Question about oscilloscope adjustment

[c4757p]

A friend just got the same oscilloscope as mine (Tek 2445A), and one channel is significantly out of vertical calibration (looks like the preamp was replaced). I have the necessary equipment to perform the vertical adjustment procedures as specified by the service manual. I do have one simple "definition" question about the procedure, though, that hopefully somebody familiar with these can answer. I need to substitute a good digital function generator for the "calibration generator". (It has proven to easily have the amplitude accuracy to do this.) When the instructions say "Connect a 0.5V, standard amplitude signal from the Calibration Generator ... via a 50 ohm BNC cable",  what exactly is this signal? DC? Square or sine - and at what frequency, and with or without a DC offset? And does anybody know if it will have the 50 ohm terminator engaged at this point? (Obviously it makes a difference for the amplitude, and it does not display the settings on screen at this point. I could probe the jack with an ohmmeter, but I don't really trust it not to switch it in and out as it sees fit.)

[SeanB]

Probably a square wave at 1kHz, with no DC offset. It probably will be terminating into 50R to do so, so your source will be delivering into 50R. will need to have good low rise and fall times with minimal overshoot.

[c4757p]

Thanks! That was quick. Perhaps a dumb question, but by "amplitude" in this case it means center-peak, right? I know this is the usual definition of 'amplitude', but a lot of function generators (mine included) only seem to take peak-peak, so I wondered if this was a common definition used for this kind of thing.

Is 10ns or so good enough rise and fall? It's not exactly an "avalanche transistor" edge, but it's still many orders of magntude beyond the 1ms period. And I can't discern any ringing or overshoot at all when delivered into 50 ohms. Very clean waveform.

[SeanB]

Probably peak to peak, as most waveforms are invariably specced to that.

[c4757p]

Ah, OK. Thanks again. (In retrospect, the amplitude question should have been obvious once you told me the 50 ohm terminator will be engaged. It wants a 10V signal, and if that were center-peak that would exceed the maximum power rating of the terminator.)

One more question, this time purely out of curiosity - hopefully you or someone else here can guess at this, because I'm stumped. I've started the routine, and at one step, it performed a sort of automatic adjustment of the vertical cursor alignment. Now, as far as I can tell, the MPU has no way of knowing what actually appears on the screen, and some other steps of the adjustment appear to confirm this. It blanked the screen and played a little dot around, and I swear to god it looked like it was performing a slightly modified binary search looking for the graticule lines. Unlike the other steps, at no point did it require me to move the dot for it onto a specific point. Does anyone know what it's doing here?

[SeanB]

It is doing a confirmation ( for you) of the DAC stages, so that you can see that it is a monotonic display and output. It could be measuring the DAC via the input ADC to check for non linearity which would fail the calibration sequence.

[c4757p]

Hmm. Surprisingly (I was sure it would be 50 ohm and not offset, just wanted to check), that's not right! It did not like the signals I was giving it one bit (kept flashing LIMIT on the screen), so I dug up the manual for the recommended calibrator (Tek PG-506). Surprise! Default output is into 1Meg and the signal is chopped DC between 0V and the reference voltage. (Running a T-adapter off the jack to a voltmeter verifies that it is not terminating the signal.) The one thing we were right about is that it's a square wave at 1 kHz! It likes these signals much more...