chip and replace the original IC333 device,
- Putted back the YA board in the unit and power-up from main,
- Then the display is green '30' thus requiring a SRAM initialization (it's logical since I've populated IC333 with a blank device ;) ) ,
- I pressed S202 for 5 second and after 1 second, the display shut off :) (IC333 SRAM initialized with success),
- I push 'ON' on the unit remote and the unit started again :)

- Enter Service Mode and check the unit serial number and timers values : all are zeroed :( (as expected, in fact :rolleyes: )

But unit seem's working nice exept that displayed image is too dim.

I've let the unit for running one hour. After that time, the displayed image started to shake again and finally the unit shut off itself with 8A orange error. So it's seem's that replacing the IC13 device did not solve the I²C communication issue after all :eek:

Because my workshop is quite cold these days (4°C), the unit seem's starting to have communications error only when unit gets warm :rolleyes:
Yesterday, I did not wait enough time :rolleyes:

Okay so back on YA original failure then. IC13 was not the source of the failure. This means that now it's probably IC421 (Sony CXD305--127R DAC controler chip) and this is a bad new indeed :(

John

IC13 was not the source of the failure. This means that now it's probably IC421 (Sony CXD305--127R DAC controler chip) and this is a bad new indeed :(

John

No, I would not suspect IC421 because of its function. The problem could be anywhere on the I²C bus line. And as I've indicated before, this reads out as an ES damage problem to me. And the best way to run this problem down is to pull-up (disconnect) each pair of 100ohm data coupling resistors as much as possible along the bus line. Meaning the ones that would not hinder the main functions of the set.

John:
I was afraid it was the SRAM chip, and didn't want to mention it for fear of jinxing your YA board.
The interesting thing is that my G90 has just developed image shakes as well - never had them before I put it in the cold and replaced the chip. I wonder if the cold is doing it - or if something else needs to be burned on the chip. in any case - I'll be watching your progress with great interest...E.

The problem could be anywhere on the I²C bus line. And as I've indicated before, this reads out as an ES damage problem to me. And the best way to run this problem down is to pull-up (disconnect) each pair of 100ohm data coupling resistors as much as possible along the bus line. Meaning the ones that would not hinder the main functions of the set.
Why not Mike, but as said before, replacing whole YA board with the one from my first G90 unit let the second unit works fine. Then, it's logical to say that the I²C bus failure comes ONLY from the YA board I²C lines. Thus, not from the rest of the I²C lines within the unit (to DD/DE/DF and BA boards) ...

If you look on the I²C lines layout on YA board, they are not as long as you think (only 40mm or so). I've checked these 'shorts' tracks links between U13 and CN421 DIN96 connector, checked the 100R resistors on the lines and saw nothing wrong :confused: Then, I'm thinking that the only remaining device before U13 is IC421 itself :rolleyes:

John

I agree, but without looking at the manual, I'm thinking that IC421 was not the only bus line chip on the board. If it's the only one on the board - yes, it's suspect.

I was afraid it was the SRAM chip, and didn't want to mention it for fear of jinxing your YA board.
The interesting thing is that my G90 has just developed image shakes as well - never had them before I put it in the cold and replaced the chip. I wonder if the cold is doing it - or if something else needs to be burned on the chip. in any case - I'll be watching your progress with great interest...E. Yes Ed, in fact, the unit had 8000 hrs chassis timers, so I can imagine that the Sony original Dallas chip may be quite old and this could partially explain why there was such 'd5' error :rolleyes:

About cold environment, I don't know. But I'm sure that my unit shake the whole image (sometime on one tube only) only because of I²C registors writing errors (bad selected address within the I²C frames). I always seen that the image starts to shake a lot just 2- 5sec. before shutting unit down with EBQ/EBH relevant errors

John

I agree, but without looking at the manual, I'm thinking that IC421 was not the only bus line chip on the board. If it's the only one on the board - yes, it's suspect.True Mike, There is another three I²C buses on YA board (generated from IC301 I/O port controler). These three buses are labeled "I2C-SCL/SDA1, I2C-SCL/SDA2 and IFB-SCL/SDA" :

- I2C-SCL/SDA1 pair is used by BA board chroma decoders / converters ICs but not on Brightness/Contrast/bias/gain/gamma adjustments circuits (also on BA board),
- I2C-SCL/SDA2 pair is unused,
- IFB-SCL/SDA pair could be used by IFB slots B/C if needed.

So I don't think that these buses are relevant to the seen problem ;)

John

Sir sound slike you are getting close. It's down to the YA Board now is it?

Yes, Don

I'm quite sure that the failure is only relevant to YA board since I swapped this board with a working one and the unit started to work fine on all aspects (see above page 1).
John

Yes, Don

I'm quite sure that the failure is only relevant to YA board since I swapped this board with a working one and the unit started to work fine on all aspects (see above page 1).
John

Could I ask if the tubes were also perceived dim with the replacement YA board....

Could I ask if the tubes were also perceived dim with the replacement YA board.... Yes, you can ask :D and no, they were shooting normal G90 unit luminance since Brightness and Contrast settings in the video memory #1 were correctly handled by BA boards DAC converters (because they were properly sended within I²C frames ;) delivered by the working YA replacement board).

John

I know who I'm sending my boards to if one of my G90's fail............

Cliffy

Can't wait to see what it turns out to be.

John:
When IC333 starts to go (or goes completely as in your case), the factory resets don't write correctly to the SRAM. Why not take the SRAM chip off the good board, and put it on the old board to see if it corrects the dimness problem. It worked for me. Best - Ed

Thanks for this idea Ed,

In fact two reasons why I didn't swap IC333 chips between YA boards :

- I didn't wanted to loose all my first unit adjustments recorded inside the Dallas SRAM. My first G90 unit is the one I use every day to watch movies :D

- The dimness problem is NOT due to SRAM content but simply because BA board bias/Drive/Brightness DACs (IC609/611 on BA schematics page 4/12) are not feeding with correct I²C values ;)

I should say that my dallas chip choosed to fail during my YA board investigations but I don't think it is a coincidence since the Writing error i've got should rized at any time I would have decided to reload the factory setting inside the SRAM (i.e. even , if the unit was functionnal ;) )

John

John:
For all its worth - your original sram data would not be touched by changing chips unless you decided to overwrite it in the second g90. I've changed chips several times without ill effect. Best - Ed.

The tinkerer in me loves this thread. Best of luck in fixing this. I don't think I would ever buy a CRT, but I think I can understand what part of the attraction is: the DIY aspect of keepng them going. I wish I had the time!

Okay guys,

After getting back from my two weeks vacation in Martinique Island (what a wonderfull stay :) ), I took some more time to continue to investigate the YA board issue.

I tried to probe the operation of the IC421 (CXD305-127R) DAC controler with scope.

I first probed the IC421 Chip Select (CE\ signal @pin #51) to monitor the main CPU writing activity to this chip when changing parameters (i.e. Contrast/brightness, ...). I also probed the D-SCL signal which is the I²C clock output from the DAC controler chip.

This was interresting to see if every Main CPU parameter change requests orders are well interpreted by DAC controler (then if it is well generating the I²C frames accordingly).

What was my suprize when I saw that Main CPU R/W request (i.e. CE pin activity) appears to be almost all the time (and especially when CONT/BRIGHT keys are depressed) BUT I²C SCL activity is most of the time completely flat :eek:

Even it is explained in G90 theory of operation that the I²C frames generation is asynchronous to Main CPU requests (because parameters changes must be done during VBI only) : I saw that most of the time, the I²C frame are NOT generated at all for a while and sometime generated 10 to 20 seconds after !

My feeling is that there is definitely something wrong with IC421 behaviour, no ??

I'm affraid I must now find a spare YA complete board for replacement :( :(

John

John,

I don´t know if it is available, but it is listed here:
https://www.ued.net/ued/addItems.do?itemCode=SNY875239188

If you want I can give a call to Sony Germany also...

Axel

Thanks a lot Axel,

I'm currently trying to order one piece of this IC and let you know if I will get it or not.

John

All,

Have you seen this : http://195.2.37.154/indexes/pi/EB/Jun_2006/00042041.pdf ?

Sony said that "The new YA board does not contain the ICs, which are no longer
available as service parts."

And IC421 is in the part list :( :( ...

....
Seems one advantage of Sony in comparison to NEC or Barco - the availability of relatively cheap spare parts is gone... :(

John, try to insist on it .... there was the rumor that Sony promised the availability of spare parts for 10 years ....

Axel

John:
Is it possible that the same chip is used on the G70? Boards for that pj are much easier to get hold of..... E.

John,

There is some contention about that technical bulletin being a hoax.

Talking "technical bulletins"...is there a source for these for the G series? I tried a search of the URL in the link...but no joy...also Sonys FTP site requires a password....can somebody point me in the direction of?

Talking "technical bulletins"...is there a source for these for the G series? I tried a search of the URL in the link...but no joy...also Sonys FTP site requires a password....can somebody point me in the direction of?

I would like to know this also. I know that someone here must have these bulletins, and I'm surprised they haven't surfaced yet.

Sony wants something like $4K for their technical "library".

If indeed the tech bulletin is true, then frankly I think not only should Sony get a big bitchslap across the face, they deserve to go under IMHO. I think it's high time the penalty should be for some large electronics manufacturer to go under. Maybe it will shock some of the other manufacturers to start making QUALITY items that are properly Beta tested and supported after the sale is complete.

I wonder if the Japanese gov't would give massive tax breaks like we do here in North America to allow a large corporation that is sinking to remain alive and to restructure or whether it's so cut throat over there that they simply let giants die so that others may take their place?

There is some contention about that technical bulletin being a hoax.

I seriously doubt it's a hoax. Especially since Sony has had blank PC boards on their parts list since I can remember.

Though this particular board does have some parts missing, it looks better than some of the ones I've received after thinking I would get a complete board from their parts depot.

That problem is ES (electro-static - cause, external to PJ..) damage. I knew that when there was an open SMD resistor that was not broken. Plus, everything posted on this problem indicates it.

I have a G90 service manual somewhere on a CD. I'll see if I can find it, and see what I can add here.

I've had problems many times like this, and my approach was to always pull-up the data lines and look for resistive shorts either on the chip, or the ESD protect device before the chip. But in the case, the ESD protection seems to be in the chip itself, and in most ES damage devices, the load can be detected with an ohm meter by pulling the data lines and testing for leaks right at the chips data feed.

Also, that same DAC controller chip may have another part number, and may not even be made by Sony. So it's possible that the chip is still around, but with another part number on it.

Is it possible that the same chip is used on the G70? Boards for that pj are much easier to get hold of..... E. Thx for the idea Ed but I already checked this and the G70 do not use this device but rather multiple separates CPU to provide multiple I²C buses.

Axatax : the technical bulletin is surely NOT a hoax, I've found it from www.sonysupport.com official Website by typing the "VPH-G90" sentence in their model search engine. Then I've got "REUSE ICS WHEN REPLACING YA BOARD 302006088 19/06/2006" link on the list. The issued date of this TB is just 8 months old ;)

Also, that same DAC controller chip may have another part number, and may not even be made by Sony. So it's possible that the chip is still around, but with another part number on it. I doubt Mike : this device seem's to be a pure Sony designed specific IC (and probably only for G90 projector). Also agree with you about ESD failure(s) on this unit but according to SM, the only part between IC421 and output connector is IC13 (TC74HCT541AF). This leave IC421 itself...

All : I still not have any feedback from UED.net about CXD305 device. I can order it and put it in my shopping cart. but at the end of the process I get "Sorry, international orders cannot be proccessed on the website at this time" for an unknown reason (tried three times)... Can someone from US try to proceed shopping this part and see if he can order it or not ? Thx

John

John:
According to the UED web page - the item (number 875239188) is not stocked. E.

True Ed, but according to https://www.ued.net/ued/addItems.do?itemCode=SNY875239188 , the "The projected shipping date of this item is 03/01/07."

So this means that they have to order it from another supplier (Sony).

John

....
Seems one advantage of Sony in comparison to NEC or Barco - the availability of relatively cheap spare parts is gone... :(

John, try to insist on it .... there was the rumor that Sony promised the availability of spare parts for 10 years ....

Axel


Umm......what planet have you been on???? Sony G series have always been the hardest to get parts for. I'm tripping over NEC parts - I'm giving half of them away to one of the other Aussie CRT guys.

OK, John:
I have ordered the part for 95$. If and when it arrives, I will send it to you. Let's hope that SONY has some squirreled away somewheres! BEst - Ed.

Thanks Ed, that's kind of you :)

I really appreciate but unfortunaly I've just received one answer from Harry Benson @ued.net and he wrote :( :
I checked with Sony and this part is discontinued unfortunately. Sorry I couldn't help.

I'm affraid you'll get the same answer after your order. We'll see.

I'm condamned to buy a brand new YA board or maybe a defective one (to recover IC421) if I could find such board around here :o

Any idea of a brand new YA board price in USA ? Here is France it's $2416 :eek: :eek:

John

Thanks Ed, that's kind of you :)

I really appreciate but unfortunaly I've just received one answer from Harry Benson @ued.net and he wrote :( :
I checked with Sony and this part is discontinued unfortunately. Sorry I couldn't help.

I'm affraid you'll get the same answer after your order. We'll see.

I'm condamned to buy a brand new YA board or maybe a defective one (to recover IC421) if I could find such board around here :o

Any idea of a brand new YA board price in USA ? Here is France it's $2416 :eek: :eek:

John

John there is a G90 for sale "buy it now" on fleabay with TOASTY tubes, maybe a few of you should get together and chop it up for a parts machine....

Okay,

Here is the summary of the 'ex' Curt's G90 unit after finishing servicing the unit :

- Found BA board Green Bias issue : R603 (100K 0.5% SMD0805) : found it opened. Resistor replaced : Solve the Green BKG CONTROL issue. Cost : free. Time = 3 hours.
- Ya board I²C buses issue : found DACs controler IC421 (Sony CXD305-127R) that should be defective. Hardly tried to find a replacement part on the net : no more available. Send the board to Ed. (glassact <- Big thanks :) ) to let him resend it to Sony repair sevice center at San Jose. Board fixed by Sony (IC421 and IC431 were replaced with part that came from Sony Japan). Total YA servicing cost : $1813 including Shippings :eek: No choice :mad: .Time = 15 hours.
- Found another BA board issue on B/C RGB inputs : Green channel grayscale is compressed after 60 IRE (non-linear). Issue located on IC2 MAX4121CSA (2:1 300MHz MUX). IC2 replaced -> issue solved. Cost : $5 . Time = 2 hours.
- Found EA board issue on H-KEY and H-PIN adjustments (that made nothing on the screen). Issue located on IC2 MC1495BP (four quadrant multiplier) which has pins 8/9/10/11 wih low impedance shorts. IC2 replaced -> issue solved. Cost : $5. Time = 2 hours.

As curt's feeling, this unit should been striked with lightning on Main as the failures I found are relevant to such effects on electronic chips...

Fortunatly, I could swapp the board with my first unit to ease the location of the failures :rolleyes: .

I've fully retubed the unit (thanks to Chuchuf for the Red, thanks to Fireball66 for the Blue and thanks to Greg Eisemann for the bare green P19LUG HR)

Changed Green 'C' Element with Marquee one for green color purity (thanks to Tim).

Added one IFB-DVI-V2 card in B slot.

I've spent 2 hours on adjusting D65 grayscale tracking with Spyder2Pro and Colorfact 6 Pro.

The unit was sold and installed yesterday : picture is stunning on 2.7m wide screen with HD-DVD from Toshiba HD-XE1 :)

Total cost for servicing the unit : 8500€ - Sold: 6500€ : this was not a good deal :( but it helps me to teach the complete electronic of the G90 : now I'm skilled in it ;)

Cur't you've been cleaver enough to sold this unit to me :o it was holding a lot of differents electronic failures :( The YA board one was the most costly one; unfortunatly for me :mad:

John

A financial loss, but a technical victory. It's great that you got it working again at all. Sony's policies are frustrating, but at least they have not yet stopped servicing the G90 altogether, as I originally thought.

-Mark

WOW!

Nice work John. It's really hard to fix a unit when there are so many different issues. I always find myself second guessing if I'm missing something. Like you say, now you are an expert though ;)

I used to own a business repairing Pinball (flipper) machines. I'd buy trashed ones at auction and rebuild them. Only choice was board level repair. Usually most units had dozens of bad ICs... The challenge was always the fun part.

craigr

Okay,

Here is the summary of the 'ex' Curt's G90 unit after finishing servicing the unit...

John

Hey Craig! You need to change you location in your profile now that you have you new place !

Athansios

PS Check out some screen shot i posted of the mArquee you calibrated scroll down a bit.

http://www.avsforum.com/avs-vb/showthread.php?t=784331&page=2

Congrats!

Very good work John!

- Found BA board Green Bias issue : R603 (100K 0.5% SMD0805) : found it opened. Resistor replaced : Solve the Green BKG CONTROL issue. Cost : free. Time = 3 hours.
- Ya board I²C buses issue : found DACs controler IC421 (Sony CXD305-127R) that should be defective. Hardly tried to find a replacement part on the net : no more available. Send the board to Ed. (glassact <- Big thanks ) to let him resend it to Sony repair sevice center at San Jose. Board fixed by Sony (IC421 and IC431 were replaced with part that came from Sony Japan). Total YA servicing cost : $1813 including Shippings No choice .Time = 15 hours.
- Found another BA board issue on B/C RGB inputs : Green channel grayscale is compressed after 60 IRE (non-linear). Issue located on IC2 MAX4121CSA (2:1 300MHz MUX). IC2 replaced -> issue solved. Cost : $5 . Time = 2 hours.
- Found EA board issue on H-KEY and H-PIN adjustments (that made nothing on the screen). Issue located on IC2 MC1495BP (four quadrant multiplier) which has pins 8/9/10/11 wih low impedance shorts. IC2 replaced -> issue solved. Cost : $5. Time = 2 hours.

Sounds nearly like a never ending story. I am glad for the happy end...

Axel

ERROR 66 on G-90

Have a G90 that suddenly developed a problem. Used it one night inputting 1080i off cable. Image was gorgeous. Trouble free. Next morning, the projector was screwed. Without input signal in its warm up cycle, I could see low res scan lines. Warm up post was shaky and distorted like there was some kind of internal oscillating problem. Switched projector off and on but no correction. Menu also had a similarly messed up image. At first there was no error message at all. Gradually, after a few attempted start ups I got an Error 66 reading. The manual indicates that the most possible problem is in the E Board, related to horizontal deflection (H. OUT). Having problems tracking down this board. It's part number A1346772A. Anyone know where I can get one? Anyone know if the G-70 projectors use a similar board? Any caveats on replacing the E Board in this projector?

Occasionally the projector can boot normally and output a steady signal in its current state. However, as soon as I cut the input signal the image destabilizes again. Or, if I press something on the remote a steady signal might suddenly destabilize. Now, when I press the menu key in the warmup cycle the screen goes blank. I'm not even seeing error messages on the led diagnostic anymore.

Don't know what first caused this problem but my guess is an electrical surge through the cable box to the projector. I didn't have the cable on a surge protector and I noticed the cable box was a bit disoriented too. I had to reboot the cable box to get it back to normal. This clued me in on the possible cause. Belatedly, I have the cable on a surge protector.

Posting two images here to demonstrate what the boot looks like.

Looks like a H-sync problem to me. I'm no Sony expert though.

I would that that any H-def problem would make it shut down completely.

Well, that's what the E board addresses. Here's what the G-90 manual says (referring to the "66" ERROR protector code),

"This protect works when the horizontal deflection is inactivated. The most possible failure is H.OUT of E board. If not, failures might be found in H. Drive series or the negative power source series. Look into DA/EA/ED/PD board."

I don't want to make it seem that all problems stem from the SVRAM chip - but it may be the problem. When the SVRAM chip runs out of on-board power, error codes abound - none of which are caused by defective boards. I have had d5 errors, 39 errors, 86 errors, and others come up - as well as bad images similar to yours. It seems that the SVRAM chip controls the electronic convergence functions of the G90 - and when the chip goes, the electronics go off scale, leading to numerous unrelated problems.
Worth considering, particularly as the problem suddenly occurred. Best - Ed

Good point Ed,

Had the same kind erratic errors displayed on the G90 two digits display when the YA board had the I²C buses communication issues (bad data sended). If the NVSRAM chip content is damaged, the data sended to the DACs cards are corrupted and generate numerous errors that are not relevant to their cards but rather to YA ...

John

John and Ed,

Both of you have noted problems peculiar to the G-90 and its SVRAM chip which gives false error signals in the boot protector circuit.

If this is the true issue here and not the board, how is this corrected? Is the SVRAM chip replaced or reprogrammed? Do you have a part number on that?

Thanks,

Howard

http://www.avsforum.com/avs-vb/showthread.php?t=841021

On G90, NVSRAM chip reference is DS1245Y-120 but you can use -90 or lower access time as well.

But based on your pictures, it seem's that the horizontal deflection board have a failure that NVSRAM content can't generate in your case...

John

John,
I have a G90 with an orange 89 error code. I swapped the EBH board per the manual with no success. I then swapped the DC board incase I was reading it upside down. This also had no success. I tried the YA board and it fired right up. Where did you send your YA for repair?
James

PS The IFB-DVI you sent me is awesome, thanks!

James, it's Sony in California. CHeck earlier in this thread, it's posted I think.

Be prepared though, as per this thread the price to repair it will be about $1700 USD, the price of what Sony used to charge for a new one. You might see if John will take a stab at it.

Umm, anyone see a trend here with G90s? Not good!

Umm, anyone see a trend here with G90s? Not good!

Don't let Cliff or Clarence see this!! :eek:

wallace

I had the orange 89 error with the bad SVRAM chip.
Look, folks - this is getting a bit silly. The chip has a maximum lifespan of 10 years. They cost 45$ to replace. Before you send the machine or board to SONY - make sure it is not the chip.
Once the chip starts to go - all bets are off. The pj goes out of spec and things begin to show up on the error chip. That seems to be the nature of the machine. Even if the chip is not the problem - it needs to be replaced if the g90 is over 8 years old. So says I.
Best - Ed

Which reminds me Ed - did you get the $45 from Adel? I asked him to send it and forwarded your details.

Mark

Umm, anyone see a trend here with G90s? Not good!

Yikes... This was going to be my "Dream Projector" now I am having second thoughts.

Mike

Ed,
I am willing to try your idea if you don't mind giving me a little help. I am not too familiar with board level repair, but if it is just a chip i can pop out and replace, then I can do it. Where can I get one, and would I need to do anything to it before putting the new chip in?