OK I have made up my mind.
Trying to recreate a resistor with the same weird tempco is too difficult to get right. I would never be sure that it restored proper operation of the original design, because I am not competent enough.
I think it's easier to start from a blank sheet and do an entire new replacement board with a modern Vref. This way I control everything, use known parts with knows specs, available datasheet etc.
It's also a marvelous excuse/opportunity to do a little bit of design, use a Vref which I have never done before, and of course give me some more training on Kicad, which I need. Every opportunity to do design little boards on Kicad is good to take here. Plus, it will look cleaner/tidy once done, that trying to Frankenstein the old board.
Plus, I find it cool to go the extra mile and design custom stuff to save an old TE. It's my way of showing it some love and respect...
So... new board it will be, with a "modern" little Vref chip. Don't know much about them... luckily Dave Jones recently did a video on these : "Jelly bean Vref ".
While this is being sorted out, I can get the meter going as you said with any cheap and dirty solution. Hell I could simply replace these two resistors with regular modern carbon resistor and call it a day, should "work". Bye bye tempco, but at least it would allow that board to produce a voltage of appropriate value. Stable it won't be, but still, it's gonna be good enough to test the meter, allow to fully troubleshoot / repair it, etc.
Hell I could even just shove two wires in the meter and supply the Vref using my external lab power supply !
![Laughing :-DD](https://www.eevblog.com/forum/Smileys/default/smiley_laughing.gif.pagespeed.ce._hfWAz_QHO.gif)
Yeah..... this meter is becoming more and more of a project ! It's a time to make its own thread before I am too far into it !
![Laughing :-DD](https://www.eevblog.com/forum/Smileys/default/smiley_laughing.gif.pagespeed.ce._hfWAz_QHO.gif)