Rejustage my Fluke 341A

[bastl_r]

I have an older Fluke 341 here that has too large jumps in the second decade, which serves as the basis for adjusting the first decade, both in the positive and negative direction.
It may be that the specified deviation is maintained, but the resistances of the third decade are consistently too high and I end up with a deviation of approx. +0.01%! Measured with my Prema 5017 and in parallel with an HP3457A.
I would like to correct this without having to recompile all the resistances.
For the third decade, I thought that it would be sufficient to add high-impedance parallel resistors to the existing resistors. This would have the advantage that any EMF voltages at both soldering points would be equalized i think.
For the resistors in the second decade, which deviate too much downwards, I would try to install low-impedance SMD resistors in the tracks, which would have to be separated.

Is there a NOGO or any hints as to what else I could do or can I do something different?

Enclosed are two tables with which I have calculated the correction resistors and the results.
I made the measurements in the 100V range, as recommended in the manual.

Regards Armin

[bastl_r]

I know. The 341 is not the high-end device.
Does nobody have an opinion on the project?
Or are my thoughts on the procedure plausible?

Regards

[Kleinstein]

The weak point seems to be the 2nd decade, with quite some difference betwen the resistors.
If one decides to adjust the 2nd decade, one would likely take the 3rd decade as the starting point / reference, as this part is staying as it is.
From the data shown the 3rd decase does not look that bad, at least if it matches the 4 th.
This could in most cases mean adding some series resistance added to the 2nd decade and than readjust the 1 st decade.  I see not sensible way to get around readjusting the 1st decade. With a modern meter as a help this should not be such a hard part - easier than soldering trim resistors.

The very fist part would anyway checking if the switches make good contact - added resistance from the switches is a very plausinble failure point and can be changing over time.

Instead of low value SMD resistors one could consider getting some manganin wire.  The wire can be trimmed to length can can have very low thermal EMF. For wire of some 0.28 mm nominal I got some 15 ohm per meter and this a managable length for the 0.1 - 0.5 ohm trim resistors.

[coromonadalix]

on a 343a  i had many capacitors to change,  1st  it was in the chopper amplifier section, it was unstable

rotary switch cleanup with deoxit

check the shaft extenders if they are not broken or fissured

[bastl_r]

Are the switches that susceptible?
They have sliding contacts which should actually clean themselves to some extent.
Why I have hope: I have already read a lot of negative things about oxide cleaners and wouldn't have any here.
I think I'll try using paper first. Never use sandpaper!

The shaftextenders of the switches all still look very good and have no visible cracks.

I have already replaced some of the smaller capacitors and all the large high-voltage capacitors. One of the latter died two years ago, causing the calibrator to fail completely.

[bastl_r]

Instead of low value SMD resistors one could consider getting some manganin wire.  The wire can be trimmed to length can can have very low thermal EMF. For wire of some 0.28 mm nominal I got some 15 ohm per meter and this a managable length for the 0.1 - 0.5 ohm trim resistors.

Manganinwire would be the best choice. I know.
But if I calculate a 0.66Ohm resistor with TK100ppm/k, the value increases by 0.002066Ohm with a 10° temperature change.
In series with a 20000.072Ohm resistor, its value increases to 20000.072066Ohm.
This can only be seen at the last digit on an 8.5-digit multimeter.
Thats good enough for me and my old Calibrator. 

Or do you expect a much larger error if I solder SMD resistors instead of manganese wires onto an interrupted conductor track?

regards Armin

[Kleinstein]

At least 2 types of manganine wire are available in small quantety (some 8 EUR for 5 m) via Ebay in Germany.  If one needs several odd values, this may be easier than buying SMD resistors, that can be hard to get for all values. The wire would allow easy tweaking of other values if needed in more iterative steps.
The problem with the SMD resistors would not be the TC, but more possible thermal EMF and simply getting odd values. 0.56 and 0.66 (as 0.68) ohms may however actually available.
For just 2 values the SMD reistors could still be easier.

[bastl_r]

Hi
Thanks for your opijion.
I saw  0,25mm manganinwire on ebay(from ukraine).
For a 0,56Ohm resistor i will need a lengt of 5.6cm. For this i need some space an it must be more precice soldered than 0,5mm(1%). I don't think i can manage that.
For lower resistances it will be much mor complicated.

Regards Armin

[Victorman222]

Yeah it's hard with smaller wires. If a bigger wire is used then it can be adjusted by removing some material accurately with a file. Many old wirewounds i seen were made so that they have copper wire overlapping the manganin wire at the place where they are soldered, maybe that allows some fine adjustment by moving the spot where they are soldered (see picture). Also for smaller wires strain relief becomes more important.

[TheDefpom]

I have done some videos fixing my 343A, and yes the switches do need cleaning, as well as replacing ALL of the electrolytic caps, I also had issues with some of the switch mounting posts being broken, so I 3D printed replacements (available on my thingiverse page).