EEVblog #170 - Agilent FAIL!

[pmrlondon]

Well, that is some proper weirdness there with that meter. I hope it turns out to be something simple, at least in explanation, as odd faults like this can be a pain to track down.

The stuck readings remind me of my Precision Gold WG021 - that could be odd like that sometimes - but it was manual range and more obviously wrong when only the decimal point moved.

[Bored@Work]

It looks like a microcontroller in that thing is running wild. Of course this assumes there is a microcontroller in it. My first suspicion would be something wrong with the brownout detection or reset circuit.

[Kiriakos-GR]

Just watched the video, and actually three times..

Well I am a bit confused .

I have see the pictures of PetrosA teardown ,  and still I do not have understand , if the LCD is on a separate PCB ?
or how it connects with the PCB.

From my own experiences , if there is an connection problem , between the main PCB with the secondary for the display,
or even if the rubber zebra is not in place 100%  , or even if an common hair of our head, drops in the path of the rubber zebra , it could lead to issues.

I have not get my own sample yet ( still in mail ), and I can not offer any test results.
   
Dave , just recheck the unit ...  inspect the connection ( display & PCB ) and use an magnifier glass too.

 
 

[saturation]

I am evaluating the same meter myself and have run into problems.  Nothing permanent and major, but certainly problematic for such a high priced meter, at a price were there are other competitors with no known issues that would make one skeptical about its long term reliability for making accurate measurements.

In the middle of testing, once or twice in 10s of hours of testing, while I switched from range to range, it did go berzerk briefly as the switch hung between ranges, then became right again once it was in seated in the range.  I also had it once read hundreds of Vac in the low Z range as I quickly switched from range to range and it was again, not seated fully in the range setting, and the leads were not connected.  I would not have thought much of these 'glitches' as I haven't been able to repeat it and its only happened, maybe no more than 3x, but reading Dave's comments and video, I'll now concentrate on seeing if I can repeat them consistently.

But this is with several days and multiple hours of testing, and it working right now.  Not sure what the cause is, and I continue to stress it.

I have not taken it apart.

The importance to me is, is before I take this DMM on a mission critical trip, it better live up to basic accuracy that is rock stable.

I'm glad to see that in Dave's test unit, there is no problem with Vdc calibration.

[Mechatrommer]

where is eevblog #170. is it missing? or is it just me?

[elCap]

I have seen the same thing on my 1272 meter but didn't think much of it at that time. I will for sure test it more tonight! I haven't taken my apart.

(you can find the eevblog #170 on youtube, appears there before here)

[PetrosA]

I haven't experienced this issue on my 1272. It did freak out when I first got it, but that was the warning beeper for leads, not jumpy readings (I wrote about it in my teardown thread IIRC). Since mine had come direct from Malaysia and the problem only occurred within the first 10-15 minutes of use, I wondered if it might be some corruption from security scanners EMF or something else during transport. I don't think it had anything to do with tearing down the meter...

[insurgent]

where is eevblog #170. is it missing? or is it just me?

Hey Mecha,

The video is here:

[EEVblog]

where is eevblog #170. is it missing? or is it just me?

I upload onto Youtube first, often the night before it makes it to eevblog.com (upload and go to bed).
A little tip, all the Youtube videos get automatically added here:
http://www.eevblog.com/episodes/
So you'll see it there straight away before it makes the main page.

Dave.

[Mechatrommer]

o ok thanx. i used to go to main page, download and play it locally. mosty i will keep in my HDD. youtube sometime lost in the cache files or i have to dig it or broken download/play is hurting as the net here is slow and BW limited.

[Floyo]

I watched the video just now and my unit has exactly the same problem, tough switching it on and off "fixes" the issue most of the time. The first time this happened I was measuring ohms and I was like "wtf is happening here, shitty resistor or something". Then it turned out to be this glitch, readings all over the place for no apparent reason, exactly as Dave described.

I tried to replicate the problem just now, but no luck.

I think it might me a software glitch, at least I hope so that owners of the meter can simply update it. If it where to be a hardware problem Agilent should do a recall (A).

Now let's hope that Agilent looks in to this problem quickly and is open about what the problem actually is, I really appreciate companies being open about their mistakes (and things they get right for that matter).

[EEVblog]

Thanks for the confirmation Floyo.
Agilent are looking into it.

Dave.

[Mechatrommer]

beta testing in progress! the plague already infected hardware!

[Nermash]

Let's see if Agilent will fix it like Fluke fixed the GSM issue with the 87V...

[firewalker]

Thanks for the confirmation Floyo.
Agilent are looking into it.

Dave.

So it seems it wasn't you taking it apart. No esd or so...

[Mechatrommer]

...like Fluke fixed the GSM issue with the 87V...

they do? afaik its only fixed in 28II if i'm not mistaken the name is.

edit: just by coincident browsing the new replied thead here.

Hey Kiriakos, wasn't it your FLUKE that got "killed" by your cellphone? Did FLUKE fix it?

Yep , actually it got replaced with one of my choice, because there was not an quick fix about the 87.
And also got 8 Fluke fuses for free .
I got truly impressed , and I will always favor them, for their immediate response. 

[EEVblog]

Let's see if Agilent will fix it like Fluke fixed the GSM issue with the 87V...

Yeah, never heard back about that one!
Agilent have a good opportunity to show up Fluke here!
Agilent fixed the hardware problem in the scope LAN module ASAP which was very encouraging.
So I think they'll take this one pretty seriously.

Extech, BK Precision, and Agilent have come to the party before, so lets see what Agilent have to offer again this time around.

Dave.

[Kiriakos-GR]

First of all I like to say something to Dave, about my suggestion above , about the recheck ,
I do trust the man , and it was not an sign or behavior due disrespect, or of having doughs for his opinion.

I did not received my unit, even today.

So far I had no plan to tear down the unit ..
But now I will do .

Even so I will keep my word to Agilent , about reporting to them directly ,
anything as remarks. ( for the first 30 days ) .

The story with the Fluke 87V , ended as Circus show , from young ones who started to play with it and their GSM phones.
I do not wish to see, the same thing to happen again.
No matter if this specific device , hides or not , an imperfection under the hood.

So far, we had see many people owning HH of Agilent ,
and there was no even one similar issue , with all those models.

And so dear friend Shafri , I do not see any " plague " near by ..  

And if you do again an such of joke about this DMM ,
I will tell to all the world , that your UNI-T does not know how to measure " Peak to Peak " LOL

Kiriakos ..


Edit:
The Fluke 28II , never had an GSM issue .
Inside haves other fresher microprocessor than the 87V .   


 

    

[Mechatrommer]

And so dear friend Shafri , I do not see any " plague " near by .. 

i mean, beta testing usually involved in software. but i dont know if its happening in hardware as well, but it seems starting like that. except they dont call dave a beta tester. but the worst thing is... i dont want to say that. and i dont mind if you say uni-t cannot do peak to peak my friend thats by design and i admit its in cheapo range of price. it served my purpose. i will not use it to measure oscillation except pure mains sine. i hope Agilent take this seriously and hopefully not Malaysia made/assembled. if it is, then they must take it seriously otherwise i will be ashamed!

ps: i have future plan to build a product/hardware, but before selling it out, i think i will ask few people as "beta testers" to prove the product non-flawed

[saturation]

It didn't look like the GSM issue is a quick fix.  Given that the phone has to be fairly close to the DMM for the interference to occur, its unlikely to be an major issue in real world use; I can see if a technician were to receive a phone call with his phone placed ontop of the DMM used in speakerphone mode while actually working with the DMM, but if he moved it 6 inches or so away, it would stop.



By contrast, the unstable megaohm measurement issue I raised with the 1252a is still with the 1272a, since its the same design, is a practical problem.  Measuring megaohms without a filter on the Agilent can be difficult.

Let's see if Agilent will fix it like Fluke fixed the GSM issue with the 87V...

Yeah, never heard back about that one!
Agilent have a good opportunity to show up Fluke here!
Agilent fixed the hardware problem in the scope LAN module ASAP which was very encouraging.
So I think they'll take this one pretty seriously.

Extech, BK Precision, and Agilent have come to the party before, so lets see what Agilent have to offer again this time around.

Dave.

[Kiriakos-GR]

And so dear friend Shafri , I do not see any " plague " near by .. 

and i dont mind if you say uni-t cannot do peak to peak my friend thats by design and i admit its in cheapo range of price. it served my purpose. i will not use it to measure oscillation except pure mains sine. i hope Agilent take this seriously and hopefully not Malaysia made/assembled. if it is, then they must take it seriously otherwise i will be ashamed!

ps: i have future plan to build a product/hardware, but before selling it out, i think i will ask few people as "beta testers" to prove the product non-flawed

Aaaa you are aware that I like you as friend , and you are taking advantage of it ..
Keep shooting I am bullet proof .. 

Unfortunately more delay added about getting my sample,  as it will arrive from Germany to Greece,
and it is not even send yet , due a mix-up.

Lets hope that those good for nothing Germans, they will send and all the additional accessories too. 

[Kiriakos-GR]

its unlikely to be an major issue in real world use; I can see if a technician were to receive a phone call with his phone placed ontop of the DMM used in speakerphone mode while actually working with the DMM, but if he moved it 6 inches or so away, it would stop.

And so what if it stops ?
I do not wish to have such a unit, in the field with me , and that's final.

By contrast, the unstable megaohm measurement issue I raised with the 1252a is still with the 1272a, since its the same design, is a practical problem.  Measuring megaohms without a filter on the Agilent can be difficult.

I hold my horses on that , until to see the complete video review of Dave .
And I would become an believer if I see your own video about it.

But even if you do one , it will show , that the new layout of the buttons ,
will had another degree of response at about how the display flickers.

Bottom line ,  if the operation of the buttons does not effect the measurement,
no one cares about flickering , the Fluke flickers a bit too.
And its not capable to measure above 50MOhm ..

I would be interested to see the Fluke one to measure 150M Ohm ,
and to have the same stability as it does at below 50M Ohm.

But I bet that I will not see that , in this Fluke series ,
probably because they can not jump that high !!    

[saturation]

After what you experienced, I expected many Fluke 87V to be bricked.  But more than just yourself have tried it, the video I posted is an example, Dave tested it, and I did so on mine, none have been bricked.

However, given the it actually did happen, I concur with you, that its of higher risk.  I use the Series 85III when travelling outside the USA because its more reliable.

its unlikely to be an major issue in real world use; I can see if a technician were to receive a phone call with his phone placed ontop of the DMM used in speakerphone mode while actually working with the DMM, but if he moved it 6 inches or so away, it would stop.

And so what if it stops ?
I do not wish to have such a unit, in the field with me , and that's final.

[Kiriakos-GR]

After what you experienced, I expected many Fluke 87V to be bricked.  But more than just yourself have tried it, the video I posted is an example, Dave tested it, and I did so on mine, none have been bricked.

Just one word about that ... I had the LCD back light on , when I did my tests ... because I had in my other hand the digital camera,
so to take a picture of it ..

Does this detail matters ? so to be totally killed ? ( I do not know , and I do not care to find out )

However, given the it actually did happen, I concur with you, that its of higher risk.  I use the Series 85III when travelling outside the USA because its more reliable.

Good for you ..

[adam_lumpkins]

I am willing to bet that dave did not make a mistake on the rebulid.  I would hope he would have checked the zebra strip on the L.C.D.  I am gravitating towards an E.S.D problem.   Good job Dave on the video love it!