Tektronix 2465A repair: general questions on calibration

[Bill Woodbridge]

After some adventures in recapping and fixing a faulty illumination board, my 2465A (through-hole A5 board with Keeper, not Dallas) is now functionally working, it seems.  Unfortunately during the course of that work I damaged the calibration data beyond repair.  I’ve now replaced the Keeper battery with a clone unit (the amusingly named ‘Energy Very Endure’ LTC-7PN from Farnell, as here in the UK we can’t get the original Keepers shipped easily), and tested the RAM in as far as it will correctly write and recall dummy data for hours and power-up count (CAL05).  The rest of the data is still random rubbish, awaiting the full calibration, and as expected I still have a calibration checksum error showing on power-up (TEST 04 FAIL 13 XX<C620>).

So I now want to take the plunge into a full calibration - as someone said in one of the forums I’d rather invest time learning more about scope performance and calibration than in trying to hack RAM data that is never going to give me calibration constants appropriate to my particular machine.  I’ve been lucky to buy nearly all the called-for test equipment at reasonable prices, on the basis that I’ll resell it once the job is done so I’m really treating it as equipment hire (I can give a full listing in due course if anyone is interested).  I’ve done quite a bit of pre–reading in the manuals (Tektronix original and the military version) and online, but I have a few overall questions before starting please that I haven’t found definitive answers for, in an attempt to avoid frustration and needing too much Energy Very Endure.

1.   Does each CAL routine save its data individually when exiting, so the machine can then be powered-down if desired and work can resume on the next routine another day?  Or must all the CAL routines be done in one single powered-up session? (I’m aware the routines have to be done in the correct order, apart from CAL08 that should precede everything, that they all have to exit cleanly, and that  the NO CAL / CAL jumper must only be moved with the power off)

2.   (Possibly linked to the previous question): at what stage does the calibration data checksum get recalculated and written back into RAM, thus hopefully clearing the checksum error?  As mentioned above, I’ve been through CAL05 and its new data stores correctly, but the checksum error is still present after that.  Is it the case that because the majority of my data is still rubbish at this stage (complete with many parity errors no doubt) I’m never going to get a good checksum until all the routines are complete?  I don’t want to invest hours doing calibration only to find that at the end of it all I still have a checksum error that would require repeating some or all of it.

3.   CAL08 (the CRT adjustments), which I’ve dipped my toe into already, presents the sub-tests in a slightly different order to the one in the service manual appendix I’m using, dated 6/5/87 for serial numbers B016276 and above (my serial no. is 131499, a Guernsey unit, the leading 1 being the Guernsey code so how do I relate 31499 to the B(eaverton) number range in the manual?).  Is there a later revision to the service manual to reflect this?  I am now afraid that in the more complex CAL routines (eg 01) I’m going to get completely lost if the sequence differs from the manual.

4.   Part of CAL08 instructs: ‘Adjust Z-Axis drive (R949) full CCW for maximum display intensity’ (if R949 is fitted: mine is).  But I’ve found no further reference to readjusting this control anywhere else in the CAL procedures.  What is the point of an adjustable R949 control if it should be always set full CCW regardless?  Did the designers realise at some later stage that making it a pot rather than a fixed resistor wasn’t necessary after all (hence the caveat that it might not be fitted at all)?

Finally I am generally nervous that the Service Manual seems to make it a point of honour not to include any pictures of the displays during CAL, and to a newcomer some of the wordy descriptions sound very difficult to interpret in terms of what is on the screen, what feature of the trace needs adjustment, and where it should ideally be adjusted to.  But I’ll hold off on questions about all that until I’ve actually tried it, perhaps it all becomes clear once it’s on the screen. 

I did start, and finish, another thread on my previous repair efforts with the 2465A, but I’ve been advised open a new one as things move to calibration.  Thanks as always for any help.

[DLJ]

Hello Bill.

I am very interested to see how you get on with the calibration of your 2465A, as I am in a similair position. I have a 2455A that I have repaired but now need to calibrate. I do not have access to the official calibrators so have been trying to work around that, with only limited success so far.
So please keep posting about your progress and good luck.

David

[Bill Woodbridge]

Thanks for your kind words David, I’ll certainly keep posting my amateur experiences on progress, pitfalls etc if it would be useful.  But I’m no substitute for the 24xx gurus here and on tekscopes@groups.io, so I’d encourage you to read them first.  They’ve probably answered my four questions already somewhere, and if so it’s my fault for not doing enough online digging.

I’m still waiting for some of my calibration kit to arrive so it will probably be next week before I start.  As I said, the plan is to sell it again at the end of the day, so if you’re in the market at that stage … 

[Kwakerman]

Obviously the calibration accuracy of your calibration kit will affect the overall results.  Luckily the service manuals for most of the plug-in are available so it is possible to do a 'confidence check' just so that you know your new old kit is about right.  On the PG506 for example you can set the service switch to DC and measure the various output levels with a DMM to check they are ok and on the TG501 you can measure the master oscillator frequency and if that is correct then the pulse timing should be correct

[Bill Woodbridge]

Thanks for the tips.  Yes, I was trying to ignore the inconvenient truth that my calibration equipment itself is long out of calibration!  My reasoning was that nevertheless it was still likely to be more accurate than the 2465A currently with a completely random set of calibration constants.  But I’ll certainly look at the PG506 voltages, as you say, and the TG501 master frequency.

The whole project is going slower than expected, because the next challenge is to repair the calibration kit itself!  I now have a motley assortment: a fully working PG506, two partially working TG501s and two partially working SG503s, plus the TM503 mainframe.  So far I’ve started with the easiest challenge: sorting out stuck pushbuttons on one of the TG501s.  I hope I won’t need to come back to the forum on this subject, as they’re all much more straightforward than the scope itself (electrically at least, physically they look like a real challenge to disassemble enough to faultfind while keeping them powered up)  and that the 501 problems are on the digital/low speed side of things rather than with the more exotic high-speed board.