@helloekco
To expand on Davids post. If you haven’t got them, Patchett TV in the Data library, 4 volumes is an excellent source of information. Book three has AGC and Flywheel sync in detail.
Frank
I've got to the bottom of the problem with the discriminator and the line frequency.
I checked the waveforms recommended by @till, they looked OK and in sync with one another.
I compared the waveforms again at both of the diodes of the discriminators, and for some reason there was in fact a difference in the magnitude that I'd missed before (I've discovered that one of my scope probes seems to be way off with the other, perhaps I didn't have the voltages calibrated properly on each channel of my scope before). Having found that difference, I changed the resistors R103 and R104, as there was a difference in circuit resistance across both sides, and although the resistors once removed both measured high and non-matching, that didn't fix the issue, so I removed the rectifier MR2 / MR3.
Once I'd pulled the rectifier out, testing it showed that it seemed to be completely open circuit; I've no idea how it would have worked at all in that state, so I can only assume that the heat from desoldering it damaged it. Not being able to find any specification of the type of the rectifier, and searching the web for its markings ("D3-2-IYZ") turning up nothing, I decided to try whatever I had, and fitted a pair of 1N914 signal diodes. This did the trick - the line now runs at the correct frequency and the voltage at the grid of V17B is close to zero. However, the previous problem remains and has in fact grown worse - the horizontal hold control is very, very critical and there is horizontal foldover - in fact the picture is offset to the left and stretched out on the right.
The picture below shows two scope traces - the top one is from the output of the discriminator at the junction of the two diodes and the bottom one is the anode of the line oscillator V17A. It looks to me like the line oscillator is distorting, as though there's a power sag as the voltage increases; however there is also what looks like a little oscillation just before the sync pulse occurs on the top trace.
I did try swapping V17 with the AGC valve but it made no difference (I do have a couple of 6/30L2s somewhere but seen as there was no change I didn't bother to dig them out).
One more thing I should have done is to check the trace from the line oscillator with V17B shorted out, to remove the effect of the flywheel.
I finally got back onto this TV after a long break. I have actually been working on it in the background but I didn't really have anything to report until now, plus I need to ask for some opinions.
After making progress as described above, the picture on the TV gradually deteriorated. I've had various problems to deal with that confusingly seem to have developed during my work on it!
The first of these that I can recall was C95 going leaky and its neighbour R82 going very high - the two components responsible for integrating the flyback pulse into the output of the discriminator. The result was a general loss of sync.
Having replaced those components, the hold still wasn't great, and gradually during investigation, the picture lost width and brightness. I was puzzling over this one for a good while, since I had already checked / replaced V17, the line oscillator, but sure enough it was that that had gone low emission.
So having replaced V17, I am now back to a full width, bright picture, but the horizontal hold is quite critical, and worst of all there is severe jitter and tearing at the top of the frame - but only when the set has been running for a few minutes. I've accidentally discovered here, that a trailing test lead attached to the junction of R99 and R101 and draped over the front of the CRT fixes the problem! Obviously this isn't a workable solution, so my question to everyone was going to be, what do you think the electrical effect is that this is having on the timebase and how can I properly recreate this.
However, I've always had this nagging fear that the fault described in the service bulletin for the set (kindly uploaded and mentioned by @freya), namely that of shorted turns in the sync transformer, might be present. I'm now reasonably convinced that is the problem that I have. I've removed the transformer from the set (NB - I am assuming that "sync transformer" in the aforementioned bulletin refers to T6, the flywheel discriminator's transformer), and using @crustytv's excellent guide to ringing a transformer, I've obtained the following:
Apologies for the lighting - hopefully it shows that there appears to be virtually no ring whatsoever. Now, I have read that laminated core transformers, which I'm assuming this is, can give very poor results even when good. However, this does look to me like it's pretty duff. What do others think - has anyone tested a similar transformer to compare? The very last thing I want to do is pull the transformer to bits to rewind it if it is in fact OK.
Just to check, with regards to the ring test - am I correct in thinking that the square wave 2v probe calibration output of my scope is fine, even though the frequency is well below the transformer's resonant frequency of (presumably) 10kHz - because it's a square wave? Also, that no parallel capacitor is necessary?
I did think of substituting an electrically similar transformer for testing. I was thinking along the lines of a 230v to 115v-0v-115v mains isolating transformer, but these are only available in large and expensive sizes (the only one I have is used to power the set whilst I work on it).
Thanks.
I seem to have lost the ability to edit my posts (or I'm being blind and cannot see how to do it), but I meant to say in the above "I was thinking along the lines of a 230v to 230v-0v-230v mains isolating transformer". I've also realised that this could be any typical valve amp mains transformer, but I don't know if I have one lying around.
EDIT (!) - I can edit this one! Something odd is going on, as though I have two profiles or something.
EDIT (!) - I can edit this one! Something odd is going on, as though I have two profiles or something.
The ability to edit a post lasts for fifteen minutes after it is first posted then times out.
@cathovisor Thanks, I didn't know that.
@helloekco I've actually made a mistake myself: the edit time for a post is ten minutes, not fifteen and is covered in the forum FAQ here.
OK, here is the ring test from the sync transformer with its iron core removed:
Now I really don't know what to think... That looks to me as though it might be OK.
What do others think?
I'm going to also try the neon method of testing the coils.
...I can't get anything out of the transformer with a neon test - but I wonder whether this little transformer doesn't have enough inductance to do so. I'm not totally convinced it has a problem, but due to my testing the primary wires have broken alarmingly short 🙁
I might have a transformer, but would need to see a picture of the original.
Sorry, not been able to locate them yet, will keep looking.
I ended up rewinding the sync transformer successfully, and the picture is now where the TV was before I undid the modifications that were bypassing the flywheel part of the sync circuit. The horizontal hold works but is still a little critical (perhaps normal for the earlier revision of the circuit, before Ekco made changes, that my model has?). The main problem remaining though is the smearing of the image, particularly on dark areas of the test card, and of what is either foldover at the left, or again some kind of smearing:
These scope traces show the waveform at the grid to the video amplifier (lower) and the inverted waveform at the grid to the sync separator (upper), with a vertical band test card, scope y deflection adjusted for easy comparison:
It looks to me as though there is a slow response somewhere around the video amplifier valve, causing the steps of the waveform to become rounded.
Picture smearing can be caused by impaired low frequency response in the video amplifier. The same condition can degrade the frame sync waveform.
Till Eulenspiegel.
Thanks @till, I'll update here when I get to the bottom of it.
I forgot to post these shots of the 'scope with the rewound transformer - to hopefully be of help to anyone wondering what a good one of these should look like in a ring test. The first one shows the rewound coil before the core was reassembled:
This one was taken after reassembling the core:
Compare these with the images I posted of the previous ring tests.
