Author Topic: Uni-T UT171B Teardown (Updated)  (Read 36623 times)

0 Members and 1 Guest are viewing this topic.

Offline MaxlorTopic starter

  • Frequent Contributor
  • **
  • Posts: 565
  • Country: ch
Uni-T UT171B Teardown (Updated)
« on: January 29, 2015, 02:08:35 pm »
Let's have a look at the Uni-T UT171B, one of Uni-T's new high end models: .

First things first: let's take it apart! Btw, I'm an amateur, not an expert, so forgive me if I don't notice everything, or get some things wrong. Helpful pointers are always welcome.
  • The soldering of the through hole components is a bit uneven. There's the one extreme, way too much solder (see that large solder ball directly above the COM hole) on one of the pins holding the 10A fuse, and in some other places (below COM and V holes) the solder doesn't even fully cover the exposed copper. But it's not a problem, the connections are solid.
  • No idea what that chip is, it doesn't have an manufacturer logo on it.
  • Note the LEDs/photosensors around the A and mA jacks, they're for warning you if the leads are plugged into the wrong jacks.



  • The main processor is an EFM32 Cortex-M3, the display processor is separate (and again, contains no manufacturer markings.)
  • The measurement section has its own shield, in addition to the shielding in the case. Nice! Slightly disappointing: one of the shield's tabs isn't soldered down.
  • The white 2-pin connector is for the lipo battery. The PCB also contains two unused pads for a battery connection, which are probably meant for the UT171A.
  • The input protection, which is discussed curiously often here, seeing how it's an electronics and not an HVAC forum, seems solid. The meter claims 600V Cat IV, and since it claims being ETL listed*, I believe it.
  • The input jacks are quite well made. It's a tube design, not the slotted design that Uni-T uses on their cheaper models. The mA and A jacks contain milled slots on the sides for the above mentioned lead detection system. And unfortunate consequence of that is that you have an openining to the inside of case there, meaning that if you were to blow up the meter somehow, you'd get hot gas in your face. The strain relief is excellent, since the metal strips leading to the PCB are slightly bent, so they'll absorb any movement of the jack and not put strain on the solder joint.
  • The fuses are both 1000V rated. The smaller fuse is an 800mA one. Both fuses are accessible through opening the battery compartment.
  • So, the PC interface is bidirectional this time, which means reverse engineering it will require slightly more effort. Still, shouldn't be very difficult.



I'll add more pictures later. For now, some more observations:

The case feels very nice. The rear is rounded, and the bail sits flush in the case when not folded out, so it feels good in the hand. It's a large meter, but thanks to this, it feels very comfortable to hold. The bail is very large, and the is flattened at the bottom so a large amount of rubber makes contact with the desk when the bail is used. This makes the meter very stable with the bail out, even on a slippery surface. :thumbsup:

Underneath the bail there's the regulatory info. The most interesting bit is certainly that it's ETL Listed (4007682). I checked the ETL database, and the UT171B isn't actually in there yet. But since the UT181A is, and the UT171B is very new, I assume it'll appear there soon.

Next up, the display, probably what many of you are curious about. Yes, it really does look pretty much exactly like it does in Uni-T's picture at the top. The contrast is excellent, the black parts bleed no light at all. Unfortunately, the plexiglass covering the display hasn't been given any antireflection treatment, which means that in bright environments, it'll pick up glare which will make the display hard to read. The display does depend on its backlight for readability, it doesn't reflect any light. It's not that bright, but depends on the contrast to work. What this means in practice: indoors it's excellent, it looks a lot better (clearer, higher contrast) than both the LCD and VFD display on various devices around here. If I hold it so the currently overcast sky is reflected in the display cover, it gets quite hard to read though.

There's no noticeable fading when the various display segments turn off or on, it happens instantly.

It's a TN display, so what about viewing angles? There are only small changes in contrast or brightness looking at it from all angles except one: looking at it from above, at between about a 30° and 45° angle, there's spot where you get this color inversion effect that all TN displays seem to have. Now, this thing has no colors but is just white (the green APO and red lightning symbols are created by a piece of colored film on the display,) so it just goes dark. It's a bit unfortunate, but too bad: it seems to be much more common to look at the display from the bottom as the meter is lying on the table, and from that direction the contrast and brightness actually increases slightly.

Let's discuss the battery. It's a 2 cell 1800mAh LiPo. I've done a couple of measurements, the current from the battery in various modes is as follows:
OFF: 25.5uA
Auto-OFF: 7.8uA
ON, max brightness: 35.5mA - 50 hours runtime
ON, med brightness: 21.0mA - 85 hours runtime
ON, min brightness: 13.5mA - 130 hours runtime

Switching between the modes doesn't influence the current by more than 0.5mA, nor does switching on the PC interface. Curiously, in Auto-OFF mode the meter uses much less power than when switched off with the selector switch. I'm guessing this is because in OFF mode, the battery charging circuitry is actually connected to the battery. It doesn't matter though, because even in OFF mode it'll take several to drain a full battery.

The meter switches to low brightness mode after 10 minutes by default (configurable), so for extended measurements and data logging, I'd say 100 hours is what you can expect. If you need more: just disconnect the battery and hook up an external power supply to that white connector on the PCB!

Recharging the battery happens through a very small wall wart that outputs 10V DC on a barrel plug, which then connects to an adapter that covers all 4 input jacks. They do that to prevent you from using the meter with the wall wart connected. Actual charging happens through the COM (-) and mA (+) jacks, so if you don't have the adapter wall wart handy, a lab power supply connected to those jacks will work too.

Now, the leads. They are rubbish. I believe that they fullfil their ratings (Cat IV 600V with the end caps on), but the PCB wire is quite stiff, and the finish on the ends seems very uneven so it's quite difficult to get a solid, stable reading on them in resistance mode for example. The tips are sharp enough, but certainly not the sharpest I've ever seen. Actually touching both tips to a small piece of metal works much better for getting a stable resistance reading.

Last, let's finally turn it on! Actually, I don't have time right now to do thorough testing, so I'll just note some quick points:
  • The display update rate of 5 updates/sec seems about right. Can't tell how quickly the bar graph updates, but it certainly seems fluid I'll test it later
  • I notice not a hint of overshoot doing the on/off tests with the lab power supply. After two readings (so 0.4 seconds!) I have a stable result.
  • Auto-ranging is properly quick... except in AC+DC mode. It takes around 10 seconds to get a reading on my power socket... and it says 225V in both displays, which doesn't seem right. I'll do more tests on that.
  • Continuity... with the provided leads it's hopeless. It's latched, and seems to be medium fast (100-200ms reaction time maybe?) when the leads get a good connection. I'll test with some fluke leads later.
  • The setup menu is pretty barebones: Only six settings are available: default brightness, USB on/off, beep on/off, display brightness timeout, poweroff timeout, wipe memory. I'd have expected a few more options. What is nice though is that you can enter  and exit the setup menu at any time, without having to turn the meter off.

Alright, that's it for now. I'll add more later. If you want me to look out for anything in particular, let me know.
« Last Edit: January 31, 2015, 04:04:10 pm by Maxlor »
 
The following users thanked this post: pxl, dophuc

Offline omgfire

  • Regular Contributor
  • *
  • Posts: 125
  • Country: ru
Re: Uni-T UT171B Teardown
« Reply #1 on: January 29, 2015, 02:36:50 pm »
The white 2-pin connector is for the lipo battery.
Let's discuss the battery. It's a 2 cell 1800mAh LiPo.
Do they place balance circuit inside battery?
 

Offline MaxlorTopic starter

  • Frequent Contributor
  • **
  • Posts: 565
  • Country: ch
Re: Uni-T UT171B Teardown
« Reply #2 on: January 29, 2015, 02:43:13 pm »
Do they place balance circuit inside battery?
I'm not entirely sure there is one, one'd have to cut open the pack to be sure. I'm not willing to do that. I can feel something rectangular under the wrapping though. I don't think it's a cell's individual protection circuit, since it's only one one of the two cells.

If it isn't in the battery, then there is no balancing circuit, seeing how there are only two wires to the board.
 

Offline teslafan

  • Regular Contributor
  • *
  • Posts: 75
  • Country: us
  • Nubie
Re: Uni-T UT171B Teardown
« Reply #3 on: January 31, 2015, 03:00:23 am »
eevblog XL Spreadsheet shows abt. $200 to buy, ebay $351, actual purchase cost = ? Looking forward to further testing.

Thanks
« Last Edit: January 31, 2015, 03:02:36 am by teslafan »
 

Offline MaxlorTopic starter

  • Frequent Contributor
  • **
  • Posts: 565
  • Country: ch
Re: Uni-T UT171B Teardown (Updated)
« Reply #4 on: January 31, 2015, 04:03:45 pm »
Ok, let's continue with the review! First, some exterior pics:



The UT171B comes in a nylon bag. In it you'll find the meter itself, the probes, a temperature probe (beaded K-type), the wall-wart and an adapter for charging the battery, and a CD with the software.



The device is large, but not huge. Compared to two other meters that are popular on these forums, you'll see that it is taller but not thicker, and with the beforementioned rounded back it feels good in the hand.



I've made a series of photos of the display at different angles, with an UT61E next to it for comparison. What you can see is that the UT171B's display has higher contrast, but reflections on the protective cover can be problematic. Also, you can see the color inversion effect.  In the last pictures, I'm standing hunched over the meters to photograph them, blocking a lot of light in the process. Whereas the standard LCD display suffers from this and becomes hard to read, these low light conditions help the EBTN display of the UT171B.



On to some measurements! I have here a chinese voltage standard which I think is accurate to 0.1mV, let's see what the meter says!. In DC mode, the meter is specced at 0.025% + 5 counts accuracy for the 6V and 60V ranges we'll be using.



UT171B -- Reference -- Error
2.4988V -- 2.49894V -- 1 count
5.0015V -- 5.00167V -- 1 count
07.500V -- 7.50014V -- none
10.002V -- 10.0024V -- none

For resistance measurements, I have some precision resistors accurate to 0.05% (Vishay PTF56 and PTF65 series.) The accuracy of the meter for measuring resistance is specced at 0.05% + 2 counts. Unfortunately, those resistors aren't really good enough to determine whether the meter is in spec or not. They will let me observe how the meter drifts over time though. For now, let's just see what the meter measures:

UT171B -- Resistor Value
 49.99Ohm -- 50.00Ohm
1.0002Ohm -- 1.0000Ohm
9.998kOhm -- 10.000kOhm
100.00kOhm -- 100.00kOhm

This is a datalogging multimeter with internal storage for up to 10000 measurements. When you long-press the STORE (Rec) button, you enter datalogging mode. It asks you for a logging interval between 1 and 240 seconds, and then for how long to log for in minutes. That duration is limited to the number of minutes it takes to fill the memory of course. Inputting the numbers happens with the arrow, +/- and OK keys and is easily done.

When it's logging a data point, the [STO] indicator lights up briefly. It also displays the remaining time in the secondary display. If the secondary display is used for another function (e.g., frequency in Vac mode,) the display will alternate every second or so.

Nice: auto power off is disabled in data logging mode, and is the automatically reenabled when datalogging ends.

Not so nice: you can review the logged values in the meter by endering RECALL mode and using the arrow keys. It does not tell you what your logging interval was however, so you need to remember that. The UT171C is better in this regard, since it has a clock and stores timestamps with the values, at least according to Uni-T's website. Unfortunately, this leads to a problem when using the software:



What I've done here is data log for one minute at 1 sample/sec (fastest speed), stopping at approximately 13:52:40. While it's datalogging, the meter keeps sending the currently measured values to the software at about 5 measurements/second, and the software helpfully plots those values. Afterwards, I recalled the stored values using the software, and here's the problem: You can't tell the software what the measurement interval was, it proceeds to assign the current time to the replayed data points, so you end up with 5-ish datapoints per second. Indeed, you can see that in the graph there, those two dips on the right are the replay of the one of the left; they're condensed but shouldn't be!

The workaround of course is to export the tabular data, completely ignore the DateTime column and assign your own time offset using the recording interval you chose (hope you remember it!)

And yes, transferring recorded data is slow. If you need to transfer all 10000 data points, expect to wait about 30 minutes.

Otherwise, the software seems to work. I tried switching through the modes, and it seemed stable and functional. It has various buttons that let you change anything you could change on the meter using the buttons, but it doesn't let you virtually turn the rotary wheel of course. So those red labels in the screenshot are pushable. In case you're wondering, "Value Model" simply means relative mode. It lets you actually enter the offset value manually, which is nice.

The values in the software are updated at display update speeds, this unfortunately includes the bar graph, so you get 5 updates per second for that too.

Btw, the adapter Uni-T supply isn't actually a simple USB/Serial adapter like I expected, instead it shows up as an USB HID device. This makes reverse engineering the protocol slighly more difficult and means you can't just hack up a simple python script to do your own data logging, pity. The advantage is the software just works when you press "Conn", no need to figure out which serial port is the right one. It also means that at this point I see nothing to indicate that the software supports more than one connected meter. Might be possible by connecting one meter first, then starting a second instance of the software and then connecting the second meter. The software does not work under wine.

Some observations about some secondary functionality:
  • Relative mode and Min/Max mode disable autoranging.
  • This meter does averaging in Min/Max mode.
  • Min/Max mode does not increase measurement speed as it does on some other meters. Also, only the secondary display is used for displaying the Min/Avg/Max results, so you lose a digit of precision there.
  • There is no auto hold or delayed hold like on the UT61E, but since there's no overshoot, Min/Max mode can serve as a replacement.
  • Peak hold only works in AC mode and really can only be used for recording AC waveform peaks.
  • The bar graph update rate seems to be 20Hz. A sine of that frequency will result in the bar graph showing a constant value in DC mode.

Ok that's it I think. If there's anything more you'd like to know, let me know.
« Last Edit: January 31, 2015, 10:02:34 pm by Maxlor »
 
The following users thanked this post: dophuc

Offline Lightages

  • Supporter
  • ****
  • Posts: 4313
  • Country: ca
  • Canadian po
Re: Uni-T UT171B Teardown (Updated)
« Reply #5 on: January 31, 2015, 04:21:54 pm »
Thanks for the review and tear down. It is too bad the logging isn't that great. I am not sure I could get used to the reversed LCD but it looks like a very good meter with proper protection.
 

Offline teslafan

  • Regular Contributor
  • *
  • Posts: 75
  • Country: us
  • Nubie
Re: Uni-T UT171B Teardown (Updated)
« Reply #6 on: January 31, 2015, 08:53:10 pm »
Great review, very detailed, not perfect but a very good dmm, IMHO, thanks. :clap: :-+
« Last Edit: February 03, 2015, 01:34:42 am by teslafan »
 

Offline MaxlorTopic starter

  • Frequent Contributor
  • **
  • Posts: 565
  • Country: ch
Re: Uni-T UT171B Teardown
« Reply #7 on: January 31, 2015, 10:18:43 pm »
eevblog XL Spreadsheet shows abt. $200 to buy, ebay $351, actual purchase cost = ?
It cost me $170 on Aliexpress, plus $60ish for shipping via DHL. I would expect that prices will stay below $200 once international distributors add it to their inventory, and prices in china are probably going to drop a slight bit.

It is too bad the logging isn't that great. I am not sure I could get used to the reversed LCD but it looks like a very good meter with proper protection.
Yeah, pity there's no timestamp logging, it also means that it'll be difficult to keep to logging series apart. There are workarounds, like switching to a different mode and storing that as a marker. The main issue with the wrong time intervals in the PC software might get fixed in an update though.

As for the display, felt slightly weird at first, but it's growing on me quickly. The high contrast makes it so much easier to read at a glance, and it gets better as lighting gets worse.

Btw: the meter has a 60 nanosiemens range. Can anyone tell me what the usecase for that is? The applications where high resistance values appear that I know of typically also require high voltage to get a sensible measurement, whereas this meter only puts out 0.5V in conductance mode. It does work, two oscilloscope probes together give a reading of 55nS.
« Last Edit: January 31, 2015, 10:20:24 pm by Maxlor »
 

Offline teslafan

  • Regular Contributor
  • *
  • Posts: 75
  • Country: us
  • Nubie
Re: Uni-T UT171B Teardown
« Reply #8 on: February 03, 2015, 01:36:14 am »
eevblog XL Spreadsheet shows abt. $200 to buy, ebay $351, actual purchase cost = ?
It cost me $170 on Aliexpress, plus $60ish for shipping via DHL. I would expect that prices will stay below $200 once international distributors add it to their inventory, and prices in china are probably going to drop a slight bit.

VG, thanks.
 

Offline nanofrog

  • Super Contributor
  • ***
  • Posts: 5446
  • Country: us
Re: Uni-T UT171B Teardown (Updated)
« Reply #9 on: February 03, 2015, 03:53:42 am »
Thanks for the review.  :-+

I am not sure I could get used to the reversed LCD but it looks like a very good meter with proper protection.
Protections do look good, and I think I could get used to that display rather quickly.  :) But the A variant does offer a standard LCD display. C variant is OLED, so I'd stay away from that one.

Pricing could stand to come down a bit if they really want to compete with Brymen's BM86x series (seeing pricing of ~$320 for the 171B, not $230), but hopefully will settle down once the supply stablizes.
 

Offline EU1

  • Regular Contributor
  • *
  • Posts: 62
  • Country: ua
Re: Uni-T UT171B Teardown (Updated)
« Reply #10 on: February 16, 2016, 01:12:33 pm »
FYI: to enter calibration mode turn the DMM off, press and hold RANGE+STORE+REL and then turn it on.
« Last Edit: March 03, 2016, 12:00:49 am by EU1 »
 

Offline crispy_tofu

  • Super Contributor
  • ***
  • Posts: 1124
  • Country: au
Re: Uni-T UT171B Teardown (Updated)
« Reply #11 on: February 16, 2016, 10:41:12 pm »
Very cool, thanks for the review!  :-+
 

Offline zaoka

  • Frequent Contributor
  • **
  • Posts: 376
  • Country: us
Re: Uni-T UT171B Teardown (Updated)
« Reply #12 on: February 16, 2016, 10:51:31 pm »
UNI-T working on quality thats for sure. I am sure in few years we would have high-quality meters designed and made in China.
 

Offline EU1

  • Regular Contributor
  • *
  • Posts: 62
  • Country: ua
Re: Uni-T UT171B Teardown (Updated)
« Reply #13 on: March 02, 2016, 11:59:56 pm »
Continuity... with the provided leads it's hopeless. It's latched, and seems to be medium fast (100-200ms reaction time maybe?) when the leads get a good connection. I'll test with some fluke leads later.
50ms in the diode test mode, 70ms in resistance measurement mode, which is still not too fast.
Bar graph reacts much faster.
« Last Edit: March 03, 2016, 12:02:51 am by EU1 »
 

Offline papabol_24

  • Contributor
  • Posts: 44
Re: Uni-T UT171B Teardown (Updated)
« Reply #14 on: June 03, 2016, 07:27:31 am »
Hello

Whats the resolution of the resistance measurement? thanks. :-+
 

Offline MaxlorTopic starter

  • Frequent Contributor
  • **
  • Posts: 565
  • Country: ch
Re: Uni-T UT171B Teardown (Updated)
« Reply #15 on: June 03, 2016, 01:28:47 pm »
0.01 Ohms (but with a stated accuracy of 0.5% + 10 counts.)
 
The following users thanked this post: papabol_24

Online joeqsmith

  • Super Contributor
  • ***
  • Posts: 11629
  • Country: us
Re: Uni-T UT171B Teardown (Updated)
« Reply #16 on: June 04, 2016, 12:58:51 am »
Looking at the fused side of the board, the front end layout looks a little better than the 181.   

Offline Ciber SLasH

  • Newbie
  • Posts: 1
  • Country: ru
Re: Uni-T UT171B Teardown (Updated)
« Reply #17 on: December 11, 2017, 06:12:35 pm »
Maxlor
Help please: i need dump U5
 

Offline andyred

  • Contributor
  • Posts: 10
  • Country: hu
Re: Uni-T UT171B Teardown (Updated)
« Reply #18 on: May 28, 2018, 07:36:10 am »
FYI: to enter calibration mode turn the DMM off, press and hold RANGE+STORE+REL and then turn it on.

If you have it then share it, it seems nobody knows about this calibration procedure than you, not even uni-t support or the sellers.
 

Offline samko.pulpan@gmail.com

  • Newbie
  • Posts: 1
  • Country: sk
Re: Uni-T UT171B Teardown (Updated)
« Reply #19 on: September 16, 2019, 05:20:17 pm »
Great review Tahnk you. :clap:
 

Offline EU1

  • Regular Contributor
  • *
  • Posts: 62
  • Country: ua
Re: Uni-T UT171B Teardown (Updated)
« Reply #20 on: October 24, 2022, 07:25:23 pm »
Has anybody noticed uneven display segment brightness on UT171B?
It is worth noting that the display was not wholly uniform when the DMM was new, but now, after 8 years of use (not heavy use), it looks worse.

To run the display test: turn off the meter, press and hold the yellow button, and turn the multimeter on. The test pattern is displayed until you release the yellow button.
Here is how it looks on my unit (on photo it looks a bit better than in the real life):
« Last Edit: October 24, 2022, 07:28:12 pm by EU1 »
 

Offline electr_peter

  • Supporter
  • ****
  • Posts: 1300
  • Country: lt
Re: Uni-T UT171B Teardown (Updated)
« Reply #21 on: October 24, 2022, 07:53:36 pm »
Image suggests that some LCD segments have poor connection on zebra strip. You can try disassembling and cleaning it. But before that, does LCD becomes better when you apply some pressure on LCD? If yes, that zebra strip needs cleaning.
 

Offline EU1

  • Regular Contributor
  • *
  • Posts: 62
  • Country: ua
Re: Uni-T UT171B Teardown (Updated)
« Reply #22 on: October 25, 2022, 12:24:04 am »
There is no apparent effect when I press the display or PCB, but the display assembly is solid thus likely it transfers very little pressure to the zebra strip itself.

After looking through the Internet I found several UT171B photos with similar unevenness.

Below is a photo from a Chinese seller, so I presume that the multimeter in this photo is new.
I would say, DMM on the photo looks similar to mine or just a little better (during measurements the display of my DMM looks not so uneven as it looks during the test when all segments are active).

What is interesting, bright and dim segments pattern on that photo matches the pattern of my display...
So I'm wondering if this is "normal" behavior for UT171B?
 

Offline electr_peter

  • Supporter
  • ****
  • Posts: 1300
  • Country: lt
Re: Uni-T UT171B Teardown (Updated)
« Reply #23 on: October 25, 2022, 06:45:50 am »
Similar segments are dim on the display which strongly indicates issues with connection. Outer plastic probably shields zebra strip from pressure, so you would have to disassemble DMM to check it properly.
 

Offline EU1

  • Regular Contributor
  • *
  • Posts: 62
  • Country: ua
Re: Uni-T UT171B Teardown (Updated)
« Reply #24 on: October 25, 2022, 07:09:45 pm »
I found several photos of new UT171B suffering the same problem. Moreover, exactly the same segments are dim on different units, so I'm wondering does anybody have a UT171B with a properly working display?

I checked the electronic part of the driver and its power supply and it works correctly as designed.
But it looks like 3.6V is too low for this LCD. Likely, due to zebra strip aging the problem worsens over time.
UT171A has exactly the same LCD driver and likely the same zebra strip but another display, and on all photos its display looks uniform.

I didn't remove the LCD yet, because the zebra strip sits correctly in its place and I'm afraid it may become worse if I just refit it. Unfortunately, I don't have a new strip of the same size.
 
The following users thanked this post: Sergius71


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf