|
Post by gjaky on Feb 10, 2018 15:53:36 GMT -5
This is something I'm struggling from the very first time since I have my "new" 2010 VDC Marquee 9500LC, as forefront the unit came from a TOFT simulator, yes, and has ~16000hrs usage with ~8000hrs on tubes. So there is some sort of interference seen on the projected picture, the effect very much looking like 480i with massive scanlines, but these "scanlines" are not synced to either line (H or V) frequency so it has a definite waving effect too, and are present on any resolution. But this noise onlyappears to affect the vertical plane, but not the Horizontal.
I've easily narrowed down that this is not coming from the video chain as I've tested with the same results the original VDC VIM+VNB combo, the classic line up (02p VIM + ...3339 VNB) and of course with both my mod boards. I've had once similar on a NEC 9PG xtra and there the "noise" came from the dyn. astig driver.
Long story short, I found that the noise only appears on screen when the convergence yoke is plugged in. The most obvious idea is to recap the whole CVA, but it is not that funny to think about it, so let's chat a little bit first.
The board does employ Rubycon caps just like the ancestors, they all look good, although I haven't pulled a single one to measure, so in fact they can be bad, but why only vertical then? Also, isn't there some grounding issues around those boards, if so what was the deal?
|
|
|
Post by thelasttubewarrior on Feb 12, 2018 8:23:45 GMT -5
I have seen this exact issue as you describe it and swapping out the LVPS instantly fixed it.
The original power supplies were made by MicroEnergy and they were great. After ME stopped making them for VDC, VDC found alternative suppliers and none of them have been as good even though they are identical, design-wise. Pop open two units and compare them side by side and I think you'll find that some inferior components will be found in the newer Chinese made supplies.
I'm not saying there can't be other sources of noise but definitely look at the LVPS. As anybody with any decent knowledge of electronics must eventually learn, no circuit can perform better than the quality of its power source allows.
It costs less than 50 dollars in parts to recap a tired MicroEnergy power supply with top quality capacitors. It's about a two hour job.
I would rather rebuild an old MicroEnergy supply than live with a fresh Chinese supply.
|
|
|
Post by gjaky on Feb 12, 2018 9:27:21 GMT -5
While I haven't take any measurements on the rails yet, but this is what I was thinking too, because the CVA has only 10uF caps per rails and failure of those could not make such dramatic effect, so LVPS is my first suspect. Earlier I inspected the power supply (while I installed my modifications to the standby section) and did not see any obvious sign of wear anywhere, but I think I'll take a closer look on the +/-24V rail puffer caps. Unfortunately I don't have any other LVPS at hand to play with.
|
|
|
Post by barclay66 on Feb 12, 2018 9:31:23 GMT -5
Hi,
Were You possibly referring to the 'ground strap fix'? I've downloaded the relevant posts from 'somewhere else'...
Kind Regards, barclay66 Attachments:Ground Strap fix_2P.pdf (583.24 KB)
|
|
|
Post by barclay66 on Feb 12, 2018 9:43:59 GMT -5
Hi,
BTW: The +/-24V rails are being used on the FGM, the +24V go to the HDM and the -24V to the belly fans. Any of these could feed some 'dirt' into the supply. If this 'dirt' had the right frequency, it could create interference with the convergence amplifiers and their pulsating current draw...
Kind Regards, barclay66
|
|
|
Post by barclay66 on Feb 12, 2018 9:49:23 GMT -5
Hi,
BTW2: I never understood why they would use 250V caps on the CVA although the supply voltage is only 24V per rail. Any ideas?
Regards, barclay66
|
|
|
Post by gjaky on Feb 12, 2018 10:39:17 GMT -5
Thanks Barclay,
Yes, that ground fix I was thinking about, but I did not remember what it really was.
I believe they used 250V caps to have "LOW-ESR" -or better said high ripple current- characteristics there (as the convergence waveform has current peaks at the scan frequency), that part of the schematic is basically untouched since 1992, at that time probably there wasn't too many low-esr caps to select from. *Edit: I am wrong here, with the Marquee 9000 they in fact used 470uF 50V caps there, still I think this was an ESR related decision.
I'll have a dense week, not sure if I could take any measurements what's going on with the pj.
|
|
|
Post by gjaky on Mar 11, 2018 16:45:36 GMT -5
I checked the +/-24V lines, and saw nothing obvious... I also took a look in the LVPS for the caps, again, even this Chinese made LVPS employs Nichicon caps in almost everywhere, with a few exceptions, one was the +/-24V line of course, where "Capxon" caps were populated. thes of course are not recognised by their superior quality, and neither were Low-ESR caps, as one would expect in an SMPS. So I replaced these with EPCOS low-ESR caps (EPCOS - B41888C7477M000). Nontheless the interference remained the same as before...
|
|