Products > Test Equipment

What is the internal resistance of the Uni-T UT139C (on the mV range)?

<< < (6/6)

Sredni:
Ok, I am now convinced that the UT139C has big enough an input resistance to be able to read reliably the base-emitter voltage a BJT exhibits with an open base. Turns out... plastic cases like TO-92 suck big time when residual charges can be a problem.
Instead of shielding the whole setup, using only battery and going in the middle of a desert to perform my measurements, I just got hold of an old BC311 BJT transistor in TO-39 metal case. And I also used better probes, not those toys I had bought on Amazon for ten bucks.

Turns out that, when C and E are shorted with a 20cm wire, the UT139C only reads -4mV across base and emitter, reliably in a noisy environment. This is the rectified AC picked up by the setup. When powered with 8.3V battery, the measure is really fast, like I saw the first night in the middle of my garden. For this BJT it gives 308 mV. Clean, reliable, repeatable.

So, I hooked up my AC powered power supply, routing cables around the table (picking up all sort of noise) and... here are the measurements


https://i.postimg.cc/ZRBmjWD6/image.png

The higher voltage ones were very fast: 5-10 seconds tops and they almost stabilized ; the sub-volt ones required some 20 seconds to become steady.
The curve shows the expected exponential behavior as can be shown by stretching it to cover the ideal case


https://i.postimg.cc/yx4MPvRz/superposition-sim-data.png

(All exponentials can be stretched to superpose)
The value of VEB at the same voltage of the battery is lower (287mV vs 308mv) but this has to be expected because the AC noise superposed to the static value is danglilng from the DC value because the coupling capacitance and the EB diode form a clamper.

I am now beginning to think that the measurements I got the first time were actually 'correct' (within the error envelope of instruments and improvised procedure) and I was able to obtain them so fast because I had the transistor case touching a moist garden tile that removed stray charges, leaving only the charge driven by the external source.

So, my preliminary conclusion is: plastic cases suck.


AVGresponding:
You could just take it apart and reverse engineer the front end...   :popcorn:

Sredni:
I can barely see those tiny 0000 parts!  8)
And besides, maybe the FET frontend is inside the chip...

Anyway, I did a few more measurements and all data point to the UT139C having an internal resistance comprised between 80 and 110 gigaohms. I tried few more transistors, including a fake 2N3055 in TO3 case and then went back to the first 2N2222 in plastic TO92 case. This time I put it inside a metal box and, despite the probes still being exposed to a lot of AC interference I was able to get the following measurements at 19.5°C


https://i.postimg.cc/FHpWY6GL/image-2.png

(I did it in two parts and there seems to be some memory effect for the measurements under 3V)
Then I redid the simulation changing the beta of the default 2N2222 in LTSpice to 262, the beta I measured for the specimen I used.
And here is how the data fits the simulation (here shown with open base, and with base shunted by a resistor from 1gig to 71gigs)


https://i.postimg.cc/Xv7CvHDX/superposition-sim-data.png

And here is a detail of the simulation for values of the shunt resistance comprised between 80 and 120 gigaohms, in steps of 10 gigaohm


https://i.postimg.cc/xTDNmTFw/superposition-sim-data-detail.png

I would call this a good agreement. A good FET input preamplifier would offer an input resistance of 100 gigaohm (see for example the high-Z preampli proposed by Bob Pease), so this is not totally implausible.
(But I might have some negative offset that, once taken care of, would lower the experimental data to a curve of maybe 70-90 gigaohm).

One thing is sure though: the internal resistance is so high that I cannot determine it with the divider method (using my instruments).

Sredni:
To close the topic: the internal resistance of the UT139C in the mV range is not documented in the English version of the manual, but - I just found out - it is in the Chinese version. It's in the asterisks: the 60mV range has an internal resistance bigger than 100 Mohm, while the 600 mV range has more than 1 Gohm (and from what I can tell is several tens of gigaohms at DC).

The information is given as a footnote, and it appears it was not translated in the English version of the manual.

Navigation

[0] Message Index

[*] Previous page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod