|
Post by tschaeikaei on Feb 15, 2016 15:11:51 GMT -5
Oh and again i broke something. During a mechanical setup i left the HDM uncooled. The main heatsink didn't get hot but Q20, the only part on the board having his own heatsink, said puff. Everything went dark, h-fail, i took the board out stole some parts from another HDM and finally got it to work again. Everything works but the main width setting (PIC-2 in the on screen menu). Strange thing is: if i start at picture width 0%, ramp up to 100%, everything behaves normal till about 90%, then only the number increases, the picture doesn't grow any wider. If i decrease from 100 to 0%, nothing but the numbers change between 100 and 65, then all of a sudden the picture shrinks and everything between 65 and 0% is again as it should be.
What i changed so far: Q20, R62, D21, D22, R1001. All these because of the initial fail of Q20. The resistors are meant as fuses and were ...exploded. I checked a quiete large numbers of parts including everything in the modulator circuit (right side third of page 3).
After realizing the width setting did not work properly (i had reassembled the projector completely at that point), disassembled erything again, took the HDM out and tried another HDM, another daughter board (HDB) and so on. I can definitely say that it is the HDM itself, not the daughter board or any other part of the projector.
Then i changed U17 and U20 on the HDM and tried again. As it was, strange behavior of the width setting. Key and Pin seem to work well, besides i did not ramp them up fully to 0 or 100%.
The next part i would change could be the multiplier circuit U18, but i don't like the idea of changing countless parts.
In addition, i cannot measure anything on the live HDM, because the projectors are both at the ceiling and i don't have anything to keep the loose boards in place while running.
So will you help me again? Remember i have two other HDMs here, one of them working and the other partially harvested.
Thank you in advance, Julian
|
|
|
Post by tschaeikaei on Dec 31, 2015 17:45:25 GMT -5
Do you mean 20-30µF each or 20-30µF both together. They are in parallel, aren't they? Regards, Julian
|
|
|
Post by tschaeikaei on Dec 31, 2015 12:09:16 GMT -5
I have a bunch of 390µF 400V electrolytics lying around. It would be better with 450V types, but yeah. Is there higher voltage than the nominal 390V to be expected? They are Hitachi HP3 type and used. According to the datasheet, 0,275Ohm at 20°C (room temp) and 100Hz. I expect the ESR to be rather good (in comparision to other caps datasheets). What capacity do you use? Regards, Julian
|
|
|
Post by tschaeikaei on Dec 30, 2015 1:28:14 GMT -5
I've done the resistor & cap mod on one of my HVPS today. The 560k were at 620k. Didn't measure more than that, also changed the 1µF and 0.47µF caps. And two 10 Ohm resistors on the daughter board. The big caps on your pics Nash, are they simply in parallel to the 0.47µF caps? Why tantalum? I guess it would be wiser to work at the LVPS 390V output and maybe lower impedance there than adding huge caps
that will only helo for a few miliseconds if more current is drawn.
Since there are so many different HVPS versions, i even ordered some resistors (e.g. R1) not beeing present on mine (Rev. B4). But nice results. Very visible noise lowering. Next project will be the LVPS, all I've done there is a fan mod, there will be more work, i'm sure. Feel free to link some info about HVPS and LVPS mods. Regards, Julian
|
|
|
Post by tschaeikaei on Dec 29, 2015 18:11:10 GMT -5
Late answer: I tested the +-15V again, not at the FGM itself, it is hardly accessible now, because I've bolted it into its metal cage for better heattransfer. But I checked at the astig board connector and the multipin connectors at the VNBs. +15,1V vs. -14,9V. Not exactly perfect, but seems ok. Regards, Julian
|
|
|
Post by tschaeikaei on Dec 15, 2015 7:49:19 GMT -5
I'm still a bit disappointed that the vcc voltages, especially 15 and -15v are so much off from nominal. Is this normal? I'm going to crosscheck with the other projector. But afaik, they come directly from the LVPS. Have to check for bad solder connections or worn out capacitors maybe. As far as I recognise, ripple was about 0.1V which I think is not too bad but improvable. What do you think? Regards, Julian
|
|
|
Post by tschaeikaei on Dec 14, 2015 9:41:39 GMT -5
Yeah I worry more about x-rays than about 1 or 2 miliamps. Even if it's 34.9kV. Good to see that it helps. Regards Julian
|
|
|
Post by tschaeikaei on Dec 13, 2015 14:28:08 GMT -5
VIDEOOk, after struggling with youtube there it finally is. I hope this helps a bit, redfox. Taught that may be better than a list. Regards, Julian PS: you know, english is not my mother language. That's why i don't find the right words sometimes. Be patient
|
|
|
Post by tschaeikaei on Dec 9, 2015 17:53:09 GMT -5
Ok, i fixed it. Had to change U3. Had to go down all the signals till the output of U3. Mostly using the scope. Oh and i need to get a scope with digital numbers readout. Should be much easier. I've gone through the whole chain, following the TOP_PIN signal. Noted every step first on a known good FGM, then on the failing one. I have a list and sketches of the signals i measured on the different ICs. Is it of interest to publish it? If I was gone through this way, it would have been faster. 1. Measure the output of U7. Yes-> got to step 2 No-> Vcc? Inputs? 2. Measure the output of U3. Yes-> go on further on the convergence amp module No-> Vcc? Inputs? No Vcc: Can't be.. the projector would not start if 15 or -15V were not present. No input: Check H_RAMP, H_PAR, TOP_MOD, BOT_MOD All of them ok? Check U7 pin5 and U15 pin5. And so on. Funny that it is not possible at all (even on 0.1V/DIV) to see any signal at the OPAMPS inputs. Nothing, even when crosschecking with the working FGM (and I have 2 of them now ) Oh and that damn electronics specialist dealer does not have anything. Not even the TDA8444. I paid 5€ for two 11W resistors and 12€ because i lost my parking ticket. And so on. Regards, Julian And thank you very much for helping me to understand and fix this, Mr. Sir Mastertech!
|
|
|
Post by tschaeikaei on Dec 8, 2015 7:13:41 GMT -5
Yes. That's the first thing I checked. Regards,Julian
|
|
|
Post by tschaeikaei on Dec 8, 2015 4:56:02 GMT -5
I'm at the university and can't do anything than read and think (which should be helpful in some cases Just for my own brain: If a specific output is not present, there are four possible reasons. First, there is no input (to the FGM board). I kind of tested that using the other FGM in the projector. This worked. But there could be something bad in the signal chain between Mainboard connector and the IC. Second: one or more IC supply voltages are not present. We checked that. They're fine. Third: broken IC or broken part of an IC. Check if inputs are present and see if I can measure something on the output. Oh and before you ask. I have an oscilloscope. Fourth: something between the IC's output and the board connector to the motherboards. No that I know which IC does what i can understand what i'm doing. I'll buy some TDA 8444 on my way home and check. Might get an MLT04 and MC34074, to. SDA and SCl are coming from the CLM, right? They're digital I²C bus signals that get converted to analogue waveforms by the TDA8444 . U7 is the interesting one in this case. Those waveforms get amplified by the ICs U9 and 11 MC34074 Quad OPAMPthen mixed together with the top and bottom "MOD" (?) signals that are created on page 5 of the schematic, meaning on the FGM board. If i interpret the datasheet right, you can't multiply a value with zero and get an output. So i should measure at e.g. pin 3 AND 4 of U10 to see if i get BOTH inputs. My conclusion on this is: U7 or U9 and U11 are bad. If the TOP_MOD signal wasn't there, TOP_KEY would not work, to. Same for BOT_MOD (BOT_KEY would fail, but this works, to). So forget them both. U10 could be partially defective,to. Feel free to correct me! Regards, Julian
|
|
|
Post by tschaeikaei on Dec 7, 2015 19:43:39 GMT -5
Yes, speaking about pressing Geom then 2 on the remote for keystone. It is top and bottom at the same time, affecting the whole picture. But i think you know what this setting does. My projector reacts normally in that regard. Seems only vertical settings are affected. Vertical convergence and vertical geometry settings are not working.
Regards, Julian
|
|
|
Post by tschaeikaei on Dec 7, 2015 17:28:21 GMT -5
Yeah i did count the pins in the wrong directions. Ok, measured again: U9 p4 +14.35V p11-13.35V U10 p5 +4.5V p14 -4.9V U7 p1 +11.9V
I mistranslated nearly everything. Horizontal convergence works, as i can move the raster left and right. It does not move vertically. In geometry menu: - (1) size h and v work, - (2) keystone works, - (3) side pincushion works, - (4) top and (5) bottom pincushion do not work, - (6) horizontal bow does work -(6) vertical bow doesn't work - (7/8) both linearity adjustments work - (9) horizontal skew works, vertical not
Sorry for the wrong terms!
Regards, Julian
|
|
|
Post by tschaeikaei on Dec 7, 2015 13:15:53 GMT -5
Ok, here is what i measured. +-15V at U9: -13.2V at pin 4 and +14V at pin 11 +-5V at U10: -4.5V at pin5 and +4,5V at pin 14
-2.494V at R93.
I guess that somehow the 15V lines get overloaded a bit. Or is it normal, that they're off so much?
Regads, Julian
@redfox: yes i know that. i just wasnt dure if they renumbered the board after that mod.
|
|
|
Post by tschaeikaei on Dec 7, 2015 6:20:31 GMT -5
Did you want to link a schematic?
It appears to be the same as the schematic I linked. I don't know for sure. Measured resistors r1 to r20. Some of them seemed to be fusible's. But no, they're all ok. Regards, Julian
|
|