UNI-T UT89XE: 20000 manual ranging counts of pure fun - impressions & teardown

[Cymaphore]

Hi,

after my last impulse purchase of a crap DMM (see here, the "killer") this time I intentionally searched for a bizarre device. I tried to find the most ridicilous DMM possible (right before the level of the bluetooth speaker dmm). Like the "best" manual ranging chinese miracle device possible.

I found it and I purchased it. After my last good experience with UNI-T (3 years ago, had a rebranded UT insulation tester that burned down at normal usage) I decided to give them another chance.

The UNI-T UT89XE. 20.000 Counts of manual ranging ridicilousness. This toy cost me 46€. Yes, no transistor testing included, and yes, it contains fuses, but I had to go for it anyways.

Automatic backlight, peak, min/max, continuity with led indication, diode with "beep"-indication, conductivity measurement, frequency, capacity, current, voltage (truerms), resistance, temperature... Mold-on soft case. Autoranging on capacity and frequency. Fuse access without opening the case. Crappy test lead jacks. What a marble. Claimed DC accuracy 0.05%+5, comes with Type K probe, standard Unit-T test leads. Rest can be found at the UT89XE Product page and the user manual.

And yes, I will now stop buying those chinese toy meters. That's it. :-)

It has some alibi input protection with 250V fuses and a couple of small PTCs and even some clamping. I wouldn't trust the rating. Looks like bare minimum, but ok for non-mains voltage low energy elv hobby usage.

DMM-Chipset is Cyrustek ES289B, some sensitive components are even under a big metal shield. It has a trimmer for manual adjustment :-)

And it's manual ranging! MANUAL! Like in the old days! But with 20.000 counts! I already feel like over 9000% more competent when I have to cycle on millions of options and ranges. What a great toy.

Anyways, I did some quick comparisons on the desk to my BM789:

At ~23°C against my reference:

2.5V: 2.498V (BM789: 2.4982V)
5.0V: 5.000V (BM789: 5.0004V)
7.5V: 7.499V (BM789: 7.499V)
10V: 10.001V (BM789: 10.001V)

100R: 99.74 (BM789: 99.96)
1k: 0.9987k (BM789: 1.003k)
10k: 10.000k (BM789: 9.997k)
100k: 100.02k (BM789: 100.02k)

At ~230V 50Hz mains voltage it deviates about ~1V from the BM789 and beeps regularly.

Temperature measurement looks mostly comparable to what the BM789 reports for room temperature and when putting them both next to the open window. I took some electrolytic from the box, labeled with 2.2mF, the UT89XE measures 2.074mF in about 4s (BM789: 2.15mF).

Some nearby diode gave a reading of 491mV (BM789: 465.3mV). Diode testing gives 2.9V output voltage and lights up some LEDs I had around.

Continuity test is very quick, latched and gives red LED feedback on continuity.

Display is nice, viewing angle is ok, bit bad from the very top, backlight works well. Switches on automatically when it's dark and goes off after some timeout if it's light enough again. Auto power off of the device worked.

Crazy toy.

Best regards,
Martin

[Andrew_Debbie]

Slot in the PCB too.

[Fatih Aydogan]

Uni-t renewed all its models a few years ago and added a probe mount to the back of the devices (photo 2). So is the probe mount useful?

[Fungus]

Impressively weird.

[Avelino Sampaio]

I got it for $52.00 (including shipping) in the AliExpress Anniversary Sale. Its current price is around $60.00.

I will get rid of the ANENG multimeters, which already have problems with the selector switches. What I don't like about the AN870 is that it's not accurate to the last digit, making its 20000 counts useless. I rarely need absolute precision on the least significant digit, but I need it to be minimally functional when needed. The UT89XE has its least significant digit with good accuracy.

[Avelino Sampaio]

Most multimeters cannot read capacitors on the picofarad (pF) scale. The UT89XE is not doing well with values below 1nF. See in image 1, the 47pF capacitor is being read as 14pF. The trick used to perform the correct reading is to add a 1nF capacitor in parallel. Use the relative button to zero the 1nF capacitor reading and then read the capacitor to be verified. See in image 2 the new reading of the 47pF capacitor, with a good accuracy (48pF). In image 3 I have the reading of a 10pF capacitor.

[Avelino Sampaio]

It does not have the Relative option for reading resistors. The solution is to use a short test lead.

Perfect reading for a resistance of 1.74R.

[coromonadalix]

isn't it  the triangle logo  under the peak function ??

https://www.uni-trend.com/uploadfile/2021/0225/20210225113938579.pdf

they do tell its a REL  function ??  section VII

BUT  capacitive measurement ?  muh ?

[Avelino Sampaio]

I went to test the RMS frequency range and I was pleasantly surprised. Using the 20 volt scale allowed me to get the following readings:

Image1: Generator set at 1volt RMS, 1khz and offset at 8v
image 2: In the frequency of 3.3khz the AN870 already reaches the limit of -3db.
image 3: At 10khz frequency the UT89XE still has a very good reading. The AN870 is zeroed.
image 4: At the frequency of 20khz it still offers a reasonable reading.
image 5: At the frequency of 77khz it reaches -3db.

[Avelino Sampaio]

Carrying out new frequency tests, above the effective width of the meter (45 - 400Hz). In the 200mv range I used a 470nf capacitor to block the DC voltage (offset). My goal was to check the bandwidth limit down to -3db.

Image 1: Generator set to square wave 10mv RMS, offset at 5v and frequency at 400Hz.

Observe in the following images, the values obtained for the frequencies 10khz, 20Khz, 50Khz and 75Khz(-3db).

Although there is no RMS accuracy in this frequency range, it can still be useful for checks and repairs in the audio area, to check for ripples (noise) from sources and I believe it can even be used to measure ESR.

[Avelino Sampaio]

Raising the RMS voltage to 100mV, the frequency range increased to 110Khz(-3db). This is very similar to AD736's answers.

Note: The RMS converter used is ES5 from Cyrustek.

[Avelino Sampaio]

As for the frequency meter, below is a table with minimum peak to peak voltage values, to obtain error-free readings.

up to 160khz -------------- 230mVpp
from 161 to 400khz ------ 250mVpp
from 401 to 700khz ------ 270mVpp
from 701 to 1.2Mhz ------ 290mVpp
from 1.2 to 2.7Mhz ------- 310mVpp
from 2.7 to 17.8MHz ---- 330mVpp
from 17.8 to 18.2Mhz --- 350mVpp
from 18.2 to 18.6Mhz --- 370mVpp
from 18.6 to 19Mhz ----- 390mVpp
from 19 to 19.3Mhz ------ 410mVpp
...
at 24Mhz (maximum generator) --- 780mVpp

Note in the image the reading error with the value below 230mVpp

[Gyro]

Slot in the PCB too.

I can't work out what a slot in that position is actually achieving, there's no issue with creepage there.

[Avelino Sampaio]

Follow the link of the DMM-Chipset datasheet: ES289B. There is a good example circuit to study, see in the image.

[Avelino Sampaio]

In the range of 2v (higher resolution) the RMS accuracy is very good, in the limits of the effective region of the meter (45~400Hz).

Image 1: generator set at 1v RMS, 45Hz, sine.
image 2: frequency set to 400Hz, with good RMS accuracy
image 3: at 9.3Khz drop to -3db.

[Avelino Sampaio]

DC current test at 1µA

[Avelino Sampaio]

Bad image but you can get an idea of the components protected by the metal box.

[Avelino Sampaio]

Conductance test:

image 1: First measuring a 30M resistor. Measured value of 30.6M. Calculated value of conductance: 1000/30.6 = 32.679

Image 2: Value read from conductance

[rsjsouza]

Avelino, this is a nice showdown of the UT89XD. It seems to be a much better value than the UT89X that I got a few months ago just to perform some evaluation. Regardless, the mechanical build and display are lightyears ahead of the AN870 indeed, but the durability of the rotary switch is still something that remains to be seen, since a manual range requires much more work from the switch.

[indman]

Avelino, this is a nice showdown of the UT89XD.

UT89XD и UT89XE - These are completely different models in terms of characteristics, do not confuse them.

[rsjsouza]

Avelino, this is a nice showdown of the UT89XD.

UT89XD и UT89XE - These are completely different models in terms of characteristics, do not confuse them.

You are obviously 100% correct 

[Avelino Sampaio]

Today I decided to test if the least significant digit, in the 200m DC range, provides minimally reliable values. This range doesn't have much resolution, going down to 10µV. I used a 50µA current source to generate 20µV and 50µV voltages and the result was that I can trust the reading. The AN870 did well too, despite its last digit being irritatingly restless.

[Avelino Sampaio]

Regardless, the mechanical build and display are lightyears ahead of the AN870 indeed, but the durability of the rotary switch is still something that remains to be seen, since a manual range requires much more work from the switch.

Rafael, the vintage M840D is 22 years old with me and has only a small bad contact of the key, in the range of 2m of the DC current. Mechanically, I trust the Uni-t experience a lot. I'm getting rid of the Aneng gauges, as they have problems with the selector switches. This retro design of the UT89XE won me over and its ergonomics are incredibly good.

I will keep my vintage M840D (2000 counts) in operation as it is in good health and it surprises me how it still has good accuracy. It uses the famous 40-pin ICL7106. I'll leave some images, by way of curiosity for newbies.

[Avelino Sampaio]

Hi

Finally a new quality multimeter (BM869S), purchased at Welectron.
Here's a table I made, comparing different measurements.
Very satisfied with the UT89XE. Its ergonomics are incredibly good.
I already got rid of the Aneng 8002 and soon I intend to get rid of the aneng AN870.

[Martin72]

An Error Has Occurred!
You are not allowed to access this section

Appears when I try to download the table file..