Odd D9200 behavior

[lilshawn]

got a WG D9200 in. it's had blanked out clicking with the SMPS in overload, so brought it in to be worked on.

replaced some power supply caps and got it back up and displaying again.

except after it stays on for about 2 minutes and starts warming up, the bottom half of the screen starts getting tiny little horizontal scanline jitters that get worse and worse...  also oddly, a slightly darker horizontal line about the height of a single scanline forms across the width of the screen exactly in the middle just as these jitters start getting to their worst... before long the entire screen is randomly jittering.

i replaced all the capacitors around the IC403 deflection processor IC, but it didn't seem to help any.  Adjusting the monitor brightness down seems to alleviate the jitters somewhat, they are still present, but much less so....and turning it way up makes it much worse. another odd behavior is that the jitters will get REALLY bad then all of a sudden it almost stops before it starts getting worse again, as if whatever is causing it has suddenly drifted back into specification again, before drifting back out again.

seems the Wells Gardner approach to generally anything image distortion wise in their troubleshooting guide is to replace IC403. i don't want to just shotgun order a new KA2147 processor without exhausting other options first. anybody else encounter this know what it is? seems like something is causing the horizontal scanline sync timing to be randomly early/late started instead of the timing being even.

link to a short snap of what it looks like:
https://photos.app.goo.gl/gE7j5PsBoiLEGwDr5

and a youtube of the progression of the video image degradation

[buttersoft]

Your video is set to private. Not that it's likely i'll be able to help, but even so

[lilshawn]

my bad. i thought i set it to unlisted. i fixed it.

[buttersoft]

the video kinda makes it look like the set is starting to love sync, at least at the timestamp where the distortion begins it does. Could it be a low supply voltage on the IC you mention, or components around it?

[lilshawn]

looks like main VCC U403 (S1D2147A01) is expecting 10.8 to 13.2 spec'd and it is at 11.62 at startup, voltage does not change a single bit when it starts the jitters.

looks like i'm going to have to bust out the oscilloscope.  hopefully it still works.

[lilshawn]

well, i dug out the old scope and checked the hsync going into the IC and it's rock solid. so the sync going in is not the problem.

next i'll have the check the H-out from the chip and see what it looks like. if it's good, it's further down the chan in  the actual drive section... but if the output is dirty, it's the chip

unfortunately, now it's the end of my workday... so i'll be coming back to this next week...CAUSE ITS QUITTIN TIME

to be continued....

[lilshawn]

dug into this a bit more this morning.

checked waveform 12 (h-contol out from IC404) it's good and rock solid.
checked waveform 16 (h-control after sblkout bias has been applied? ...something like that safety blanking (xray protection switch basically)) and it's rock solid.

moved to horizontal yoke drive system

checked waveform 14 output from IC404 (smps control IC) and output is fine at cold startup ... got a weird dropout halfway through the waveform... but picture still seems to look okay. (could have been my connection, I had to extend my probe ground through the chassis with a croc clip wire so HV could have been affecting it)

https://photos.app.goo.gl/FdUtsJhuCQ32a28k7

.. waveform then sketches out when the screen does.

https://photos.app.goo.gl/DUNUkNXaYJCKAMqZ8

i'll probably just swap this IC404 out with another just to be sure... never know these days.

checked waveform 4 (IC404 input for current compensation) and it looks solid enough enough when cold, but it sketches out too when the screen does.

https://photos.app.goo.gl/W9cqsoexY3i2A4X97

moving backwards through the system looks like the input to pin 1 of IC404 is being (eventually) driven by T402's primary side. but not before some kind of "Locking" signal.

so i guess the next place to start looking is whatever this locking signal is coming from (apparently T802 and the main SMPS power supply drive IC circuit... for... reasons?) to make sure it's staying stable  and then i guess whatever is driving the primary side of T403

i'm eyballing Q413 and Q314... but we will see.

... to be continued. i gotta do something else for a while.


edit: sblkout is Safety Blanking output. (Activated during frequency changes, when X-RAY input is triggered or when VS is too low.)

also waveform 14 is current compensation circuit feedback into IC404

[lilshawn]

looks like that weird dropout I saw was not just a ground connection, as it appears to have manifested in another place, the pulse going out to "locking" signal. it started out good, and i was messing with the scope looking at the waveform closer up and how it changed when the picture geometry change... and it coming back to it i noticed the dropout appeared.

https://photos.app.goo.gl/S2JvhKfVPuQSxv5Q6

i guess i didn't take a pic of it, but i got a clip of it as it's sketching out, and it can be clearly seen as a dip in the beginning of the large peak.

https://photos.app.goo.gl/RpgqYahnPg1oYfBk8


so it also looks like "locking" is some kind of power supply feedback into the main power SMPS IC. not sure if the dropout is from IC404 or the h-yoke predrive feedback through T403, since it's looped around and connected together so it's cause is definitely coming from somewhere around here.

some kind of "hs" pulses (EDIT: Horizontal Size PWM signal) also come out of the main processor IC at pin 20 before heading over to IC404 i doubt it's the cause since it's tied to the blanking and B+ control pins (pins 22 and 23) of IC404 as well as the vertical dc bias control pin 33 and it seems to work in the vertical perfectly, so the signal SHOULD be good. though i still wont discount it since it's related the the waveform 4 I measured as sketching out and had that weird dropout in it. i'll measure right at pin 20 of the main processor IC and see how it looks.

if the main IC output on pin 20 looks good, i will probably just shotgun the area and source a replacement IC404, Q413, Q114 just to spite and see what happens. i mean for the sake of a couple bucks of parts, not a big deal.

if not, i have a trashed chassis here i can yank the main IC off and drop it in to test.

if it's neither... i guess i keep probing

[lilshawn]

i started out removing and swapping IC404, i haven't tested it yet, but I thought more about what was going on with what circuits where working perfectly and which ones seemed to have the problem. so i decided to step back for a hot minute.

have since discovered pin 1 with waveform 4 is a current compensation circuit for the yoke drive... and even though it does eventually connect up with the microprocessor HS signal, it's also tied into the E/W compensation circuit, and the E/W circuit and pincushion appears to be working as designed... either that or the amount of change in the signal is inconsequential to those circuits.

I've printed a section of the schematic and started highlighting and marking what was good and what was bad crossing things off the list of things that work... basically leaving that which does not.

I seem to have been left basically the block of components around IC404.

so i'm guessing the problem does indeed lie within IC404, and possibly D426, D425... or possibly also Q404. I'll be checking R438 (150k) R489 (20k) and probably swap a couple capacitors C472/C460/C469 since i have piles of those and the 1n4148 diodes. might as well go over basically this whole block...

[lilshawn]

Today, I finally got around to this project again and I threw a handful of parts at this IC404 area to see if any have an effect on the issue.

I replaced IC404, D425, D426 and went through and double checked the resistance values of all the resistors. (didn't find any out of spec) The only resistor i found odd was R415 (22 ohm) and it appears at one time it was a little warm, having some slight discoloration of the casing. It supplies a power rail to this circuit, it's a really low value, and it measured a little on the low side anyway (21.6 ohms)...so i'm not too worried about it since really fried resistors start going UP in resistance.

unfortunately it still has the issue, though somehow it is not as bad? it seems to run quite a bit longer now before starting to sketch out (had to run the monitor 15 or 20 minutes before even starting to see the jitters start to appear)... and when it does, it doesn't get nearly as bad as it did with the whole screen getting affected. It does still have that weird scanline dropout line right in the middle of the screen, and the jittering confined below this line.

this image was taken after about 20 minutes of runtime while i've been typing up this post....

https://photos.app.goo.gl/DXyM2MW6WycT9NLg8

you can see the line is still there... and these jitters look maybe about as bad as it got while it's running. (it's hard to get a pic of CRTs). it could be a fluke, and it's being extra nice today... or my poking and prodding and bending of components to try and read silkscreens to find stuff could have pushed something back into spec.

maybe at lunchtime i'll pick up a can of air duster and turn it upside down and give a few things a blast of liquid CFC and see if it smartens back up if something in particular gets cold again.

other than that... next to get the shotgun is... Q404 C472 and then another run to check. failing that, Q405, Q419.

welp, back to work....

[lilshawn]

maybe at lunchtime i'll pick up a can of air duster and turn it upside down and give a few things a blast of liquid CFC and see if it smartens back up if something in particular gets cold again.

so, I did in fact grab a can of duster (a couple days later... boo other things )

and also been sidetraked with other projects... but

i started out gingerly spritzing a few areas and select components I thought it might be, to see if it changed at all. and it did not.

then i started blasting some of the entire areas i thought might be responsible... still doing it with nothing changing.

i decide eff it, just absolutely hose literally every square millimeter of the board with liquid refrigerant all over the place, all at once, till the whole board was frosty... like i literally was waiting for the board to short out. (it didn't)... but i'll be damned if it's still doing it like it DGAF. (screen did distort geometry a bit from timing components being sub zero and being way out of spec and some being superconductors now.)

blasted the yoke down... it has to be that right?

no. change. at. all. BRUHHHHH.  where the hell is this forsaken jitter coming from!?!? it's not my output signal... and i did rule out sync doing it with the scope before. i can hear the variance in the scan timing when it does it, so it's being produced in the board somewhere.

it doesn't do it when it's "cold"...it does it when it's "warm"... but also does it when it's COLD? so... not heat related? time related? charge related? something not affected by heat.

i dunno man.

[grantspain]

does it do the same without video signal?

[lilshawn]

i am currently testing without signal as I type this.

oddly, I couldn't get a "no signal" box to display on the screen. I initially thought that it didn't have one, or that maybe it shutdown and went to sleep because of no signal input... and right now the monitor is on the floor, and i couldn't see the power LED on the board from where i was positioned over it.

i pulled the plug thinking about what to do and indeed it was still on. so i plugged it back in and was just going to crank up the flyback till i could see a raster and go from there. i turned the flyback up a little and there in the darkness was the no signal box.

it turned it up actually quite a bit. i did turn it up enough so the black is still black and the no signal boxes are bright.

so it was adjusted "correctly" for 31k but just didn't have enough at this resolution mode... i mean i expect to have to move some stuff around when changing resolutions...but not normally the flyback screen control... usually that's all handled separately. never touched the sub-brd control...i dunno if that has anything to do with it...never had to mess with it. just a sub brightness adjustment i gather.

in any case, this monitor never even ran this at anything but 31k (it's a big buck hunter pro monitor.. i'm pretty sure it's 31k vga) so i don't know how it's supposed to act in 15/25k resolutions.

my computer signal i was feeding in is from my desktop, 640x480 60hz so it's for sure at 31k. I did hook up the scope to the horizontal video signal coming in and it was rock solid with no evidence of noise or anything. the horizontal pulses that the flyback shares from the control IC are solid too.... it's just something in horizontal deflection doing it.

I assume this mode it's in is 15k. it's been running about 15 minutes now as i've been typing this. it hasn't manifested any jitters yet. the weird line in the exact center scanline at the middle of the screen has. i don't know if it's related or just something with it's vertical driver IC (i have seen some of those do something like this before. it's not a real concern if it's not related. the jitters make the monitor unuseable.

https://photos.app.goo.gl/CmQXs7z28PBSEwv37

so now i'm wondering is there maybe a capacitor or diode or something that is activated in the 31k mode that is making the screen really bright in 31k mode and by proxy, causing the jitters? i mean, assuming the jitters don't appear here in 15k mode.

i'll let this run a while and check in on it in a bit and update what happens.

thanks grant  always helps to have another set of eyes looking at a problem. i get blind to obvious things often, especially when it's been this long i've been looking at something.

EDIT:

after about 45 minutes or so i'm calling it good in this "no signal" mode it's running in. there is no evidence of jittering. Just the weird line. I do have to run out of the shop for a few hours and i'd rather not leave this unattended... so for now, i'm shutting it down. the saga continues.

[grantspain]

when signal is applied does the on screen display work?

[lilshawn]

yes, the OSD does work. 100%

During my o-scope testing, i had gone in and adjusted a bunch of stuff to see what affected the waveforms on the scope trace i was looking at. (what things made the wave level bigger, longer in duration, changed the slope etc.)

i only have to bust out the scope every few years, so i try and learn as much as i can when I do.

i was curious if you know of any kind of adjustment in the service menu mode that might have gotten changed that might cause this, i know nothing of the meanings of most of the settings in there so i never bother messing about in it.

[grantspain]

there are some in extended factory menu i have no idea about also, pretty sure there is one that does something hv

[lilshawn]

hmmm. interesting. I'll have to take a peek at that setting and see if changing it affects "the jitters" (i won't change by much... there are scary things in there.) i wish wells gardner had instructions regarding the service menu... maybe i'll shoot them an e-mail and ask if they have anything.

I am currently on my days off away from the shop, (stupid covid screwed everything up for us... still to this day) but, i will give that a go when i go back in again.

thanks for the idea. i'll let you know how it goes.

[lilshawn]

looks like my factory mode only has color gain adjustments with the addition of bias control and turning off or on the "no signal" message.

i think the 9400 has the extra extra controls including the Hv setting you are thinking about.

the jitters does affect the OSD overlay. so it's not a signal generation issue. it's a horizontal drive/timing issue.

[grantspain]

you tried a known good flyback?
I have had some do similar to this, notably on the weiya/rodotron tri sync

[lilshawn]

the horizontal timing pulses (waveform 12 which is the h-control pulse from the KA2147 IC) feeds into a flyback drive and up into the yoke drive... since it tees off and go to the drive circuit for the horizontal yoke and the drive circuit for the flyback, they share a common pulse drive and that waveform is perfect.

in a nutshell the 2 systems are barely interconnected, and the waveforms in the flyback drive are perfect.

that said, i think i have a new flyback on a chassis i tried fixing a while back, but i think it might be a 9400 chassis. i'll have to take a dig for it.