First of all, a big Thank You to Jay_Diddy_B for this project. It is a piece of art.
So, a short history. Since few months I went back to my love, electronics. Since many years I had not time and my job was not as an electrical/electronics engineer, and my passion was on hold, till this pandemic crisis erupted. Now, I had time, and I reorganized my workshop as it was converted, along of the years, in the warehouse of useless things around, which missed to be throw away. I started to design and build or build or just repair different electronics things.
This project was a challenge as none of my initial PCBAs, versions of current level, worked as expected. With the last one, 1 mA version, I went back to oscilloscope, books to remember what I learnt in University and did not practice, and paper and pencil as I just decided to learn Micro Cap simulation and it was not enough to solve the issue. The conclusions:
1. Schematics with R10 = 60 ohms and R8 = 6 ohms (first circuit diagra in the post) does not work with the scale of Jay_Diddy_B. Actually, he mention in one post that his design is based on 3 ohms and not 1 ohm or 10 ohms he presented in the charts. It works with 30 ohms and 3 ohms.
2. My version of 1 mA had an error comparing with the mentioned scale of 10%. I reduced R10 value (30ohms)with 10%, which means 27 ohms and it is sharp aligned with the values on the scale for each measurement.
3. If you would like to use the above mentioned scale and do not build it, which I strongly advice as, in my opinion, its repeatability and accuracy are unexpected good, then my suggestion is to replace R10 with a trim pot and adjust its value by measuring different resistor against the scale. Attention, you must do the zero after each adjustment of the R10 value. With R10 = 27 ohms I obtained full alignment with the scale for any value I measured, on the right and left side of scale.
4. To create the print of scale of Jay_Diddy_B my solution was to convert the picture, he published herein, in vector by Inkscape and then resized to the meter dimension, keeping the aspect ratio. Just resized the picture, due to the pixel manipulation, gave to me errors in dimensions.
5. The different current versions are optimized by the ratio between R15(R20) and R14(R19). Based on my calculation the optimized ratio for proportional increase is 3.63, which is the ratio of 50 microA R15/R14=R20/R19= 3.63 (I am referring to the annotations from the first circuit diagram from the post).
6. Checking the calculations by measuring few current versions I built, following the above rule, with a good approximation, to define a certain current value version, you just need to divide the value of R14(R19) of 50 microA with new current multiplication ratio and obtain the new value of R14(R19) for that current value. Multiply the new value of R14(R19) with 3.63 you find out the value for R15(R20). This is an approximation but it worked on all five current level PCBAs I built. My 50 microA PCBA version is able to deliver 140 mA, my PCBA version of 250 microA published herein by Jay_Diddy_B, R15(R20)=680 ohms, R14(R19) = 160 ohms delivers maximum 246 microA (330ohms/160 ohms= 2.06 and 140microAX2.06 =289microA but this is not following the ratio of 3.63!), my version of 1000 microA R15(R20)=220 ohms, R14(R19)=68ohms delivers 676 microA (330 ohms/68 ohms = 4.85 and 140microAX4.85=679 microA, ratio is 3.24). The aproximation is verified also for the other versions I built but the text it is already too long.
7.Based on this approximation I will build, in order to check if it is still valid the version of R14(R19) = 40 ohms, R15(R20) = 160 ohms R16(R21) = 4 ohms. It must deliver maximum 1100 microA according the above calculus approximation.