Hioki AF-105 needle too low

[jeroen79]

I got an old Hioki AF-105 meter.
It turns out that the needle is falling just short of the measured value in all ranges.
Adjusting the trimpot gets it a bit closer but not far enough.

In the schematic I see a 7.8k resistor an 1k trimmer in series with the coil and a 30k resister in parallel.
(along with protection diodes and  capacitor)


A quick survey shows that all resistors are within spec.
So I think that the meter's coil has become less sensitive somehow.

What would be the best way to fix this?
Replace the 7.8k series resistor with a lower value? (so more current goes through the whole)
Or replace the 30k parallel resistor with a higher value? (so more current goes through the coil)

[tkamiya]

How much off are they?  They were NEVER known to be accurate.  +/-5% error full scale is common.  What this means is that say you are in 100V range.  You could read 95V or 105V and that'll be within spec.  That 5V difference exists all the way to zero.  It sounds ridiculous by today's standard, but analog days and in that era, that was pretty common.

My "GUESS" is, it has always been the way you see it.  My suggestion is not modify it and trying to fix it. 

I had lots of Hiokis and Sanwas.  Those were two of famous/big Japanese multimeter makers.  They were all like that.  We, in fact, never used those multi-meter to measure anything.  We knew what we read were proximate values and treat it as such.

[floobydust]

The meters can be accurate, it's usually parallax error and gravity that makes them seem bad. Ensure the mechanical zero adjust is correctly set.
What is typical:
Trimpot in parallel to the meter movement (30k resistor) - cal the full-scale meter current i.e. 20uA
Trimpot in series with the meter movement (500R trimpot) - cal meter coil DC resistance i.e. 15,000 ohms

I would use a DC power supply and series resistor like 100k and go for full scale on the 0.03mA DC range.
That is closest to direct access to the meter's coil (it includes internal 30k shunt resistor). Use a DMM in series to measure the loop current. If the meter is out, then the coil 30k (33k in some pics) resistor needs to be adjusted or a clamp diode is leaky, assuming the 30k ohm shunt is accurate.

After that, you calibrate the meter's total DC coil resistance, not sure what it is for the AF-105 guessing 15k. That is the 500R trimpot setting the dividers' load resistance including the 7.8k and meter coil... or you can pick a DCV range and calibrate to that full scale- assuming that range's resistors are accurate. If it isn't, then other ranges would be worse for accuracy.