Handheld meter robustness testing

[joeqsmith]

The Keysight has been running with the two sample rates without a single glitch.  However, I think I now know how to replicate the problem at will.   I need to do a little more work to get some proof but my confidence is increasing.

[joeqsmith]

Now we are getting somewhere.   

[Trader]

I got to run it on the signal generator; up to -30dBm I got about 19MHz; at at -20dBm I got 25MHz, -10dBm I got about 27MHz and at 0dBm I got 31MHz.

Same here, 31.000MHz is the max freq with accuracy I got it.

[Trader]

We are using a 40 ohm resistor and the meter sends up  something like: +4.00108000E+01.  There's no missing data that would get us to what we are seeing.    So far it appears random and can be reproduced without Keysight's software, drivers or interface cable.   

I changed the test program to switch between two sample rates every 10 samples to see if that has any effect.   

Personally, I blame this problem on those that voted for this meter!!  You knew dam well I wouldn't just unbox it and give it five stars!!   

Joe and Defpom videos showed that this DMM is so damn good and accurate, I really don't care about the 31MHz limitation, this is a secondary feature, just to compare, the Fluke 289 is only up to 1MHz, 1Vpp.

That Log glitch maybe could be fixed via Firmware, not a big deal since most people only use the Max/Min for a couple of seconds or minutes.

BTW, Congrats to Defpom, awesome tests, try to get a Fluke 289 to do the same tests.

[joeqsmith]

We are using a 40 ohm resistor and the meter sends up  something like: +4.00108000E+01.  There's no missing data that would get us to what we are seeing.    So far it appears random and can be reproduced without Keysight's software, drivers or interface cable.   

I changed the test program to switch between two sample rates every 10 samples to see if that has any effect.   

Personally, I blame this problem on those that voted for this meter!!  You knew dam well I wouldn't just unbox it and give it five stars!!   

Joe and Defpom videos showed that this DMM is so damn good and accurate, I really don't care about the 31MHz limitation, this is a secondary feature, just to compare, the Fluke 289 is only up to 1MHz, 1Vpp.

That Log glitch maybe could be fixed via Firmware, not a big deal since most people only use the Max/Min for a couple of seconds or minutes.

BTW, Congrats to Defpom, awesome tests, try to get a Fluke 289 to do the same tests.

    Outside of a basic checkout,  I haven't done anything with it.   I would hope all products would meet the manufacture's claims.  In the case of the U1282A, we are already seeing there are problems.   Is their lack of attention to detail at this level an indicator of what's to come?   We know from the last Keysight meter I looked at, their choice of materials led to a spring failure.  The input protection was also on the weak side compared with other products. 

I won't pretend to speak for the majority  but certainly agree that for you personally, much of what I show may not be a big deal.   I would hope the majority buy products that meet their personal needs and based those decisions on the specifications the manufacture supplies.  My goal is to go beyond their specs and push the meters to failure to get some idea how electrically robust they are.   Again, may hold little interest to you personally but there are a few of us who like to see how the products compare on a level playing field. 

[Trader]

My goal is to go beyond their specs and push the meters to failure to get some idea how electrically robust they are.

I follow your videos, yes, I have an interest in these minimal details and in the electronic robustness and safety. I already have "an idea" (based on other videos and 7+ years on market) that the U1282A will be good on these tests.

I just think that some issues don't deserve so much punishment or a brand judgment if those features are not so used or important. I didn't see anyone judging the entire Brymen company just because of the issues in the 869s, or the entire Gossen Metrawatt because of the Critical Failure in the M248B, even high-end Flukes have some issues.

Rarely someone would buy a DMM just because can measure 100MHz frequencies. The log glitch is more important, maybe could be fixed by firmware.

I hope that Keysight uses your tests to improve their product, thank you so much for your efforts, and keep moving forward with the new tests.

[Fungus]

I'm not sure where these Keysight meters fit into things.

If I wanted a really "analytical" meter I'd go with the Fluke 289. It shows a lot of information on screen and doesn't seem to do anything badly.

If I anything over 100kHz (or even 20kHz...) I'd probably be using an oscilloscope, not a multimeter.

For anything under 100kHz I'm not sure why anybody would pay 3x more than a Brymen.

That's just me though. What would the owners say is the killer feature of these meters? Why did you buy one?

(not trying to derail the thread, I promise I'll sit and listen quietly)

[Trader]

I'm not sure where these Keysight meters fit into things.

If I wanted a really "analytical" meter I'd go with the Fluke 289. It shows a lot of information on screen and doesn't seem to do anything badly.

If I anything over 100kHz (or even 20kHz...) I'd probably be using an oscilloscope, not a multimeter.

For anything under 100kHz I'm not sure why anybody would pay 3x more than a Brymen.

That's just me though. What would the owners say is the killer feature of these meters? Why did you buy one?

(not trying to derail the thread, I promise I'll sit and listen quietly)

I'm not sure if Fluke 289 has more features than Keysight U1282A, but both complement each other.

U1282A 4-batteries last more than 1000 hours, the 289 6-batteries won't last 10% of that.

This question also can be applied to some Gossen's, Chauvin Arnoux (Metrix), and even Greenlee, Voltcraft, etc...

ps.: Discontinued Fluke 189, 4-batteries last only 72 hours (maybe 1-week of work, LOL).

[Trader]

I also recommend these 2 excellent reviews of Keysight U1282A:

https://community.element14.com/products/roadtest/rv/roadtest_reviews/109/roadtest_the_keysigh_6

https://lygte-info.dk/review/DMMKeysight%20U1282A%20UK.html

[HKJ]

I'm not sure if Fluke 289 has more features than Keysight U1282A, but both complement each other.

To compare features on high-end meters I made this: https://lygte-info.dk/info/DMMHigh-end%20UK.html

[Fungus]

U1282A 4-batteries last more than 1000 hours, the 289 6-batteries won't last 10% of that.

Datasheet says "up to 800": https://www.keysight.com/us/en/assets/7018-04867/data-sheets/5992-0847.pdf

What I mean by "analytical" is things like min/max measurements. The 289 will show min/max/actual on screen simultaneously (or min/max/avg). That seems a lot more useful than pressing buttons to see each one.

Look at the image you posted - the 289 just shows the info much better:

[rsjsouza]

Look at the image you posted - the 289 just shows the info much better:

It is a tradeoff, where display complexity brings convenience at the expense of battery life and contrast (at least on the Fluke). The U1282A is closer to your typical portable multimeter (7-segment numerical display) than the graphical ones, so it is not necessarily a detriment to it.

The fact it has the interface cable included (expensive options on Gossen, Fluke and perhaps others) and a multitude of logging and triggers is one of its raison d'étre. Also, it is mechanically quite robust (Dave went through a canyon with it) where I suspect the other two would fail miserably (understandable, and they were not designed for such conditions). Add to this the external remote control, the waveform output, the 30+MHz frequency measurement and you can get a fit for it.

[Trader]

Datasheet says "up to 800": https://www.keysight.com/us/en/assets/7018-04867/data-sheets/5992-0847.pdf
Look at the image you posted - the 289 just shows the info much better:

Datasheet says 60k counts, but is 66k.  Battery by sure is more than 1000 hours, ai never replaced the original batteries in 6 years (but I'm not sure how many hours I used it, I have other DMMs too).

I said: they complement each other very well. That's a particular screen about mV and Hz, other screens are different, and you can do some setup.

Some people like the Continuity/Resistance Tone, which is practical.

I'm not saying to buy a U1282A, there are cheaper and very good DMMs like Brymen 869s, but, is nice to have and work with a U1282A, the accuracy is astonishing.

[Trader]

Maybe HP (Agilent, Keysight) is good in making low power consumption and very (for decades) accurate DMMs.

http://powersupply.blogs.keysight.com/2012/06/battery-drain-analysis-of-handheld-hp.html

[AVGresponding]

We are using a 40 ohm resistor and the meter sends up  something like: +4.00108000E+01.  There's no missing data that would get us to what we are seeing.    So far it appears random and can be reproduced without Keysight's software, drivers or interface cable.   

I changed the test program to switch between two sample rates every 10 samples to see if that has any effect.   

Personally, I blame this problem on those that voted for this meter!!  You knew dam well I wouldn't just unbox it and give it five stars!!   

Joe and Defpom videos showed that this DMM is so damn good and accurate, I really don't care about the 31MHz limitation, this is a secondary feature, just to compare, the Fluke 289 is only up to 1MHz, 1Vpp.

That Log glitch maybe could be fixed via Firmware, not a big deal since most people only use the Max/Min for a couple of seconds or minutes.

BTW, Congrats to Defpom, awesome tests, try to get a Fluke 289 to do the same tests.

The 289 is a pretty old design by now, and I'm not sure it's really the direct competition for the 1282. Maybe the 87V MAX (that always makes me think of an old Yamaha muscle-bike) would be more appropriate?

Re: Rotary switch feel comparisons, I don't have a 1282 but I do have a 1273, and though it isn't terrible, it's not as nice as a Fluke 87V or 289. I'd also agree that the 869S is a bit stiff, and mine is far from new...
The best feeling HPAK rotary switch I've used is on my Agilent U1401B, which wasn't even made by Agilent... My old HP 971A had a better feel to the switch action, though it's less comfortable to use as it's small and excessively squared off IMO. Surprising how they can get something so fundamental so wrong.

[Trader]

When measuring VAC, I like that shows big numbers for Voltage and small for Frequency.

Keysight and Fluke are brands built over decades, even today we can see very used and abused Flukes and HPs with perfect accuracy, with no worries about calibration.

But, as I said, it's nice to have and use a Keysight U1282A, but there are other high-end and cheap DMMs.

[AVGresponding]

When measuring VAC, I like that shows big numbers for Voltage and small for Frequency.

Keysight and Fluke are brands built over decades, even today we can see very used and abused Flukes and HPs with perfect accuracy, with no worries about calibration.

But, as I said, it's nice to have and use a Keysight U1282A, but there are other high-end and cheap DMMs.

Just to say that my used (and presumably abused) 1273A does have an accuracy issue on DCV and ohms ranges, which I have not yet tracked down (no visible damage, only got as far as ruling out the protection circuitry to date). So they are not indestructible by any means...

[Trader]

So they are not indestructible by any means...[/b]

I didn't say "indestructible" or "perfect", All products could require Repair services, the difference is the Frequency and Percentage of defective products.

Suppose a company buys 10 DMMs, after 3-years, we can see how many will be inoperable, and I guess Brymen won't last like Flukes and Keysights.

I'm sorry about your problem.  The U1273AX has an interesting feature "-40 to 55 °C operating temperature", this is useful for outdoor work in Cold and Hot places, very few DMMs can operate in that range.

Fluke 87V, 289; Keysight U127XA, U128XA is "-20°C to + 55°C" and Altitude 3000m.

Brymen 869s is only "0°C to 45°C" and Altitude below 2000m!!! LOL

[AVGresponding]

So they are not indestructible by any means...[/b]

I didn't say "indestructible" or "perfect", All products could require Repair services, the difference is the Frequency and Percentage of defective products.

Suppose a company buys 10 DMMs, after 3-years, we can see how many will be inoperable, and I guess Brymen won't last like Flukes and Keysights.

I'm sorry about your problem.  Your DMM has an interesting feature "-40 to 55 °C operating temperature", this is useful for outdoor work in Cold and Hot places, very few DMMs can operate in that range.

Fluke 87V, Fluke 289, Keysight U1282A is "-20 °C to + 55 °C".

Brymen is only "0°C to 45°C"!!! is a pussy for internal use only LOL

It's a bigger operating temperature range than me, that's for sure!

I will get around to fixing it (I suspect one of the MUX's), it would be a shame not to after I went to the trouble of replacing the OLED display, which looks very nice btw.

[Trader]

For anything under 100kHz I'm not sure why anybody would pay 3x more than a Brymen.

I was rechecking the Brymen 869s specs and seem lower compared to Fluke 289 and Keysight U1282A.  It's more comparable to Fluke 87V and Keysight U1272A:

https://static.eleshop.nl/mage/media/downloads/Brymen_867_869_manual.pdf

[joeqsmith]

Sorry for the late reply. It's been a long day
....
Thanks for sharing these. The glitch looks similar to what I experienced before on a Keithley 2700. Other than that the drift doesn't look so bad. Definitely better than my Gossen calibrator. Did you connect the resistor directly to the banana socket?

I ran the battery life test tonight and as promised, I tried it in both DCV and resistance modes.   With the 40ohm load, I was able to lower the battery voltage until the meter turns off without any change in the measurement.  The same was true with a DCV source.    I have seen this be a problem before so didn't hurt to check.   

I have some other tests I want to run before we start the destructive testing so if there is anything else you would like me to look at, let me know.   

[Chance92]

I ran the battery life test tonight and as promised, I tried it in both DCV and resistance modes.   With the 40ohm load, I was able to lower the battery voltage until the meter turns off without any change in the measurement.  The same was true with a DCV source.    I have seen this be a problem before so didn't hurt to check.   

I have some other tests I want to run before we start the destructive testing so if there is anything else you would like me to look at, let me know.

The stability of this meter looks pretty good. Thanks for taking the time to do these tests. Have you figured out the reason behind the glitch?

[Fungus]

Battery by sure is more than 1000 hours, ai never replaced the original batteries in 6 years (but I'm not sure how many hours I used it...)

 

[rsjsouza]

Sorry for the late reply. It's been a long day
....
Thanks for sharing these. The glitch looks similar to what I experienced before on a Keithley 2700. Other than that the drift doesn't look so bad. Definitely better than my Gossen calibrator. Did you connect the resistor directly to the banana socket?

I ran the battery life test tonight and as promised, I tried it in both DCV and resistance modes.   With the 40ohm load, I was able to lower the battery voltage until the meter turns off without any change in the measurement.  The same was true with a DCV source.    I have seen this be a problem before so didn't hurt to check.   

I have some other tests I want to run before we start the destructive testing so if there is anything else you would like me to look at, let me know.

We had gone through some simplified battery life tests before and I show the numbers below including the U1282A
https://www.eevblog.com/forum/testgear/hear-kitty-kitty-kitty-nope-not-that-kind-of-cat/msg2100013/#msg2100013

[joeqsmith]

I ran the battery life test tonight and as promised, I tried it in both DCV and resistance modes.   With the 40ohm load, I was able to lower the battery voltage until the meter turns off without any change in the measurement.  The same was true with a DCV source.    I have seen this be a problem before so didn't hurt to check.   

I have some other tests I want to run before we start the destructive testing so if there is anything else you would like me to look at, let me know.

The stability of this meter looks pretty good. Thanks for taking the time to do these tests. Have you figured out the reason behind the glitch?

Again, keeping in mind the meter was sitting out in the open on my desk with no attempt to control the temperature.   

In the following message, you can see how I am now able to reproduce the problem with ease without their software, cable or driver.   I would say the root problem is with the firmware.   Sadly, as I have said, Keysight will not open a dialog with me and I see no reason to send anymore time with this particular problem.  Just be aware of it.     
   
https://www.eevblog.com/forum/testgear/hear-kitty-kitty-kitty-nope-not-that-kind-of-cat/msg4358551/#msg4358551