CurtPalme.com Home Theater Forum

[barclay66]

An episode about the work on my Marquee 9500 Ultra. Feel free to share or laugh about my despair...

This Tuesday evening after a long day at work I wanted to measure and analyze the noise on the power rails with my newly acquired oscilloscope (HP 54510A, 250MHz, 1Gs). As I do have some minor noise in the picture I first tried to measure the +85V line at the blue tube's neck board connector on the upper mother board.
Unfortunately I slipped off with the probe tip and it momentarily made connection to the adjacent -5V line.
*BAM* and the PJ shut down. I know, I'm a complete dumbass trying to attach a probe while the PJ is powered on. It did take me some time to cool down from being pissed about my thoughtless action.
To my surprise I was able to power up the Marquee again using the remote. The CLM light went to green, no error light on either CLM and LVPS. The VDM went humming and the high voltage crackled into life. But no picture. The remote stayed responsive although the only thing I could observe would be the ability for powering up/down. I first measured the voltages at the LVPS connector under the blue tube and got this:

+85 rail -> +84.7V
-85 rail -> -84.8V
+5 rail -> +5.0V
-5 rail -> -6.1V

My first guess was that the main damage most probably would have occured on boards connected to the -5V rail because they were exposed to a much larger voltage with reversed polarity. From a technical document ("Marquee Power Supply Troubleshooting") I could read that the following boards are connected to -5V: The neckboards, CLM (including daughter boards), VIM and VDM.

As the VIM's relays didn't click (I'm using a Moome card, so it's Input #2) I assumed that something could be wrong with either the CLM or the VIM but swapping both showed no effect so I put the original boards back in.
On the VIM I tried to trace some voltages/signals:

- +15V, -15V, +5V and -5V were present after the on-board fuses
- The I2C bus showed activity after power up
- No video signal present after the input relays

Next thing I did was to place a jumper wire so that the input relays would be forced to Input #2 and voilá: The picture came back! Very faint and red only. I had to crank up contrast and brightness to the max in order to really see something. Geometry was off and obviously all of the Marquee's settings had been defaulted. I attribute this to swapping the CLM as the other CLM had a different software version (older). Unfortunately without green I couldn't see anything of the menus so I swapped the red & green mini RCA connectors on the VIM.
All of the menus were present so I could switch to Input #2 and remove the jumper wire. After that I found that I had video on all of the three mini RCA outputs.

After that I swapped the blue neck board - now I had blue and red working, swapped the green neck board - still no green, swapped the red neck board - no change. At that moment all neck boards were replaced with spares. I tried to adjust the G2 levels and I was able to get a visible raster on green too - but no picture content. Additionaly, what left me puzzled was that the blue channel would shut down whenever the picture had too much intensity. So when switching through the test screens, blue would fade away on the grey scale screen and the full-white screen and would reappear at the following grid pattern. The picture would completely disappear when adjusting brightness below the on-screen value of 96. I measured the G2 voltage on the red tube while changing brightness and it stayed rock solid.

At that moment I was really lost. There were some things I could be sure of:

- The CLM seemed to work perfectly. The daughter boards weren't harmed as their output signals were present.
- The VIM seemed to work perfectly. It delivered the expected output to the neck boards.
- The spare neck boards seemed to be OK as I swapped them between tubes and got no change.
- The VDM seemed to work perfectly. Vertical deflection was present.

That left me with the LVPS. I got the impression that the -5V regulation circuit had been damaged which is why I was measuring more than -6V. This imbalance towards the perfect +5V may have had an adversary effect to all OP-Amps and DACs expecting a symmetrical supply.

Next evening I tried out with an untested LVPS spare. That fixed it! Unfortunately I didn't want to use this LVPS because it neither had the filament mod nor the fan mod done. In addition all of the capacitors were stock and therefore quite old. So my only option was trying to repair the original LVPS.

With some help (You know who You are!!!) I was able to determine that only the -5V rail of the LVPS had been damaged. And the only part affected could be the three-terminal regulator (7805A, U26). I removed it and could measure that it had "converted" itself to a resistor with ~800 Ohm. Replacing the regulator fixed it. So this was the only damaged part *sigh of relief*. Swapping back the original neck boards finalized this episode. I'm really amazed how robust these machines are.

Now back to my original plan...

Regards,
barclay66

[Nashou66]

Nice work and isn't it great we have people on this site who have the know how and willingness to help out!!

Thanks goes out to that person!!!

Nashou

[gjaky]

[barclay66]

Hi,

Yes, it's a little weird but they are using a positive regulator and its GND connector is the -5V output. I put in a LM340 5.0

Regards,
barclay66