UNI-T model UT-70D

[mzacharias]

Well, my curiosity got the best of me. I'd been keeping an eye on this model (although with suspicion) because the specs etc looked impressive. When a used one came on eBay for 40.00, (albeit without the PC connection kit) I couldn't help myself. I know Uni-T meters are somewhat controversial in these parts but I thought I'd give my first impressions of the UT-70D. A full review will follow soon.

First, the not-so-good. The continuity check is not latched, but quick. It "barely" will light a standard LED on diode check. The display was obviously designed for dual display but instead uses the smaller upper right display to show the current range selected, not used for any active reading. Maybe a step-up model has the full dual display capability. And the max voltage range is 8000 volts! I haven't tried anything more than about 500, and don't plan to; 8000 would exceed it's ratings anyway.
The operators manual on PDF can be a little misleading in places, notably on the Auto Hold function. It will function as a true auto-hold, though, no problem. The capacitance range is maxxed out at 100uF.
It does frequency measurement to 1 mHz and duty cycle but does not do temperature.

The input jacks are solid cylinders secured to the board by threaded screws and lock washers. Very solid. There are a couple of MOV's and a spark gap for protection, no "blast shield". The fuses are 5x20mm ceramic types, and the whole case must be opened for access to the fuses and battery.

Auto ranging on Ohms range is pretty quick, especially considering it's ranging down through from an 80 megohm range. Screen updating is relatively quick in 8000 count mode, but only about twice per second on 80,000 count mode. That's right - it can supposedly go 80,000 counts at .05% basic DC accuracy.

It's almost identical sized to the 87-5 but the thicker holster makes it "feel" larger. The tilting bail is "OK" but even when not in use the meter is tilted slightly. I really like that.

The function switch turns pretty easily and has a good, distinct, slightly metallic "ball bearing" feel to it, it's really a joy to use - not spongy or ambiguous in any way.

I have tested it's basic accuracy against my new 87-5 and I can report that there is only the slightest variance, about what you'd expect being adjusted on a different day in a different lab.

It does not specifically do microamps but with 80,000 counts on a milliamp range you can get within about 1 digit of what the 87-5 will do.

It uses a 9 volt battery and consumes about 2 mA in normal operation, about 80uA in sleep mode (after 30 minutes). I have not yet tested the level at which the "low battery" indication comes on.
The included probes are rated at 1000v or 20 amps. (!)

I'm not going to say it's "as good as a Fluke" and it's not the heirloom quality of a Fluke but it's obviously a very good, competent multimeter, and all most people would need, at about 1/3 the cost of an 87-5.

I own several very good meters but this will be one of my favorites.

I got a little carried away here. Sorry about that, as I say a full review will follow soon.

[saturation]

Nice, thanks for the mini review.  I don't recall do we have pics of this DMM?

One simple test you can do against the 87V is temperature tolerance; after you've compared the 2 DMM at room temperature, put both in the refrigerator for an hour.  This simulates using it in the outdoors in winter, for colder winters put it in the freezer.

After you remove it from the fridge extensive condensation will occur, and will simulate severe humidity too.

After an hour, remove it and quickly compare its performance again.

For heat, its trickier as it could damage the meter.  You can wait for a full summer day or check your oven under warm, if temp don't exceed 100F or the rated upper limit of your DMMs.  Then let the meters cook for an hour and compare.

Since these DMM are made to be used in the outdoors or so, that simulates the colds of winter and the heats of summer.  If you fear your meter may be damaged by these tests, don't try it.  But this is easily what you pay for in the expensive DMMs even if they are not ruggedized like the Fluke 28II.

[Kiriakos-GR]

I got a little carried away here. Sorry about that, as I say a full review will follow soon.

I do not need an full review , I believe you ... 

The 95% of the people who will get one, they will be totally happy with it.

And when they will come to the level , to need one DMM that looks like fortress ,
or have on it scientific functions like  P-P  ( peak to peak) ,
or is specialized for modern electrical motors with totally different power up options ,
Then .. AND ONLY THEN ... they will get one of the top range.

Enjoy your redhead   

( Shafri where are you ? )   

[PeterG]

Anyone got a UT71E? The specs look good on paper.

Regards

[mzacharias]

Yeah, I've been looking at that one, too (UT-71E). Does power measurements, and at about $1000.00 less than the Gossen. Of course, I don't need that type of measurement in my work as a repair technician. I'm mainly concerned with total  power / current, whether reactive or not. I've never had a power-factor related repair in my career.

[nukie]

Regarding the 40,000 counts UT-71D, I have for a few years now, it's continuity lag a little, on 40000 counts mode, it takes 2-3 seconds for the voltage reading to settle, especially the least significant digit. If you care about that then you probably shouldn't get this meter because after a while the voltage accuracy would drift. Another problem is when you measure a noisy DC voltage, the result is not repeatable. It doesn't happen to my flukes or cheapo meters. I haven't measure the noise yet, I will look into it when I am free. On the other hand, the 4000 counts mode is blazing fast, very impressive. Inside is a nice 5ppm MAX6190 bandgap reference.

70D measures down to 1uV 
71C/D/E measures down to 0.01mV

[Kiriakos-GR]

Another problem is when you measure a noisy DC voltage, the result is not repeatable.

Define what you describe as noisy DC voltage ?

[nukie]

I get results such as
 
0.0271 / 0.0270
0.1936 / 0.1937
0.8910 / 0.8911
1.2485 / 1.2487

Yes it's within specs, but other meters don't do this. Other meters always produce repeatable results.

Another problem is when you measure a noisy DC voltage, the result is not repeatable.

Define what you describe as noisy DC voltage ?

[Kiriakos-GR]

Other meters always produce repeatable results.

I think that you confusing stability over environmental temperature,
with dirty DC .

And dirty DC , can be found only in variable speed controllers  and motors .

About multimeters , their stability as devices are more than good enough for the 95% of the tasks.

If you need to worry about stability ,you should get one current clamp that uses the Hall effect.
The cheap ones =  low stability against environmental temperature , they are like gimmick .
And so you are forced to pay the 250 - 300$ price point,
so to enter in the level that the instruments acts truly stable .

Now imagine that at the 400A or 600A  , the 2% error  =  8 or 12A  of error ,
and this difference it does matters allot but it is acceptable .  

The average, low cost ones,  does  3-5% at an specific and narrow temperature range,
with base the room temperature.
( and this could work , in you have on your own home one large motor at the center of it )    

But I have not find yet, an factory with air condition units over their electrical motors 
And so the temperature range VS the error factor could be 10% or more .        

[nukie]

What has current clamp to do with measuring voltage. We are talking about DC voltage here right?

I also mention, I have a few 0.01mV resolution cheapos($50 range) too, they produce repeatable results. Always spot on same results, accuracy wise is another story. The key word here is repeatability.

If I punch 2+2 in to my calculator I get "4". The next time I punch the same numbers I get "3.92". That's 2% error, is that acceptable? Okay that's pretty bad example because that's not how binary system in a calculator works but I hope you get what I mean. The problem here is how the A/D algorithm was implemented in the UT71 series meter. It's like a meter which cannot decide which way to go. However, the good news is, resistance, current, Farads, produce repeatable results. Okay I am a great nikpick but it doesn't stop me from telling you not to buy it. I see it as a potential meter when I was cheap.

Finally, the power supply is a linear HP E3610A.

For now, I am getting interested in the 80,000 counts UT70 series meter  

[alm]

And dirty DC , can be found only in variable speed controllers  and motors .

Really? So everything else is pure DC? I can drop all those decoupling caps? What about power supplies to high-current digital circuits (eg. a high-performance CPU)? A poorly filtered SMPS? 12V supply in a vehicle while the motor is running? Power supply in airplanes? I believe '28VDC' in airplanes may vary between 16VDC and 40VDC, with all kinds of spikes and dropouts.

I'm sure there are countless other examples.

About multimeters , their stability as devices are more than good enough for the 95% of the tasks.

95% of your tasks.

[mzacharias]

Still working on a review but there is a potential EPIC FAIL on the UT-70D. I was testing the point at which the low-battery indicator came on. Initially I was encouraged that the indication triggered at 6.2 volts. This would indicate a very good useful battery life on this model. On the second test, the indicator came on at 5 volts! After that it wouldn't come on at all! The display dropped out at about 3.3 volts. Naturally I was concerned, so I repeated the test several times but the indicator never came on again. I then tested at what level the UT-70D accuracy started to suffer and discovered it began to lose accuracy at about 7 volts, although not by more than a couple tenths of a volt until the "battery" was under about 6 volts. I suppose there could have been some issue with my methodology versus the UT-70's power monitoring which isn't obvious, but my test was pretty simple. Time will tell, I guess. Will keep you all posted.

[nukie]

It was user error   nothing wrong with the UT-71D input impedance.

In addition to what I have mention previously, the UT-71D also has another pit fall that breaks my day.

When I connect the Fluke 187 to a voltage source of a 47K voltage divider, I could see the voltage drop as soon as the UT71D is connected in parallel. 10M Ohm input impedance??!?!?  That's bullshit specs! That's NOT even turned on. When I turn it on to voltage reading, the voltage on the Fluke 187 drops even more. This does not happen when I connect a Fluke 189.

Can someone with UT-70 series do a similar test? I wish to know if it happens across different 'families'.

Still working on a review but there is a potential EPIC FAIL on the UT-70D. I was testing the point at which the low-battery indicator came on. Initially I was encouraged that the indication triggered at 6.2 volts. This would indicate a very good useful battery life on this model. On the second test, the indicator came on at 5 volts! After that it wouldn't come on at all! The display dropped out at about 3.3 volts. Naturally I was concerned, so I repeated the test several times but the indicator never came on again.

Are you doing this with a power supply? If yes I suggest that you try limiting the current to a few miliamps to mimic a used battery. A freewheeling power supply can deliver more current than a dying battery.

[Bored@Work]

10M Ohm input impedance??!?!? 

Do the math and tell us the result of your calculation:

[alm]

Measuring the input impedance with another DMM should be an easy test. Calculating it from your measurements should also be straightforward once you know the Thevenin equivalent of your voltage source.

[Kiriakos-GR]

And dirty DC , can be found only in variable speed controllers  and motors .

Really? So everything else is pure DC? I can drop all those decoupling caps? What about power supplies to high-current digital circuits (eg. a high-performance CPU)? A poorly filtered SMPS? 12V supply in a vehicle while the motor is running? Power supply in airplanes? I believe '28VDC' in airplanes may vary between 16VDC and 40VDC, with all kinds of spikes and dropouts.

I'm sure there are countless other examples.

When I reply , I do not race against the Wikipedia ... and I just offer an major example.
Thanks for adding additional examples , but I am the one who will get the cigar at the end .. 

95% of your tasks.

Do not get in there,  I am the wacko who had mess up with many and different tasks,
by far more many, than some ones who deals with routine checks.

[alm]

And dirty DC , can be found only in variable speed controllers  and motors .

[...]

When I reply , I do not race against the Wikipedia ... and I just offer an major example.

Emphasis mine.

You may want to look up the meaning of the word 'only' when you have a spare moment.

Thanks for adding additional examples , but I am the one who will get the cigar at the end ..

You can have as many cigars as you want, they're not for people under sixteen anyway .

[Kiriakos-GR]

You can have as many cigars as you want, they're not for people under sixteen anyway .

Thanks mate .. 
Even so give to your father, my congrats for his son.
Your head it does spin nicely about electronics, lets hope that you will not build weapons as grownup !! 

[mzacharias]

UT-70D DC input resistance: 11.13 megOhm direct measured with 87-5

[Kiriakos-GR]

What has current clamp to do with measuring voltage. We are talking about DC voltage here right?

It was an nice example .

FYI  the current clamp as probe ,
it does need to be connected on a very accurate DMM,
as it has an output of mV DC,
and needs also an accurate input resistance of 10MOhm.
The extra counts on the DMM extends the resolution of the clamp it self.

I Just read the last message of mzacharias ,
UT-70D DC input resistance..

If you are correct about it , I can not use it with my clamp probe !! ( and this is an real problem )
The UT-70D it could be programed to compensate the 11.13 megOhm , and show the correct readings.
But my clamp does not , it expects 10MOhm spot on , at the other end.  


 

 

[mzacharias]

I Just read the last message of mzacharias ,
UT-70D DC input resistance..

If you are correct about it , I can not use it with my clamp probe !! ( and this is an real problem )
The UT-70D it could be programed to compensate the 11.13 megOhm , and show the correct readings.
But my clamp does not , it expects 10MOhm spot on , at the other end.  


 

 

I think if you measure any "10 meg" model it will measure the same. The 10 meg rating is nominal.

[nukie]

Opps I posted too early before confirming my finding. The Uni-T UT-71D has a 15.1Mohm when switched off and 11.10Mohm when switched on. The voltage drop I experienced earlier is normal for a 10Mohm meter when connected in parallel. So two 10Mohms in parallel makes 5Mohms input impedance. I ran the test again it also happens to Fluke.

I don't know why the voltage did not drop earlier maybe it was a higher voltage I couldn't remember anymore.

*me reminds himself to measure small voltages with Gohm meter

[mzacharias]

I Just read the last message of mzacharias ,
UT-70D DC input resistance..

If you are correct about it , I can not use it with my clamp probe !! ( and this is an real problem )
The UT-70D it could be programed to compensate the 11.13 megOhm , and show the correct readings.
But my clamp does not , it expects 10MOhm spot on , at the other end.  


 

 

I think if you measure any "10 meg" model it will measure the same. The 10 meg rating is nominal.

Fluke 87-5 input resistance measured on UT-70D is 11.10 meg. Measured on a Tektronix TX-3 is 11.13 meg. The Tektronix input resistance is 11.00 meg measured using the 87-5. I think any nominal 10 meg input impedance meter should work fine with your current clamp.

[Kiriakos-GR]

About the input impedance :

The truth are that I got the Fluke 87V first = Good quality gear ,
and then ..  I got the expensive current clamp probe ( AC/DC ) , and even the 40KV HV probe.

As I was and still do , feel some how safe ,
that in such professional equipment, the called impedance compatibility, it is given from the start.

Today, instead of the 87V , I have the 28II , that serves best, my own working style. (out door use )
And it is identical in the electrical specs with the 87V.

After getting the 87V ( and in the six months that I had it ) , I got also three Fluke bench type DMM's,
and the one that I favor most is the 8050A due the high resolution.

I did some measurements today , just to confirm for one more time ,
that my DMM's are spot on.

Picture No 1 :
The 8050A is connected with my super accurate resistors decade box with 999K as load,
at the 20MOhm range ..
( I have no idea when ever got calibrated as last time )
Even so , I got an almost spot on reading.  

Picture No 2:
The 8050A measures now the  impedance of the 28II ,
at the same range that it was tested with my decade box.
And the 28II is set to DC mV range , the range that I use with my current clamp probe.
Words looks poor , when the picture speaks by it self.  
 
(8050A 20MOhm  + / - 0.25% +3 digits )  




      

[Kiriakos-GR]

Low Battery indicator problem

My opinion : remove this video from youtube before you become an laughing stock.  

2) Send one email to Uni-T and ask one simple question,
like :  Please tell me how often the device performs an battery check , I need to know the time interval.  
Does it check the battery condition only at boot ?  

3) As soon you got the new information , make a video that confirms it.
And avoid the FAIL description ,  because currently you are the one who failed to test properly the device.

This is my opinion, and nothing will come out by arguing about it .
If it works best for you , just ignore it.

Edit: The message that I was replying to, it got deleted. ( Wise move )