• Please review our updated Terms and Rules here

TRS-80 Model II CRT Blurry with Horizontal Lines

VeryVon

Miner 2049'er
Staff member
Joined
Oct 21, 2021
Messages
325
Location
Land of the Cheesesteak
I have a TRS-80 Model II here in pretty good condition. I couldn't get the 8" Shugart going (yet) but managed to load CPM from 5 1/4 floppy. I'm going to pull the video board out and give it a good once over, possibly recap.

Has anyone seen screen video artifacts like this before / know possible cause? Thanks

1670032422581.png
 
I think you can locate the problem by checking the Vertical Sync Signal on the Model 3.


Larry
 

Attachments

  • M3-Vsync.png
    M3-Vsync.png
    174.8 KB · Views: 3
Thanks ldkraemer, I've noticed the Model 3 and especially the Model I seem to have more service & troubleshooting manuals than the II? For example the Model I "Video Display Service Manual" is very detailed. I'll have to see how much if any of that can be transposed to this crt. I can see now why the Model II has been referred to as the "system collected by the masochists of the TRS-80 world" :LOL:

p.s. for example look at this from the TRS-80 Video Display Manual. Amazing!

1670073690536.png
 
Veryvon,
OOPS, I misread the II as III. I was wondering why you asked about the II version. SORRY.

DM sent.

Larry
 
I have seen a similar problem. (not a TRS-80)
It was a problem with loss of low frequency response in the video amplifier.

Looking at a line of video I see:
start of line unblanking the video is grey. When the characters start the line gets lighter with each character and saturates white.
when the characters stop I see a drop slowly back to the grey level.

sync and sweep look normal.

I would put a scope on the video amp and take look.

joe
 
It is a video signal frequency response issue in the VDU most likely. Caused by the degradation of electrolytic caps, in coupling stages, in the video amplifier chain. When you find these, I would suggest replacing them with film capacitors.

If I could see the schematic I could suggest what to replace.
 
Hunting through the VDU schematic sent by Idkraemer, the video output stage in this set is super simple (see attached).

Check with the scope on points A & B, that there are sharp edged rectangular signals, and check that the resistor at C has the correct value and check the emitter bypass capacitor at location D. AS a start, if we cannot see the issue here we will have to look at the circuitry feeding the contrast control, but if that is digital the fault wont be there.

It is also worth scoping the CRT's cathode (x10 probe), just on the off chance (it is fairly rare) that the CRT has developed heater to cathode leakage and is corrupting the frequency response that way, but it does not happen very often.
 

Attachments

  • VDUx.jpg
    VDUx.jpg
    223.3 KB · Views: 14
The internal mail system bounced for some reason, With regard to C301 in the circuit fragment you attached:

If C301, a signal coupling capacitor, if it was defective, open or high ESR, it would just leave the 220k in circuit to pass signal , that in conjunction with the CRT's capacitance would create a low pass filter, causing the smear and loss of high frequency detail in the picture.

The circuit is drawn like it is a polarized electrolytic capacitor. If it was a film type it would highly unlikely have failed.

So I'm not certain that it has failed, or if it is actually an electro or film type, but it is worth checking. I would probably replace it with a film cap of the same uF and voltage rating, or a higher voltage rating if that was what I could get.

I'm still not certain which is the exact schematic for your set. But the basic plan would be to trace the video, with the scope , from its output from the computer, all the way through to the CRT cathode. It should be pretty easy to find where the high frequency response gets degraded.
 
Last edited:
@jlang @Hugo Holden Update on this. I have the RCA video board in my Model II. I've replaced C301 & C302, no change. Put a scope on either side of the video amp points A & B, see below. The input doesn't look rectangular. Has me wondering if the problem lies on the video card?

index.php


Point A:
<inaccurate trace removed, see a few posts down for new ones>

Point B:
<inaccurate trace removed, see a few posts down for new ones>
 

Attachments

  • VideoInputSchematic.PNG
    VideoInputSchematic.PNG
    107.1 KB · Views: 57
Last edited:
Yes, the video on point A is very abnormal and the signal looks like it has been integrated and it is consistent with the mess on the screen.

The video signal needs to be traced back to its source on the video card.
 
I removed the original scope images above because they were incorrect. Pretty sure I was zooming in on noise there.

Here are the updated scope shots:

CCLK
1672152409791.png

VSYNC
1672152444414.png

HSYNC
1672152459134.png

Video Signal Before Amplification (Point "A")
1672152486187.png

Video Signal After Amplification (Point "B") - The bottom trace is the cursor flashing on & off
1672152531948.png

Close up of Point B #1
1672152604605.png

Close up of Point B #2
1672152621471.png

Flashing Cursor:
View attachment PXL_20221227_143800732.mp4

@Hugo Holden
 
Last edited:
Can you measure the frequency or period of those pulses seen on point B ? they might just represent coupled H rate pulses rather than graphics signals and the gain of the scope turned up too high for the voltage there, generally the modulation at the CRT cathode would be in the 30 to 40V peak to peak vicinity...so set the scope& probe combination to maybe 10V per cm and use AC coupling for that measurement.

If that other wave that was recorded was part of the graphics/cursor signal, it is abnormal with a inverted exponential profile, but the scope input might be over-driven.

Make sure to fill the whole CRT screen with graphic/characters of some sort , so there is plenty of video signal to see on the scope trace. And report the scope's vertical voltage scale per/cm and the timebase per/cm settings.
 
Last edited:
@Hugo Holden, This is about as many characters I can get on the screen at the moment:
1672232890283.png

The period of the signal (at Point B) looks to be 150 microseconds or 6.6 Hz frequency. Amplitude is reading about 4V.

Point A
Volts / Div: 2
Sec / Div: 50 microseconds
1672235854599.png

I switched to the x10 probe and set Volts/Div to 10, here's what Point B looks like with .5 msec per Div:
1672236976323.png
 
Last edited:
Yes the signal is a mess corresponding to what you see on the CRT screen.

Stay with the x10 probe. Set the scope to something like 5V/div and test the signal at the Base of the video output transistor again (point A on post #10). You could speed up the timebase to maybe 0.1mS/div to spread the pluses apart more.
 
The signal input to the video amplifier transistor base circuit looks very abnormal and even worse than what is on the collector circuit and it is not a proper digital drive voltage corresponding to the text. The VDU may well be ok and it is just receiving a corrupted signal.

Just double checking, where are you connecting the scope probe's earth clip to, is it a good ground ?

I suggest following this signal back into the computer & video card.
 
Just double checking, where are you connecting the scope probe's earth clip to, is it a good ground ?

I suggest following this signal back into the computer & video card.
The probe's earth clip is connected to a ground pin on the video card itself. I can double check that though.

Agree about tracing the signal back. I've already started that process. I'll report back when I have some more info. Thanks!
 
The probe's earth clip is connected to a ground pin on the video card itself. I can double check that though.

Agree about tracing the signal back. I've already started that process. I'll report back when I have some more info. Thanks!
It might be worth double checking that the ground system of the VDU and that of the video card are both properly connected together too. What is the signal link between them like, is it coax cable with the earth sheath connected at both ends, or some other arrangement ?

I notice from the photo that the VDU is out of the case/computer's housing, how did you link it back up to the computer, was it with the manufacturer's existing wiring or something you made ? I don't know the computer, but they might have acquired the earth to the VDU body from the case it was in. One quick way to tell, leaving the scope earth clip where you had it, is to connect the scope probe to the ground of the VDU and see if you see a signal there.

These sorts of earthing issues could explain the corrupted signal.
 
Last edited:
I can show you what it looks like on the bench (see below.) All connections between the computer and the VDU (an edge connector) and between the VDU & CRT are the manufactures existing wiring. I have to carefully disconnect & reconnect them pulling the VDU.

I double checked the grounds with a continuity tester, looks good between the keyboard/video TTL logic board, VDU and chassis (see red "GND" markings on the picture below.) And I'll see if I can check them for noise as well.

Grounds:
  • The top ground on the computer card is the negative end of a decoupling capacitor. I have a logic tester connected to it.
  • The bottom left ground on the VDU is a ground lug, probably there for testing as I am now
  • The bottom right ground is the chassis
1672437275633.jpeg

Top view of VDU / CRT Connections:
1672437499680.png
 
Back
Top