Fluke 17B accuracy?

[vtl]

I've just tweaked my 17B to provide better accuracy. The Basic DC volts reading is adjusted with just 1 pot for all ranges. I'm not familiar with how calibration procedures work but the manual seems to tell the user to only adjust the pot until its on the cusp of 0.5% accuracy and no better? Or maybe thats just my literal intepretation

[Achilles]

well, not that irrelevant. It is common to export higher grades and keep lower grades for inland consumption to keep a good reputation on international markets.

And the other way is also common. This area produces a special alcoholic beverage. The good stuff is all sold locally, and the horse piss version that no one here would touch is exported.

Yeah, but you can get B+**di there, too . Export stuff at least seems to have higher %vol. And well, I just take nice stuff......no big marketing brand.....rhum agricole de Martinique 

@VTL: Maybe they want no so high accuracy to keep the customer open fir higher spec stuff which is usually more expensive. Great that it works. So if you have some good referencing source, take it as accurate as you can get. It would be interesting for a long tim insight how good they keep that calibration.

[samgab]

I didn't expect that of a Fluke meter, even if it is a low end model.

Expect what of a fluke meter? We have three low cost devices in a dubious test situation. What is to say any of them are correct?

The error is less than 1% the quoted accuracy of the 17B is  DC Volt : 400mV / 4V / 40V / 400V / 1000V , +/-1.0% + 10

I see some people just invent figures and quote them to sound like they know what they are talking about:
The actual quoted accuracy spec of the 17B in the DC 4.000/40.00/400.0/1000 volts range is +/-(0.5% +3 counts LSD).
Only the DC 400.0 Millivolts range is quoted at +/-(1.0% +10 counts LSD).
Even so, the meter (it's my meter, by the way) was within spec.
5.0000 Volts with +/-(0.5% +3 counts LSD) means any reading between 4.95 and 5.05 Volts is within spec on that (40.00V) range.
So it made it ...just.
But I think it shows somewhat of a lack of care and attention to detail to sell a meter that far off accuracy.
Yes, if one has a "within spec is good enough" attitude, then it's okay. But with that sort of attitude, one doesn't maintain a company reputation for accuracy and care for detail.
But, it is a relatively low cost meter. So whatever. I've adjusted it now and it reads correctly.
If I was a Digital multimeter manufacturer, I would not be satisfied selling someone a meter that read 4.96V from a 5.0000 V source, even if it was "within spec".
If I was a Fluke employee sitting on the manufacturing line responsible for calibrating the meters as they go out, with a high precision calibrator in front of me, and a device came through reading just a tad within spec, but that far off accuracy, I would not be content to let it go through without correcting it. 5 Volts is 5 Volts, regardless of how many zeros you put after the decimal point.

The TI chip accuracy is quoted at 1%. I doubt very much the millispec part is used, I can see the rest of the parts are no better than 1%. What is the accuracy repeatability of the rest of the device?
Sure it's checked against a laboratory device. That is before shipping storage, operator manhandling etc. Were the temperature, humidity etc identical to those When was the calibration pot was last tweaked (knocked)?

More off the top of your head spouting...
The TI 5050 standard grade chip is quoted at 0.1% inaccuracy MAX. But that isn't really the main concern in this instance.
The device is aged for >250 hours then it is carefully trimmed using a precision 10 turn trim pot to exactly 5.0000 using a calibrated 8.5 digit HP 3458A  DMM that has an accuracy of 0.0008%.  (Tektronix Certificate of Calibration # 4787923 valid through January 19, 2012).
The temperature was recorded at 21 degrees C at the time of calibration of the voltage reference.
So more important than the accuracy of the TI chip, is the stability, or level of drift, which is stated at 8ppm/degree C MAX temperature drift, and 5ppm/1000 hours long term stability. But much of this possible time related drift happens in the first 250 hours (as proven under testing and shown in the data sheet), which is why Doug - who makes the voltage reference device - ages the chip for 250+ hours prior to trimming.
The rest of the components are high quality parts but don't affect the voltage reference, if you know how the chip works.
Also, I don't know how you "can see  the rest of the parts are no better than 1%", when if you actually looked at the device, and knew what you were talking about, there are several resistors that are clearly marked to be of 0.01% accuracy (either in writing on the brown resistors, or with a violet coloured 5th band).
The date of calibration of the voltage reference I was using was exactly one week prior to the date the photos were taken.
The temperature at the time the photos were taken was also 21 degrees C  (+/-1 degree), the same as that at the time of calibration.
As for transit, the device has had a touch of enamel put on the trim pot adjuster after calibration to prevent change, and it is still in place un-altered.
It was carefully packaged and shipped.
There was no "operator manhandling" thank you.
Yes the temperature was within a degree of when it was calibrated 1 week prior, and no, it hasn't been "knocked".
The DMMCHECK device is guaranteed to be accurate to within 0.01% (plus or minus 500uV) for a minimum of 6 months, but as it has just been fine tuned, it is much more accurate than even that at present.

Seriously if a couple of sub $100 meters and a mail order reference board are that close in a knock about workshop test what is the problem? 
The Fluke could be spot on, there could be error each way or thing could be as pictured.

The Fluke was not spot on, but it is now that I've adjusted it. Now it matches the readings of my 87 V.
The reason a difference of .04 V can be of concern to me, is when charging li-ion cells.
The difference between charging a li-ion cell to 4.20 V and 4.24 V can have a big impact on the long term life of the cell. They really are that sensitive.
Safety of the charge is also impacted if a reading is out by even such a small amount.
The same is true with a .04 V difference with lead acid batteries, The difference in voltage between nearly 0% SOC and nearly 100% SOC is only about 0.6 V in some cases, so a .04 V difference can be important, especially if working with individual cells.

Anything requiring better accuracy shouldn't be using this class of device.

Quite right, and that is why I bought a Fluke 87 V as well.
A hobbyist can perceivably have need for good accuracy, but not have the funds to spend on something like a Fluke 87 V or something else in that class.
There is no need for elitist snobbery, it simply makes sense when on a very truncated budget to attempt to get the best item one can afford.
If all one can afford is a <$100 meter, and one wants to get the best possible DMM one can within those constraints,
in the past, one could be forgiven for thinking that by getting a Fluke there was a certain promise of care and accuracy in manufacturing and setting them in the factory before shipping.
Clearly, with these cheaper, Chinese manufactured Flukes, that isn't the case.

Having said that, the build quality of the 17B seems really good, it has good components, and is nicely soldered and well put together and designed.
It was just let down by the lack of a tiny turn of a couple of trim pots. A lack of care and attention to detail.
Yes it was within spec, but surely they should make an attempt to have the units leave the factory "ACCURATE" not just "WITHIN SPEC" It's the matter of a slight turn of the trim pot. and makes a huge difference to the end user, who may not have access to the high accuracy calibration devices that they have at the manufacturing location.
If then, after a few years it's reading slightly out, well one might expect that, but this device was only a couple of Months old, and should have been accurate, in my opinion.

/rant.

[Uncle Vernon]

I see some people just invent figures and quote them to sound like they know what they are talking about:

I see a 2 post jackass spouting shit. The figures quoted were directly of the sales page. Want to argue about you half a percent on a cheapo meter take it up with them or someone who cares.

The actual quoted accuracy spec of the 17B in the DC 4.000/40.00/400.0/1000 volts range is +/-(0.5% +3 counts LSD).
Only the DC 400.0 Millivolts range is quoted at +/-(1.0% +10 counts LSD).
Even so, the meter (it's my meter, by the way) was within spec.

Then may you be eternally happy with your $100 gem.

But I think it shows somewhat of a lack of care and attention to detail to sell a meter that far off accuracy.

But you purchased it anyway and were blissfully happy till some one else point out that it may not be the sharpest tool after all.

Yes, if one has a "within spec is good enough" attitude, then it's okay. But with that sort of attitude, one doesn't maintain a company reputation for accuracy and care for detail.

But you bought one, didn't you, could be they had the product and it's target market sussed.

But, it is a relatively low cost meter. So whatever. I've adjusted it now and it reads correctly.

Meter doesn't say what you want then mess with it till it does.

If I was a Digital multimeter manufacturer, I would not be satisfied selling someone a meter that read 4.96V from a 5.0000 V source, even if it was "within spec".

Oh I don't know if you were making dollars, selling meters to people like you it may soften the blow some what.

If I was a Fluke employee sitting on the manufacturing line responsible for calibrating the meters as they go out, with a high precision calibrator in front of me, and a device came through reading just a tad within spec, but that far off accuracy, I would not be content to let it go through without correcting it. 5 Volts is 5 Volts, regardless of how many zeros you put after the decimal point.

You think sub hundred dollar meters are actually hand calibrated in the factory. Muhahaha that's funny.

More off the top of your head spouting...

Well come on Einstein where is your supporting evidence to say I'm wrong and your POS meter is actually hand crafted by blonde virgins using only the finest of hand selected and individually tested top spec parks. 

The TI 5050 standard grade chip is quoted at 0.1% inaccuracy MAX.

And this by your logic proves the entire instrument must manage similar accuracy. Did you work this out in your clue free zone?

The device is aged for >250 hours then it is carefully trimmed using a precision 10 turn trim pot to exactly 5.0000 using a calibrated 8.5 digit HP 3458A  DMM that has an accuracy of 0.0008%.  (Tektronix Certificate of Calibration # 4787923 valid through January 19, 2012).
The temperature was recorded at 21 degrees C at the time of calibration of the voltage reference.

Then it's banged about in a (hopefully) padded post pack exposed to all manner of climatic aberrations during transit before arriving in the post box of everyday chronic fiddlers, allowing them to perform bizarre calibration  procedures at varying room temperatures with the aid of serious looks and kitchen scissors.

So more important than the accuracy of the TI chip, is the stability, or level of drift, which is stated at 8ppm/degree C MAX temperature drift, and 5ppm/1000 hours long term stability. But much of this possible time related drift happens in the first 250 hours (as proven under testing and shown in the data sheet), which is why Doug - who makes the voltage reference device - ages the chip for 250+ hours prior to trimming.

Yeah 250 hours sitting about at Doug's place ensures the device is totally impervious to all the rigours of transport and misuse by smarmy upstarts. You could probably even use one underwater and it's still be spot on eh?

The rest of the components are high quality parts but don't affect the voltage reference,

especially since they were all installed by nude virgins, under soft lighting, right?

Also, I don't know how you "can see  the rest of the parts are no better than 1%", when if you actually looked at the device, and knew what you were talking about, there are several resistors that are clearly marked to be of 0.01% accuracy (either in writing on the brown resistors, or with a violet coloured 5th band).

Am I supposed to swoon now?

And of course the manufacturer of those parts will guarantee this some time after assembly at Doug's place, uncertain transit and storage and operation in your NATA certified bedroom lab?

The date of calibration of the voltage reference I was using was exactly one week prior to the date the photos were taken.

And everyone knows calibrations must be accurate at times while a certificate is current, right? Underwater, swamped with RF, in any heat or humidy. Right?

The temperature at the time the photos were taken was also 21 degrees C  (+/-1 degree), the same as that at the time of calibration.
As for transit, the device has had a touch of enamel put on the trim pot adjuster after calibration to prevent change, and it is still in place un-altered.

Magic stiff that enamel if the'y only pour it into pots during manufacture then everything could be in calibration forever.  And I do trust you checked that temperature against a universal must be correct temp cal board.

It was carefully packaged and shipped.
There was no "operator manhandling" thank you.

Yep Blonde virgins accompanied the entire transit to your door. And you used all your skills as a practicing know-it-all to ensure things were just dandy in all weathers at your house.

Yes the temperature was within a degree of when it was calibrated 1 week prior, and no, it hasn't been "knocked".
The DMMCHECK device is guaranteed to be accurate to within 0.01% (plus or minus 500uV) for a minimum of 6 months, but as it has just been fine tuned, it is much more accurate than even that at present.

OH I have no doubt and it will accurately indicate the temperature six inches from your thumb plus or minus theodd draft through your bedroom lab.

The Fluke was not spot on, but it is now that I've adjusted it. Now it matches the readings of my 87 V.

So you potentially have two out of cal meters now accurate to plus or minus half to one on the doug scale?

The reason a difference of .04 V can be of concern to me, is when charging li-ion cells.

This is nice for you! Zero error everywhere when it's you and batteries, right. NASA would be impressed.

The difference between charging a li-ion cell to 4.20 V and 4.24 V can have a big impact on the long term life of the cell. They really are that sensitive.

And the extra life on the first 50  batteries will go part way to recouping the cost of your Doug board.

Quite right, and that is why I bought a Fluke 87 V as well.

And there was me thinking you bought it because people laughed at you for having a 17B.

A hobbyist can perceivably have need for good accuracy, but not have the funds to spend on something like a Fluke 87 V or something else in that class.

Just like amateur race car drivers can need a Ferrari yet live out their dreams in nan's old Honda Civic.

There is no need for elitist snobbery, it simply makes sense when on a very truncated budget to attempt to get the best item one can afford.

Which still does not explain how shelling out around a hundred bucks on a badge engineered promise brought you any closer to being able to eventually purchase of a meter you could rely on.

If all one can afford is a <$100 meter, and one wants to get the best possible DMM one can within those constraints,

Curious you had such a constraint and yet found funds for Doug's wonder board.  Purchase price of the 17B plus the board would have got you straight to the buy price of a better entry level meter.

in the past, one could be forgiven for thinking that by getting a Fluke there was a certain promise of care and accuracy in manufacturing and setting them in the factory before shipping.
Clearly, with these cheaper, Chinese manufactured Flukes, that isn't the case.

Did the meter perform within it's quoted specifications? And yet despite this and the kid zone price you expect it to somehow be better. How very realistic of your.

Having said that, the build quality of the 17B seems really good, it has good components, and is nicely soldered and well put together and designed.
It was just let down by the lack of a tiny turn of a couple of trim pots. A lack of care and attention to detail.

Detail you chose not to pay for. You do realise what individual hand calibration would do to the price of that 17B? No you wanted the impossible and refuse to include the eventual cost of the wonder board into your equation.

Yes it was within spec, but surely they should make an attempt to have the units leave the factory "ACCURATE" not just "WITHIN SPEC" It's the matter of a slight turn of the trim pot. and makes a huge difference to the end user, who may not have access to the high accuracy calibration devices that they have at the manufacturing location.

Whine, whinge, whine, but never admit you aren't prepared to pay the cost of such factory calibration. Besides whats to say it wasn't spot on until your third 1m drop of the thing.

If then, after a few years it's reading slightly out, well one might expect that, but this device was only a couple of Months old, and should have been accurate, in my opinion.

And as a prospective Multimeter production coordinator you'd know right?

In fairness to Doug I haven't too much reason to doubt his device. What I do doubt is the ability of pedant and whiners to use it accurately especially when they choose to ignore so many aspects of measurement environment, accuracy and repeatability.

The fact remains the dodgy Chinese Flukes mentioned have all been within their quoted specification. Yet somehow post purchase this is not good enough.
Did you read the specification prior to purchase? If you didn't and your whining you're a moron!  If you did and you're still whining then you're an unrealistic tool!

[vk6zgo]

Geez! you're a grumpy old sod,Unc!

Sometimes it seems like you & "Bored at Work" are competing for "Grump of the Year!"

That said,I find myself bemused by this "accuracy fetish" so many people seem to have.
In most cases,if you are trying to fix something,& you check the 5 volt rail,if it reads 4.96 volts you are quite happy & head off elsewhere,looking for the problem.
After all,a logic level of 4.96 means the same thing as one of 5 volts,or 5.4volts for that matter!
In fact,--SHOCK!!,HORROR!!,I usually use an old analog Oscilloscope set to DC coupling to check such voltages.

Obviously,there must be situations where extreme accuracy is necessary,but out in the big,bitey,nasty world inhabited by Technicians,rather than Development Engineers,these are very much fewer than imagined.

VK6ZGO

[Uncle Vernon]

Geez! you're a grumpy old sod,Unc!

Why thank you! I'll take that as a compliment!
It's taken years of practice and time spent setting know-it-all sprogs right again.   

Sometimes it seems like you & "Bored at Work" are competing for "Grump of the Year!"

Not a problem till the world champs we each represent different geographical regions.

That said,I find myself bemused by this "accuracy fetish" so many people seem to have.

Bemused is one word for it, I can think of a few others.

In most cases,if you are trying to fix something,& you check the 5 volt rail,if it reads 4.96 volts you are quite happy & head off elsewhere,looking for the problem.

Anyone with any brains would be, some will still choose to argue that point with you, despite their being ill equipped to comprehend simple realities.

Obviously,there must be situations where extreme accuracy is necessary,

None of which involve amateur geniuses armed with hand-helds and only some of the necessary clues.

but out in the big,bitey,nasty world inhabited by Technicians,rather than Development Engineers,these are very much fewer than imagined.

Too true! The world is far from exact or constant.  What annoys me is the routine accuracy fetish while at the same time ignoring half the factors of influence. It's not just clueless sprogs either, in the corporateworld I see Flukes that are in reality used as little more than continuity testers and mains indicators routinely sent of for NATA lab calibration.  Of course the test leads are never included or checked despite the rigours of their user environment.
I see temperature meters lab calibrated to plus or minus 0.25°C to measure the temperature of rooms plus or minus the influence of a selection of user thumbs.
Then for the next six months those instruments survive, within the limits of the environmental and G-Force extremes of a plastic  toolbox within a company van. But hey theyhave a sticker so must be accurate to within a gazilionth of a degree for exactly six months.  The term wankers doesn't come close enough to describing those who perpetuate this pointless stupidity.

I'm convinced that many companies issue their staff with meters so that employees will have both their hands out of their trousers, when viewed by passers by.