06-08-2012, 11:22 PM
Update: Aug. 6th.
So let's re-cap: I had no O/P on the lowest freq. range but one: that's range B plus a few concerns about the calibrator.
The "no O/P on range B" wasn't too difficult to track down, but it was necessary to discover how the osc. module for this (and any other) range features in the overall scheme of things: osc. generation to final R.F. O/P. And, I will add, that discovering that arrangement, by reference to the manual, was not easy. The cause was a failed coax cable termination at the osc. board: same type of problem as is commonly found in the filter boards, A7 and A8: perished insulation. However in this case (a coax termination) the Marconi chosen method was to connect a flying insulated wire to the braid of the coax and insert a cylindrical copper (?) sleeve over that junction and crimp that sleeve - then add a rubber insulating sleeving over all that. The inner of the coax had perished and the braid where the sleeve was fitted was very badly tarnished: clearly a form of chemical reaction had taken place. Now armed with that discovery plus the similar problem with the A7 & A8 filters, I set about examining various other coax terminations (of which there are a very great deal in this unit). And yes, many of them were 'on their last legs'. So some preventative maintenance was necessary.
Here are some pictures of the coax termination problem:
[attachment=5508]
[attachment=5509]
As for the calibrator, well, I'm still thinking about it. I have enclosed some pictures of 'scope waveforms:
1 MHz xtal osc. at TP1: [attachment=5505]
O/P of limiter at TP2: [attachment=5506]
O/P of 100 kHz divider at TP3: [attachment=5507]
And here are some pictures of the calibrator board itself and a couple of general views showing that board and the A7 and A8 filter boards - with new inductors fitted:
Calibrator board:
[attachment=5510]
Two general views:
[attachment=5511]
[attachment=5512]
You can see from the photo of the calibrator board that all the resistors are mounted vertically. What isn't shown is that of those ends that are 'scopeable', many of those ends are at 0v. / chassis. That does not help in fault diagnosis. With the exception of transistors that I had changed (or removed for testing), the transistors have their cases very close to the pcb surface: thus getting a probe in is not possible. Thirdly, the aluminium frame around the board (one-inch high) is also a restriction on access. And finally there is the simple issue of density of packing. All those factors make fault diagnosis with conventional bits of test kit very difficult. Moreover, not that this is a criticism of Marconi's design, etc., but the documentation I had to work with - especially the cct. diags. - is of a very poor print quality.
► Warning:
general moan and rant follows.
'Quality':well I have my reservations about 'quality' with this item.
For a start, the technical manual is poor: there is no other word that is adequate. Inconsistent forms of reference, contradictory and confusing layout of drawings and the general master block diagram is simply inadequate for its intended task.
As far as the overall electro-mechanical design is concerned, on the basis of the very evident 'module assembly' style that Marconi chose for this item, it is perfectly clear that the maintenance philosophy was one of 'replace the faulty module'. O.K., that's fine as far it goes - but it comes with a rather large and expensive additional price tag: the large amount of extra wiring and cabling to connect all the modules together, especially when all the necessary 'switchery' is included. Not that I have anything against the module build approach as such - provided it isn't carried to unnecessary extremes - but in my opinion, in the TF2002B, Marconi took that approach far too far. Moreover, the physical layout of those modules does not seem to bear a logical relationship to the path of the associated signals - or which there are many: the module layout seems quite random. In addition to that, in terms of cct. design, in some places things are clearly taken to an unnecessary extreme; in others, the design is a bit patchy - or worse. Then there is the issue of specialised components: a typical Marconi ploy (almost as bad as Tektronix!). I found that where it was necessary to strip off the sleeving of ordinary stranded copper wire to make a soldered joint, the wireman had not twisted the individual copper strands together prior to soldering. That is poor workmanship. When that solder joint is unsoldered, it then becomes necessary to chop off the splayed out strands and re-strip. And since in many places there is inadequate wire left to do this and then re-make the connection, problems arise; not good; not good at all.
So overall, what do I think of it as a quality engineered product: very little.
But by what I have seen of other Marconi products in the date range mid 1950 → mid 1970, it is typical. But a lot better than the majority of Marconi T & M equipment that I have used and repaired after that period, most of which is virtually unrepairable at the component level, and in some, the build and design quality is simply appalling. 
► O.K. moan over.
Next step is to bolt the whole issue back together in its case, a quick check for basic functionality and put it on 'soak test' for a few days: tomorrows task; shouldn't take long.
And then we'll see.
Al.
So let's re-cap: I had no O/P on the lowest freq. range but one: that's range B plus a few concerns about the calibrator.
The "no O/P on range B" wasn't too difficult to track down, but it was necessary to discover how the osc. module for this (and any other) range features in the overall scheme of things: osc. generation to final R.F. O/P. And, I will add, that discovering that arrangement, by reference to the manual, was not easy. The cause was a failed coax cable termination at the osc. board: same type of problem as is commonly found in the filter boards, A7 and A8: perished insulation. However in this case (a coax termination) the Marconi chosen method was to connect a flying insulated wire to the braid of the coax and insert a cylindrical copper (?) sleeve over that junction and crimp that sleeve - then add a rubber insulating sleeving over all that. The inner of the coax had perished and the braid where the sleeve was fitted was very badly tarnished: clearly a form of chemical reaction had taken place. Now armed with that discovery plus the similar problem with the A7 & A8 filters, I set about examining various other coax terminations (of which there are a very great deal in this unit). And yes, many of them were 'on their last legs'. So some preventative maintenance was necessary.
Here are some pictures of the coax termination problem:
[attachment=5508]
[attachment=5509]
As for the calibrator, well, I'm still thinking about it. I have enclosed some pictures of 'scope waveforms:
1 MHz xtal osc. at TP1: [attachment=5505]
O/P of limiter at TP2: [attachment=5506]
O/P of 100 kHz divider at TP3: [attachment=5507]
And here are some pictures of the calibrator board itself and a couple of general views showing that board and the A7 and A8 filter boards - with new inductors fitted:
Calibrator board:
[attachment=5510]
Two general views:
[attachment=5511]
[attachment=5512]
You can see from the photo of the calibrator board that all the resistors are mounted vertically. What isn't shown is that of those ends that are 'scopeable', many of those ends are at 0v. / chassis. That does not help in fault diagnosis. With the exception of transistors that I had changed (or removed for testing), the transistors have their cases very close to the pcb surface: thus getting a probe in is not possible. Thirdly, the aluminium frame around the board (one-inch high) is also a restriction on access. And finally there is the simple issue of density of packing. All those factors make fault diagnosis with conventional bits of test kit very difficult. Moreover, not that this is a criticism of Marconi's design, etc., but the documentation I had to work with - especially the cct. diags. - is of a very poor print quality.
► Warning:
general moan and rant follows.'Quality':well I have my reservations about 'quality' with this item.
For a start, the technical manual is poor: there is no other word that is adequate. Inconsistent forms of reference, contradictory and confusing layout of drawings and the general master block diagram is simply inadequate for its intended task.
As far as the overall electro-mechanical design is concerned, on the basis of the very evident 'module assembly' style that Marconi chose for this item, it is perfectly clear that the maintenance philosophy was one of 'replace the faulty module'. O.K., that's fine as far it goes - but it comes with a rather large and expensive additional price tag: the large amount of extra wiring and cabling to connect all the modules together, especially when all the necessary 'switchery' is included. Not that I have anything against the module build approach as such - provided it isn't carried to unnecessary extremes - but in my opinion, in the TF2002B, Marconi took that approach far too far. Moreover, the physical layout of those modules does not seem to bear a logical relationship to the path of the associated signals - or which there are many: the module layout seems quite random. In addition to that, in terms of cct. design, in some places things are clearly taken to an unnecessary extreme; in others, the design is a bit patchy - or worse. Then there is the issue of specialised components: a typical Marconi ploy (almost as bad as Tektronix!). I found that where it was necessary to strip off the sleeving of ordinary stranded copper wire to make a soldered joint, the wireman had not twisted the individual copper strands together prior to soldering. That is poor workmanship. When that solder joint is unsoldered, it then becomes necessary to chop off the splayed out strands and re-strip. And since in many places there is inadequate wire left to do this and then re-make the connection, problems arise; not good; not good at all.
So overall, what do I think of it as a quality engineered product: very little.
But by what I have seen of other Marconi products in the date range mid 1950 → mid 1970, it is typical. But a lot better than the majority of Marconi T & M equipment that I have used and repaired after that period, most of which is virtually unrepairable at the component level, and in some, the build and design quality is simply appalling. ► O.K. moan over.
Next step is to bolt the whole issue back together in its case, a quick check for basic functionality and put it on 'soak test' for a few days: tomorrows task; shouldn't take long.
And then we'll see.
Al.
