Predicta Holiday - First power up attempt

[Electronic M]

X means line cut in some models. Dashed means line added in some.

How much vert overscan do you have? More overscan will accentuate the lines. I always tune for sharpest scanlines. You should not notice them from normal viewing distance (Which should be 10'-20' for a set that size). It is not supposed to look good when your face is near the screen...Ever watch a 90's projection set or a 40" color CRT from 5' or closer?...That is the natural look of NTSC when too close to the screen.

[madlabs]

Not sure how much vert overscan I have. How do I check?

I am looking at it from far enough away. I can't believe it but I can't find a 1M pot laying around loose. I may see if I can salvage one from some junk. But with 430V on the focus jumper something is wrong. Most likely I wired something wrong under the chassis.

But, with the jumper connected to ground it is too fuzzy, with it floating it is pretty good and when connected to the too hot side it seems like the scan lines are a little too much. I can't believe it but I can't seem to find a 1M pot. I'll get to tracking down the reason for the too high focus voltage.

[Electronic M]

To measure vertical overscan take a grid test pattern and view it on a LCD and count the bars/squares separating top from bottom, then feed the same pattern into the set and count what you see. If say there is one inch/division on screen and you are missing two evenly then you have one inch overscan on each side (top and bottom that I mean).

You should try to avoid more than 1" of image overscan as it will make the scanlines more spread out and easy to see.

Optimal configuration is to get the scan lines as thin as possible with the focus and get them close together/overlapping with the vertical size/linearity controls (by minimizing overscan). One thing that can accentuate scan lines even when overscan is at minimum is failure of the set to properly interlace on scan I think the term is "double scan" or "line pairing" or something similar (some member had a roundy with that issue years back)....

If you have a lower voltage source than the one you think is too high you can connect a 100k-1M pot at the ends to the two sources, and the center tap to the focus lead.

[madlabs]

Didn't have enough time to track down the too high voltage on the focus jumper. In any case I am going to try a pot on the focus and see if I can dial it in just right. I got the CRT protector buffed out and it looks good. I hope to put the set back together this week.

But, before I do I want to go through the adjustments again now that I have a new CRT in the set. I am still unclear on the best order and procedures to do the adjustments. For example, until Tom told me about it above, I didn't know about the vertical over scan. Is there a similar thing for horizontal over scan, adjusted by the width adjuster? Is there a proper order to run through the adjustments?

And the wire wrap connections that are still on the controls, should those be soldered? They look a little suspect even though they seem to work.

Thanks!

[Electronic M]

Yes there is horizontal over scan which is adjusted with width...If you plan to adjust both do the horizontal first...Many chassis use B+ boost derived from the horizontal to drive the vertical so changes to H can effect V.

[madlabs]

Thanks. What do you think about the old wire wrap connections? Seems like a potential source of problems and I am tempted to solder them all.

[Electronic M]

If undisturbed I usually don't worry. If I unwrap one in the course of work I usually re-wrap a few turns then solder.

[madlabs]

So, I got a 1 meg pot and put it on the focus and it works well, nice to have a little more control. I'm working on tracking down the last few bugs. I started looking for a loose connection, if you stamped your foot or tapped the set you got picture issues. Finally tracked it down to a bad solder joint in the tuner connector plug. Must have been loose from the git go. Nice to have that one fixed.

I finished polishing the CRT protector and put it back on. It came out very nice and clear, I used 1500 grit wet sandpaper and Novus 3&2 and spent about three hours on it. I had never actually seen a Predicta playing with the screen cover on. I have to say that while the set looks much better with the shroud and CRT protector on, I wish it wasn't so dark and greenish in tint. Still, the picture looks good and the set looks proper. Putting the shroud on is a mild PITA and I nearly dropped the screen protector!

Over all, I am very happy with the set. There is just one niggling problem and then I can call this set done. The problem centers around vertical hold. Not only does it lose lock sometimes but it jiggles or bounces. I spent some time fooling around with all of the adjustments and got a test pattern looking fairly reasonable but it seemed as if the lock and jiggle were worse. De-tuned the adjustments and it seems a little more stable. Maybe.

Just to recap (rim shot) I have re-capped the set, replaced the couplates with ones made by Bob, swapped out tubes, wiggled tubes and wires. I did replace the mica cap with a four cap combo to get the right value. The CRT is a NOS 21FAP4A and it looks great when it's locked. Raising and lowering input voltage to the set a little makes no difference. I'm running the set at full brightness but the problem is unchanged if I lower it. Contrast seems to have less effect than it used to. If cranked all the way right I get a pulsing flash on the screen accompanied by an audio pulse.

How do I go about troubleshooting this? Help me get this last thing out of the way and I'll get out of y'alls hair!

[reichsrundfu]

Re the vertical sync, try swapping out the sync separator tube. I had experienced some issues on my Barberpole (same chassis) and that cleared it right up!

George

[madlabs]

George, I tired that. No joy. It is driving me crazy, this thing would be done!

[madlabs]

Went to bed last night feeling defeated. Woke up determined to fight. Since my Tek 465M is failing, I decided to take a chance and hook up my cheap digital scope to the set. I started with taking waveforms 6 through 9, which are appended below, except for W6. W6 calls for grounding pin 7 to ground (chassis, correct?). When I did that the CRT displays a bright horizontal line so it was killing the vertical and the scope was flat lined.

As you can see, W7 and W8 look pretty good. Although if I am reading the schematic they should be at 30hZ and they are at about 60hZ. W9 looks terrible. Could K1 be responsible? That is the one couplate I did not replace.

V9 tube voltages seem close enough:

2 = -19
3 = -19
6 = 115
7 = -7
8 = 11.8

All voltages taken with a modern 10m input meter. The wave forms were taken with the set giving the best picture I can get. The schematic does say they should be taken with 50V PtoP at the CRT. I did not verify that but am going to sneak into the CRT area. Of course I had put the dang shroud back on so it isn't easy. I take it that should be between the chassis and pin 7 of the CRT?

Thanks!

[madlabs]

Made a CD with a test pattern yesterday as I had discovered that the DVD player I am using gives a much better picture than the HDMI to composite converter that I have. I'll look into getting a better one. Got the picture tweaked in much better. I can't do the over scan adjustments as Tom suggested exactly as I don't have another set with a composite input. But it is better.

But the jiggle is KILLING me. I've licked all the other problems but can't seem to figure this one out. I have read through and tried the things in the two tv repair books I have.

[madlabs]

Here is a link to a video of the problem in action. It varies, there is almost always a little wiggle and then it gets bad and then the picture rolls. Interesting, there are contrast effects that you see through the camera but not with the eye. I apologize in advance for the poor video quality and the ham radio chatter in the background.

https://youtu.be/uTbcTB9KQcY

[WISCOJIM]

Originally Posted by madlabs

https://youtu.be/uTbcTB9KQcY

"This video is unavailable."

[madlabs]

Hmmm. The link works for me. This is the first time I've ever uploaded a video to youtube so I may be doing something wrong.