BLIT display debugging advice

[Josh Dersch]

On 10/26/2014 8:09 PM, Josh Dersch wrote:
> On 10/25/2014 10:17 PM, Don North wrote:
>> On 10/25/2014 8:14 PM, Josh Dersch wrote:
>>> I picked up a Teletype DMD 5620 (aka BLIT) terminal a couple of 
>>> weeks back and I finally got the keyboard working this week; in the 
>>> meantime a problem with the display has cropped up.
>>>
>>> See here for a video: 
>>> http://yahozna.dyndns.org/scratch/blit/WP_20141023_001.mp4
>>>
>>> (apologies for the quality and that it's rotated 90 degrees... you 
>>> get the general idea).
>>>
>>> For those without video-playback capability, essentially after 
>>> warming up for awhile (generally after a couple of minutes) the 
>>> screen starts shaking horizontally; like the horizontal position is 
>>> oscillating rapidly.  This causes "fringes" on the horizontal edges 
>>> of the picture, as this oscillation occurs multiple times during the 
>>> course of a single rescan. Generally it gets worse and worse the 
>>> longer it's been on, though sometimes it will stabilize for a little 
>>> while.
>>>
>>> I thought it might be a bad solder joint and I started prodding 
>>> around with a dowel when the set was "cold" but I was unable to make 
>>> the screen jump at all while doing so.  I also cleaned the 
>>> horizontal position and horizontal oscillator pots with some contact 
>>> cleaner to no real effect.
>>>
>>> I am not particularly good with CRTs, working on them is not my 
>>> favorite thing, something about the high voltages I guess :). Anyone 
>>> have any suggestions?
>>>
>>> Thanks,
>>> Josh
>>>
>>
>> It appears the the screen is jumping in some integral number of 
>> pixels horizontally, like 32 or 64. I don't see any vertical scan 
>> line displacement.
>>
>> I would check that the digital horizontal sync (and blanking if 
>> available) are rock solid, and that the digital pixel stream is gives 
>> a stable eye, and
>> it solidly locked to the horizontal sync signal.
>>
>> Based on the short little movie I would suspect more of a digital 
>> issue than a monitor analog issue, at least for starters.
>>
>> Don
>>
>>
>
> Thanks (to both you and JWS) for the tips.  I'll see if I can get a 
> scope hooked up to the sync signals sometime this week, though the way 
> the behavior progresses (it gets steadily worse and the amplitude of 
> the oscillations gets greater) makes me think it's probably not a 
> problem with the signal coming from the display logic -- a digital 
> failure I'd expect to stay approximately the same, something that 
> increases in magnitude over time seems more like an analog thing, but 
> maybe I'm jumping to conclusions :).
>
> At any rate, it doesn't appear to be a thermal issue -- I powered it 
> on this afternoon and it's now immediately showing the issue, much 
> worse than before; to the extent that I can hear a not-very-pleasant 
> whining from the flyback that makes me nervous to run it for any 
> significant length of time as I'm afraid something (like the flyback) 
> might get damaged.
>
> I may attempt a recap at some point (when I find time -- I don't know 
> why I keep getting into new projects when I have no time for them...) 
> since it's likely due anyway.  There are a lot of capacitors to 
> enumerate and stuffed into a relatively tight space. Did I mention 
> that I hate working on CRTs? :)
>
> Thanks again,
> - Josh
>
>

Just a a quick follow-up here:  I recapped the BLIT this weekend. 
Recapping the monitor's power supply had no effect but after doing the 
main PCB it seems to be humming along nicely.  I also redid the logic 
power supply while I was at it.

Now to fix the mouse...

- Josh