Uni-T UT61E slowly returns to zero in all AC modes. Should I keep it or return?

[Serg65536]

It takes 2 minutes for my multimeter to drop down from several volts (or several hundreds milliamps) AC to 0. Probes are short circuited while waiting. First minute it takes to go down to 0.6mV and second - to actually show zero reading.
Is this normal operation of TrueRMS chip or should I return my new UT61E meter back? 

[siealex]

Are you sure that your equipment does not induce current on the test leads? Instead of short-circuiting the test leads, remove them and short-circuit directly on the meter. Does the behavior stay the same?

[Serg65536]

Are you sure that your equipment does not induce current on the test leads? Instead of short-circuiting the test leads, remove them and short-circuit directly on the meter. Does the behavior stay the same?

Other meters (including UT61B) drop to zero very quickly. No noise sources turned on. Alligator clip leads are good, resistance is stable 0.09Ohm.

[Fungus]

It takes 2 minutes for my multimeter to drop down from several volts

Sounds broken to me.

[mwb1100]

Just to make sure I understand - you're putting the meter in VAC mode and shorting the leads?

My limited testing shows that my "higher end" meters (87V, BM786) tend to take about 10 - 12 seconds to drop down to something around a 1mV reading, and they don't ever get to zero (or at least it takes more seconds that I was willing to wait around for - which wasn't anywhere close to 1+ minutes).  Though this isn't universal - my BM235 drops to zero within a few seconds.

My cheaper, no name meters tend to immediately drop to zero.

If you look at the specs for VAC, I think you'll see something about readings having much less accuracy at below 10% of the range.  For example, the 87V manual says, "Below 10 % of range, add 12 counts".  12 counts in the 600.0mV range of the VAC mode is 1.2mV - and that's on top of the 4 counts specified for the range in general.

The specs for the BM786's 600.00mV range simply say, "Accuracy specified from 10% to 100% of range".  In other words below 60mV all bets are off.  The specs for the UT61E say the same thing.

I suspect that you are right that the TrueRMS functionality has something to do with it, but I honestly don't know.  Note that your UT61B isn't True RMS.

I'm pretty sure there was a thread about this behavior in the last several months, but I couldn't find it.

I was about to say that your UT61E is in good company, but then I realized that your meter starts out at several volts - mine start out at something less than 100mV, including a UT61E I have on hand.  It takes my UT61E almost 10 seconds to drop to 5mV when I short the leads.   How long does it take yours?

I'm not sure whether or not having an open circuit reading that high is an indicator of anything.

[Serg65536]

...
I was about to say that your UT61E is in good company, but then I realized that your meter starts out at several volts - mine start out at something less than 100mV, including a UT61E I have on hand.  It takes my UT61E almost 10 seconds to drop to 5mV when I short the leads.   How long does it take yours?

Thank you for the reply. I've made video comparison with other meters:

[Fungus]

Just to make sure I understand - you're putting the meter in VAC mode and shorting the leads?

My limited testing shows that my "higher end" meters (87V, BM786) tend to take about 10 - 12 seconds to drop down to something around a 1mV reading, and they don't ever get to zero

That sounds about right.

Starting at several volts and taking a whole minute to get under 1mV though? Doesn't seem normal to me.

(I guess it could be if you live under a power line and surrounded by enormous motors...  )

The real question is how long does it take the reading to settle when you're connected to an AC voltage?

[Fungus]

Thank you for the reply. I've made video comparison with other meters:

It looks normal in that video...

What happened to "several volts" and "a minute"?

[Serg65536]

What happened to "several volts" and "a minute"?

Sorry, it took about 50s to go down from 13V AC to 1mV. This time I've used simple 12V AC output wall power supply.
Here is the second video:
https://youtu.be/CMYpBbLL56U

[J-R]

The manual says AC TRMS measurements are only applicable from 10% to 100% of range.

[Serg65536]

The manual says AC TRMS measurements are only applicable from 10% to 100% of range.

I'm not talking about True RMS measurements, I'm talking about any measurements hidden below those decaying numbers. 

[J-R]

The manual says AC TRMS measurements are only applicable from 10% to 100% of range.

I'm not talking about True RMS measurements, I'm talking about any measurements hidden below those decaying numbers. 

All the AC measurements are True RMS measurements, so the 10% minimum would apply.

[2N3055]

TrueRMS chips have filters on output. Otherwise you would see numbers pulsating at low frequencies.
Filter means capacitors, they have to discharge..
There is compromise between lowest frequency you can measure and discharge time.
It is normal. Meter is not broken.

Meters that have limited BW or those that directly sample digitally will not have same behavior.

[Serg65536]

Doesn't seem normal to me.

And here is another glitch: it jumps back and forth continuously if 9V battery is measured in AC mode.
https://youtu.be/wRiDpvmHxmo

[Serg65536]

...
Meters that have limited BW or those that directly sample digitally will not have same behavior.

And what about another glitch: my UT61E jumps back and forth continuously and beeping if 9V battery is measured in AC mode?
https://youtu.be/wRiDpvmHxmo

[J-R]

The manual does not state any AC accuracy specifications below 45Hz.  So applying an input below that frequency is not a valid input for the DMM's AC modes.

[Serg65536]

The manual does not state any AC accuracy specifications below 45Hz.  So applying an input below that frequency is not a valid input for the DMM's AC modes.

No 45Hz or similar limit is listed in my manual. But "no DC component" requirement is there. And frequency <1kHz.
So, to make a conclusion: AD737 is a good True Pain in the A$$ chip! 

Thank you everybody!
The meter remains in my laboratory. I don't know of any other model that measures resistance this fast anyway.

[Aldo22]

No 45Hz or similar limit is listed in my manual.

Yes, it does say that.
Bandwidth (Hz) :45Hz~10kHz

I just wonder if it's true.
I just tested this out of curiosity on the An870 and it measures accurately down to about 8Hz. Below that, the measurement becomes unstable.

[J-R]

Page 48 of the official manual lists the AC V specifications for the UT61E, which cover the ranges from 45Hz to 1kHz and 1kHz to 10kHz.  It also mentions 10 counts of "a residual reading".

Inputs not listed in DMM specifications are typically considered undefined, as in the behavior is unknown.  But you could run tests to see for yourself how it behaves.

[myf]

Hello !

So "Uni-T UT61E slowly returns to zero in all AC modes" but seems (on video) "very quickly (1 sec. or less) go to the right values".

But usual (electronic linear with capacitors and OpAmp) filters have same rise time and relaxation time in electronics, it is not the case on this DMM.

Numeric Kalman filters have an (recursive) linear formula : Result(n+1) = A.Result(n) + B.NewMeasure(n), it does not match for this DMM.

Do you have any idea about the (low-cost) way used in this case ?

Thanks a lot (and excuse my poor english).

F. from France.   

[2N3055]

Hello !

So "Uni-T UT61E slowly returns to zero in all AC modes" but seems (on video) "very quickly (1 sec. or less) go to the right values".

But usual (electronic linear with capacitors and OpAmp) filters have same rise time and relaxation time in electronics, it is not the case on this DMM.

Numeric Kalman filters have an (recursive) linear formula : Result(n+1) = A.Result(n) + B.NewMeasure(n), it does not match for this DMM.

Do you have any idea about the (low-cost) way used in this case ?

Thanks a lot (and excuse my poor english).

F. from France.   

AD737 datasheet explains it in detail..

[Serg65536]

I just tested this out of curiosity on the An870 and it measures accurately down to about 8Hz. Below that, the measurement becomes unstable.

ZOYI ZT219 rebranded as: zotek ZT219, Aneng AN870, richmeters RM870, Zoyi ZT303 (the smaller version).
All of this meters are based on DTM0660
"Maximum display count: 4000/6000 (9999 for frequency and capacitance)" from the datasheet http://www.kerrywong.com/blog/wp-content/uploads/2016/04/DTM0660DataSheet.pdf
So no 19999 real counts on this multimeters is possible. Only through multisample average! This means: you are obliged to wait...
And no real mV resolution also.

[Kleinstein]

Hello !

So "Uni-T UT61E slowly returns to zero in all AC modes" but seems (on video) "very quickly (1 sec. or less) go to the right values".

But usual (electronic linear with capacitors and OpAmp) filters have same rise time and relaxation time in electronics, it is not the case on this DMM.

Numeric Kalman filters have an (recursive) linear formula : Result(n+1) = A.Result(n) + B.NewMeasure(n), it does not match for this DMM.

Do you have any idea about the (low-cost) way used in this case ?

Thanks a lot (and excuse my poor english).

F. from France.   

The fitler with some RMS chips is nonlinear and it takes considerably longer to go down than up. The time visible for the last few digits to go down still looks too long. Because of the relatively large capacitor needed some meters use electrolytic capacitors and these can have rather high dielectric absorbtion on top of the aready slow time constant.
Besides the filter at the RMS chips there is also usually an AC coupling capacitor to measude only the AC part - this capacitor can also give settling, especially in the 9 V to short test.

The lower frequency limit is usually not a brick wall, but a more gradual decay like a 1st or 2nd order high pass. One part is the input capacitor and the other is the fitler at the RMS chip that at low frequencies makes the chip no longer follow exact RMS response (less weight to the peaks). The first thing is usually the accuracy suffering. 

It is normal for analog RMS chips to take some time to settle, especially the step down. The response in the video is still too slow.

Some cheap DMM use digital RMS and this way have nearly instant response. However this usually comes with a limited bandwidth.

[Aldo22]

I just tested this out of curiosity on the An870 and it measures accurately down to about 8Hz. Below that, the measurement becomes unstable.

ZOYI ZT219 rebranded as: zotek ZT219, Aneng AN870, richmeters RM870, Zoyi ZT303 (the smaller version).
All of this meters are based on DTM0660
"Maximum display count: 4000/6000 (9999 for frequency and capacitance)" from the datasheet http://www.kerrywong.com/blog/wp-content/uploads/2016/04/DTM0660DataSheet.pdf
So no 19999 real counts on this multimeters is possible. Only through multisample average! This means: you are obliged to wait...
And no real mV resolution also.

I didn't mean to say that the An870 is better, I was just interested to see if the UT61E really gives up below 45Hz.
I mean, 5Hz below the most famous frequency! 

[Veteran68]

I just tested this out of curiosity on the An870 and it measures accurately down to about 8Hz. Below that, the measurement becomes unstable.

ZOYI ZT219 rebranded as: zotek ZT219, Aneng AN870, richmeters RM870, Zoyi ZT303 (the smaller version).
All of this meters are based on DTM0660
"Maximum display count: 4000/6000 (9999 for frequency and capacitance)" from the datasheet http://www.kerrywong.com/blog/wp-content/uploads/2016/04/DTM0660DataSheet.pdf
So no 19999 real counts on this multimeters is possible. Only through multisample average! This means: you are obliged to wait...
And no real mV resolution also.

Interesting. I had not seen this called out before. I always assumed the 20K count of the AN870 and it's brethren was legit. I have an AN870, have never used it much, as it sits in a niche better served by my higher-end meters (unlike the more compact Aneng/Zoyi 8008/8009 formfactors that are quite handy for most portable/cheap meter uses cases). It also has a rather cheap quality feel for its size, which doesn't instill confidence when handling. But I see it referenced often as a big bang-for-buck meter with a purported accuracy and precision that rivals much more expensive meters. Seems suspect if what you say is true.

[Aldo22]

Interesting. I had not seen this called out before. I always assumed the 20K count of the AN870 and it's brethren was legit. I have an AN870, have never used it much, as it sits in a niche better served by my higher-end meters (unlike the more compact Aneng/Zoyi 8008/8009 formfactors that are quite handy for most portable/cheap meter uses cases). It also has a rather cheap quality feel for its size, which doesn't instill confidence when handling. But I see it referenced often as a big bang-for-buck meter with a purported accuracy and precision that rivals much more expensive meters. Seems suspect if what you say is true.

As far as I understood, Serg65536 didn't mean to say that AN870 doesn't give precise results, only that it calculates them by averaging and therefore you have to wait a little.
Right?

[Veteran68]

Interesting. I had not seen this called out before. I always assumed the 20K count of the AN870 and it's brethren was legit. I have an AN870, have never used it much, as it sits in a niche better served by my higher-end meters (unlike the more compact Aneng/Zoyi 8008/8009 formfactors that are quite handy for most portable/cheap meter uses cases). It also has a rather cheap quality feel for its size, which doesn't instill confidence when handling. But I see it referenced often as a big bang-for-buck meter with a purported accuracy and precision that rivals much more expensive meters. Seems suspect if what you say is true.

As far as I understood, Serg65536 didn't mean to say that AN870 doesn't give precise results, only that it calculates them by averaging and therefore you have to wait a little.
Right?

Waiting aside, I'm pretty sure an averaging of samples is by definition not going to be as precise as a real-time measurement. That's what I was saying, not that it wouldn't be still be adequate for many uses. Just as an averaging AC meter is still useful to those who don't require TRMS measurement.