DMM input impedance measurement

[opa627bm]

Hi all,
I was curious about how a handheld DMM is designed and how input division and protection was implemented so I bought a fluke 107 just to take it apart and trace back everything. (since it is cheap and very simple）
When I tried to measure the input impedance, I got about 11MOHM from 34401A, 34470A and DMM7510, but when measure with 3478A I got about close to 18Mohm.
When I connect the 1M and 10M resistance standard to 3478A the read out is spot on.
Is it possible that during the resistance measurement mode, the output voltage/current is smaller than other DMMs so it reads way off from other DMMS? （or it is too high and tipped some active disconnect ?)

I am going to cal the 3478A in the 2 wire mode since I received the calibration short for it. but I heard the 30MOHM range cannot be calibrated on some DMM, is that true ?


also, I noticed there is a piece of "Glass" looking plate on the PCB , and I can see finger print on it, not sure what the purpose of that is for....

Image credit from Z

Thank you
Regards,
Li

[TheDefpom]

If I remember rightly the 3478 has selectable input impedance of either 10M or 1G, that might alter the result

[bdunham7]

The input impedance of the F107 will change a bit with different voltage ranges, the 11M probably means it is on its lowest range, right?

As a first step, turn the F107 on and measure its impedance again with the HP3478 and one or more of the other meters and note the voltage reading on the F107 for each.

[opa627bm]

The input impedance of the F107 will change a bit with different voltage ranges, the 11M probably means it is on its lowest range, right?

As a first step, turn the F107 on and measure its impedance again with the HP3478 and one or more of the other meters and note the voltage reading on the F107 for each.

HI,
It is always in DCV mode. But somehow reads a lot larger on my HP3478 than other meters (20MOhm Vs 11Mohm)
But when I set the 3478 in ohm 2 wire and measure the input impedance of DMM7150, I am getting spot on 10Mohm and on DMM7150 is reads 1 V.
Regards,
Li

[bdunham7]

It is always in DCV mode. But somehow reads a lot larger on my HP3478 than other meters (20MOhm Vs 11Mohm)
But when I set the 3478 in ohm 2 wire and measure the input impedance of DMM7150, I am getting spot on 10Mohm and on DMM7150 is reads 1 V.

OK, but what voltage does the Fluke 107 read when you connect the various meters to it to measure its input impedance?  And can I safely assume you are always using 2W ohms mode on the 3478A ?

[GigaJoe]

"finger print" - on glass looking plate, is a unique signature specifically for a fluke meters, of the MASTER , who assemble the multimeter. 
Its well-known fluke slogan - "One Master - One Multimeter". 
similar to Mercedes AMG engine assembly.
So you are proud owner of THE FLUKE ONE !

[opa627bm]

loool, now it reads 11.05Mohm from my 3478A. The only difference is the battery.
before the battery on 107 was not installed and it reads high, but since you asked me to take the reading on the 107 side I put the battery back and it reads about 1.02V on the 107 side and resistance on 3478A reads 11.05Mohm.
This is kinda strange... maybe there is an active protection on the 107?
all measurement was done on 107 DCV only.. this is strange...

[opa627bm]

"finger print" - on glass looking plate, is a unique signature specifically for a fluke meters, of the MASTER , who assemble the multimeter. 
Its well-known fluke slogan - "One Master - One Multimeter". 
similar to Mercedes AMG engine assembly.
So you are proud owner of THE FLUKE ONE !

I do have one made by Tobias Shuttles (Think that's how it spells, signature is hard to read) M177, Love that thing lol

This glass plate is just strange. maybe it is a shock indicator ?

[Andreas]

This is kinda strange... maybe there is an active protection on the 107?

Hello,

definitively no but:

many handheld DMM use a input voltage divider which changes the input impedance with the range. (the 11 Meg instead of 10 Meg indicates this).
(the voltage selector does not connect from a fixed string to the input of the ADC but
there are several switches to ground which tap one of the low side divider resistors to ground).

see for example the CADDOCK 1776-C48xx resistor series.

So the input impedance of the multimeter is dependant on the selected voltage range.

with best regards

Andreas

[bdunham7]

loool, now it reads 11.05Mohm from my 3478A. The only difference is the battery.

 

It wouldn't have occurred to me to measure the input impedance of a DMM with the power off.  Of course, you do want that impedance to be at least the normal value in case the meter is connected to a test point before being powered on, so maybe not so crazy.

Now to explain the difference,  you have to do the same test using your other meters and see what voltages they use to test ~11M.  I'm guessing they'll all be higher than 1V.

[Conrad Hoffman]

Meter inputs may not be entirely benign. I'd do a reality check by measuring a known source through a high value resistor and calculating from that based on the readout error.

[Kleinstein]

Measuring the input impedance of a meter in high Z mode (e.g. > 10 Gohms specs) can be tricky. It starts with using a suitable definition. It is not just a resistance between the terminals, but also an additional bias current and possible resistance to the case / guard if present. The input resistance can also change with voltage.