Craftsman 82139 Multimeter Teardown

[MWisBest]

Found this meter inside my dad's toolbox and thought I'd pop it open to take a look inside. It's a fairly cheap meter, it can (or could) be had for $30, but it is CAT II 600V rated and can measure capacitance, frequency/duty cycle, microamps, and temperature, in addition to the usual stuff. 4,000 counts.

(Note: Click on any of these pictures for hi-res versions)

The included boot/case is quite nice, and the stand is stable enough to use the range switch without tipping over.


Separate battery and fuse compartments, self-tapping screws only. The fuse compartment screw holds the entire unit together though, so the separate cover is useless anyway! The case has curves along the sides, and the sides overlap a lot. Overall it's very solid in the hand and does not twist or creak with force. There is also some shielding on the back cover.


The main IC is an FS9711-LP1, essentially identical to the FS9721 series. Compared to the (arguably more common) LP3 version, the LP1 sacrifices RS232 output for some additional measurement functions (adds temperature, doubles maximum capacitance and frequency range). The input terminals are solid (non-split) type. Fuses are glass 250VAC type, but can be replaced with ceramic versions, e.x. Littelfuse 216 and 285 series. The uA/mA range is a shared jack, but the current ranges are all the way on the right of the dial, so accidents are very unlikely.


The MOV is only there to clamp the thermocouple jack. The 10A current shunt is messy. The range switch is smooth, but the wipers go over some traces, which is never a good thing.


I do have one problem with this meter: accuracy. It's not a very precise meter, 1 to 2% plus quite a few digits. That is understandable in this class, but it could be improved with more precise resistors.

The resistor network is simply not precise enough. You can't calibrate out the difference between ranges, so you have to cal somewhere to get a decent average accuracy across ranges. Also, lol at the tiny MLCC in C15 with the large pads on the board

[MWisBest]

Overall the circuit follows the examples in the FS9721_LP1 datasheet closely. http://www.ic-fortune.com/upload/Download/FS9721_LP1-DS-20_EN.pdf

Going to try whacking in some precision resistors to fix the drift between ranges. Ultimately the accuracy of the IC comes down to the resistors, and it seems like they've used standard 5% tolerances.

I did verify that the ENTX pin doesn't turn on RS232 output, the display does have the segment for it and the datasheet goes back and forth between RS232 existing and it not existing, so I was curious. Would be nice to have the hFE tester but the changes required to do so aren't worth it IMO.

[MWisBest]

Sorry to bump my own thread again, but I noticed something odd and was hoping for some ideas. I tried measuring the voltage of a li-ion battery, and noticed that the auto range went to 03.61V instead of 3.61XV. Manual ranging down displays it fine. After observing closely, I saw the meter overshoots to 5.9V before going down to 03.61V, which is just above the range-down value of 3600 counts. What could be causing this overshoot? Can it be solved somehow?