Pre-war Television Fault

[Panrock]

I've just heard from a past customer. His HMV 904 has started to play up. He writes:

"I am afraid the HMV has developed what I suspect may be a capacitor issue.
I have attached some photos, which show the set after initial power-up and then after around 50 minutes.

During that 50 minutes there are the occasional (2 or 3) audio pops and simultaneous picture blinks but then it settles back...but then after about 45/50 (in total) minutes the video signal disappears as per the second picture. If the set is left for a few hours the sequence repeats."

Any ideas about what could be causing this? On the face of it, it looks like a frame fault, reverting to the bottom of the picture. But could it be something else? Your help here is appreciated.

Steve

PS. This set is fitted with a 5FP4 tube. The trapezoidal distortion on the picture is believed due to the camera shooting angle and the shaded bar is an artifact.

[FERNSEH]

The various HT supplies should be monitored with an oscilloscope.
Using a dual trace 'scope one channel should set to DC and the other to AC.

DFWB.

[Panrock]

Thank you David. It would indeed seem a good idea to first gather information about the various power rails, and how they are affected when the fault occurs.

This is a fault that only happens when something is breaking down after a period of time running and probably getting warm. The remnants of the picture shifting to the bottom points to DC getting into the frame scan coils. So C75, a DC-blocking coupling capacitor (previously re-stuffed and unfortunately rather inaccessible for dismounting and changing) feeding the scan coils, would be my first port of call. The popping noises could indicate a 'voltage' related breakdown is involved, and that is present in the anode circuit. C74 might warrant a check too?

Steve

[murphyv310]

Hi.
Looks as if the frame is deflecting downward and possibly reflecting back to the upper part of the screen. C75 would be my first port of call too.

[Panrock]

A warning, in case anyone is considering ever working on one of these 'table' sets... things are really cramped in and they are an absolute death trap... worse than the mirror-lid models.

There are tags present at mains EHT potential, completely unshrouded, and easily touched at the underneath of the chassis. Just take off the bottom cardboard cover for access! Not only this. At least one of them is uncomfortably close to another point at 'normal' potential - risking a spark-over. I inserted additional insulating material here.

Steve

[Jac]

That's good advice Steve.

I did the same with mine:



Jac

[Panrock]

Originally Posted by Panrock

So C75 .... unfortunately rather inaccessible for dismounting and changing)

See pointed-to items to see what I mean. The chassis has to come out.

Steve

[Panrock]

The previous photos were from my records of the original restoration.

The set was brought back to the workshop by the customer yesterday. Here it is. What a weight!

Now let's see what's what...

Steve

[Panrock]

The insides have now been extracted from the cabinet. They were a tight fit.

Coarghh what a weight! I'll say that again. It's finger pinching time. The first two pictures show the chassis (less the HT rectifier, for easier handling).

Unfortunately both the HT and the EHT transformer live at one corner of this unit. Together, they are so heavy it is completely impossible for me to raise the chassis at this end. But they make a great stabilising weight once the chassis is placed on its side, since the area of attention is underneath at the other side.

Our quest is to extract and replace C75: an 8uF electrolytic in an amber coloured cardboard box. It's somewhat inaccessible! The last two pictures show the actual appearance and equivalent appearance in the manual...

Steve

[stevehertz]

Given its inaccessible position, that's sure to be the problem.

[stevehertz]

That's not the original (proper) CRT in there is it? What is it?

[Sideband]

I don't know the set or anything about it's component layout but if the cap is so difficult to get to, would it be possible to bypass the one fitted by 'hanging' a temporary replacement in place of it. It'll prove the fault one way or the other and may save having to extract a difficult-to-get-at component if the temporary replacement doesn't cure the problem.

[stevehertz]

Originally Posted by Sideband

I don't know the set or anything about it's component layout but if the cap is so difficult to get to, would it be possible to bypass the one fitted by 'hanging' a temporary replacement in place of it. It'll prove the fault one way or the other and may save having to extract a difficult-to-get-at component if the temporary replacement doesn't cure the problem.

If the cap is faulty it's either O/C or S/C (or resistive ). In the case of the former, connecting a new cap across it will identify if it's faulty or not. In the case of the latter it most probably won't unless one end of the original cap is disconnected, in which case access to the cap (one side) will still be required.

[Panrock]

Originally Posted by stevehertz

That's not the original (proper) CRT in there is it? What is it?

As stated in post No.1, it's a 5FP4. This was obtained from Steve McVoy in the States and is considered the nearest practical substitute. It has a flatter faceplate than the original Emiscope 3/1, but no ion burn. The original 3/1 had a grid-cathode short after warm-up (and ion burn!), which made regular viewing impracticable. It, and the holder, were returned to the customer for safekeeping.

Originally Posted by Sideband

I don't know the set or anything about it's component layout but if the cap is so difficult to get to, would it be possible to bypass the one fitted by 'hanging' a temporary replacement in place of it. It'll prove the fault one way or the other and may save having to extract a difficult-to-get-at component if the temporary replacement doesn't cure the problem.

That would be one way to go, though I'm minded to extract the old box and re-stuff it anyway. I simply don't remember if I re-stuffed it originally or re-formed the original.

Originally Posted by stevehertz

If the cap is faulty it's either O/C or S/C (or resistive ). In the case of the former, connecting a new cap across it will identify if it's faulty or not. In the case of the latter it most probably won't unless one end of the original cap is disconnected, in which case access to the cap (one side) will still be required.

Correct.


Steve

[Sideband]

Originally Posted by stevehertz

If the cap is faulty it's either O/C or S/C (or resistive ). In the case of the former, connecting a new cap across it will identify if it's faulty or not. In the case of the latter it most probably won't unless one end of the original cap is disconnected, in which case access to the cap (one side) will still be required.

Well yes...I should have stated that the old one should have at least one end disconnected....I sort of thought that that would be understood. It's the sort of thing I would do in the case of an 'awkward' component.

[Panrock]

Originally Posted by Panrock

... though I'm minded to extract the old box and re-stuff it anyway. I simply don't remember if I re-stuffed it originally or re-formed the original.

It turned out had re-stuffed it, with a 10uF 450vw electrolytic.

Here we see the original electrolytic's case (less its mounting clip) after it has been extracted, following un-hitching of various things getting in the way.

I have now ordered its replacement from RS: 377-8914 - an 8uF motor-run type polypropylene capacitor rated at 470v AC. It should fit in the cardboard case quite snugly. Better suited - we hope - to feed those inductive scan coils.

Steve

[stevehertz]

Did it measure faulty? I assume so being as you're fitting a replacement.

[Panrock]

No. There wasn't much point in running a cold test of the electrolytic, since the original fault (as reported) only would appear after some time of running. Also I take the view that, having gone this far, it's better to make the most of the opportunity to replace the original component with something better anyway.

Whether this has actually removed the fault, we'll find out later. However, it's the first thing to replace.

Steve

[stevehertz]

Makes sense Steve, I was just wondering if it showed as being faulty.

[Panrock]

The new capacitor has gone in today, and everything on top of it re-assembled. I shall also be replacing the (fairly hard working) frame output anode load resistor tomorrow, as a precaution. This 85 year-old set is (would you believe it) under guarantee and I don't want it going wrong again.

Steve