Author Topic: EEVblog #489 - Agilent 34461A Multimeter Review  (Read 29806 times)

0 Members and 1 Guest are viewing this topic.

Offline EEVblog

  • Administrator
  • *****
  • Posts: 29694
  • Country: au
    • EEVblog
Re: EEVblog #489 - Agilent 34461A Multimeter Review
« Reply #50 on: June 28, 2013, 11:07:46 pm »
It's a shame because if there was an easy way to actually use the 10,000 samples/sec at 6.5 digits it would be a fantastic tool for logging and characterizing things. Unlike an oscilloscope the readings are calibrated to a much higher standard and have much more resolution and dynamic range.

You'll never get 10,000 sps at 6.5 digits, the top speeds are rated for lower resolution like 4.5 digit.
 

Offline orin

  • Frequent Contributor
  • **
  • Posts: 434
  • Country: us
Re: EEVblog #489 - Agilent 34461A Multimeter Review
« Reply #51 on: June 29, 2013, 03:26:40 am »
After Dave's teardown last week, I said I just got to get me one of them...  After a fight with MetricTest's web site/order system/email misconfiguration (sending email from an IP with no reverse DNS), it arrived Tuesday - 10% discount, but got it back and more in taxes and shipping costs to boot.

Anyway, I've been monitoring a Fluke 731B that I'm fixing up.  See picture if it makes it... 

Like Dave in the review at about 1:00:00, I have a trend chart showing 9.999 9 10.000 0 and 10 for the voltage axis!  The 10.000 0 should be 9.999 95, but it has been rounded up.  The 10 is really 10.000 0, but I guess that since it is 'exact', they don't show the fractional part.  They need to rethink this labeling IMO - sometimes it gets too cute - all labels should show the same number of digits.  I'm with Dave when I see the 10... WTF?

The stats functions seem to quit at 24730 samples.  I didn't get the screenshot and the number of samples reset when I returned the DMM to local from the web interface so I can't attach a picture, but it got stuck at 24730 samples.  I took the snapshot when I got home and it stopped at 24730 soon thereafter.

The 10,000 samples in memory must be a minimum.  When I took the screenshot, the web interface reported over 10,200 samples in memory and I could read them with the web interface.

FWIW, an in-cal 3456A is reading 10.000 07 connected in parallel with the 34461A.  All well within spec.

I've got to say the web interface is neat.  I can monitor it from the other side of the house.  If I were to open the right ports on the router, I could monitor it from work.  I've yet to install Agilent's IO library and DMM monitor software...

Yes, the UI needs some work, but not bad for a 1.0.

Another FWIW, when it first arrived, I powered it up at work and monitored a power supply with the trend chart.  A few hours later, we had a power glitch/brownout.  The 34461A stopped triggering and stopped updating the display.  I did not have to power it off to recover.  I must say that the 5370A that I had monitoring an oscillator and transmitting results over a Prologix GPIB ethernet adapter also came to a screeching halt, so it was a pretty gnarly glitch.  However, it should have reset if it couldn't handle the brownout or even if the brownout was a potential problem.  Sitting looking stupid isn't the right answer.  My PCs on a UPS didn't even notice the glitch.

Now does anyone, preferably in the NW USA want to buy two 3455A DMMs (rhetorical question - test equipment follows the N + 1 rule)?  Something has to go now according to the powers that be.

Orin.
 

duskglow

  • Guest
Re: EEVblog #489 - Agilent 34461A Multimeter Review
« Reply #52 on: June 29, 2013, 03:31:44 am »
Where in NW USA and what do you want for one?
 

Offline Alexei.Polkhanov

  • Frequent Contributor
  • **
  • Posts: 683
  • Country: ca
Re: EEVblog #489 - Agilent 34461A Multimeter Review
« Reply #53 on: June 29, 2013, 03:48:01 am »
A common problem with managing several devices with clocks is to keep them synchronized. If you got an automated test setup and get data/measurements with timestamps from the T&M devices it's really annoying if they are several seconds or even minutes apart. In that case it would be great to have support for NTP (current T&M got LAN interfaces and IP stacks) to synchronize the clocks.
This is what Data Logging / Data acquisition systems are for. For example this one from Keithley http://www.keithley.com/products/data/datalogger/?mn=2700, National Instruments has whole range of gear for that. Also there is external trigger input/trigger output on almost all equipment.
If you look at this DMM (34461A) it has <1us (microsecond)  for external triggering even if it only capable of 1ms period between readings. I would not just get readings from different devices and then try to match them together by timestamps.
 

Offline orin

  • Frequent Contributor
  • **
  • Posts: 434
  • Country: us
Re: EEVblog #489 - Agilent 34461A Multimeter Review
« Reply #54 on: June 29, 2013, 03:59:36 am »
Where in NW USA and what do you want for one?

I sent a PM.
 

Offline casinada

  • Frequent Contributor
  • **
  • Posts: 596
  • Country: us
Re: EEVblog #489 - Agilent 34461A Multimeter Review
« Reply #55 on: June 29, 2013, 06:21:59 am »
Besides cosmetics and looks that everybody seem to be complaining about, how about performance compared to the 34401?
The 34401 had some problems measuring low AC voltage but it used an off the shelf true RMS converter chip. Newer Agilent multimeters use different techniques to measure RMS voltage.
The low voltage problem is described here:
http://www.gellerlabs.com/34401A%20AC%20zero.htm
Can anybody with both devices run some comparative tests?
Thank you
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 29694
  • Country: au
    • EEVblog
 

Offline casinada

  • Frequent Contributor
  • **
  • Posts: 596
  • Country: us
Re: EEVblog #489 - Agilent 34461A Multimeter Review
« Reply #57 on: June 29, 2013, 06:45:59 am »
:)
Yes, I found that out before I could re-post. I just remembered the limitation on the 34401, I didn't know he run a test on the 34461 as well.
Thanks Dave
 

Offline ElectroIrradiator

  • Frequent Contributor
  • **
  • Posts: 614
  • Country: dk
  • More analog than digital.
Re: EEVblog #489 - Agilent 34461A Multimeter Review
« Reply #58 on: June 29, 2013, 10:54:58 pm »
Does anyone know why Agilent chose to use WinCE for their newest lines of test gear over a RTOS like for instance QNX?

Not everybody keep their test equipment under power 24/7, and I cannot get used to the idea of having to wait nearly a minute after power on. Imagine if your car, home appliance or medical device took that long to power up... :scared:
 

Online Monkeh

  • Super Contributor
  • ***
  • Posts: 5945
  • Country: gb
Re: EEVblog #489 - Agilent 34461A Multimeter Review
« Reply #59 on: June 29, 2013, 11:01:47 pm »
Does anyone know why Agilent chose to use WinCE for their newest lines of test gear over a RTOS like for instance QNX?

WinCE is an RTOS.

Quote
Not everybody keep their test equipment under power 24/7, and I cannot get used to the idea of having to wait nearly a minute after power on.

Then you may as well not bother with precision equipment.

It's only a minute, people. This is not a handheld DMM. Cope.
 

Offline baljemmett

  • Supporter
  • ****
  • Posts: 666
  • Country: gb
Re: EEVblog #489 - Agilent 34461A Multimeter Review
« Reply #60 on: June 29, 2013, 11:25:21 pm »
Not everybody keep their test equipment under power 24/7, and I cannot get used to the idea of having to wait nearly a minute after power on. Imagine if your car, home appliance or medical device took that long to power up... :scared:

My television - a two-year-old Sony - takes about 40 seconds to boot to the point where it can actually be controlled.  It's a pain in the arse, really.  Mind you, what it's booting during that time is Linux-based so I guess the moral is that it's not the OS that makes it slow, just whoever set out the boot sequence when building it.
 

Offline John Coloccia

  • Super Contributor
  • ***
  • Posts: 1199
  • Country: us
Re: EEVblog #489 - Agilent 34461A Multimeter Review
« Reply #61 on: June 29, 2013, 11:51:36 pm »
It's always been a complete non-issue for me.  I walk into the lab, turn on power supplies, power meters, OSAs, spectrum analyzers, whatever....I sip my coffee for a minute, complain a bit with the nearest co-worker about the latest idiot who's clueless, and get to work.  This is pro equipment designed for pros, and things like boot time, screen angle and things like that are pretty irrelevant.  Stuff that annoys me is performing out of spec, strange GPIB lockups in the middle of the night so all my testing is garbage, etc etc.

I'm not saying that it wouldn't be BETTER in some sense if this other stuff was better, but it's really not even on the radar when I'm selecting equipment.  Shoot, it takes me longer to remember where I was from yesterday than it does for any of this stuff to boot up.
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 29694
  • Country: au
    • EEVblog
Re: EEVblog #489 - Agilent 34461A Multimeter Review
« Reply #62 on: June 30, 2013, 12:16:21 am »
Shoot, it takes me longer to remember where I was from yesterday than it does for any of this stuff to boot up.

That sums it up nicely!
 

Offline c4757p

  • Super Contributor
  • ***
  • Posts: 7805
  • Country: us
  • adieu
Re: EEVblog #489 - Agilent 34461A Multimeter Review
« Reply #63 on: June 30, 2013, 12:23:40 am »
You know, I take back my complaint about the boot time. I thought it'd be annoying, but it just occurred to me that my old Tek analog scope (with digital controller and a very long POST) takes about the same amount of time to boot, and I still prefer it to the newer DSO that's almost instant-on. Would I prefer that it started faster? Yes. Does it affect whether or not I like it? Not one bit.
No longer active here - try the IRC channel if you just can't be without me :)
 

Offline SeanB

  • Super Contributor
  • ***
  • Posts: 15094
  • Country: za
Re: EEVblog #489 - Agilent 34461A Multimeter Review
« Reply #64 on: June 30, 2013, 07:07:29 am »
WinCE because they already have the test procedures, the build libraries and the firmware in use on other devices, so it is not a big cost and learning curve to use another OS. The cost for them is essentially zero to use it, anything else will have a bigger cost in time or development.
 

Offline mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 12001
  • Country: gb
    • Mike's Electric Stuff
Re: EEVblog #489 - Agilent 34461A Multimeter Review
« Reply #65 on: June 30, 2013, 09:52:24 am »
Just found a hidden function!
In any help screen, press the second function key. This enters a mode that lets you browse through all help messages.
And if you hold the button for help on this function....
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 

Online amyk

  • Super Contributor
  • ***
  • Posts: 6483
Re: EEVblog #489 - Agilent 34461A Multimeter Review
« Reply #66 on: June 30, 2013, 10:22:24 am »
Then you may as well not bother with precision equipment.

It's only a minute, people. This is not a handheld DMM. Cope.
...and maybe you should be pleased that it lets you use it after only a minute, when it takes an hour to warm up to specified precision.
 

Offline ElectroIrradiator

  • Frequent Contributor
  • **
  • Posts: 614
  • Country: dk
  • More analog than digital.
Re: EEVblog #489 - Agilent 34461A Multimeter Review
« Reply #67 on: June 30, 2013, 10:26:17 am »
It's always been a complete non-issue for me.  I walk into the lab, turn on power supplies, power meters, OSAs, spectrum analyzers, whatever....I sip my coffee for a minute, complain a bit with the nearest co-worker about the latest idiot who's clueless, and get to work.  This is pro equipment designed for pros, and things like boot time, screen angle and things like that are pretty irrelevant.  Stuff that annoys me is performing out of spec, strange GPIB lockups in the middle of the night so all my testing is garbage, etc etc.

I'm not saying that it wouldn't be BETTER in some sense if this other stuff was better, but it's really not even on the radar when I'm selecting equipment.  Shoot, it takes me longer to remember where I was from yesterday than it does for any of this stuff to boot up.

Several issues comes to mind here:

*) In many cases with top shelf equipment you have physical limitations on the power on delay, like ovens needing to warm up etc. Here there is often little point in having the UI operational after a few seconds. If you actually need the precision, then you gotta wait for things to warm up correctly (assuming you actually do turn off your frequency standards and the likes, of course). In this case the UI boot delay is not an usability problem.

*) However, many people actually do move their test equipment around on the shop floor, wheeling it around down in the cellar next to the particle accelerator etc. That is why many of the smaller pieces of test gear from Agilent have rubber corners and a handle. In these situations you frequently do power cycle the equipment several - or even many - times a day. If you work in product calibration, then your test setup may consist of, say, a bunch of power supplies, DMMs, counters and more, all wheeled around on a cart. Here the boot delay frequently doesn't serve a physical purpose, and it gets to be a real pain in the rear, for a whole bunch of reasons.

*) My home lab is probably not a good example to go by. However, most labs I have worked in, seem to have been temporary affairs (for some suitable definition of 'temporary'), frequently being both fairly small and with questionable A/C - where it exists at all. In these situations you tend not keep your whole collection of stuff powered up all day long due to heat build up, and potentially noise as well. Thus, when you realize you need the DMM, you get to wait for it to boot.

WinCE because they already have the test procedures, the build libraries and the firmware in use on other devices, so it is not a big cost and learning curve to use another OS. The cost for them is essentially zero to use it, anything else will have a bigger cost in time or development.

I suspect few would question Agilent's wisdom in focusing on a single OS across their product range. What I am asking is whether people have heard or seen Agilent mention that WinCE has technical merits, which warrants the UI boot delay? If Agilent had chosen, say, QNX instead of WinCE, then a fast UI boot would have been an option. So our DMMs could be ready in seconds, and the 'scope with the complex power-on chaining and self calibration routines will be ready when the hardware is ready.

To put it differently, then I personally prefer instruments with LED, VFD and CRT readouts, compared to LCD screens. This due to the higher illumination intensity and increased contrast, which is available when using these technologies. However, upgrading to a full color LCD screen as seen with the 34461A is a valid design choice IMHO. By doing so you get the possibility of making an UI both more complex/powerful yet still easier to use/operate, compared to being limited by, say, the VFD readouts.
 

Offline SeanB

  • Super Contributor
  • ***
  • Posts: 15094
  • Country: za
Re: EEVblog #489 - Agilent 34461A Multimeter Review
« Reply #68 on: June 30, 2013, 10:58:24 am »
They could cut boot time a lot by having a ready to run frozen image that is decompressed to ram, registers in expected stuff set up and then control is transferred to it from a small basic bootloader. You can get this to go in less time than it takes for the LCD controller to come out of reset from power on. It works well for embedded stuff that only sends data out but you need extra if you want to have an ethernet stack, as you need to set it up.
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 29694
  • Country: au
    • EEVblog
Re: EEVblog #489 - Agilent 34461A Multimeter Review
« Reply #69 on: June 30, 2013, 11:13:43 am »
*) However, many people actually do move their test equipment around on the shop floor, wheeling it around down in the cellar next to the particle accelerator etc. That is why many of the smaller pieces of test gear from Agilent have rubber corners and a handle. In these situations you frequently do power cycle the equipment several - or even many - times a day. If you work in product calibration, then your test setup may consist of, say, a bunch of power supplies, DMMs, counters and more, all wheeled around on a cart. Here the boot delay frequently doesn't serve a physical purpose, and it gets to be a real pain in the rear, for a whole bunch of reasons.

I've built countless mobile production and service test jigs. Most of them that require gear such as this (i.e. high end/high precision) will inevitably have a PC based control and logging system on the trolley as well. So the boot time of the PC and program usually dominates.

Quote
*) My home lab is probably not a good example to go by. However, most labs I have worked in, seem to have been temporary affairs (for some suitable definition of 'temporary'), frequently being both fairly small and with questionable A/C - where it exists at all. In these situations you tend not keep your whole collection of stuff powered up all day long due to heat build up, and potentially noise as well. Thus, when you realize you need the DMM, you get to wait for it to boot.

That's what hand held meters are for. Or a bench meter without such high end system functionality that is instant-on.

And that's the kicker. This a complex system multimeter with oodles of power and functionality, it is not designed for the push-on and take a simple measurement crowd.
If any buys it for that and them proceeds to complain about the boot up time, then they are being rather silly.
 

Offline John Coloccia

  • Super Contributor
  • ***
  • Posts: 1199
  • Country: us
Re: EEVblog #489 - Agilent 34461A Multimeter Review
« Reply #70 on: June 30, 2013, 11:59:05 am »
Dave nailed it with the PC requirement too.  Anyhow, we used to do stuff like this all the time.  I wasn't doing testing and calibration, but the nature of my work required moving setups around and things like that.  I went to the store up the street and bought a $50 UPS.  All this stuff should be protected somehow anyway.  Then I can unplug the setup and wheel it anywhere I want with no drama.

So I hear what you're saying, but as a practical matter it's still a non-issue.  Maybe there's that one guy in a million whose job it is to walk around all day long, on 3 different floors, and do nothing but take a measurement on 30 different test stations.  I would suggest that a bench meter is maybe not the right tool for the job.  Maybe something like a Fluke 726 would be a better choice.
 

alm

  • Guest
Re: EEVblog #489 - Agilent 34461A Multimeter Review
« Reply #71 on: June 30, 2013, 12:12:55 pm »
It's positioned as a replacement for the 34401A, which had a much faster boot time. The 34401A was also the entry level bench meter in the HP/Agilent range for a significant amount of time (before the Escort-designed meters and 34405A/34450A came along). It was not a high-end/high-precision meter within the HP/Agilent line. The 34401A was and is common as basic bench meter in many labs, and I can certainly see people putting it on a cart without sophisticated analysis software.

You may argue that the 34461A is not a good replacement in this application, but I don't think Agilent is positioning this meter purely as a system meter with all the fancy plotting features. In my opinion, the long boot time is a sign of lack of attention to detail, especially given the CPU power available compared to the 34401A. All the time-outs that are expiring (which are presumably still there in the production firmware since it takes almost as long to boot) show that they did a poor job on configuring WinCE. If we complain about the boot time of low-end Chinese scopes, then it would be unfair to hold Agilent to lower standards.
 

Offline Bob S

  • Contributor
  • Posts: 23
Re: EEVblog #489 - Agilent 34461A Multimeter Review
« Reply #72 on: June 30, 2013, 02:32:39 pm »
There has been much said on the topic. I am not a fan of anything Windows but....
If you are in need of such a precision meter, you should let it warm up before you use it. In this way the boot time is of no concern. I have this meter and can only say that using it so far has been worth the wait.
 

Offline AlfBaz

  • Super Contributor
  • ***
  • Posts: 2007
  • Country: au
Re: EEVblog #489 - Agilent 34461A Multimeter Review
« Reply #73 on: June 30, 2013, 03:20:11 pm »
Considering that having trending, histograms, and other such features that normally require a pc running software to do, I think the boot time is an improvement!  :box:
 

Offline robrenz

  • Super Contributor
  • ***
  • Posts: 3035
  • Country: us
  • Real Machinist, Wannabe EE
Re: EEVblog #489 - Agilent 34461A Multimeter Review
« Reply #74 on: June 30, 2013, 03:58:37 pm »
A nice use of Trend Plot is to identify your meters actual warm up time by shorting the inputs and turning the meter on and instantly turning on trend plot in DCV. The chart will show the Logaritmic? warm up curve and you can determine visualy what your meters actual warm up time is instead of using the manuals generic time. On my 8846A I find it is about 30 minutes instead of the stated 1 hour. I would guess the 34461A would be similar.


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf