EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: Iwanushka on April 03, 2014, 11:06:41 pm
-
Hi guys and gals!
I'm new here so please don't beat me up too hard :)
A few weeks ago I bought Agilent U1252B and U1272A :-+, first day both DMMs tested perfect, but the other day I saw quite a strange thing on U1252B: if I short test leads (which should be around 0.4R) it shows -0.03R, and yes readings are the same at all times and places :palm:, at first it was showing something around +0.04R; my U1272A still shows 0.04R and looks fine. So did my DMM went crazy or those things are calibrated to compensate for test leads and if they are still its a bit fail? :--. It little bit drives me crazy 'cause DMM for 650$ can't keep track of ohms... :-//
UPDATE1: NULL/Rel function is turned off, so it's not compensating for some random value
UPDATE2: as you guys suggested played around with a few resistors and managed to kill second DMM |O, I will update the post after I calm down a little....
UPDATE3: both DMMs before playing around: U1272A shorted leads 0.034R, U1252B shorted leads -0.03R.
1) Measured 100R, 1K and 10K vishay 0.01% 2ppm resistors and compared results with DER EE DE-5000, all meters are a few least significant digits off from the nominal -> well in spec.
2) measured 0.3R resistor with U1252B -> 0.3R, then with U1272A -> 0.34R, then with DE-5000 -> 0.33R, so far so good.
3) measured 0.3R again with U1272A (like 1min after I pulled it out of DE-5000) and I was like what the hell it shows 0.3R where those 0.04R went.
4) shorted leads on U1272A boom -0.8R and recovering (going to positive side), switched to mV range -0.04mV :palm: After reset and a few on/off cycles it's working just fine :)
Looks like it's recovering but what the hell is going on :-//? I always knew that HP was total :bullshit:, but I still can't believe that measuring a 1 cent resistor can kill 650$ meter and not one but two of them.
Agilent do I need to short freakn' resistors before measuring them :-DD?
P.S. Please excuse my English, it's quite a terrible mix of America/British English with some Engrish ;/
-
Does it read resistors correctly, eg 10, 100, 1k 100k?
Does ac and dc volts, current work correctly?
-
PS Is the 'null' reading enabled?
-
Mine behaves exactly the same way, and it's normal.
Touch the leads together, and wait for them to stabilize. Once the lead resistance value has settled, you hit the NULL key, and it will display zeros, perhaps the least significant digit fluctuating between zero and one (NULL hit just before it's truly settled tends to cause the latter condition). Not only does this give you zeros, but more importantly it's what those zeros mean (test lead resistance is removed from the measurement value).
Hope this helps. :)
-
Mine behaves exactly the same way, and it's normal.
I have never used any Agilent meters so forgive my newbie question, but are you saying that shorting the probes resulting in a negative ohm reading of 0.03ohms is normal?
Or am I misinterpreting something?
-
Updated first post with more findings.
-
Something is still a bit funny, I have not had any problems with a U1253B (for about 3 years), the Agilent lead resistance I get about 0.04 Ohms without nulling.
Reliable and repeatable measurements to better than 0.1 Ohm accuracy is a real art, I think you have to have a pure copper circuit due to electrochemical potentials. See Volt Nuts. I think a difference of 40mOhm for repeated measurement as pretty good.
There was an Agilent software update for my U1253B ( for a battery charging issue and I think a AC RMS issue), have you looked to see if you have the latest s-ware version.
Another thought to try is I recall there is a factory setting restore in one of the menus, in case some gremlin has got in somewhere!
The other thought is could the leads be contaminated and a small electrochemical potential is giving the incorrect readings?
-
Something is still a bit funny, I have not had any problems with a U1253B (for about 3 years), the Agilent lead resistance I get about 0.04 Ohms without nulling.
Reliable and repeatable measurements to better than 0.1 Ohm accuracy is a real art, I think you have to have a pure copper circuit due to electrochemical potentials. See Volt Nuts. I think a difference of 40mOhm for repeated measurement as pretty good.
There was an Agilent software update for my U1253B ( for a battery charging issue and I think a AC RMS issue), have you looked to see if you have the latest s-ware version.
Another thought to try is I recall there is a factory setting restore in one of the menus, in case some gremlin has got in somewhere!
The other thought is could the leads be contaminated and a small electrochemical potential is giving the incorrect readings?
I will try to reset them, I have 3 pairs of brand new agilent leads and bunch of el cheapo leads, readings are always repeatable except for the least significant digit, but things went wrong after checking that 0.3R resistor on u1272a today and it is still recovering now it shows around 0.025R and slowly rising, voltage on shorted leads is getting better too.
regarding firmware both of them have latest versions 2.04 for u1272a and 2.17 for u1252b
-
Yes it is normal. You do not know how the machine was calibrated. Let's say it was calibrated using wires of exactly 1 meter long. At that step it reads exactly zero. Now you connect the testprobes. And the loop is slightly less then 1 meter. You will get a negative reading.
If the loop is slightly longer it will read a bit above.
All meters, irrespective of manufacturer behave that way. If they stay at zero it is because they arent accurate enough. Cable length has an impact on the reading.
The only way to get rid of cable impact is to do a 4 point measurement (so called kelvin connections)
If your machine cant do that you use the null button.
-
Yes it is normal. You do not know how the machine was calibrated. Let's say it was calibrated using wires of exactly 1 meter long. At that step it reads exactly zero. Now you connect the testprobes. And the loop is slightly less then 1 meter. You will get a negative reading.
If the loop is slightly longer it will read a bit above.
All meters, irrespective of manufacturer behave that way. If they stay at zero it is because they arent accurate enough. Cable length has an impact on the reading.
The only way to get rid of cable impact is to do a 4 point measurement (so called kelvin connections)
If your machine cant do that you use the null button.
Hi, strange thing is that out of the box u1252b was measuring ~0.4R on shorted leads now its -0.04R I don't believe that length of cable changed over night.
Second finding on u1272A, short leads after it recovers (takes ~an hour) it will be 0.03-4R, measure voltage on leads its like -.004mv, connect any resistor, then short leads boom it shows 0.00R or even something negative up to -0.xR, hit mv range it shows 10x voltage, looks like it's building a charge inside.. leave it alone for an hour it's back to 0.04R and voltage on shorted leads drops by 10x did reset turned on/off a few times now it works fine :)
P.S. I don't care about the accuracy, but I care about repeatability, how you can trust instrument that show's random things.....Also I never seen this behavior on Fluke and/or One-Hung-Low gear
-
Everything else looks spot on, so maybe "contacts got better over time".
-
Yes it is normal. You do not know how the machine was calibrated. Let's say it was calibrated using wires of exactly 1 meter long. At that step it reads exactly zero. Now you connect the testprobes. And the loop is slightly less then 1 meter. You will get a negative reading.
If the loop is slightly longer it will read a bit above.
All meters, irrespective of manufacturer behave that way. If they stay at zero it is because they arent accurate enough. Cable length has an impact on the reading.
The only way to get rid of cable impact is to do a 4 point measurement (so called kelvin connections)
If your machine cant do that you use the null button.
Vincent, can you tell us more about that? ??? Seems very interesting :D
-
I have never used any Agilent meters so forgive my newbie question, but are you saying that shorting the probes resulting in a negative ohm reading of 0.03ohms is normal?
I was thinking in terms of the value showing prior to settling or nulling the value displayed. I even pulled mine out to confirm the issue using the stock Agilent leads (left Auto ranging enabled) prior to posting.
To clarify the values in my case, I get +0.04Ohm readings once it's settled fairly consistently. Never got -0.03Ohm, but before it's settled, there can be instances where the - (minus) sign does illuminate for just a moment on rare occasion, and don't recall any digits displayed at that instant (i.e. rub the leads together/don't have the best contact). Values can jump over a wider range prior to settling (0.01 - 0.12Ohm IIRC w/ stock Agilent leads; Probemaster's come 0.01 - 0.04Ohm, then tend to settle in the 0.01 - 0.02Ohm range).
Do keep in mind, this is fast meter for the counts (60x/sec on the bar graph, 5 - 6x/sec on the displayed value in Auto ranging IIRC), but isn't an absolute speed demon for settled values.
Turn off Auto range, and it settles a bit quicker, and can't recall ever seeing the - (minus) sign ever display. Better leads tend to improve matters as well (better contact, so the meter settles just a bit quicker).
I will try to reset them, I have 3 pairs of brand new agilent leads and bunch of el cheapo leads
You might want to try cleaning the probe tips with a little alcohol or similar, and see if that helps matters. I'd also get another good set of leads to check with (Agilent's are not junk by any means, but there are better ones out there IME). FWIW, Probemaster sets I have perform better than the stock Agilent leads.
BTW, did you try pulling the battery for a few minutes as a reset method?
Might be worth a shot if you didn't.
-
See how consistent the readings are on a plain banana jumper plugged into the meter directly.
-
Quick update,
First of all thank you all for your input, tried around 6 different leads with 0.03-4R (measured with u1272a) and a few with ~0.4R. With all low resistance leads best I can get is -0.01-2R which is stable, with all leads u1252b measures 0.03-4R less than u1272a, tried to remove the battery, soft reset, cleaning tips/clips has no impact on readings. Going to build banana jumper.
-
See how consistent the readings are on a plain banana jumper plugged into the meter directly.
Hooked banana jumper on u1252b:
Attempt #1: it stabilized at -0.07-8R (least significant digit is not stable but that's expected) then it went crazy and I turned it off after it reached around -0.12R
Attempts #2-3: it stabilized at -0.07-8R (least significant digit is not stable but that's expected) for 30-60sec and then started jumping, everything happens instantly so it's hard to see what's going on but looks like it went to -0.03R and then to -0.13R and then back to -0.07-8R, after every 30-60sec the same thing happens again.
u1272a stays stable at -0.047-8R, no crazy things happening.
-
I think you're supposed to remove the inline batteries in the test leads....
(LOL)
if you are getting negative ohms, the test leads are not adding enoug resistance. have you tried longer leads? that might be the easiest way to 'zero' this, really.
-
Could it be a dodgy battery? Intermittent contact within battery?
I don't know the the 1252, but the 1253 has to be set for the battery voltage (in one of the menus) could it be set incorrectly and overloading V references???
PS Remember an idiot can ask more questions than the wisest man can answer. (HiHi) ;D
-
Could it be a dodgy battery? Intermittent contact within battery?
I don't know the the 1252, but the 1253 has to be set for the battery voltage (in one of the menus) could it be set incorrectly and overloading V references???
PS Remember an idiot can ask more questions than the wisest man can answer. (HiHi) ;D
I will replace current 8.4V battery with alkaline 9v will see what happens, in settings battery is set to 8.4V, so no problems in configuration.
Update: tried a few alkaline 9v batteries no change.
-
Vincent, can you tell us more about that? ??? Seems very interesting :D
simple: we are talking 0.03 ohms here. simple testleads are in that range.
lets say the meter was calibrated using 1 meter long leads.
stick in longer leads ( or thinner leads ) and you will get a larger positive offset.,
sitck in shorter leads ( or thicker leads , conductor wise of course). and the meter will read a smaller number, possibly negative.
so it is possible to see such behavior.
that's why we have null buttons.
if you want to measure stuff in the 0.0x ohms range you need to apply kelving sensing techniques.
-
simple: we are talking 0.03 ohms here. simple testleads are in that range.
...
So in the factory they used some long/thin leads to hook up this multimeter to their calibration standards. And that makes the offset, the wires. I thought they would compensate for the wires in the resistance standard, but when I think about it, wires probably add more error to the measurement than the resistance standard.
I guess nothing is wrong here, but I find it weird. We shouldn't care. 4 wire measurements exist for a reason. :-\
-
simple: we are talking 0.03 ohms here. simple testleads are in that range.
lets say the meter was calibrated using 1 meter long leads.
stick in longer leads ( or thinner leads ) and you will get a larger positive offset.,
sitck in shorter leads ( or thicker leads , conductor wise of course). and the meter will read a smaller number, possibly negative.
so it is possible to see such behavior.
that's why we have null buttons.
if you want to measure stuff in the 0.0x ohms range you need to apply kelving sensing techniques.
Thank you for the information on how these things are calibrated. My main concern is not that it measures -0.03R actually I would not care it if even measured -5R, I can always use NULL functions if it's stable. My main concern is that DMM out of the box was measuring +0.03-4R depending on the leads, after that it spent a week or two in it's carrying case turned off and now it's -0.03R,now I can't depend on it and who is going to take responsibility if it drifts significantly in next 6 months and I end up killing some expensive gear or even worse my health...
P.S. I have a few 5years+ DMMs that still measure spot on shorted leads from the first day.
-
The 1252B data sheet says the 500ohm scale error with the NULL function is 0.05% +10 counts of the least significant digit. So when shorted, then NULLed, wouldn't anything +/- 0.1ohm be withing spec? That 10 counts of possible error is 10x the resolution of the meter on the 500 ohm scale. IMO the meter shows you more resolution than you should believe.
I understand that you are worried that it is drifting, so why not ask Agilent customer service for an answer, you paid for that. I have had good experiences with them responding to my questions.
As a comparison my FLUKE 87V which is 0.2% + 2 counts, flickers +/- 0.1 ohm (+/- 0.02ohm in High Res mode) with a shorting bar at the meter after REL (same as NULL).
-
My 1273A reads -0.044ohm and my fluke reads -0.01ohm