|
|
Post by crtbilly on Nov 11, 2018 19:06:54 GMT -5
Well.... Replaced ic1 but no diffrence. I have some replacements for Q1 and Q2 coming wish I could identify the ZD'S.
What a pain in the ass this thing is to work on.
|
|
|
|
Post by crtbilly on Jan 8, 2019 12:58:03 GMT -5
Ok, got back to this thing. I'd like to get it off my bench but know if I do it will sit on the shelf collecting dust and never be put back together.  So I thought that I'd try testing the optocouplers. This is in standby 0v 72v \ / [ 2581A.] / \ 0v. 70v 0v 70v \ / [°2581A] / \ 0v 70v 0v 70v \ / [°2581A] / \ 0v 70v ----ON---- 0v 29v \ / [ 2581A.] / \ 5.1v 55v 5.1v 55v \ / [°2581A] / \ 5.1v 55v 5.1v 55v \ / [°2581A] / \ 3.8v 55v I read another article saying replacement worked also suggested PC123 as possible replacement but wasn't used in the fix. Let me know if anyone has similar readings. Thxs |
|
|
|
Post by mastertech on Jan 8, 2019 18:34:47 GMT -5
Ha, talk about bad timing. I just last week put my RS45 back together since it was working properly and there did not seem to be any interest here in furthering info. I did how ever take a video of the spark gap firing on the ballast when attempting to light the lamp but have not bothered to upload it.
1 of those 3 optocpouplers is for the return signal which is why it is installed in the opposite direction of the other 2 so it would be better if you marked your charts above which optos were which by location number and with an attached image for reference.
|
|
|
|
Post by crtbilly on Jan 8, 2019 20:17:49 GMT -5
Damn.....I'm know for my bad timing I'd be interested in seeing the vid if you get around to uploading it. Yeah I'm working on a photo to show the findings better. I tried to mark pin 1 with the . And ° symbols and the orientation is as the appear on the board. Oh and Happy New Year everyone! |
|
|
|
Post by mastertech on Jan 8, 2019 20:29:29 GMT -5
|
|
|
|
Post by crtbilly on Jan 8, 2019 21:02:51 GMT -5
Yeah, I saw this. Need to hook up my logic analyzer and see what I can capture.
|
|
|
|
Post by mastertech on Jan 8, 2019 22:57:27 GMT -5
That was not my point. My point was to just jump your ballast and see if it works. |
|
|
|
Post by crtbilly on Jan 8, 2019 23:29:42 GMT -5
I like your style..... just get er done! I will check the UART commands and see if its even getting the "lamp on" or anything else. I will keep everyone posted.
|
|
joffe
New Member
Posts: 3
|
Post by joffe on Jan 21, 2019 15:55:10 GMT -5
Hello there, I stumbled over this thread and find this the most interesting in the whole internet regarding the RPB-0526GA lamp ballast that brings so much hazzle to JVC's X3/X7/X9 beamer series. Short history: I collected a very cheap X3B with defective lamp, aware of the lamp/ballast problem found in the net. Interesting challenge I thought and was mainly driven by curiosity to try 3D over a beamer. I own a JVC-HD550 and very like those JVC beamers.
Anyways, turned out that the X3 lamp has blown after 1000hrs, moved old light bulb from HD550 to X3 housing and gave it a try, it worked. Ordered an Osram P-VIP 180-230-1.0 cE19.5 bulb which is listed to be compatible. It worked but it was immediately obvious that brightness compared to HD550 (1000lm @hi) is not reached by the X3 (1300lm @hi). Both beamers with same distance and picture size I measured brightness in the center of both pictures using the white test-screen from the service menu. Got 202lx vs. 125lx. And the lamp of the X3 is beeping annoyingly at around 800Hz to 1kHz in low current mode.
Well, I suspect that the ballast is not working properly so I teared it out an get it operating on top of the X3, now able to measure some stuff. This may bring this posting to some useful purpose, the voltage levels for zener diodes ZD1, ZD2 and ZD3 as mentioned by stridsvognen and crtbilly - all three measure 18.0V to 18.4V so they are likely 18V zeners.
I guess this is enough for the first posting, in the next posting I'd like to describe what I am trying next and share more thoughts about how the RPB-0526GA is working and why I think it has some design flaws. Maybe there are some questions or things to measure, let me know.
|
|
joffe
New Member
Posts: 3
|
Post by joffe on Jan 24, 2019 20:52:59 GMT -5
Back again after some more testing. To check lamp and optical path for brightness a new ballast would be nice. Since an original RPB-0526GA would cost here around 360 Euro I was looking for an alternative and had a look at Jan's posting about Kerry Wong's video playing with an Osram P-VIP ballast. Crawled the net for all available docs about the P-VIP ballasts and can say they are amazing. Found a new one on ebay as spare part 55.JD30Q.002 for an Acer X1213P beamer, 29 Euro incl. shipping. According to the docs it should be possible to directly connect it to the X3. After some checks on the schematic and the ballast I connected it and powered up. It works \o/. The lamp is nicely driven with 190W in low and 240W in high mode, totally silent, no beeping or chirping.
But there is one downer: the Osram ballasts can use the enable line named SCI/Sync as a sync signal where the ballast synchronizes the lamp switching frequency and waveform to. The X3 indeed supplies a squarewave signal based on the selected framerate, it is 200Hz for 25&50fps, 192Hz for 24fps and 240Hz for 30/60fps. The Acer-configured Osram ballast does not like the maybe 2x too fast signal from the X3 and switches to internally 100Hz waveform for the lamp. Since the waveform is not strictly rectangular and designed for a rotating color wheel, the fixed 100Hz interfere with the shown framerate. It is a funny slow modulation for 50fps, barely noticable for 60fps and flickery for 24fps. Unfortunately there is no information available how to re-configure the Osram ballast parameters permanently. They can be changed on the fly using a UART control mode but that would require an inserted PCB that configures the ballast on every start. I think the Osram ballast can be configured in a way that low power, high power and waveform settings are set in EEPROM and remain persistent. That would be the holy grail and a perfect replacement. Another option would be to write own firmware for the AT90PWM microcontroller . Overall a very nice ballast with 2/3 component count compared to the Rubycrap. If anyone has some information regarding Osram ballast config, please let me know.
Now back to the RPB-0526GA and try to find the problem. Observing the waveform to the lamp, which Osram did so nicely, Rubycon has break-ins and overshoots. My first check was the widely mentioned LNK363 low voltage supply for the driving electronics. It works fine, stable 18.4V but according to the datasheet the wrong chip to use? According to the datasheet it should be used with a transformer and connect the FET-source to GND. But they use it as a buck converter with inductor L2, maybe the LNK363 should be replaced by an LNK306 which resembles according to datasheet the application used on the ballast. Maybe typical way of chinese component sourcing...
The next stage is the lamp voltage generation as a buck converter to bring down the 390V input to the required 20-80V. This is mainly done by the M51995, T1, Q6, L3, D5 and C28. This generated DC voltage is then converted to AC by the full H-bridge hybrid SLA2601M (IC6). The lamp voltage and polarity switching is generated by an H8/3690G microcontroller. It seems that the MCU measures lamp current and adjusts lamp voltage accordingly. The lamp AC switching is synchronized to the enable-input squarewave signal. In low current mode the lamp supply frequency is half the input frequency with some fancy short pulse switching on each edge. When polarity changes there is a short 200µs/500µs cycle, maybe the secret lamp saving sauce. On high lamp current it gets more funny. For 0.5s the frequency is 1/4 of the input frequency followed by 2s equal the input frequency. If anyone could confirm this behaviour please let me know, maybe this is a defect of my ballast.
Anyhow, in my case the lamp voltage is not stable so I focused on C28 of the lamp voltage generation. Removed it and replaced it by a 2,2µF/350V low ESR electrolytic cap that I had at hand (watch polarity). The lamp voltage is pretty stable now and lamp power is 180W/230W and brightness is way better. Overall chirping is now reduced but slight lamp voltage ripple on polarity change is still left. Maybe the M51995 switching regulator or the SLA2601M bridge still have issues. I guess I'll put the X3 together again and keep experimenting with the Osram ballast as a final replacement.
Hope no one fell asleep during reading and sorry for the rambling . |
|
|
|
Post by crtbilly on Feb 5, 2019 11:03:20 GMT -5
Thanks Joffe for the zener readings. Could you look on your sub pwb assy board and tell me what led's you have light. Well I tried what master suggested and supply 3.3 volts to pin 4 and nothing. I tried to capture communication on rx and tx but wasn't able to read it. I used settings from wongs blog but still could see data properly. Will keep trying. The picture I attached show mine as it's attempting to light the lamp. Green Green Not lit Red On my scaller board I have led1400 red and led800 green.  |
|
|
|
Post by stridsvognen on Mar 14, 2019 9:47:13 GMT -5
Anyone had luck repairing the Balaster.?
Seen that the X75 lamps should fit in the older models, so wondering if the newer type balaster might work as well.
|
|
|
|
Post by omegasteffy on May 3, 2019 11:54:27 GMT -5
Is SA1 the spark gap?
|
|
|
|
Post by mastertech on May 3, 2019 21:17:13 GMT -5
Yes
|
|
|
|
Post by stridsvognen on May 15, 2019 11:26:54 GMT -5
I ordered 2 new ballasters from JVC for 227$ each, as i got a X3 with defect ballaster as well, hope to get them both up running when the ballasters ariv.
|
|
