looking for decent multimeters

[robrenz]

M. András

One thing to be carefull about if you are buying a new meter especially since you are concerned about length of calibration.
You want to be sure you will be getting a meter that has just been built and calibrated or at least freshly calibrated before it ships to you. When I was shopping for my 8846A in the  US several places had them in stock but when I inquired about how long they had been on the shelf some had been there over a year. When I asked what about calibration the answer was it is as is, meaning the year specs were technically expired. I finally found tequipment.net where the meter is ordered from fluke and built to order obviously with fresh calibration (took 2 weeks).   Also ask for quotes of their best prices. tequipment.net shows the retail price of US$1425.00  but they quoted me US$1296.00  If you buy one in the next 60 days they will include the 2x4wire Ohms test leads http://www.tequipment.net/Fluke8846ASpin.html  I realize you are in Europe but the 8846A is built in the US so it would have to be shipped from the US to you anyway. 

[M. András]

there is an official distributor here in Hungary for the fluke where i asked for the 289 price and you can get that from stock, not many Hungarian company/distributor stocks damn expensive stuffs, they do calibration services too, but i dont know if they calibrate it when you buy the instrument or just gives the sealed box which was hell knows when was packed and shipped here from the US, i will ask them when i have time to go to their store, anyway the 289 is 1.5x more expensive here compared to digikey and tequipment prices, and of course 25% vat included in that 1.5x price (i love the direct 1:1 usd/eur conversion rather than the eur/usd to huf conversion ) i dont know how much would be import these from the us cos i cant find any information at the local customs website, i didnt find a category for test equipment but i find 1 for nuclear reactors

[M. András]

fluke 289 was 583 usd+ 25%vat about month ago at the distributor calculated at the current exchange rates, from farnell 662usd+25% vat and the eur model is 612usd+25%vat
8846: 1554usd+25% vat, 8845: 1175usd+25% vat, 8808: 776usd+25% vat.
agilent: 34401: 1016usd+25% vat, 34405: 720usd+25% vat

just for a quick price chart here in this country inside the EU

at the distributor they quote 58usd baseprice and 2usd for every measurment points(dunno what do they mean under this) for calibration for a multimeter

[robrenz]

at the distributor they quote 58usd baseprice and 2usd for every measurment points(dunno what do they mean under this) for calibration for a multimeter

min, mid, and max for a range would be three measurment points x 5 ranges would be 15 points just for DCV then do that for again for ACV, DCI, ACI, Capacitance, frequency, resistance.  You could ask for just max of each range to minimize cost.

[M. András]

at the distributor they quote 58usd baseprice and 2usd for every measurment points(dunno what do they mean under this) for calibration for a multimeter

min, mid, and max for a range would be three measurment points x 5 ranges would be 15 points just for DCV then do that for again for ACV, DCI, ACI, Capacitance, frequency, resistance.  You could ask for just max of each range to minimize cost.

thats a lot 

do the fluke instruments comes with any sticker or cal certificate from the manufacturer when you buy them like the agilent ones? anyway i guess they do a full calibration anyway before they pack each of the instruments into their boxed

[robrenz]

do the fluke instruments comes with any sticker or cal certificate from the manufacturer when you buy them like the agilent ones? anyway i guess they do a full calibration anyway before they pack each of the instruments into their boxed

Pictures of:

• Cal sticker
• Tamper proof sticker
• Calibration docs
• Misc. screen shots (on next post)

Look at the Cal. data and see how much more accurate the meter is than the one year spec. You have read the Fluke calibration pdf and this Calibration shows how conservative that 2.6 standard deviations is.  The screen looks better than the pictures, its hard to take pics of that display.  The frequency pic. is blurred because of long shutter speed and digits bobbling.  Notice the 10 nanoOhm internal resolution on the trend plot picture. Second display is showing in pics that use it.

[robrenz]

Remaning pictures.  Ok, I guess I am a 8846A Fanboy

[grenert]

Look at the Cal. data and see how much more accurate the meter is than the one year spec. You have read the Fluke calibration pdf and this Calibration shows how conservative that 2.6 sigma is. 

Well, I'm sure it's a fine meter, but they really should have included the 24-hour spec with the data, which would have been a more appropriate comparison, since this was probably done right after calibration.  Showing this "one hour" data point with the one year spec inappropriately exaggerates the accuracy.  It's probably well within the 24 spec anyway, so they might as well make it easier to see (they publish 24-hr specs in the manual).

Thanks for sharing the images 

[robrenz]

at the distributor they quote 58usd baseprice and 2usd for every measurement points(dunno what do they mean under this) for calibration for a multimeter

So the Cal I showed has 125 measurement points so that would cost you 308usd. So if a dealer wants to sell one off the shelf you should either get a fresh Cal for free or a 308usd discount.

[robrenz]

Look at the Cal. data and see how much more accurate the meter is than the one year spec. You have read the Fluke calibration pdf and this Calibration shows how conservative that 2.6 sigma is. 

Well, I'm sure it's a fine meter, but they really should have included the 24-hour spec with the data, which would have been a more appropriate comparison, since this was probably done right after calibration.  Showing this "one hour" data point with the one year spec inappropriately exaggerates the accuracy.  It's probably well within the 24 spec anyway, so they might as well make it easier to see (they publish 24-hr specs in the manual).

Thanks for sharing the images 

My point here was more about how meter specs are a statisticaly based such that such that in flukes case they use 2.6 standard deviations such that 99% of meters will meet the specs as produced and after a year of additional aging/component drift ( http://support.fluke.com/calibration-sales/Download/Asset/2547797_6203_ENG_B_W.PDF ). That also means there is a very good chance your meter is way more accurate than the worst case Specs .   Which is shown in this particular Cal (.  I would Think the same thing would apply to Agilent or Keithley procucts. 

[alm]

I'm sure the other reputable manufacturers also apply guard banding to make sure their products meet their specs, but Fluke is the only one I've found to explicitly state that the specs are 99% confidence intervals. This is where the 2.6 * SD figure comes from, a two sided 99% confidence interval of a normal distribution is equal to the mean +/- 2.6 times the standard deviation.

[M. András]

at the distributor they quote 58usd baseprice and 2usd for every measurement points(dunno what do they mean under this) for calibration for a multimeter

So the Cal I showed has 125 measurement points so that would cost you 308usd. So if a dealer wants to sell one off the shelf you should either get a fresh Cal for free or a 308usd discount.

i will ask the distributor when i will buy it, it would be good if they would calibrate it for free or even check the cal and if necessary they will calibrate the necessary ranges, i calculated the price for the 289 which is more likely to be bought first almost the half of the meter original price . i found another calibration labs whit fix prices for a full calibration for approx 200usd for multimeters

[grenert]

The reading rate and offset ohms are the only pluses I see for the 34411

I sent an email to Fluke asking if they had something like offset ohms.  This was their reply:

Yes the zero function works in 4 wire ohms.  Short all your test lead together, the press the zero button.  You will see the word "math" appear on the screen.  This takes the lead resistance out of your measurement.  Remove the short then take your measurements. 

That doesn't seem quite right to me, because you still have potential thermal EMF at the lead/resistor junction that won't be compensated for.  Could someone tell me if my interpretation is correct?

[alm]

Maybe you should ask for offset compensated ohms? Or just how to compensate for offset voltages, like thermal EMF, in resistance measurements.

[robrenz]

The reading rate and offset ohms are the only pluses I see for the 34411

I sent an email to Fluke asking if they had something like offset ohms.  This was their reply:

Yes the zero function works in 4 wire ohms.  Short all your test lead together, the press the zero button.  You will see the word "math" appear on the screen.  This takes the lead resistance out of your measurement.  Remove the short then take your measurements. 

That doesn't seem quite right to me, because you still have potential thermal EMF at the lead/resistor junction that won't be compensated for.  Could someone tell me if my interpretation is correct?

As far as I know zeroing the meter does nothing for thermal EMF.  You need to use current reversal on 4 wire kelvin connected setup. Take one reading, reverse the current take another reading and average them, that eliminates all thermal emf votages in the measurement. I made a setup for my 8846A to make the current reversal easy see reply 15 on this post https://www.eevblog.com/forum/index.php?topic=5460.msg71401#msg71401 
Offset Ohms takes a voltage measurement with no drive current, then takes a measurement with drive curent and take the difference between them.

This keithley book is excellent and covers all this and more
The whole book is excellent.

http://www.keithley.com/knowledgecenter/knowledgecenter_pdf/LowLevMsHandbk_1.pdf

[M. András]

i read somewhere here there are some autocalibrating bench meters, what are the entry level cost of these?

[alm]

I believe autocal was a Datron brand name that stood for closed case calibration, ie. no need to adjust trimmers, only software constants. Many recent DMMs support this.

You may also be referring to artifact calibration, calibration with just a voltage reference and a reference resistor. As far as I know the HP/Agilent 3458a is the only DMM that supports this, expect to pay a least a couple of thousand dollars on the used market, and more new.