View Single Post
  #11  
Old 10-30-2026, 02:24 AM
TinnedTrace TinnedTrace is offline
VideoKarma Member
 
Join Date: Oct 2026
Posts: 4
Yea, that's why I want a proper multisync. 24Khz doesn't really work on most monitors.

Some more progress, I think I've found which part of the protector is firing.

The protector has three inputs feeding the Q2006/Q2007 latch: the HV branch (D2002/ZD2001), the primary overcurrent branch (Q5F1/R5F3) and the FBT secondary leak branch (Q2001/R2005). I lifted one leg of each in turn. With R5F3 lifted it still trips. With R2005 lifted it starts every time. So it's the secondary leak branch. Its sense node (R5U3/ZD2002/C5U1) takes 1~2 s to come up to about 15 V, which is longer than the latch delay.

Then I put the scope on B+ (C5G3) at a cold start, with R2005 lifted. Right at switch-on there's a short spike to ~118 V, then it humps up to ~99 V for about a second before dropping back to ~55 V. Soft-start reference on IC5H2 pin 3 looks normal, ramps to ~4.6 V in about half a second.

The interesting part: I also scoped C2001 (the rectified flyback pulse from FBT pin 8, the same thing the HV protector looks at). That one just ramps up slowly to ~27 V over about 1.2 s and doesn't follow the B+ hump at all. So the HV isn't actually overshooting. For the first second the flyback just makes a much smaller pulse than it should for the B+ it gets, and the loop pushes B+ up to make up for it. No white flash, picture fades in normally, so I don't think it's beam current.

So it looks like something on the horizontal side is weak or in the wrong mode for the first second after power on. Next I'm going to check:
- H frequency at switch-on (is it 31.5 kHz right away or does it start somewhere else)
- Base drive on the HOT (Q561) during start-up
- The driver supply cap C554. The recap put 1000uF 50V there where the M700 has 470uF 25V, but the fault was there before the recap.
- Band switching (Q57x MOSFETs) in the first second

Smaller things I found on the way: the chopper sawtooth on IC5H1 pin 3 flat-tops at ~10 V instead of the ~7.5 V the clamp zener (ZD5H4) should give, so that zener might be open. Going to test it and maybe swap IC5H1/IC5H2 for LM393/LM358 while I'm at it.

Has anyone seen this behavior before? Or is that B+ hump at switch-on normal for these and the protection is supposed to ride it out? Still looking for the actual M500 service manual too if anyone has it.

(Only running it with R2005 lifted for short tests, not leaving it like that.)


Update:
Found the issue. C2005 was disconnected as the via was corroded. It also doesn't do 24khz... Just looked that way as I only saw a corner of the image.

Last edited by TinnedTrace; 11-01-2026 at 11:24 AM. Reason: update
Reply With Quote