Has anyone seen this problem before or could point me in the right direction?
Its best to remove and clean all the chips on the CLM and daughter boards with DeOxit cleaner and then followed by the protectant. I bet that fixes it.
Edit: yes most certainly an HDM heat issue.
I'm having this problem with a machine after doing the HD mod's, 2000 vintage 9500Lc ultra. After doing the work, replacing all capacitors, re-flowing the big transistors, etc. I Ran the boards here for 12 hours at 1080P, everything looked perfect. Customer reports oscillation after 2 of viewing. Problem disappears if you switch from short retrace to long.
I now have the HDM back + CLM (which I didn't work on). Tested CLM for over 4 hours and HDM for 12 hours at 1080P and saw nothing wrong.
Also had customer disconnect CVA, FGM, and Stig amp to verfy it wasn't it. the only thing I haven't re-tested here is the VDM.
triplington can you please verify that all 3 belly fans are working in your machine? Especially the one directly under the HDM?
One more thing, there's a panel right above the CLM that covers the ACON board and CCM. Can you please take this off and maybe put a fan back there to blow onto the CLM?
I really think it's the CLM as well. Tested this HDM here now for 24 hours with no problems.
Last weekend I had enough and the 3-eyed monster came down from the ceiling. Initially I found that the loss of vertical sync could be changed to a stable signal by removing the white connection on VIM. But then hours later the situation was reversed where I had to reconnect the white input when there was no "horizontal sync" to get a stable display - which lasted 10 minutes or so until I lost vertical sync again. I realize fall and winter are coming, but I was quite frustrated trying to find and correct this intermittent problem. In the end I decided it was no longer worth pursuing and will put the HC3800 digital I have up in the theater.
You probably have more reason to pursue with a 9 inch LC versus my 8 inch AC version.
new as in brand new from VDC or new to you but old /used
Last weekend I had enough and the 3-eyed monster came down from the ceiling. Initially I found that the loss of vertical sync could be changed to a stable signal by removing the white connection on VIM. But then hours later the situation was reversed where I had to reconnect the white input when there was no "horizontal sync" to get a stable display - which lasted 10 minutes or so until I lost vertical sync again. .
so you could completely disconnect the Vertical BNC cable and the picture would remain stable for a while?
Both posters who had the issues live in areas where there is an ocean with in 100 miles,
And I think a marquee owner in Spain that lived on the ocean also had a similar issue.
And Ralphs goes away in the winter months when humidity is lower typically. I Know Humid air can lead to more frequent electrostatic and other electrical arching.
maybe add a de humidifier in the room for testing, since you had no issues with the AC on since it is basically a de humidifier as well that also cools the air.
The reason it happens more when the PJ gets warm is that the humidity with in the proximity of the hat components would increase since warm air can hold more moisture than cooler air, right?
Just my Hypothesis.
yes - several times - in many cases it only needs the white wire ((second from left) - but then that didn't last that long maybe 15 to 20 minutes and it would lose sync (horizontal reported with the black wire off) During the remainder of the year (s) - 9months it would be stable with all cables connected the whole time
Regarding the humidity comment - it's been way too hot here to live without the ac on full time. I am pretty sure there is very little humidity in our house and it has been kept pretty cool (76-77). The attic is sweltering hot and the pj is (was) hard up on the ceiling - hard to pass the power cord to the front through the gap.
One tid-bit I left out was the effect of running without the back heat sink enclosed. That gets quite hot and takes a fairly long time to heat up (which may be coincident with my loss of sync. When I tried this last summer it seemed to put off the onset of loss of sync - but I did not try it again this summer.
I recollect a discussion by someone about adding custom fans to the heat sync for supplemental cooling. I never did that.
What I've found after all this reading is that the one thing all 3 machines have in common is you guys are using the VIM's RGB connectors vs. an internal HDMI card. I have done all my testing here with a Moome internal. WHY that would cause this problem I don't know but last night I ran my clients HDM and CLM for 12 hours solid and nothing. I even hit them both with a hair-dryer until the metal around them was uncomfortable to touch and nothing.
That's approx. 36 hours of testing the HDM and no fricking way it's bad. The only other board it could be at this point is the VDM but I ran that here was well previously overnight at 1080P. I have a feeling that simply testing the VDM here for hours and hours more would produce more of nothing.
This is what I'm going to do, re-cap Brent's CLM and send him the moome card out my test-mule. I'm going to figure this out if it's the last fricking CRT thing I ever do.
ARggghh, a Sony problem in a MArquee, say it ain't so
If it is not that then id say bad connectors/ solderjoints on the Mother board or back plane.
Thats all thats left.
Either way I will post back after I try it out a while in a new location.
(I also noted the wrong cable above which I disconnected to get stable for a while when this happened in past - it was the black cable - last one on left looking at pj on ceiling)
One thing I also did somewhere along the line while troubleshooting was remove a resistor (or wire?) that disabled sync on green as discussed in some tech bulletin. It appeared to make no difference but I thought I would mention that as well. Other than that the boards are as received.
I checked the belly fans and they're all working OK. I removed the cover for the ACON board and pulled the board to get better air flow to the CLM board. The problem still occured after the "normal" 30 to 40 minutes. I even tried blowing on the board with a hairdryer on cold, but it didn't seem to have any effect.
I can test any boards here that you would care to send over. PM me if interested
Tim at E-Tech Phoenix
Tim at E-Tech ooo firstname.lastname@example.org ......your Marquee Pro Shop!
I'm sending my cleint his CLM and HDM which I have tested here now for 3 days. Will move on to the VIM and VDM next if problem persists.
I hope it's not what Nash suggested, Mobo or back-plane, but that's really all that's left short of power supply's.
i would tear the entire pj apart remove the mobo and clean all contacts with alcohol.
remove the vdm cva and astig from the heatsink clean the heatsink and the fets on the 3 boards with ethanol and apply new cooling paste on the fets of the 3 boards.
i would solder all the fets on the cva too just to be sure (90 solder joints or so) this is done in less then 10 minutes.
if you have a known good working lvps and hvps i would test that first.
the problem occured with CVA and STIG amps disconnected
I have one strong theory left. The one thing that would explain what could have happened from the time the boards were sent, tested, re-built, re-tested, and re-installed. The ONE thing I'm still considering is the RGB connectors on the VIM, the ones on top of the board. There's been instances in the past of these connectors on the board developing small crack's in the solder from being handled. I know they were only un-plugged and re-attached a total 6 times from my place to the clients but if the solder connection was a little shaky from the factory then this would do it and be heat related as well since the joint would swell as it warmed up. I've seen less than ideal wave-soldering on some marquee boards before. Mostly too little solder (cupped joints) or even air bubbles in the joint itself.
recall 2 has the RETRACE setting set to "Short", whereas the recall 01 has the RETRACE setting set to "Long". True enough, if I'm on recall 02 and the problem happens, if I go into the Pic menu and change the RETRACE to "Long", the problem goes away - changing the RETRACE back to "Short" causes the problem to occur. This is AFTER I heat up the HDM opening (so it's still heat related). Any clues from this?
any ideas? why would short re-trace cause the raster to become unstable after the PJ warms up and not when it's cold and why would switching to Long retrace cause the problem to go away? Is this simply the common problem with MArquee's not liking short porch settings and developing scan problems when they're not long enough?
I still think its the HDM connector on the mother board. have him move the Cage that hold the HDM, wiggle it a bit while it is doing the problem.
Now thinking more as I type, why would it only do it with Short retrace if it is a bad mother board connector , unless the signals are on different connector pins.
haven't tried another HDM yet
I wiggled the HDM cage with the problem happening and there was no change (problem didn't get worse or better)
Also, once the problem starts to occur, if I blow hot air into the openings of the HDM cage on the side of the cage, the problem doesn't get worse quickly, whereas if I blow hot air into the cage hole below the HDM heat sink, the problem quickly gets worse (I can alternately make the problem come/go by blowing hot air and cool air into the cage hole).
- have tried another HDM with no change (problem still occurs)
- the problem only occurs with retrace set to 'short'
- with no external source, the problem only occurs when internal frequency is set to #4 - 62.58 horiz and 66.67 vert (1080p?)
- lowering hsize from 69 to 50 caused the problem to go away. I found that the delta needed for the problem to go away was directly proportional to temperature. E.g., As the hdm cooled, changing hsize from 69 to 65 caused the problem to go away.
- changing the keystone did have an effect. My setting is currently 27 with the lines vertical. If i changed the keystone value plus or minus to make the vertical lines tilt, then the problem would lessen and go away. Putting the keystone back to vertical caused the problem to occur
- the problem repros 100% by blowing hot air up (ceiling mounted projector) into the HDM cage opening (takes less than a minute this way). blowing hot air in the same direction to the sides of the HDM has no effect. blowing hot air into the HDM cage opening seems to cause the hot air to go directly into the belly fan chamber (possible that the fans then recirculate the air to another component, or a bad connection inside that area?)
- blowing hot air into the VIM chamber does cause the problem to occur, but takes longer than blowing hot air into the HDM cage opening (probably related to the hot air going into the opening at the back of the VIM chamber where the belly fans are located)
- blowing hot air into the LVPS has no effect (also, no lights are lit on the LVPS when the problem occurs). I've also removed the LVPS, cleaned out the dust, and cleaned the connector that slides into the backplane and the backplane connector too, with no change.
- blowing hot air into the CLM opening (from rear of projector or sides) has no effect on the problem (so, it's more than likely not the CLM)
next step may be to test/swap the LVPS, but it doesn't seem to be affected by heat so may be a dead end
does this ring a bell with anyone?
Blowing hot air into the HDM Cage causes the issue quicly,
Blowing hot air into the VIM which is located right below the HDM connector
causes the issue but at a slower pace.
I say there is a loose or broken solder joint on the mother board for the HDM, a cracked
PCB trace on the mother board, a cracked or loose solder joint on the HDM connector itself.
First resoled all the connector joints on the HDM. if issue persists move onto the Mother board. this will require a complete tear down of the PJ.
- wiggling the hdm cage has no effect. I actually pushed on the hdm card with no effect
- heating the connector side of the HDM has no apparent effect
- putting heat into the side of the HDM cage where the coils are has little effect (it requires longer application of heat)
I can put heat into the hdm cage opening and have the problem repro in under a minute. If i heat the ViM opening, it takes 3-4 times longer for the problem to occur.
Definitely a mystery...