how about UT70D multimeter as one of the higher end China made multimeters?

[Kiriakos-GR]

Here is the proof , that the UNI-T does have 0-1000 scale !! 

Congratulations, the bar graph scale does work correctly at one range setting! I'm impressed.

Yes it does         

[saturation]

I'm not sure why you had issues with your Fluke 87V for a simple measurement, my 20 year old 85I reads true today as it did when new per the original calibration data for series 1; I just spent some time checking its calibration with some new equipment I have that was calibrated in March and April this year.  

I saw the 70D video in the past, while it could have just be a 'fluke' if you read the comments on the video, he was not alone.  

There are other favorable reviews elsewhere though, particularly those looking for a low cost data logging DMM, which is helpful if it were reliable and read accurately.   So I have some interest in a meter than works like Fluke but isn't in its price range, but I'm not convinced you'd want to invest time and energy collecting important data with this Uni-T.  I wouldn't hesitate to buy it for low voltage or DC use in the field, but I think you can do that with a cheaper DMM.

The spec sheet is a bit misleading:

http://www.uni-trend.com/UT70D.html

If this is true, for an 80,000 count DMM its accuracy is more closer to 8,000 count.  The last digit has little value.  

For e.g.: specs report 0.05%+10 digits in the 8V range [ but the manual Kiriakos shows that at 8V its really 20 digits.]

8.0000V ~ 7.9960-8.0040,  + 10 digits  ~ high as 8.0050.  That's quite a bit of variation unless you ignore the last 2 digits.  Likewise, the reported accuracy of the ACV and other ranges to a high of 20-120 digits, reinforce the last 2 digits unreliable, at worst.  

The manual page 34 shows what Alm already suggested, its full scale represents the same scale as the digits, so in this example, it would be 8v, and the scale represents 8/10 for every digit.  Luckily most folks use the analog scale as a relative reference or real time trend, rather than to measure a value.

The penultimate voltage range is 800V then finally 1000V, which seems like a unusually small increment given the way a DMM scale is made.  I think this is more to keep in line with the limits of CAT III safety which limits input at 1000V AC or DC, for transients up to 8000V.

FWIW you won't get this issue with an reputable DMM, like a Fluke that you'll have for a lifetime.

I do not have faith  in this "marketing statement" .
especially by having in mind , the negative and poor behavior of Fluke , to any one who got an used meter....

If I was an manufacturer of DMM's, I would use also acceptable "middle cost" FUSE solutions.

Comparing apples with apples ,  the 300$  UNI-T  will have no deference from the FLUKE 87-5 .

Oh yea this fantastic video ... will blame the UNI-T  for a life time ..  

I will have to make my own video , to blame the FLUKE 87-5 about it poor performance as frequency meter ..

I was testing my own tone generator , and I had to amplify the signal with 5W amplifier so to get reliable readings ..
With 0-600mV  input level , it should be hyper sensitive as frequency meter too !!

Or ...
I had to use the output of the tone generator as input on my amplifier ( square signal) , so to have acoustic output,
and another output of my generator ( sine signal ) , as input for the Fluke !!  

Pictures here ..  
https://www.eevblog.com/forum/index.php?topic=629.0

I have 50/50 % , love & hate feelings for the top Fluke , thats why I consider my opinion as less biased.  


    

[Kiriakos-GR]

I'm not sure why you had issues with your Fluke 87V for a simple measurement...........  

For the price that it costs , I have no tolerance with such quirks.  

Or even with the other issue that it does at High resolution , when the voltage gets under the 19.999
it turns normally to the normal resolution { 20.00 },  but I have to drop 4 volts , so to enter again at the high resolution.
Thats not right ... simple as that.  
It should turn to high resolution , just by a millivolt drop from the 20 volts.

If this one costs 500 EUR , and it does not do it right , how much I need to spent for true perfection?   1000 EUR ?  

If so, no one haves the right to blame the economic UNI-T .

[alm]

If this one costs 500 EUR , and it does not do it right , how much I need to spent for true perfection?   1000 EUR ?

Nothing is perfect . I'm sure much richer people have spend a lot more money to find that out.

If so, no one haves the right to blame the economic UNI-T .

But it doesn't hurt to make potential buyers aware. This isn't a 'who has the biggest meter' game, or about who's right or wrong, nobody here has personally designed any of these meters. Everyone just gives his or her opinion, and if you combine them all, you might get something resembling an objective description of the meter.

[Kiriakos-GR]

Well at the price level of the specific Fluke , that is mainly my problem , I have no people to " combine "  their observations ..

Even so, I am starting to believe again at the old Greek principal " That simplicity works always best in anything "

So there is no need , for super complicate instruments , with bugs that no one are able to fix.
 
And before I let this ranting of my, to over for good ,
I will also nag  about the display  of the fluke 87-5 , due the poor viewable angle . ( up - down)
And  the super silent beeper .... all those hard plastics as multimeter case, plus the rubber in between 
the plastic covers ,  plus the rubber holster ,  cause the beeper to sound with so low volume ,
that no one will be able to listen in it ... specially at the famous " industrial environment " .
What the damn FLUKE had in mind about that ? " industrial environment " + low volume beeper !!       
 
Even at my bench I had to stop any noise source and  even the music (radio),  so to listen it .. 
I am truly thinking to drill the Fluke, so to get some sound of it ..

All my other low priced DMM's have an powerful beeper ...  how can I blame them ?
 

[slburris]

I did a few more tests of UT70D, comparing with my Fluke 87-V and Fluke 8505A 6.5digit bench meter.

Using a Malone voltage standard to test with:

http://www.voltagestandard.com/Home_Page_JO2U.html

This is supposed to by 0.01% accurate at 5V.  I got the following
measurements (note I just turned everything on, didn't let stuff warm
up, so your mileage may vary):

87-V: 5.000
8505A: 5.00022
UT70D: 5.008

Hmm, so it does indeed look like the last digit should be ignored
and the manual says it's += 10 digits for accuracy.  Not clear to
me if you could trust the last digit for relative measurements.

On to resistance.  I used a 500K resistance standard, accurate to
+- 0.05%.  Here are the results:

87-V: 500.0K
8505A: 499.93K
UT70D: 498.6K

Other notes:

I don't experience the OL at turn out as shown in the youtube video someone
posted.

The continuity test is only moderately responsive and sounds a little scratchy.
I think some meters pulse stretch the continuity results -- this one does not.

I agree that the analog bar graph is pretty useless for measurements.  It does
update lightning fast though, so I would use it just to see if the input was varying.
Looking at the manual, the graph updates 50 times/sec, and the digital display
5 times/sec.

The foil shielding is more durable than the photo would show.  It's encased in
pretty tough plastic, and is grounded via the ground input lug.

Scott

[slburris]

More measurements.

I connected the Fluke 87-V and the UT70D to line voltage and the results were:

87-V: 119.5VRMS 59.98Hz 50.0% duty cycle
UT70D: 120.0VRMS 59.98Hz 50.01% duty cycle

Next, I generated a 100Khz sine wave at 1Vpp with a signal generator

87-V: 100.00Hz, 46.4% duty cycle
UT70D: 100.01Hz, 50.65% duty cycle

Next I tried measuring two capacitors:

2200uf

87-V: 2261uf
UT70D: failed, checking manual shows this is beyond the range of the meter

0.002uf

87-V: 3.09nf
UT70D: 2.77nf

Scott

[saturation]

Thanks Scott for the updates.  Good to know results particularly knowing of the instruments you're referencing the Uni-T against.  Please keep them coming.

The 8505A is a Fluke DMM?  I think its has rated accuracy much over the 0.01% Malone reference.

On the DCV referenced against the Malone standard, that gives the 70D a DC accuracy of 0.1556%, a bit off from the published 0.05%.  Even with the 10 digits extra, it would rate at 0.06%.  The 87V DC accuracy here is 0.0044%

For the precision resistor, using the 8505A as a reference then the
87V is 0.014%
70D is 0.266%

Note, on the measurements without higher reference device, its hard to know who is off, the 87V or the 70D.  It would be great if you could measure them against the Fluke 8505A.

Caps tend to have high tolerances so what's printed on the device may only remotely be what its true nF is.  Again, best to measure it against a higher accuracy device.

Thanks for the clarification on the foil shield, its well built for it then, my err.

More measurements.

I connected the Fluke 87-V and the UT70D to line voltage and the results were:

87-V: 119.5VRMS 59.98Hz 50.0% duty cycle
UT70D: 120.0VRMS 59.98Hz 50.01% duty cycle

Next, I generated a 100Khz sine wave at 1Vpp with a signal generator

87-V: 100.00Hz, 46.4% duty cycle
UT70D: 100.01Hz, 50.65% duty cycle

Next I tried measuring two capacitors:

2200uf

87-V: 2261uf
UT70D: failed, checking manual shows this is beyond the range of the meter

0.002uf

87-V: 3.09nf
UT70D: 2.77nf

Scott

I did a few more tests of UT70D, comparing with my Fluke 87-V and Fluke 8505A 6.5digit bench meter.

Using a Malone voltage standard to test with:

http://www.voltagestandard.com/Home_Page_JO2U.html

This is supposed to by 0.01% accurate at 5V.  I got the following
measurements (note I just turned everything on, didn't let stuff warm
up, so your mileage may vary):

87-V: 5.000
8505A: 5.00022
UT70D: 5.008

Hmm, so it does indeed look like the last digit should be ignored
and the manual says it's += 10 digits for accuracy.  Not clear to
me if you could trust the last digit for relative measurements.

On to resistance.  I used a 500K resistance standard, accurate to
+- 0.05%.  Here are the results:

87-V: 500.0K
8505A: 499.93K
UT70D: 498.6K

Other notes:

I don't experience the OL at turn out as shown in the youtube video someone
posted.

The continuity test is only moderately responsive and sounds a little scratchy.
I think some meters pulse stretch the continuity results -- this one does not.

I agree that the analog bar graph is pretty useless for measurements.  It does
update lightning fast though, so I would use it just to see if the input was varying.
Looking at the manual, the graph updates 50 times/sec, and the digital display
5 times/sec.

The foil shielding is more durable than the photo would show.  It's encased in
pretty tough plastic, and is grounded via the ground input lug.

Scott

[saturation]

Maybe there is something subtly wrong in your DMM, you did buy it used from eBay as I recall?

https://www.eevblog.com/forum/index.php?topic=354.15

Also in that thread you did a voltage test, but if you have access to such equipment maybe you should do a full calibration across all the ranges.

I'm not sure why you had issues with your Fluke 87V for a simple measurement...........  

For the price that it costs , I have no tolerance with such quirks.  

Or even with the other issue that it does at High resolution , when the voltage gets under the 19.999
it turns normally to the normal resolution { 20.00 },  but I have to drop 4 volts , so to enter again at the high resolution.
Thats not right ... simple as that.  
It should turn to high resolution , just by a millivolt drop from the 20 volts.

If this one costs 500 EUR , and it does not do it right , how much I need to spent for true perfection?   1000 EUR ?  

If so, no one haves the right to blame the economic UNI-T .

[Kiriakos-GR]

Maybe there is something subtly wrong in your DMM, you did by it used from eBay as I recall?

https://www.eevblog.com/forum/index.php?topic=354.15

Also in that thread you did a voltage test, but if you have access to such equipment maybe you should do a full calibration across all the ranges.

I have no true issues with the DMM , thats why I am holding my horses  about any true debate.  

Some quirks , are software related issues , some others are about the design of it,
and about the sensitivity of it as frequency meter ,  I  own an true Japanese frequency meter from LEADER 0~150MHz, that its does not fool around.

Its all about price tag and expectations.
I was expecting more perfection from it, as " package ".

I am not surprised , that the specific UNI-T performs decently .

Even so, when we test DMM's  its best to use the same cables set    
  


  

[slburris]

Bah, my first attempt to post this got eaten!

The Fluke 8505A is a 6.5 digit benchtop multimeter.  For DC volts,
it's rated to be within 0.001% +- 8 and for resistance it is
0.003% += 1. 

As you can see, it's far better than any voltage or resistance standard I
have....

The 87V is rated DCV 0.05% +-1 and resistance 0.2% +-1.  So if all the
math is right, the 87V is doing much better than spec.

For capacitance, I don't have anything to compare against.  The only
equipment I have is the 87V and now the UT70D.  I'd like to get
a real LCR meter, but I haven't yet.  I know Dave likes the Smart Tweezers,
but I'm still looking at the Agilent U1731B for $276 from Newark.

http://www.home.agilent.com/agilent/product.jspx?nid=-34196.920242.00&cc=US&lc=eng

I'll see what other tests I can come up with.

Scott

[slburris]

OK, time for more multimeter abuse.

Probably most people use frequency counters instead of
multimeters to measure frequency, but I ran some more
tests just for fun.

Here's the setup.  I used a Wavetek 288 signal generator
to generate sine waves.  I'm feeding the output to both
the Fluke 87-V multimeter and the UT70D.  As well, it
goes to a Racal-Dana 1998 frequency counter that is
using a Trimble Thunderbolt as a GPS locked 10Mhz external
frequency reference.  I'd say the frequency measurements
of the 1998 will be extremely accurate :-)

The signal generator is set to produce 1Vpp and terminated into
50 ohms.

At 100Hz, the results are:

1998: 100.004Hz (many more digits of accuracy, but not important for this test)
87-V: 100.00Hz
U70D: 100.01Hz

Now at 1Khz:

1998: 1000.046Hz
87-V: 1000.0Hz
U70D: 1000.1Hz

Now at 100Khz:

1998: 100.00455KHz
87-V: 100.00Khz
U70D: 100.01Khz

Now at 200Khz, beyond what the U70D is specified for:

1998: 200.009Khz
87-V: 200.0Khz
U70D: 199.71Khz

Next, I pushed the limits of the frequency to see what the meters
could do.  The U70D is only specified up to 100Khz and the 87-V
up to 200Khz.

The U70D started displaying completely wrong values above about 250Khz
then eventually only displayed zeros.  The Fluke 87-V worked up to about 660Khz,
then displayed only zeros.  The Fluke never displayed any nonsense numbers -- it
either displayed the correct frequency or zero.

OK, going back to 100Khz, I started dropping the amplitude of the signal to see
where the meters would drop out.  The U70D stopped working below 0.93Vpp.
The 87-V stopped working below 0.07Vpp (no, that's not a typo, more than 10X
lower amplitude).

I'm really coming to appreciate what a nice meter the 87-V is.

Scott

[Kiriakos-GR]

Well Scott, it looks that you know, what you are doing, but I need to see pictures,
so to cheer up .  

With words I build castles .. too

The 600KHz one ... and the low 0.07Vpp

[slburris]

You know, I'm going to hold back my opinion until I see the results of Dave's $100 multimeter
test.  I wonder if he's going to test any UniTrends, and if so which ones?

I really wanted a multimeter with capacitance measurement, so that knocks some of the
cheaper meters out for me.  Ignoring that, I kinda like the VC99 for 1/3rd the price of
the UT70D, but Dave really doesn't seem to like that one.

Scott

[EEVblog]

You know, I'm going to hold back my opinion until I see the results of Dave's $100 multimeter
test.  I wonder if he's going to test any UniTrends, and if so which ones?

Yes, the UT61D, it's in the mail.

Dave.

[Kiriakos-GR]

slburris has made a detail testing there............

Sorry but I did no see it ..  other than words.

[alm]

What's the point of pictures? Unless you take great pains to show every single wire and control setting, it contains less information than written text. Either you trust someone to do a correct and unbiased test, and you accept their numbers, or you don't, and no amount of data will convince you. Pictures of the setup are quite rare in science, especially if it's standard equipment.

[saturation]

You can make a simple high pass filter to get a good measure of capacitance, and I presume you have a scope but you can do this with just Vac on the DMM:





http://www.play-hookey.com/ac_theory/hi_pass_filters.html

Since you have a reference signal source, a reference resistor of 500K, and you have an 'unknown' roughly .002 uF cap then fc = 159.15 Hz when the Vacout ~ 0.707 Vacin

Bah, my first attempt to post this got eaten!

The Fluke 8505A is a 6.5 digit benchtop multimeter.  For DC volts,
it's rated to be within 0.001% +- 8 and for resistance it is
0.003% += 1. 

As you can see, it's far better than any voltage or resistance standard I
have....

The 87V is rated DCV 0.05% +-1 and resistance 0.2% +-1.  So if all the
math is right, the 87V is doing much better than spec.

For capacitance, I don't have anything to compare against.  The only
equipment I have is the 87V and now the UT70D.  I'd like to get
a real LCR meter, but I haven't yet.  I know Dave likes the Smart Tweezers,
but I'm still looking at the Agilent U1731B for $276 from Newark.

http://www.home.agilent.com/agilent/product.jspx?nid=-34196.920242.00&cc=US&lc=eng

I'll see what other tests I can come up with.

Scott

[slburris]

I guess I'm not clear on what the pictures are being requested for. 
Nevertheless, here are some shots of everything I've been talking about.

First up, we have the units under test, the Fluke 87-V and the UT70D
sitting on top of the Racal-Dana 1998 counter, currently displaying
the 100Khz signal from the signal generator.

Next, at the bottom of this stack, we have the WaveTek 288 signal
generator set to 100Khz.  The Tektronix 2236 oscilloscope and
Tektronix 1240 logic analyzer are not used in this testing.

Next we have the Trimble Thunderbolt GPS disciplined oscillator feeding
a TAPR TADD-1 RF Distribution Amp (http://www.tapr.org/kits_tadd-1.html).
That's used to distribute the 10Mhz standard to multiple instruments.

[slburris]

More pics,  here we have two Fluke bench meters, the 8505A and the
8502A.

Lastly, we have the 0.05% resistance standard and the 0.01% voltage
standard.

Scott

[slburris]

Can you believe the Racal Dana 1998 was a $60 find on Ebay?  I think
I lucked out on this one.

Counts to 1.3Ghz.  It's not a bad counter by itself, but since it can take
an external 10Mhz reference (like good equipment should), it's much more
accurate than it was originally designed for when fed the output of the
Thunderbolt.  I've also got an HP 5328A counter which can take an external
frequency standard, but it's huge compared to the 1998.

I've seen some cheap 1.3Ghz counters on the market, like the Mastech MS6100,
an 8 digit counter.  I'm not sure how much I believe the specs.  No external
frequency input, so it can't be nearly as accurate as my current counters.
Maybe once Dave's finished destroying every multimeter on the market, he can
take apart some of the counters from China. 

Scott

[Kiriakos-GR]

About the 87-5 , I have move my interest, beyond the common leads measurements.  

More details , in specific thread .  


.

[saturation]

You have very highly capable gear.  What made you consider the UT70D?

Based on your prelim observations, $100 is an acceptable price point, but a most compeling feature against competition in its price range is its data logging RS232 feature.  Hopefully these reading are stable over time to make logging any data worth it.

I raise a skeptical flag as the spec sheet and the simple measured results show it doesn't live up to it, except in the frequency counter.  The good news is these prelim specs are acceptable but not what is publicized as its specs, compare to what Fluke reports vs what you actually measure.

For extra confidence, I'd remeasure it at a later date, and with a variable power supply check the accuracy of the added ranges against the Flukes on both Vdc and Vac to check its linearity, just as you did sweeping up the frequency spectrum, and also check its response against a square wave.

Can you believe the Racal Dana 1998 was a $60 find on Ebay?  I think
I lucked out on this one.

Counts to 1.3Ghz.  It's not a bad counter by itself, but since it can take
an external 10Mhz reference (like good equipment should), it's much more
accurate than it was originally designed for when fed the output of the
Thunderbolt.  I've also got an HP 5328A counter which can take an external
frequency standard, but it's huge compared to the 1998.

I've seen some cheap 1.3Ghz counters on the market, like the Mastech MS6100,
an 8 digit counter.  I'm not sure how much I believe the specs.  No external
frequency input, so it can't be nearly as accurate as my current counters.
Maybe once Dave's finished destroying every multimeter on the market, he can
take apart some of the counters from China. 

Scott

[Kiriakos-GR]

Sorry to see this Kiriakos, but it again suggest your your Fluke is damaged or in need of calibration.

I am not aware of what you think, that you have see ..  
But I will not start explaining here, of how this digital clamp works in detail.

I had spent hours , taking measurements , and learning how to properly use , this new setup.  
The only that I can tell , its that the clamp , has an "Auto Zero" calibration + 1.5% accuracy .
  

[saturation]

Sorry Kiriakos I did not see the picture clearly until I saw your clamp meters.  I removed my message, my apologies!

Sorry to see this Kiriakos, but it again suggest your your Fluke is damaged or in need of calibration.

I am not aware of what you think, that you have see ..  
But I will not start explaining here, of how this digital clamp works in detail.

I had spent hours , taking measurements , and learning how to properly use , this new setup.  
The only that I can tell , its that the clamp , has an "Auto Zero" calibration + 1.5% accuracy .