Handheld meter robustness testing

[gnavigator1007]

Really sad to hear these tests or at least this thread may be coming to an end. I'm mostly a lurker around the forum, but this has been one of my favorite threads for a long time. I'd personally hoped to see many more meters run. Perhaps as time goes on joeqsmith will find more he is interested in running or someone else will take it upon themselves to do more tests. One of the things I've always appreciated about Joe is his encouragement of others to run similar tests for themselves. I only wish I had the time, money, and honestly the skills to consistently do such methodical testing, but alas I'm a lowly hobbyist that is already outclassed by some of the equipment I own and too many gaps in knowledge to be beneficial to anyone. It has been fascinating, educational, and I'll forever be grateful to Joe for having shared this bit of his hobby with us. I have so much respect for him and the extent to which he has gone to satiate curiosity. It is frankly amazing.

[MacMeter]

I bought two of those Sanwas from Dave. Dare you to give one the full joeqsmith TREATMENT!

[malagas_on_fire]

Well This thread is very popular due to the various tests on the meters, methods implied, documentation provided, some funny stuff.

Sanwa's are going to the robustness test? do you're best

Can i post the results using the modeled  dummy 9V battery for current draw and voltage shut down? this time using this setup, not sure if it is  the best :

- Power supply is a homemade Joy-it DPS5005 ( http://anleitung.joy-it.net/?goods=dps5005 ) box powered by a 12V SLAB battery, setup 9V to and current limit of 100mA (maybe too low ? ). The voltage will be ajusted to check the shutdown point of DUT ( uni-t ut204A , the battery hogger) and thus ignoring the battery indicator
- Measure of voltage shutdown point will be measured with multimeter to confirm with the power supply ( rounded to 1 decimal place)
- Current will be measured with another multimeter in series with power supply and one of the terminals of dummy battery.
- All connections between power supply , current measurement and dummy are banana jacks 

[windsmurf]

I'm enjoying watching the thermal cycling videos.
Interesting how the different meters seem to react differently to temperature changes. 
Owon seemed to do poorly... wonder if that has anything to do with potential condensation on the board or components?

Also now I wonder what do those specs mean? 
DC Voltage, 869s, 500.00mV, 5.0000V, 0.02% + 2d

I saw the Brymen start at around 0.998mV, then go up to maybe 1.016mV, then down to 0.992mV

To calculate 0.02%, do you multiply that against the range? 500.00mV x (0.02/100) = 0.1mV?  Since the display shows down to 1uV, does "2d" mean + 2uV?
If measuring 1.000mV, then would it be in spec if the Brymen 869s showed anything between 1.102mV and 0.898mV?

[HKJ]

Also now I wonder what do those specs mean? 

I explain about multimeter specifications here: https://lygte-info.dk/info/DMMTolerances%20UK.html
Usual they are only valid between 18°C to 28°C, for each degree outside you must add some more tolerance.

[windsmurf]

Also now I wonder what do those specs mean? 

I explain about multimeter specifications here: https://lygte-info.dk/info/DMMTolerances%20UK.html
Usual they are only valid between 18°C to 28°C, for each degree outside you must add some more tolerance.

Thank you, that's a nice explanations page you developed!

So, then for the Brymen 869s in the video, specs for the 500.00mV range is:
DC Voltage, 869s, 500.00mV, 5.0000V, 0.02% + 2d

And since this spec is for 50,000 count range, I think I need to drop the last digit from the 500,000 count readings shown in the video (is this correct?).

Since Joe is measuring 1.00mV (assumed), according to the spec, readings should be:
1.00mV x 0.02% = 0.0002mV, and add 2 counts of 0.01mV and you get 0.0202mV, or for the meter 0.02mV.

So to meet specs, the meter needs to show something between 1.020mv and 0.980mV... is this right?

[HKJ]

So, then for the Brymen 869s in the video, specs for the 500.00mV range is:
DC Voltage, 869s, 500.00mV, 5.0000V, 0.02% + 2d

And since this spec is for 50,000 count range, I think I need to drop the last digit from the 500,000 count readings shown in the video (is this correct?).

Since Joe is measuring 1.00mV (assumed), according to the spec, readings should be:
1.00mV x 0.02% = 0.0002mV, and add 2 counts of 0.01mV and you get 0.0202mV, or for the meter 0.02mV.

So to meet specs, the meter needs to show something between 1.020mv and 0.980mV... is this right?

Yes, but again only in a limited temperature range, Joe goes well outside that range.

[joeqsmith]

From Page 17 of the BM869s manual:

Electrical Specifications
Accuracy is +/-(% reading digits + number of digits) or otherwise specified, at 23C +/- 5C &
less than 75% relative humidity.

Keep in mind that during the transition, all of the components may not track.  That difference can cause a fair amount of error and is why the meter is allowed to stabilize for a half hour.   There is a fair amount of air movement but even a half hour may not be enough to some of the meters to settle.   

Why push them way outside their rating?  I was asked this early on.  Pretty much the same reason I transient test them to failure. 

[malagas_on_fire]

Also remember in manual that the bigger temperature ranges refers to working and storage conditions

In the videos spotted some condensation which can increase humidity and could add up some digits, but brymen and uni-t are very stable in that test, so not  huge impact indeed.

Thanks for the tests provided. Did one of the meters  have become drifted after the test, for example the owon B41?

[joeqsmith]

Thanks for the tests provided. Did one of the meters  have become drifted after the test, for example the owon B41?

They all returned to normal after a day.

[windsmurf]

Yes, but again only in a limited temperature range, Joe goes well outside that range.

Yes I understand that... and yet the meter stayed in-spec regardless of the out-of-spec temperature (and condensation) condition, which is impressive.  I wonder if Flukes do as well... I'm guessing probably the industrial ones will (or maybe all?). 

[malagas_on_fire]

Thanks for the tests provided. Did one of the meters  have become drifted after the test, for example the owon B41?

They all returned to normal after a day.

Thats good news ohhh and the little pocket brymen / amprobe still functional for that testing... eehehehe ? That had a bad time for a lot of time Some of you're tests change  decision on the meter for develop things with triacs The little blue one .... otherwise would be ut61E

[joeqsmith]

Yes, but again only in a limited temperature range, Joe goes well outside that range.

Yes I understand that... and yet the meter stayed in-spec regardless of the out-of-spec temperature (and condensation) condition, which is impressive.  I wonder if Flukes do as well... I'm guessing probably the industrial ones will (or maybe all?).

Old Fluke 189

[malagas_on_fire]

Ohh crap was  the leads or input jacks ? Or is the default error message for anything bad ?

[malagas_on_fire]

Thanks for the reply joe. The meter is quite interesting from the manual perspective and also pdf with text is available.
 
Lets see how other fluke's perform... .. 87V , 17B+ and 115 if they are on the menu

[windsmurf]

Ohh crap was  the leads or input jacks ? Or is the default error message for anything bad ?

Leads or Fuse error message pops up if you're not in the amps range dial with a lead in the amps input, or you're in the amps range and you have no lead in the amps input or you have a bad fuse.
I get that when I clean the input jack on my 189 with a Q-tip with too much rubbing alcohol, until it dries out.  I think the condensation shorted the amps range input sensor.

Its interesting that the 189 specs have a very wide operating temperature range of -20°C to +55°CC, and even gives you a temperature coefficient for measurement at <18 °C or >28 °C.      Such detail is probably one of the reasons Fluke has become a standard in U.S. Industry.   
It did seem to fall out of the 20 count spec for some duration... I'm guessing due to the condensation? I wonder what parts are affected by it... I'm guessing an analog part like the PTC?  Or more generally some current leak on the board in the analog input section.  I bet some insulating coating on the board can eliminate the condensation effect.
 

[joeqsmith]

It did seem to fall out of the 20 count spec for some duration... I'm guessing due to the condensation? I wonder what parts are affected by it... I'm guessing an analog part like the PTC?  Or more generally some current leak on the board in the analog input section.  I bet some insulating coating on the board can eliminate the condensation effect.

Keep in mind that during the transition, all of the components may not track.  That difference can cause a fair amount of error and is why the meter is allowed to stabilize for a half hour.   There is a fair amount of air movement but even a half hour may not be enough to some of the meters to settle.   

Put two thermometers outside, one in a glass of water the other in open air.  One will respond slower.  This delay can cause a fair bit of error and again is why I let them settle.       

In the mV range the input impedance of a crap meter is a Meg.  The PTC is about 1.5K.  Say the PTC changes 100%.  What's the total error it causes? 

[windsmurf]

Put two thermometers outside, one in a glass of water the other in open air.  One will respond slower.  This delay can cause a fair bit of error and again is why I let them settle.       

In the mV range the input impedance of a crap meter is a Meg.  The PTC is about 1.5K.  Say the PTC changes 100%.  What's the total error it causes?

That makes sense... so definitely nothing to do with the PTC.  So in your opinion, do you think the error is caused more by uneven temperatures of the components, and less by condensation's effects on the board?   

[windsmurf]

Old Fluke 189

Hi Joe I just noticed the low battery indicater is on, on the 189. 
That gets me thinking... are battery outputs changing with the temperatures, and how much of an effect that plays in the readings...

[joeqsmith]

The 189 predates the use of NiMH.  The lower voltage causes this meter, along with my camera and a few other devices to display a low battery.   This particular 189 will display a low battery at 4.89V.  At 4.5, it  will begin to flash and at 4.0 it cuts out.  With a mV applied, it will not change a count all the way down to cutout.   The batteries installed are EBL 2300.  The current voltage is 4.865.    So, no, this is not a problem.   

So in your opinion, do you think the error is caused more by uneven temperatures of the components, and less by condensation's effects on the board?   

When it comes to trying to solve problems, I find my opinions mean very little and data is pretty much everything.      There are a few easy experiments that could be ran if we needed to sort it out.   The UT61E for example, drifted far worse.  It may have been the worse I have seen.  For less than the price of a coffee, I was able to tame it.  Then again, I wasn't considering my time having a cost.  If I did, the UT61E I have is worth a bit more than a standard one...   

[windsmurf]

When it comes to trying to solve problems, I find my opinions mean very little and data is pretty much everything.   

Yes but I'd image you'd have an inclination based on your expert knowledge.   

There are a few easy experiments that could be ran if we needed to sort it out.   The UT61E for example, drifted far worse.  It may have been the worse I have seen.  For less than the price of a coffee, I was able to tame it.  Then again, I wasn't considering my time having a cost.  If I did, the UT61E I have is worth a bit more than a standard one...

For our edification, may we know what you did to correct the UT61E drift issue?   

[joeqsmith]

It's good to know that the TOC was a waste of time. 

[windsmurf]

It's good to know that the TOC was a waste of time.

You'd mentioned this in many of your videos (also listed on post #1), so I guess it's an unbending rule .   

[windsmurf]

It's good to know that the TOC was a waste of time.

You'd mentioned this in many of your videos (also listed on post #1), so I guess it's an unbending rule .   

My bad I thought you were referencing your FAQ about not disclosing many of your modifications.... you actually do have a video on the thermal stability mod 

[windsmurf]

...A few good articles if you are interested:

http://www.grainger.com/content/safety-digital-multimeter
http://cp.literature.agilent.com/litweb/pdf/5990-4578EN.pdf
http://faculty.riohondo.edu/jfrala/fluke_multimeters_-_abcs_of_multimeter_safety_multimeter_safety_and_you_application_note.pdf

Just FYI the 2nd and 3rd links didn't work for me.
I'm not sure if this is the same info I would have found in the 2nd link:
http://literature.cdn.keysight.com/litweb/pdf/5992-3426EN.pdf

I think this is the same pdf as the one pointed to from your third link:
https://content.fluke.com/promotions/promo-dmm/0518-dmm-campaign/dmm/fluke_dmm-chfr/files/safetyguidelines.pdf