Bush TV22 restoration

[Panrock]

A heat-related intermittent connection to one of the gun elements?

Have you tried tapping the tube neck when it occurs?

Steve

[Freya]

My money is on a new capacitor going faulty or soldered joint, most likely loose the bet though.

[Boom]

As I say, I am only guessing but the most obvious and really only way that dc could reach the coils might be if the scan coupling cap' (on top of the scan coils) was leaking. It would be an easy task to replace it and see what happens.

[Keith]

Still puzzled by this problem - although the set is useable after a wait. I put a scope probe on the frame deflection yoke centre tap ( the offset seems to mostly vertical) and measured approx +200V dc in the fault condition. When the fault clears, this drops to about 100V. I would expect to see HT voltage here due to the 10K (R68 Trader sheet) up to the HT rail (is this to prevent the yoke charging up?). Clearly this should not result in yoke current as the transformer secondary is isolated. So why does this voltage change when the fault goes? Is this cause or effect?

[Boom]

I can't see the purpose of R68. Try running the set with it disconnected and if it still displays the fault that would point to the transformer or the scan coils themselves breaking down. I take it the five pin plug is clean? A weird one indeed.

[Dickie]

Sounds as if those measurements are the wrong way round! If that centre tap is at +100VDC when the fault clears, what is the HT on the other side of R68? If the HT is only 100V that's another problem (?), if its 200V where's the current through R68 going?

[Keith]

Dickie,
I thought that just after I submitted the post. If the HT rail is staying up and R68 is still 10K, there's 100mA going somewhere and R68 is dissipating 10w (which it clearly isn't).

Boom,
I can't believe the transformer is breaking down, having just rewound it - but you never know.

One other possibility strikes me. The set as received has the magnet assembly bracket pushed all the way back. This achieves a focused, centred picture in what I've been calling the "non fault" condition. Could this actually be the fault condition? Maybe I should try moving the magnet assembly forward during the "cold" condition to see if correct focus/centering/brightness can be achieved. I'll have another look tomorrow with a clearer head.

[Keith]

I must have been half asleep when I made the last post! Apart from my dodgy ohms law calculation (should have been 10ma and 1W), in fact the 100V measurement was a myth. The frame coil centre tap stays up at HT as would be expected. As Boom commented the function of R68 is not clear. I guess that it was considered a good idea to reference the winding to something and connecting it loosely to HT puts it closer to the line winding mean voltage thus reducing voltage stress.

The original cold "fault" won't come back now so can't test my theory at present. I wonder if any other TV22 owners find that the magnet assembly needs to be right back against the screws?

[Argus25]

Originally Posted by Keith

I wonder if any other TV22 owners find that the magnet assembly needs to be right back against the screws?

Keith,

The position of the field from the magnets (assuming the mechanical setting that controls the gap is about centered in its range) will only be correct if the EHT is correct, assuming, that is, that the magnets have the correct field.

If the assembly containing the magnet/s has been disassembled, and reassembled, it weakens them because the pole pieces around them act as keepers. Though if that has happened, I know a method to re-magnetize them.

But if the magnet assembly has not been touched (therefore likely their fields are normal) any they need to be in the wrong place on the CRT neck to achieve focus, this must mean that the EHT or perhaps the CRT's A1 voltage is incorrect.

Hugo.

[Keith]

Thanks Hugo,

I don't have a means of measuring the EHT (although there used to be a probe around when I was at work - probably thrown in the skip by now!). I can get a picture of around the right size and reasonably bright so maybe the EHT is not far off.

I've no idea if the magnet assembly has ever been dismantled - quite possible I suppose. I notice that the CRT neck is not central in the aperture with the adjustment centered- presumably it should be.

Generally I'm getting quite a good picture now. Apart from this "cold fault" the only annoyance is some random contrast jitter. I can see the video waveform on the video amplifier grid jumping for up to a few milliseconds at random times - if it coincides with sync pulses it causes sync jitter also. I dont think I can see it upstream of the detector so concentrating on this area.

[Keith]

Well, things have moved on quite a bit since my last post. The "cold fault" is still with me but I think it is lasting for less time (<5 mins) after first switch on so I can live with it for the moment. The random contrast/sync glitches turned out to be caused by the noisy 12V (from PC) that I was using to power the modulators - fixed by changing to a linear supply (the modulators by the way are ex-Rediffusion modules).

I finally got round to adding Peter's logic circuit to add in the half line sync pulses during the Vsync pulse. Works a treat - thanks Peter (I have no issue with TTL monostables - quite adequate for this application). This completely removes the "hook" at the top of the picture making line hold a lot less critical.

The only remaining problem is slightly ragged verticals - about twice the width of a line on test card C. These only occur when the contrast (i.e. sensitivity or front end gain on this model) is turned up to a good level - if this is backed off to give only a just perceptible picture, the jitter goes away, so presumably this is some sort of noise problem. I've tried extra decoupling around the sync separator but without any real improvement. Almost all components in this area have been replaced. Trying to look at the line ramp start wrt sync edge is difficult, even using a delayed timebase and I can't see anything obvious (however, there could be differences on alternate fields, I suppose). I'd be grateful for any suggestions.