How to identify a failing Cathode Ray Tube?

[Tom Vaughan-Mountford]

Hi all - I've safely discharged the tube and removed all of the boards relating to power/scanning. I've opened up the power supply, two very big caps - 220uf, 250v - the soldering looks original and untouched. No leaks, but the plastic covers on the top are a bit buckled. There's also a diode nearby that looks like it's been running very hot (board is brown right through to the reverse side) The caps are still being manufactured, and the specification sheet rates them to 15,000 hours - only about 2 years of continuous operation.

Elsewhere, this thing looks like a patchwork quilt of rather dubious historic repairs!

 

[Tom Vaughan-Mountford]

I've also noticed something peculiar about the cap JimB pointed out... only one leg of it is actually soldered into a track, the other leg just sticks through the board unconnected. Doesn't that make it somewhat redundant? Also spotted there's a distinct lack of solder on a joint nearby.

 

[JimB]

I've also noticed something peculiar about the cap JimB pointed out...

Have a look on the top side of the board, I am sure that you will see the second track on the topside.

JimB

 

[Tom Vaughan-Mountford]

Have a look on the top side of the board, I am sure that you will see the second track on the topside.

JimB

Cheers JimB - Quite correct, I can see it hiding now. That looks a pain to solder - but I'll know to look for faults on both sides now. I know I'm not best placed to pass judgement... but the construction of this monitor seems rather 'homebrew'

 

[KeepItSimpleStupid]

On multi-layer boards, the holes are what they call plated-thru. There is a cylinder of copper that lines the holes. Pads without components that connect layers are called Vias, Multi-layer boards can have burried or blind Vias because you can't see them. Removal of a component can damage those Vias.

Not removing the flux, causes corrosion over time. Acetone (Nail polish remover) and a tiny wire brush can help remove dried flux.

Sometimes parts with large massive leads don't get soldered properly.

 

[JimB]

That looks a pain to solder

Not a bad as it looks.
What you have there is a "Plated Through Hole", there should also be a circular pad on the underside of the board, but I guess that that has been burned off during a previous heavy soldering session.

Just do all the soldering on the underside of the board and you will be OK.

JimB

 

[cowboybob]

Doing great, TVM.

Imagine if you were dealing with a rig that was pre-PCB "point-to-point" wiring ... (with all the attendant spider webs and bug carcasses).

 

[KeepItSimpleStupid]

Actually, the scariest part is removing the High Voltage Lead and I never liked doing it. I did that when I was 12 years old or so. Probing 400-600 VDC with a non-auto-raging meter - Yuk.

Later in life I fixed stuff with 100 kVDC @ 0.1 A and 15 kVDC @ 1.5 A series regulated with a tube. I think the most fun was an Early Scanning Electron Microscope. Your picture tube electron beam is cake compared to a 30 kW electron beam that's scanned.

It might be wise to check the continuity on the top and bottom or solder the top when there is no bottom pad to solder too.

 

[spec]

Hi TVM,

You are doing a great job, and nice pictures too.

Just a couple of points:

(1) It is fairly normal for the plastic of molded components to crack due to heat/age. While this is not good, it does not necessarily mean that the component is faulty.

(2) The brown resistors shown in your pictures are notorious for going to a high value, so it would be best to check them with an Ohmmeter.

spec

 

[Tom Vaughan-Mountford]

Thanks all, I'm learning so much and I've not really started fault-finding yet! Plenty of useful insights gained just from a visual inspection. I'll let you know how I get on now that I'm progressing to component level diagnosis

 

[spec]

Thanks all, I'm learning so much and I've not really started fault-finding yet! Plenty of useful insights gained just from a visual inspection. I'll let you know how I get on now that I'm progressing to component level diagnosis

Hi TVM,

In many cases, it is the basic things: dry joint, cracked printed circuit track, broken wire, corroded contact, etc, even a connector not fully home.

I have seen many cases where hours have been wasted investigating system malfunctions and all it turned out to be was the supply voltage was down- nobody had bothered to check.

From what you say, it appears that your monitor has had a fault before and has been repaired. The question is: have they just cured the symptoms or actually cured the fault.

spec

 

[JimB]

I feel that I must give a word of caution here.

On the forum here we see many people with no knowledge or experience of electronic things, who will make a statement like:
"I have a Super Wizzo ZYZ123 which makes a funny noise. What component can I change to fix it?"
In the full expectation that from the vague description of the faults, the electronic gods here will know instantly what is wrong and can identify the single component which needs to be changed.
Unfortunately real world electronics is often (mostly) not like that.

Some will jump in and say "it is the electrolytics, change them", I believe that I half way went there myself:

As others have suggested, suspect the power supplies, especially the electrolytic capacitors.

The thing is, when faultfinding, the first thing is to physically examine the thing, look for bad solder joints. We learn with experience that components which get hot, or are large are the ones where bad solder joints are to be found.
(Components which are large and hot, well, nine times out of ten, there is your bad joint!)

Having done that, then consider the electrolytics, the bigger they are the more likely they are to be a problem. The will dry out, spill out their insides and become nothing more than a high value resistor.
These faults are obvious.
But caution!
Earlier this year a friend of mine called around and gave me a 40 year old spectrum analyser, sadly it was not working.
Following my faultfinding logic as given above it was obvious that many of the big electrolytic capacitors in the power supply were in a bad way, their end seals were split and had leaked electrolyte many years ago.
However measuring the supply voltages and checking the ripple voltages with an oscilloscope, showed that the power supplies were well within specification. So there was no point in splashing out a fistful of fivers to replace these caps when it was possible that the whole thing was scrap due to a whole load of other faulty components, many of which could be made from unobtainium.

As it turned out, I was able to get the analyser going quite nicely after much head scratching, and I did eventually replace all the electrolytics in the power supply, but did that improve anything?
No!

I guess the moral of this rambling tale is:
When fixing something, do the following, in this order:
1 Physical inspection
2 Check the power supplies
3 Hard core electronics faultfinding

If you skimp on steps 1 and 2, then step 3 is a waste of time.
And often step 3 is the only way to fix it.

JimB

 

[cowboybob]

Sound advice, Jim.

Sometimes tough to follow though (when in "help" mode), especially when you've no earthly idea about the experience of the OP, i.e., you don't want to insult them but, then again, you don't want them to kill themselves ...

 

[Tom Vaughan-Mountford]

Thanks, sound advice. As JimB says, there's no point replacing handfuls of components all at once - and personally I don't think that would be a very good learning experience for me, I'd have a solution without finding and understanding the problem.

Thankfully I've got the luxury of time to work through this methodically and I have no real need for the monitor, it's long obsolete - so I'm not playing with high stakes.

 

[cowboybob]

I like your attitude, TVM .

 

[spec]

Just another word of caution to compliment Jim's post:

Faults can be weird and often do not follow a logical pattern.

Sometimes there is no one fault, but an accumulation of problems- capacitors may be leaky, low value, high effective series resistance (ESR), contacts can corrode, extra high tension (EHT) diodes can go soft and so on. In cases like this it is necessary to replace all the suspect components and in many cases it is best and quicker to do just that.

On many occasions I have followed the logical fault finding approach without success, only to have a service engineer fix the fault in five minutes because he knows 'the' fault on a particular piece of gear.

In one instance our washing machine was playing up so I tested all of the components individually- they were perfect. In the end the domestic management commanded that a 'proper' engineer should be called in. He arrived, pulled the top of the washing machine and replaced the mains suppressor. He then buttoned up the washing machine and it worked perfectly for the next 10 years. I had tested the mains suppressor to death at work and it had no problem.

In the early days, when we were first married and had no money, our TV was a 1953, 14-inch, black and white, gold-fish-bowl, Bush. It took an hour to warm up and then we had 90 minutes of viewing before the set got too warm and the picture once again turned into a load of mush. I spent hours and hours tracing faults, and while there were slight improvements, the set still did not meet viewing standards.

In the end, I did a complete clean and visual inspection and checked all the non specialist components (not line output transformer or tube). All of the passive components, resistors and capacitors were way off, but strangely, the usual suspects, the valves (tubes) were all OK. Once the faulty components were replaced, the set was like new, and we could just turn it on and view any program, at any time, for an unlimited period- my missus was pleased- I wish I had that Bush TV now.

spec

 

[Les Jones]

Hi spec,
So you still have analog TV transmitters in your part of the country. (and still on the old 405 line standard !)
Les.

 

[spec]

Hi spec,
So you still have analog TV transmitters in your part of the country. (and still on the old 405 line standard !)
Les.

Yeah, we moved over from 25 lines years ago.

spec

 

[throbscottle]

You can get transistors too that seem perfectly ok when you check them with the multi-meter, voltages look ok, but actually have no gain.
I'd re-work all the joints regardless, not just the ones mentioned. You need a good hot iron for the larger ones. Hot and quick. Make sure both sides of the board get melted.
To start I'd be looking for a fault in the supply to the high voltage circuits rather than the hv circuits themselves.

 

[Tom Vaughan-Mountford]

Right, I've ordered replacements for the caps on the PSU board - and have desoldered a couple to test (with my new cap meter that arrived in the post today!) I viewed a couple of tutorials on YouTube about using a cap meter - there was a lot of angry debate in the comments about the merits of even bothering with anything other than an ESR. However, I've tested two identical 100uf caps from the board. One is giving a value of 93.5, and the other 00.8 which feels a significant difference. For comparison I also checked a 47uf which came back with a reading of 43.8... so I think I'm using the meter correctly. I've read that the value should be within 10% of the specified uf - so I figure the 00.8 cap is bad.