Another Westinghouse H840CK15

[trinescope]

I turned it on this morning and what I noticed is I get inconsistent results from the HV supply. Last night the convergence was pretty bad, but this morning it was a lot better, and the front panel control is still maxed out for best convergence. I measured 22KV at the regulator tube and could not get it lower than that, so I believe the HV resistors in the cage are flaky. Also I have a problem with the red video amplifier that gives the impression of a weak CRT (looks like smear), but testing shows it's the video amplifier having a rising low frequency response. I have not replaced the electrolytic video amp feedback capacitors yet and that is a prime suspect for this kind of problem. The resistors and capacitors are on order as of now. Those resistors are really expensive; I had to go with 85MEG in place of 82MEG since that was all i could find, 100MEG is readily available. Those parts together are about $37.

A question for those who have experience with 15GP22 sets: Is it normal for red to go out of focus at a relatively low brightness? That's what I have, but other than that it doesn't show any other more usual signs of a weak gun like smearing or saturation. With normal video it doesn't appear all that bad.

[trinescope]

I forgot, the red video amplifier problem is also intermittent and was not present last night.

[dtvmcdonald]

The problem with the red is that if you have even a teensiet weak CRT, it clips on BOTH high and low sides. The low side clips at the DC restorer and the high side clips
at the CRT grid. Since there are no more good 15GP22s (and if there were, I would outbid you!) the only cure is to reduce the drive to the CRT until the red gun stops
clipping at its grid. This results in a dim picture.

To test this you need an SMPTE color bar pattern that generates saturated
red, magenta, yellow, and white, correctly adjusted for saturation, and a two-input
scope that does subtract of athe two channels. You also have to put the scope probes directly at
the red cathode and grid (both for DC correctness and also seeing if it clips at short transients).

You will likely be disappointed at the picture brightness.

But once done ... the total CRT current will be so low that the HV supply will be happy
as a clam.

[trinescope]

I have a scope that can do the add/subtract of the two channels so I will try this once the adjustments and repairs are done and everything is working as well as possible. The red is such a deep color that it probably wouldn't be all that bright to begin with no matter what condition the CRT is in. As long as it gives a watchable picture it's OK with me, and I think it will after everything is done based on what I have seen so far. I know it won't run bright like a later set and even if I had a new CRT in it it still wouldn't run very bright. Most of the earlier monochrome sets weren't all that bright either, so it's OK with me. That's just the way these things were.

[trinescope]

I replaced the two HV resistors and the 4uF video amplifier feedback capacitors. Checking the old ones found the resistors out of tolerance, and I never could get a steady reading off the 100MEG one. One of the capacitors also was bad and it was the one in the red video amplifier. Now I have normal range on the static convergence control, and the video amplifiers are back to normal. I'll have to do a lot of tweaking to get the convergence better than it is now, but I think I can get it close. I am having trouble getting good purity, on these sets there is a lot of interaction between purity, convergence, and focus. The video looks pretty good and if I run it a bit dimmer than a modern set like it should be it actually look really nice so in the absence of an actual CRT test I will declare this one to be good. I think the convergence would be considered OK by an average person who doesn't work on these.

[Electronic M]

Looks good.

[trinescope]

While trying to improve the convergence last night the horizontal oscillator decided to lose lock and then went off frequency. Best I can tell is one of the mica capacitors is going bad since swapping tubes and checking resistors showed no problems there. I went ahead and ordered parts to replace all the mica caps in that circuit.

Also, the vertical centering control was having no effect, but the control had normal resistance. I checked the large capacitor on the top of the chassis and it was OK, actually measured 2200uF (marked 1000uF) with a bit of dissipation factor, but not too bad for an old part and still usable in a low impedance circuit like this. I finally found the problem, the wire going to the center tap of the centering control was never soldered from the factory and now had green corrosion. I cleaned and soldered the terminal, hopefully it will work when I reinstall the CRT.

[trinescope]

While trying to improve the convergence last night the horizontal oscillator decided to lose lock and then went off frequency. Best I can tell is one of the mica capacitors is going bad since swapping tubes and checking resistors showed no problems there. I went ahead and ordered parts to replace all the mica caps in that circuit.

Also, the vertical centering control was having no effect, but the control had normal resistance. I checked the large capacitor on the top of the chassis and it was OK, actually measured 2200uF (marked 1000uF) with a bit of dissipation factor, but not too bad for an old part and still usable in a low impedance circuit like this. I finally found the problem, the wire going to the center tap of the centering control was never soldered from the factory and now had green corrosion. I cleaned and soldered the terminal, hopefully it will work when I reinstall the CRT.

[trinescope]

I got the horizontal oscillator back to normal, also I think I have the color sync improved over what it was before. Both sections had flaky mica capacitors, the kind with the 6 color dots that normally don't give too much trouble. They tested OK on an LCR meter, but I'm guessing they didn't hold up with significant DC voltage across them. One was the 100pF capacitor across the variable cap in the horizontal oscillator, the others were the two 3900pF coupling caps from the burst transformer (hue control) to the 6AL5 phase detector.

Although the CRT seems to have good emission, I think it was a bit marginal in construction since I can't get good purity at the edges no matter what I do. The yoke ends up diagonal when purity is as good as I can get it, and that's how it was mounted originally before I took the chassis apart. I also seem unable to get the convergence right at the edges of the screen, probably for the same reason. I'm starting to believe that perfect is going to be the enemy of good here.

The color also looks better without the oversaturated red I had before. I did the matrix adjustments in the service data using a signal generator, so they are just a close approximation to using an actual color bar generator. Using color bars from a DVD the waveforms look reasonable, but for example checking the red video signal shows the yellow bar lower in amplitude than the red and magenta bars. I found similar results on the blue and green video signals as well. I think it's more likely the DVD video signal isn't providing an actual calibrated color bar pattern. It was a JPEG image I got from the web and is a pretty good approximation, but I have no way to know it's exact short of using a vectorscope and waveform monitor on the DVD player output which I don't have at this time.

Next thing is I'm going to run the chassis a bit each day or so to see how it holds up and start working on the cabinet which needs a lot of help. Also need to find a solution for the missing knobs.

[trinescope]

I'm still running the chassis out of the cabinet, waiting on me to start stripping and refinishing the cabinet since I got in a bit of practice on a Packard Bell B&W cabinet at first and did OK, and now know how to get a smooth finish.

I finally found the color sync problem, it was due to flaky ceramic capacitors in parallel with the 3.58MHz oscillator coil. I do have what looks like a bit of ringing in the picture, but it's really not noticeable most of the time. The signal looks great coming from the video detector but has the ringing at the delay line, and it seems to be the 6BK5 line driver stage that has the potential problem, or at least it was designed with this kind of response.

I'm also going to try swapping the leads on the vertical convergence transformer primary and see if that helps get the convergence a bit better. The controls only make convergence worse, and perhaps inverting the polarity of the convergence signal will allow correct adjustment. This is the part I had to rebuild myself and it's possible I got the two lead colors reversed when I assembled the transformer.

That means removing the CRT from the chassis one last time to do this and check out the video and color amplifiers as well. Then I need to clean the tuner so it will work without having to rotate the shaft back and forth for a good picture. At that point everything should be up to par.

[old_tv_nut]

Originally Posted by trinescope

I finally found the color sync problem, it was due to flaky ceramic capacitors in parallel with the 3.58MHz oscillator coil.

Hooray!

Originally Posted by trinescope

I do have what looks like a bit of ringing in the picture, but it's really not noticeable most of the time. The signal looks great coming from the video detector but has the ringing at the delay line, and it seems to be the 6BK5 line driver stage that has the potential problem, or at least it was designed with this kind of response.

If the ringing appears at the input of the delay line (plate of the 6BK5) and there are no off-value components there, then it may be designed in, a result of the 3.58 MHz trap in the cathode. If you have a DVD with an Indianhead test pattern, you can look at the resolution wedge to see if the ringing frequency is the same as the chroma trap frequency.

[trinescope]

Here are the pictures I was not able to post last night for some reason. The scope screen shots show a multiburst signal at the video detector and the input to the delay line.

[trinescope]

Originally Posted by old_tv_nut

Hooray!



If the ringing appears at the input of the delay line (plate of the 6BK5) and there are no off-value components there, then it may be designed in, a result of the 3.58 MHz trap in the cathode. If you have a DVD with an Indianhead test pattern, you can look at the resolution wedge to see if the ringing frequency is the same as the chroma trap frequency.

I really believe the response is designed in, the ringing goes away if I short out the 3.58MHz trap. Like I said, it's not all that pronounced under normal conditions and video material.

[etype2]

Nice color bars.

[kvflyer]

Originally Posted by etype2

Nice color bars.

My thoughts, exactly!