Tektronix probe adjustment adjust - Part 2 - Tek 475

[Redcat]

Hello EEVBloggers  ,

as promised in part 1 I want to share now part 2 of the Tek oscilloscope probe adjustment adjust/modification.

It is for adjusting the probe compensation signal output to be exact 1kHz and 50% duty cycle.

This time for the old model 475.
If you are interested in modifying it for the 2215A look here into part 1:
https://www.eevblog.com/forum/beginners/tektronix-2215a-probe-adjustment-adjust/

Would be nice if this can help someone.

Last time I thought, the signal in the 475 is generated like in the 2215A, but I was wrong.
I am a beginner in electronics and I could easily understand how the OpAmp oscillator of the 2215A works, but I can not understand, how it works in the 475.
There it's done with transistors - perhaps someone experienced can explain how it works  .

Luckily for me the important parts are named in the manual - they are R1502 for the frequency and R1504 for the duty cycle.
R1502 should be 43k and R1504 47k.



The mod is done easily:
First and important - UNPLUG YOUR SCOPE FROM THE MAINS, to avoid damage to you or your scope.
You can find both resistors at the bottom pcb - but they are not labled. You first need to make a little space, so that you can desolder them.
This was my first problem, as I didn't have a tiny 1mm allen key to get the screw of the intensity pot off, but my neighbour was so kind to borrow me one.



Then simple desolder them (without burning anything around  ) and substitute them with tiny little 100k precision pots (please look at the photo for the type).

I bended the outer pins of the pots outwards and bended and soldered the middle pin to one end, so that you get a variable resistor.
Was lucky that booth fit exactly in place.
Please preset them to 43k and 47k before soldering them in so that you have a good starting point.

Reassemble all and then you can start tweaking the new precision pots for frequency and duty cycle.
They booth depend on each other, so you will need to re-tweak them after one is set.
For calibration I used my UNI-T DMM, it has frequency and duty cycle measurement capability.







That's it. Easily done  . Please be careful when modding your scope, it's on your own risk.

This mod is mainly for optical reasons, as last time, as I wanted my calibration signals to be exact.
( Sorry for the photos. We have a saying for this: "The shoemaker has always the worst shoes" - as for me working in photo industry  . )

I'm wondering, why it wasn't done with adjustable pots out of factory  .

I hope you found this interesting and maybe it helps someone.

Nice greetings, Tom 

[w2aew]

I'm wondering, why it wasn't done with adjustable pots out of factory  .

The simple reason why this wasn't done at at the factory is that the probe calibration output was never intended to be a frequency reference or duty cycle reference.  It is/was intended only as a reasonable square wave that is used to aid in the adjustment of the compensation capacitors in a 10x probe.  For this job - approximately 1kHz and approximately 50% duty cycle are more than adequate.

[grumpydoc]

The simple reason why this wasn't done at at the factory is that the probe calibration output was never intended to be a frequency reference or duty cycle reference.  It is/was intended only as a reasonable square wave that is used to aid in the adjustment of the compensation capacitors in a 10x probe.  For this job - approximately 1kHz and approximately 50% duty cycle are more than adequate.

+1

If you look at the spec for a typical 'scope "cal" output it will be fairly generous - eg a Tek 2225 is  1kHz ± 20% and 0.5V ±5%. The 475 is somewhat better on voltage (1%) but still only claims to be "approximately" 1kHz.

That said the cal output on my Lecroy 9354 is actually the best fast edge square wave generator that I have at the moment - at least until I fix the PG506 I bought a while ago - up to 2MHz repetition and somewhere between 800ps and 1ns rise time. Pretty good on voltage as well since its adjustable from 10mV to 1V (5mV to 500mV into 50 ohms)

Really, it just needs to be decently square and have a fairly fast rise time.

[Redcat]

Hi Alan,
Yes you are right, and I know this (even in the first part), that it was not intended to be accurate (on my 475 it was 900 Hz). Only for a fast rising edge for probe comp adjust. It's pretty much useless for anything else with these specs.
But why not? It costs nothing compared to the rest of the scope to put these 2 precision pots in (or some more parts in the other scope).
For me it makes it look better (makes it more perfect) and might be handy for comparing something... or whatever  .
So would have been nice in my opinion . Why not make things better from the beginning?

But if you could be so kind and rough explain how this transistor oscillator works? It looks different from what you had in one of your videos (I think I understood yours...with the diodes... if I remember right? - but this looks different).
And is it better (or with faster edges?) compared to the OpAmp oscillator, or simple older technology...?

Tom