SciencewithDrJ said:
precision balances, distilled water
distilled water is not a problem. We have a 200gr calibrated scale that goes down to 0.05 gram and two not calibrated scales that go down 0.01 gram. Not top of the bill scales but I think I can thrust them enough for what I do with them.
SciencewithDrJ said:
Also, why only N and K? Phosphorus is critical also, and so is the pH measurement for healthy soil.
You are completely correct of course. pH measurement is being done. I'm in the process of validating my procedure, because I want to make sure I can thrust the measurement.
On this forum I found a thread about measuring P. I will try it out. Chances are I will need some help with it

but let's first try and see
Very happy with approximate indications. Something like 'nothing - low - average - more that average - a lot'.
And possibility of comparison is important: 'the same - a little more than the other one - a lot more than the other one'.
This will allow recording evolution of the levels over time. No need for absolute figures. It would be fantastic to have them, but really not necessary.
@jim mcnamara thank you for this document, was searching for some time now and couldn't find anything like this on my own. At first sight it does contain lots of valuable information.
Thank you both for your answers. If it was only for a couple of tests I would send out the soil samples to a local lab. They perform analysis very well but they are expensive (it starts at 75€/$87 per test). This year I'll need about 20 tests only but as of next year probably a few hundreds of tests. And that represents a lot of money. Therefore it is worth some investigation. And anyhow I'm curious about how it is being done, I would love to have a relatively simple DIY method that is reproduce-able by many for cheap.