EEVblog #832 - Keysight U1282A Multimeter Teardown

[EEVblog]

Inside Keysight's new U1282A IP67 waterproof and 3m drop proof precision dual display 60,000 count 0.025% multimeter with 800 hours battery life
http://www.keysight.com/en/pd-2613328-pn-U1282A/handheld-digital-multimeter-45-digit-up-to-800-hours-battery-life?nid=-34618.1144466&cc=AU&lc=eng

3 for 2 deal: http://www.keysight.com/main/editorial.jspx?cc=US&lc=eng&ckey=2622527&nid=-33167.920237.00&id=2622527

[kaadam]

"BD" may be stands for bead, so i suppose it will be a ferrite bead, although i don't know why it's necessary.

[Fungus]

I like the Anti-Batteriser sheilding under the battery compartment.


Heat shrunk thermistors? Because they turn to green dust when you zap them. I'm guessing the heatshrink is to hold them together.

[ECEdesign]

What is the difference between autohold and trig hold? 

[EEVblog]

"BD" may be stands for bead, so i suppose it will be a ferrite bead, although i don't know why it's necessary.

Yes, I think it's a bead too.
Can't say I've ever seen one in this application.
Possibly a last minute addition to pass EMC compliance?
They do have a awful lot of shielding, even going right up the side of the already shielded ceramic resistor divider.

[Fungus]

What is the difference between autohold and trig hold?

So long as it displays the number then it's the same, isn't it? The only difference is that it goes into memory as well.

[EEVblog]

What is the difference between autohold and trig hold?

Trigger hold allows you to log into memory, but it's not automatic, you have to push the button.

[EEVblog]

What is the difference between autohold and trig hold?

So long as it displays the number then it's the same, isn't it? The only difference is that it goes into memory as well.

In this case it is not the same as auto hold.

[Fungus]

What is the difference between autohold and trig hold?

Trigger hold allows you to log into memory, but it's not automatic, you have to push the button.

Ew!

[hayatepilot]

Just saw in the manual that this meter does indeed have a setup menu and Auto-Hold.
Setup is accessed by holding Shift while turning it on. There, one can activate the Auto-Hold function instead of Trig-Hold.

Greetings

[AlxDroidDev]

The Agilent DMMs where a lot prettier.

[gslick]

Did you get a U5404A Remote Switch Probe with your review unit?

http://www.keysight.com/en/pd-2615346-pn-U5404A/remote-switch-probe

The current promotion for the U1281A / U1282A includes the U5404A Remote Switch Probe for "free", normally listed at $60 separately.

[joeqsmith]

GDTs are slow and if there is nothing else in there for a clamp, I wonder how that thing would handle a transient with a fast edge.   

[brutester]

To get the 60000 count from a 50000 count chipset they must use more than one ADC. Luckily they have 3 sigma-delta ADC chips built into HY3131 .

They must use the data from the fast SD-ADC to get the base value and then use the slow SD-ADC to get a more-precise value by moving the FTN pin  up with a DAC or controllable voltage divider from the reference voltage.

[tru]

To get the 60000 count from a 50000 count chipset they must use more than one ADC. Luckily they have 3 sigma-delta ADC chips built into HY3131 .

They must use the data from the fast SD-ADC to get the base value and then use the slow SD-ADC to get a more-precise value by moving the FTN pin  up with a DAC or controllable voltage divider from the reference voltage.

I understand count as referring to speed not precision, so 60000 count means something like 60000 readings per second.

[Fungus]

I understand count as referring to speed not precision, so 60000 count means something like 60000 readings per second.

Nope. It's the maximum value that can be displayed on screen. The reason there's so many cheap "2000" count meters is that the first digit only needs two LCD segments if it can only count to 1999.

(...and why would anybody want 60000 readings per second on a multimeter?)

[TheSteve]

BD is a common reference designator for ferrite beads in radio equipment.

[mux]

Thumbs down to those rubber sealing... things. They're probably well-engineered and soft enough, but the issue with these removable boots is that their sealing is 100% dependent on the user putting them back properly after removing, and it looks like you do have to remove them every time a fuse or battery needs to be swapped.

It's too easy to accidentally get an edge folded over or pinch the fingers, which will result in an almost guaranteed non-watertight seal.

This is why we use greased o-rings, people. They're not meant to be removed, so you have a guarantee that they stay put in the orientation and location you intended them to be.

[DutchGert]

Dave, think u forgot the datasheet for the Multimeter IC

Nice teardown!

[Fungus]

Thumbs down to those rubber sealing... things. They're probably well-engineered and soft enough, but the issue with these removable boots is that their sealing is 100% dependent on the user putting them back properly after removing, and it looks like you do have to remove them every time a fuse or battery needs to be swapped.

It's too easy to accidentally get an edge folded over or pinch the fingers, which will result in an almost guaranteed non-watertight seal.

I wonder about the long term stability, too. Will that rubber last for many years? I hope they sell spares.

Mind you: If you need a waterproof multimeter then you can probably afford to replace them every ten years or so.

[tszaboo]

Anyone knows what is the reason for heatshrinking those resistors and PTCs?
BTW 500 dollars is a lot for a meter with these specs. I guess the only reason to buy it if you break meters every now and then due  to water and stuff. But who am I kidding? Are you really going to use it standing in water, hoping for your cat 4 rating? Seems very alien situation to me, the office  dweller.

[Brumby]

My guess would be in the case of failure involving physical damage, the heatshrink would limit the dispersal of shrapnel through the meter, possibly creating an alternate current path and shock risk.

[German_EE]

Why is it on almost every multimeter I see there is either no AC current range or AC current mode is selected using a 'shift' function, yet capacitance measurement has it's own position on the knob? It's an industrial meter and in industry there is often a requirement to measure the current drawn in an AC circuit.

[Fungus]

Anyone knows what is the reason for heatshrinking those resistors and PTCs?

joeqsmith managed to turn a few PTCs into green dust over in the "here kitty" thread.  I'm guessing the heatshrink is to stop them messing up the inside of the meter when you zap them too hard.

Although if you're regularly exploding your PTCs you might want to consider a whole new meter just in case, and wrapping them in heatshrink might hide any damage 

On a side note: How fast do PTCs react? Surely they have to physically heat up and that takes a few milliseconds at least. Joe's PTC-destroying pulses are only a few microseconds long.

And MOVs, how fast are they? Presumably a lot faster than PTCs.

[Fungus]

Why is it on almost every multimeter I see there is either no AC current range or AC current mode is selected using a 'shift' function, yet capacitance measurement has it's own position on the knob? It's an industrial meter and in industry there is often a requirement to measure the current drawn in an AC circuit.

Because in an 'industrial' setting you usually measure current with a clamp.

[Muttley Snickers]

I haven't read the manuals on these meters as yet and only had a brief look over the specs, through my eyes though they look like a very nice meter and until it was placed alongside I didn't think that the U1272A could be out done aesthetically.

I don't ever recall Agilent - Keysight having an intrinsically safe meter before and if they are chasing the Fluke 28 II market then this may be an indication of something ahead.

Many thanks for the teardown and I will load up on popcorn and soft drink tomorrow ready for the review, put a ticket aside for me.

[mux]

I wonder about the long term stability, too. Will that rubber last for many years? I hope they sell spares.

Mind you: If you need a waterproof multimeter then you can probably afford to replace them every ten years or so.

It's not actually waterproof though, it's only temporarily submersion resistant (hence no grease, etc.). There are better meters in this category, you'll likely go for those if you actually need guaranteed waterproofing. This is more of a 'oh shit I'm on an oil rig and just spilled half a ton of oil over my worksite'-jobby. Nothing against this meter, it's just a different application.

If it's proper rubber-rubber (and not butyl or something) it will survive for *decades*. Rubber only really deteriorates under the influence of fluorine, ozone and high-energy light (blue sunlight, UV), which is no issue in this meter. I have a Fluke 73 (1982) with rubber seal on the inside, it's still perfectly intact.

[EEVblog]

The inductor with silastic gunk is to prevent acoustic noise. They chose the wrong DC/DC converter that operate in burst mode or skipping mode, instead of PWM mode (or deliberately for low power consumption), so there will be hiccup noise falling into acoustic range at light load. A simple bulge will be simply gunk it up and damp any mechanical vibration in acoustic range.

Could very well be! 

[EEVblog]

Just saw in the manual that this meter does indeed have a setup menu and Auto-Hold.
Setup is accessed by holding Shift while turning it on. There, one can activate the Auto-Hold function instead of Trig-Hold.

Geeze, why can't it just operate like the U1272A?

[alho]

That rubber is quick and easy to assemble, slap a rubber mat on or dick around with two o-rings. Hands are big part of assembly cost and Taiwan Malaysia isn't cheapest place.

Let's say you are replacing batteries and drop the meter on dirty and dusty factory floor when its open. O-rings have hard to clean thin grooves where dirt can lodge while this meter looks like you can just shake it clean.

[Muttley Snickers]

Sorry I'm back again, that remote switch had me intrigued and I couldn't sleep without finding out more......

Most distributors are currently throwing the U5404A remote switch in with the meter as a deal, it retails for about $60 US alone, Element 14 are doing it but not sure about Trio Test, I expect that they probably are but cannot confirm at this time.

As Dave said in the video the U1282A multimeter is listed for $500 US, unfortunately down here they are going for $751 plus 10% GST so about $826 AUD, not that this meter is rated for it but It's not intrinsically safe for me to bring home anymore test equipment for a while after some recent purchases, sparks would fly if I came home with another meter and that 342 Keysight deal might be a problem, anyway I have until the end of February.

Trio Test and Measurement Australia
http://www.triotest.com.au/shop/keysight-technologies/3848-keysight-u1281a-handheld-digital-multimeter-5-digit-.html

Element 14 Australia
http://au.element14.com/keysight-technologies/u1282a-u5404a/digital-multimeter-handheld-4/dp/2499696

U5404A Remote Switch

[Christe4nM]

I might be picky, but I actually dislike the new button design on these meters. I much rather have the old solid/single color rubber ones without a LED behind them like on the U1272A and U1461A.

All in all it seems to me this meter is really aimed at field use and therefore its ruggedness. If you need a general electronics meter for bench use you don't get this one, but say the U1272A. So size wise, it depends on your toolbox if this one is too big. Once you put it next to the U1461A you can clearly see the reused the case from that one. Clever design reuse in my opinion.

I am glad they went with the traditional LCD as oposed to OLED. Especially when you want to use the meter outside, in daylight, readability on a normal LCD is much better than OLED in my experience. Hence I wondered why they put the OLED in the U1461A in the first place as that seems to be aimed at field work as well.

Anyway, I loved the teardown. I found myself actually leaning towards my computer screen to not miss any details

[boffin]

That rubber is quick and easy to assemble, slap a rubber mat on or dick around with two o-rings. Hands are big part of assembly cost and Taiwan isn't cheapest place.

Good thing it's made in Malaysia then, to keep the cost down.

[Mark Hennessy]

Apologies if I missed it, but I don't think it was mentioned: surely those spring-clip connections to the current terminals are to do with the Input Warning feature, rather than sensing for current measurements?

They look to be insulated from the current path to me, which would make sense - they are presumably shorted to the current inputs when a probe is plugged in. Likewise, this is not a PSU, so you don't need sensing at the terminals; rather, you need to sense the voltage directly across the shunt. Also, the large series resistors from the "sense" terminals need to withstand a high voltage when the fuses open...

[joeqsmith]

Anyone knows what is the reason for heatshrinking those resistors and PTCs?

joeqsmith managed to turn a few PTCs into green dust over in the "here kitty" thread.  I'm guessing the heatshrink is to stop them messing up the inside of the meter when you zap them too hard.

Although if you're regularly exploding your PTCs you might want to consider a whole new meter just in case, and wrapping them in heatshrink might hide any damage 

On a side note: How fast do PTCs react? Surely they have to physically heat up and that takes a few milliseconds at least. Joe's PTC-destroying pulses are only a few microseconds long.

And MOVs, how fast are they? Presumably a lot faster than PTCs.

MOVs will switch much faster than a GDT.  GDTs are in the microsecond range.   They are good for high energy, long long duration but because they are so slow, I would have something to help them out.  You could get away using a smaller MOV as it would only need to handle the energy until the GDT fires.  If they use some sort of diode clamp, there must be another layer of R after that last GTD.   Would be interesting to sketch it out and see if at least on paper that it could survive something with a fast edge.   

I have seen a few HV resistors fail in the meters and a few PTCs.  You are right, the PTCs when coated will come apart even with very low energy levels.  I have had a few where there is nothing left but the leads when I put a little more in them.   It makes sense that they heat shrink them to prevent the fragments from getting into the meter but I am not sure if that really is the reason for it.     

[gildasd]

I wonder about the long term stability, too. Will that rubber last for many years? I hope they sell spares.

Mind you: If you need a waterproof multimeter then you can probably afford to replace them every ten years or so.

It's not actually waterproof though, it's only temporarily submersion resistant (hence no grease, etc.). There are better meters in this category, you'll likely go for those if you actually need guaranteed waterproofing. This is more of a 'oh shit I'm on an oil rig and just spilled half a ton of oil over my worksite'-jobby. Nothing against this meter, it's just a different application.

If it's proper rubber-rubber (and not butyl or something) it will survive for *decades*. Rubber only really deteriorates under the influence of fluorine, ozone and high-energy light (blue sunlight, UV), which is no issue in this meter. I have a Fluke 73 (1982) with rubber seal on the inside, it's still perfectly intact.

I think OILPOL and MARPOL would be my major concerns, not the multimeter...

[Ice-Tea]

(...and why would anybody want 60000 readings per second on a multimeter?)

I figured counts relates to how accurate rms measurments would be? And up to what frequency they would be accurate?

[brutester]

(...and why would anybody want 60000 readings per second on a multimeter?)

I figured counts relates to how accurate rms measurments would be? And up to what frequency they would be accurate?

This parameter is given as bandwidth in data sheets/specs. It also shows maximum frequency of signal at which you will get a valid reading

[ECEdesign]

mikeselectricstuff mentioned on the YT comments about the absence of a full LCD test on startup.  This is available, if you hold down the "hold" key on startup the LCD display will show all the options so you can check to make sure everything is working.

[PA0PBZ]

Anyone care to comment about what they did here?

[Fungus]

Anyone care to comment about what they did here?

That's so the electrons can't go too fast down that wire.

[PA0PBZ]

Anyone care to comment about what they did here?

That's so the electrons can't go too fast down that wire.

With those sharp corners the electrons will be all over the place don't you think?

[Ice-Tea]

mikeselectricstuff mentioned on the YT comments about the absence of a full LCD test on startup.  This is available, if you hold down the "hold" key on startup the LCD display will show all the options so you can check to make sure everything is working.

I don't think one excludes the other? I'd say for accurate rms readings you need hefty oversampling, hence the high count numbers?

[EEVblog]

Apologies if I missed it, but I don't think it was mentioned: surely those spring-clip connections to the current terminals are to do with the Input Warning feature, rather than sensing for current measurements?

Yeah, that's what I meant.

[EEVblog]

Anyone care to comment about what they did here?

Clearance rule with that other node in the red circle?

[Muttley Snickers]

Wow !, I just read the manual and found a huge amount of setup options, possibly unsurpassed for configuration options in a handheld, some will like this stuff.. and others may not..

Both the U1240 and the U1280 series meters come with the IR-USB interface as standard, this I think is a good move on Keysights part, charge a little extra and chuck it in the box. We saw what happened with the U1272A where they had to post out this cable for free so people could do the firmware upgrade on the early versions, only they know how many went out and what it cost them.

On a side note both series meters appear to have an option for the continuity alert, beeper only, LED only or both, the U1230 series I believe is similar with beeper only, backlight only or both.

The U1272A has only the variable tone option and is both audible and visual via the backlight unless they have changed this in recent firmware which I have not installed as yet.

[EEVblog]

Wow !, I just read the manual and found a huge amount of setup options, possibly unsurpassed for configuration options in a handheld, some will like this stuff.. and others may not..

To have all that setup functionality and not make it obvious how to get into the setup like on the U1272A is just stupid.
Or perhaps it was the outcome of some focus group that said that field users shouldn't be dicking around with setups.

[Muttley Snickers]

I've mentioned this before in regards to the U1272A that they should provide a small laminated options menu cheat sheet tucked in behind the bail. anyway people can make their own I suppose if they feel the need.

There is already one brief Youtube video review on this meter, more of an introduction really and it's locally made.

[tszaboo]

Wow !, I just read the manual and found a huge amount of setup options, possibly unsurpassed for configuration options in a handheld, some will like this stuff.. and others may not..

To have all that setup functionality and not make it obvious how to get into the setup like on the U1272A is just stupid.
Or perhaps it was the outcome of some focus group that said that field users shouldn't be dicking around with setups.

So it is working as the U1252B. I dont think it is a big deal, just RTFM.
One other thing I saw, that there is a programming guide for the multimeter. So they give you the list of commands for the meter, so third party software can be made now. I dont think it is a too much of a big deal, I mean, it is not like we will see this built into a test rig or anything, cause the battery and the range switch, but still, could be interesting.

[VK5RC]

While a little off topic (teardown) I thought the leads were not up to KS usual standard and are a bit stiff, not their usual pliable leads.
It is a BIG and HEAVY DMM.
Is it just me or is the dual display of DC V and AC V not highly sort after, yet it is about the third option in the dual display function after dBmV etc

[Ricardo]

Dave finds 2 MAX4611 circuits on this board and wonders why 2 Voltage Detectors are needed.  Voltage Detectors = MAX 6411 
MAX4611 = CMOS Analog Switches

[Dubbie]

Does anyone know how you wake it up after it has gone to sleep? I wish wish wish that you could just push any button on meters to get them to turn back on instead of cranking the stupid dial round to off then back again to where you were. Super infuriating for bench use. My Bryman makes you do that whereas my fluke 289 has a nice on off button but a huge long boot up delay. If this meter would wake up instantly on a button press I will buy one.


Sent from my phone using Tapatalk

[K6TR]

I'm looking forward to the formal Product Review of this meter Dave. I would also like to see a review of the 1461 Insulation Tester/Multimeter. I have my heart set on Keysight coming out with a an OLED Version of the 1280 Series as I really like the OLED they use on the 1253B, don't have use for the Insulation Tester and can't justify laying out the money Keysight is asking for the 1461. They have done some things to address the short battery life in the 1253 and while a long ways from ideal 50 Hours use in the 1461 with Alkaline Batteries is a substantial improvement over the 1253B.

Enjoy your time off Dave.

[ali_asadzadeh]

Dave are they worth their price? what are the alternatives?

[VK5RC]

@Dubbie, after the u1282 goes to sleep, you can wake it up by hitting the centre blue button (it worked for me twice this afternoon), it goes back to the same function as set on the switch.
For me the biggest negative issues are 1. its size, 2. the leads are a bit non-pliable.
Positives; seriously accurate and switches on quickly, I tend to switch off between readings.

[Dubbie]

Well I feel like a dumbass. Just read the Brymen manual again and discovered it does turn on with a button press. Maybe I didn't press them long enough for the debounce. The new Agilent looks nice as well. I need one more meter. Shame it is so huge.


Sent from my phone using Tapatalk

[zaoka]

Did they fix Capacity issues they had with U1272A? Smoothing mode works now?

Also display contrast looks better than U1272A...

I like it a lot except font for digits, somehow space between digits is dense...

[EEVblog]

Dave finds 2 MAX4611 circuits on this board and wonders why 2 Voltage Detectors are needed.  Voltage Detectors = MAX 6411 
MAX4611 = CMOS Analog Switches

  

[EEVblog]

Dave are they worth their price? what are the alternatives?

Yes if it meets your requirements.
Fluke 28 II is a direct competitor in the rugged field category.
For a general bench meter I'd rather have the U1272A

[jpb]

I don't know if I've missed it, but from the data sheet and user guide it doesn't seem to have a real-time-clock unlike other logging meters (eg the Fluke 287/289 and the Gossen meters)? This probably helps battery usage but means you have to have it tethered to a pc in some way if you want to log with time stamps? You can set time intervals for logging but this is not quite the same thing. It also has a relatively small logging memory.

I'm currently looking for a high-end meter and the U1282A has good specs and seems robustly built with very good battery life but for some reason I feel that it seems to have been designed to tick a lot of spec boxes rather than as a single product. It is also more expensive than the Fluke 287 in the UK and say the Gossen Ultra (both of which though have major drawbacks like very low battery life and no analogue display on the Gossen).

Dave - have you done real world tests on the various claimed battery lives? I know it is a feature that you are hot on. I just wonder when you have the Fluke 289 only getting 100 hours out of 6 AA batteries (I think) and the U1282A claiming 800 hours (dc V) from 4. The Gossen meters rate themselves at 200 hours but from just 2 AAs.

[Muttley Snickers]

I found a couple of teasers which will have to do until Dave gets back from exhaustive testing in relation to lateral resistive integrity of sand castles.....

[saturation]

I have opted to get one 1282a upgrade and will keep my other 1272a so this thread piqued my interest.  While I like the toughness of the casing the GDT design in Keysight DMM input protection has always concerned me, not that it won't protect, but it will protect but lead to meter destruction.  Also the size is starting to get unwieldy for a field device; small, light and toughness are bigger concerns given airline costs and luggage limits.

As expected, when Joe did the test on a cheaper Keysight DMM with GDT protection, it died.  This is similar protection used on many of the Keysight DMMs.  Only a full test per model would say for sure.

The standard MOV design allows the meter to survive and take a large spike, as Joe Smith easily demonstrates.  That said, the test 87V failed under Joe's hands, but the other well designed MOV based protection have survived, e.g. other model Flukes, Brymen etc., also an 87V is shown on Fluke's marketing videos taking hits to 12kV, but that's on AC volt mode.

Starts at 17:00