Calibrating TDS340

[james_s]

I've now resurrected a pair of Tek TDS 340 scopes, both were filthy but cleaned up very nicely, the one that had a dead power supply turns out to have only logged 40 powerups, inside it's squeaky clean and looks brand new, CRT is nice and bright with zero burn. Anyway one of them had lost the calibration data and when I pulled the DS1644 from the other to read it that lost the calibration too, grr. Has anyone calibrated one of these? I looked at the service manual and the process looks relatively straightforward, it requires several precise DC voltages and something capable of producing a 1ms pulse with <10ns rise time. The calibrator recommended in the manual is >$20K which I'm obviously not going to buy to calibrate a couple of <$100 scopes. Can I reasonably hack something together out of a precision voltage regulator adjusted against my Fluke 45 and then a pulse generator buffered with a 74LVC14? Is there more to it? I have a TDS 784 that I can use to check the results so I'm not flying blind here.

[tggzzz]

Has anyone calibrated one of these? I looked at the service manual and the process looks relatively straightforward, it requires several precise DC voltages and something capable of producing a 1ms pulse with <10ns rise time. Can I reasonably hack something together out of a precision voltage regulator adjusted against my Fluke 45 and then a pulse generator buffered with a 74LVC14? Is there more to it? I have a TDS 784 that I can use to check the results so I'm not flying blind here.

That's what I have done for analogue scopes, but if you are measuring a voltage source and tweaking it, the source doesn't have to be precision! You'll have to RTFSM and probably infer what they are trying to measure and how.

For risetime tests, I use 3*(74LVC1G14+143ohms) in parallel to get a nice clean 50ohm output with a subnanosecond risetime (probably <650ps, unverified quick test 300ps). Make sure it is very well decoupled. The output voltage varies with the Vcc rail, which can be 1V(!) to 5V. If you need lower voltages, then you might use a 50ohm attenuator. If frequency isn't important, you can use another '14 as an RC relaxation oscillator.

[james_s]

Well I tried going through the voltage calibration. I ended up using a battery and a 10 turn 1k pot, monitoring the voltage on the other side of a BNC tee connected to the input. It asks for -1.5V, +1.5V, -0.8V, +0.8V, -80mV, +80mV, -15mV and +15mV which I gave it each time. Unfortunately it fails, same issue the several times I tried, it says failed, calibration constants not updated, the error log says "Cal error. CVR source gain". Anyone know what this means? The self tests pass except for Calibration and the scope works but the calibration is clearly a bit off. The annoying thing is that it was working fine until I removed the DS1644 and read it to prevent calibration constants from being lost.

I'm thinking of trying to build a more proper voltage calibrator with a unity gain buffer on the output and a switch that I can just select each of the voltages it asks for. The lower source impedance would let me have the 50 ohm terminator it says to use in the manual, although I can't see what difference the terminator would make when calibrating DC.

[Аrjen]

Did you manage to calibrate the TDS340 eventually? I also get the "Cal Error. CVR source" error when trying to do the voltage calibration with a self regulating voltage source, either with or without the specified 50  cable & resistor.

[james_s]

I set them aside for now, having hit a dead end. I'm able to calibrate the voltage eventually using a fancy voltage standard that I picked up but it is *very* picky. I have not succeeded in performing the timing calibration, I don't know why it's so fussy, a scope is not really a precision instrument anyway. I would be happy to be rid of the calibration fail at startup, the performance is otherwise adequate. The TDS340 is my "portable" scope that I use when I'm away from my bench.

I would love to find someone local who has the correct calibration gear to try, the procedure only takes a few minutes but shipping a relatively bulky and delicate instrument adds considerably to the cost and effort required.

[shakalnokturn]

I'm not 100% positive on what I'm saying, take this as an educated guess as I've just been through something similar on a non-Tek DSO.

CVR would be "calibration voltage reference" and one of the points to check (other than LVPS voltages/ripple) before going through the cal. procedure.

No clue if this adjustment is done through trimmers or menu on the TDS340 though.

The basic idea is that the scope uses a DAC, MUXed out to control various gains and offsets in the frontend. During calibration you feed the scope precise voltages, the DAC gives the wanted offsets/gains to the front-end so that the ADC will give a reading according to your precise input voltage.

If the DAC's reference is not correctly set (it's own offset and gain) the whole calibration procedure may either be forbidden by the software or give false adjustments if allowed.

[james_s]

It has to be all done through the menu, the entire guts of the TDS300 series is a small board on the bottom corner behind the front panel, most of the unit is just empty space as it shares the same chassis as the TDS400 series but is far simpler inside. There are no adjustments on the board at all, the only pots inside the scope are in the monitor.