5th Jun 2018, 11:45 am | #21 |
Octode
Join Date: May 2017
Location: St Austell, Cornwall, UK.
Posts: 1,018
|
Re: Meter Suggestions for Basic Alignment/Calibration
Hi Karl,
Yes for sure. The Specs I have quoted on this Thread are all specifically DC Current uA Range. Don't think your Attachment relates to a Fluke 87V though. The lowest range on the 87V is 600uA, not 500uA as in your attachment. Also what I have found today, is that However much of a Minefield Accuracies and Resolutions can be, it is not helped when even the Manufacturers have mistakes on their own Websites! Hope their IT Guys don't help with Production! Ian |
5th Jun 2018, 12:05 pm | #22 |
Heptode
Join Date: Jun 2010
Location: Zurich, Switzerland
Posts: 538
|
Re: Meter Suggestions for Basic Alignment/Calibration
Ian,
misstakes on websides and datasheets-manuals are relative common. May attachment is for version "IV" and you are spoken over "V"... Karl |
5th Jun 2018, 12:24 pm | #23 | |
Octode
Join Date: May 2016
Location: Oxfordshire/Bucks borders, UK.
Posts: 1,604
|
Re: Meter Suggestions for Basic Alignment/Calibration
Quote:
Got my 87V in front of me right now connected up to a current calibrator. I'm able to increment in 0.01uA steps. The adjustment on this calibrator is too coarse to try and get spot on what I want in micro amps, but the photo demonstrates where both readings are different by 0.01uA.
__________________
Avometer, vintage Fluke and Marconi collector. Also interested in vintage Yaesu and KW. Last edited by Sinewave; 5th Jun 2018 at 12:33 pm. |
|
5th Jun 2018, 12:51 pm | #24 |
Octode
Join Date: May 2017
Location: St Austell, Cornwall, UK.
Posts: 1,018
|
Re: Meter Suggestions for Basic Alignment/Calibration
Sinewave,
Thanks for posting those Photo's. They are pretty indisputable! To be honest, I'm not impressed with their Customer Service! They were nice enough, but hey, get it Right! I took on board what you said yesterday, and spoke to Fluke this Morning after checking the original Spec sheet I read and I noticed the discrepancy between that and the one on the Fluke Web Site. I pushed their Tech Guy and he said he trusts the User Manual more than their Web Site. The User Manual I downloaded this Morning does Quote 0.1uA Res. All I want is an accurate uA Meter, who would have guessed it would be so difficult! I might call Fluke back! Ian |
5th Jun 2018, 1:17 pm | #25 |
Heptode
Join Date: Jun 2009
Location: St Helens, Lancashire, UK.
Posts: 641
|
Re: Meter Suggestions for Basic Alignment/Calibration
Yes but you have converted to mA but quote your answer in uA!
you first converted 37.5uA to 0.0375mA! so your 1% is 1% of 0.0375mA, which is 0.000375mA (note milli amps) If you convert this back to uA and you get.....0.375uA I had not realiesed the 87V was only a 3&3/4 (6000 count) meter, so it is a lot less accurate than I calculated! You need to look for a 4.5 digit or better DMM to give precision to 0.1uA Then a calibration, or "specification garantee" to give accuracy to 0.1uA
__________________
SPECIALIST.....Knows everything about nothing EXPERT..........Knows nothing about everything |
5th Jun 2018, 1:25 pm | #26 |
Dekatron
Join Date: Jul 2007
Location: Evesham, Worcestershire, UK.
Posts: 4,244
|
Re: Meter Suggestions for Basic Alignment/Calibration
The Fluke 87V is a nice meter; I have several.
However, there's no need to spend that sort of money (there's no doubt the 87V is over-priced, both new and s/h). For less than £20, this has 10nA resolution: http://www.markhennessy.co.uk/budget...eng_an8008.htm Although it has gaps in the current ranges - as I explain on that page - if those are a problem, there are alternatives. But I offer this up initially to see how the idea of a budget cheap multimeter sits with you. Providing you're using this for low-energy bench use, it's an excellent meter. I use it about as much as I use an 87V, and thing it's great for the money. Whenever I cross-check the calibration with a decent meter, it's always impressively accurate. Of course, who needs an ammeter to measure current? I'd echo the suggestion to use a resistor and a volt meter - you can measure any current you like to whatever resolution you need. Bearing in mind the context here - AVOs and the like - this is a perfectly workable solution. And bear in mind the statistical trick you can play with resistors to make highly accurately shunts (assuming a Gaussian distribution, you can make a 1k 0.3% resistor from ten 10k 1% resistors - so if you can measure resistors to 0.1% accuracy, then you can make a 0.03% resistor). Cheers, Mark |
5th Jun 2018, 1:27 pm | #27 |
Dekatron
Join Date: Jul 2007
Location: Evesham, Worcestershire, UK.
Posts: 4,244
|
Re: Meter Suggestions for Basic Alignment/Calibration
The 87V has a 4.5 digit mode. You hold down the backlight button for 2 seconds to enable it.
|
5th Jun 2018, 1:31 pm | #28 | |
No Longer a Member
Join Date: Oct 2016
Location: Maroochydore, Queensland, Australia.
Posts: 2,679
|
Re: Meter Suggestions for Basic Alignment/Calibration
Quote:
I don't know if I would agree that this is really a catch 22. Very few macroscopic situations actually are, except at the quantum mechanical level trying to figure out the position & momentum of a particle at the same time. If you can generate an accurate voltage, like the 10V voltage reference IC's do and you can apply that to a known accurate precision resistance, the current would be within the tolerance or error values of the voltage source and the resistance specifications. Much better than any practical meter calibration, making another meter itself totally superfluous and any measurement by a practical meter only would introduce "calibration errors". An analogy to this situation cropped up recently in the calibration of valve testers, where it turns out the last thing in the world you need to calibrate a tester is a "calibration tube" with a supposed known transconductance. Analogous to "the last thing in the world you need to calibrate a meter is another meter with known calibration" It is far better to eliminate intermediate third party variables and create a calibration scenario that relies on first principles of voltage, current and resistance, controlling two of these parameters to a high accuracy, so as to generate the third one. Hugo. |
|
5th Jun 2018, 3:26 pm | #29 | |
Heptode
Join Date: Jun 2009
Location: St Helens, Lancashire, UK.
Posts: 641
|
Re: Meter Suggestions for Basic Alignment/Calibration
Quote:
Pretty good, but a lot less (33x less in fact) than the display resolution on this range Resolution is not accuracy! Also could you trust the claimed figures? On the 87v will it do 4.5 digit on 200uA range? As has been noted the specifications are all mixed up, and if it will do 4.5 digit on 200uA range, again its accuracy is 10x less than the resolution which makes no sence on a Fluke I dont have one, prefaring my Fluke 233's for portable work, as i work on BIG machines!
__________________
SPECIALIST.....Knows everything about nothing EXPERT..........Knows nothing about everything |
|
5th Jun 2018, 3:40 pm | #30 |
Octode
Join Date: May 2016
Location: Oxfordshire/Bucks borders, UK.
Posts: 1,604
|
Re: Meter Suggestions for Basic Alignment/Calibration
Once we're on the bench, then resolution will determine our accuracy.
If we don't have the resolution, then we can't be accurate. We know the accuracy of the meters, but for us to be accurate, we need the resolution. If we can't read in 0.01uA increments, then we can't be accurate to 0.01uA, for example.
__________________
Avometer, vintage Fluke and Marconi collector. Also interested in vintage Yaesu and KW. |
5th Jun 2018, 4:01 pm | #31 | |
Dekatron
Join Date: Jul 2007
Location: Evesham, Worcestershire, UK.
Posts: 4,244
|
Re: Meter Suggestions for Basic Alignment/Calibration
Yes, I am aware of this.
All I can say is that the samples of the various cheap meters I have bought and reviewed have exceeded their specifications - where they haven't, I've pointed it out. Others on forums like EEVBlog have been similarly or more rigorous. The meter in question uses software calibration, so is calibrated between construction and shipping. As there are no mechanical pre-sets to get disturbed by shock, etc, the divider/shunt resistors or voltage reference needs to change. It's possible that electrical abuse might do that to the former, but if looked after, I see no reason to distrust them. Ultimately, time will tell... Quote:
Having an order of magnitude more resolution than accuracy is not at all unusual. |
|
5th Jun 2018, 4:07 pm | #32 |
Octode
Join Date: May 2016
Location: Oxfordshire/Bucks borders, UK.
Posts: 1,604
|
Re: Meter Suggestions for Basic Alignment/Calibration
The 4.5 digit mode is what I used on the uA range demonstration above and as you say, is available on all functions.
__________________
Avometer, vintage Fluke and Marconi collector. Also interested in vintage Yaesu and KW. |
5th Jun 2018, 7:52 pm | #33 |
Octode
Join Date: May 2016
Location: Oxfordshire/Bucks borders, UK.
Posts: 1,604
|
Re: Meter Suggestions for Basic Alignment/Calibration
I still think get an Avo 8 mark 5 to 7 to do this.
__________________
Avometer, vintage Fluke and Marconi collector. Also interested in vintage Yaesu and KW. |
5th Jun 2018, 9:30 pm | #34 |
Heptode
Join Date: Jun 2009
Location: St Helens, Lancashire, UK.
Posts: 641
|
Re: Meter Suggestions for Basic Alignment/Calibration
If your AN8008 is displaying 98.76uA on the 99.99uA range can you tell me what range of values the actual circuit current could be?
What is the least significant digit of resolution that will give you 100% certainty in the display? What is the maximum useable resolution? Then tell me if you think this meter can be used to measure to 10nA, or even 100nA? Things are similar on the Volt ranges On the 1 volt range (999.9mV) the accuracy is +/- 8.2mV, which again means the last two significant digits have no meaning in any sensible measurment. This DMM should only be presented with a 3 digit display max, as it is not accurate enough to have a useable fourth digit on any range. (actually only the first two digits give an accurate display! ) Having the fourth digit tricks users into thinking it is more precise than it really is! The fourth digit could be replaced with a random number generator (or just blanked out), without affecting the meters performance However you get what you pay for!
__________________
SPECIALIST.....Knows everything about nothing EXPERT..........Knows nothing about everything |
5th Jun 2018, 10:45 pm | #35 |
Dekatron
Join Date: Aug 2016
Location: Colchester, Essex, UK.
Posts: 4,108
|
Re: Meter Suggestions for Basic Alignment/Calibration
Out of interest i have powered up the precision voltage ref. at 2.5v and put it across a precision 30k resistor. Expecting so see something around 83.3uA (on the entry level DT830B and also a sub £12 toolzone DMM) i wasn't too disappointed to see:
DT830B 83uA (2000uA range) 81.6uA (200uA) Toolzone 81uA ...."................ 80.6uA ..."...... As said, I am hardly ever in a position to need these ranges, but it's of interest to know what is accomplished for very little money. (I put an AVO 8/III in series with the toolzone DMM and the circuit was affected to the extent that the AVO read 76uA...which tallied exactly with the DMM. Not bad at all considering.) |
6th Jun 2018, 1:09 am | #36 |
Hexode
Join Date: Apr 2017
Location: Buderim, Queensland, Australia.
Posts: 428
|
Re: Meter Suggestions for Basic Alignment/Calibration
Ah, ...measurement uncertainty, with a touch of Heisenberg!
If it was me, I would do what some others have suggested, use a 10 Volt precision IC, and use some 0.1% resistors to generate a calculated 37.50 uA source, then check with say a good Fluke 4.5 bench style DVM or similar, or even another known good AVO 8, as also suggested previously. My own workshop reference is based on a 2.5 Volt reference IC, and the concept works well. I would only aim for 1% of FSD movement setting accuracy for practical reasons. That's all the originals were specified at anyway. Probably the biggest hurdle is deciding whether to pursue practical workshop accuracy, or Cal Lab accuracy. The former is easily achieved, the latter is very expensive and not easily attained without an understanding of all the variables and constraints, and with traceable equipment measurements. Your hair may turn grey overnight. Anyway, very good luck to you. |
6th Jun 2018, 12:47 pm | #37 | |
Dekatron
Join Date: Jul 2007
Location: Evesham, Worcestershire, UK.
Posts: 4,244
|
Re: Meter Suggestions for Basic Alignment/Calibration
Quote:
I've no idea who these questions are aimed at. The mention of the AN8008 makes me wonder if they are aimed at me, given that I mentioned that meter earlier. Use of quoting might have made this clearer. Perhaps you should take these questions to Fluke. They make lots of meters that have a basic DC accuracy of 0.5% - just like the AN8008 or AN8002. For example, the 115 that's on the desk beside me. Or the 101 that I reviewed on my website. Or the 15B+, 17B+ or 18B+. All of these are 6,000-count (or 4,000-count for the latter 3) meters. Also Keysight make the U1230 series, which are 0.5%, 6,000-count meters. Please let us know what Fluke and Keysight say when you tell them that they should be 600-count meters only. Meanwhile, I'll carry on with using my meters, enjoying the benefits we get from having lots of resolution. For example, if our lowly 6,000-count, 0.5%+3 meters tell us that we have 1.000V at a point in a circuit, we know that it might wrong by 8mV or so. But for more practical applications - where we're using a multimeter to troubleshoot or design a circuit, rather than indulging in a bit of amateur metrology (nothing wrong with that!) - that doesn't actually matter in most instances. What is often of interest though, is seeing small changes. If we measure 1.000V at one point, and measure 0.998V a bit later, we know that we have a small voltage drop - perhaps along a PCB track - that suggests that current is flowing. We might not need to accurately quantify that, but it is still valuable information that might lead us to a successful diagnosis - and would have been missed if we were using a 600-count meter. Similarly, we might need to match resistors. Think of a balanced audio input stage based around an op-amp and 4 resistors. These resistors should be matched as well as possible to get the best CMRR. Their absolute value isn't important - it is essentially arbitrary, frankly - but their relative value is critical. Using a 6,000-count meter, I can select 4 matched 10k resistors from a batch to within 10 ohms (0.1%) with complete confidence. Matching resistors to 0.1% rather than 1% makes a 20dB difference to the CMRR, which is well worth the effort. In short, it's common knowledge that there is at least an order of magnitude more resolution than accuracy, but that the extra resolution is still extremely valuable. OK, this is something that might come as a surprise to absolute beginners (and sometime university graduates - I know that because I teach them!), but it's wrong to suggest that we should completely disregard the least significant digit (and even the next most significant digit!) at all times. All of those meters I mentioned earlier are a lot more expensive than the AN8008 or AN8002, so in what way are we "getting what you pay for" in terms of accuracy? Yes, the Flukes and Keysights will be better built and hopefully safer, but having directly compared the AN800x models to Flukes and my Keithley 2015THD, you're not getting any more accuracy for the money. As I said earlier, time will tell - they might be more prone to drifting than an expensive meter - we will see. Ultimately, measurements are about pragmatism. And in this specific case, calibrating an AVO 8 (+/-1% at FSD) could quite reasonably be done with a 0.5% class DMM (perhaps using the precision resistor technique), though personally I'd want to use a 0.1% meter because calibration should be done using an instrument that is an order of magnitude more accurate ideally. Luckily though, modern multimeter specifications are very conservative - they have to be for mass-production - so using a 0.5% meter might not be all that bad in practice, providing you're not the sort of person who lies awake at night worrying about such things |
|
6th Jun 2018, 1:21 pm | #38 | |
Dekatron
Join Date: Jul 2007
Location: Evesham, Worcestershire, UK.
Posts: 4,244
|
Re: Meter Suggestions for Basic Alignment/Calibration
Quote:
From 2.5V, you need a 66.67k resistor. On the uA range, the AN8008 has a 100 ohm shunt resistor - which is nice and low because the measurement IC has a 1uV resolution - other meters might need to use 1k to measure uA to the same resolution. Adding 100 ohms to your 66.67k resistor will reduce the current to 37.444uA. Not a huge change, and not anything to worry about here But, re-do the numbers for a higher current. Let's say 100mA. The milliamp shunt for most multimeters will be 1 ohm. From 2.5V, you need 25 ohms, but with the added 1 ohm of the meter, you'll now have 96mA. Obviously, that's an extreme example just to make the point, but it's easy to overlook when aiming for high precision. It also causes fun when you put addition meters in series to verify the reading of the DUT. If you know that the additional meter is accurate, then take that as the reference and ignore the fact it's a bit lower than expected. But if your 2.5V reference and the resistors are the basis of the current measurement, beware of this trap In addition to the shunt resistors, the fuse will add something to the total resistance, as will PCB tracks and leads, of course. And different meters might use different shunts. All in all, it's why I favour measuring the voltage across a known resistance to determine the current - it removes the internal resistance of the meter from the equation Oh, I have a Fluke 45 - nice meter to use, but only has 100nA resolution. The Fluke 87V/187/189/287/289 (and Aneng AN8008 ) hand-held meters manage 10nA. But the 45 is usually better VFM on the s/h market. Cheers, Mark |
|
6th Jun 2018, 2:11 pm | #39 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,902
|
Re: Meter Suggestions for Basic Alignment/Calibration
I just had to pinch myself and remind myself that this is all just for checking the calibration of a simple moving-coil meter movement.
David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
6th Jun 2018, 3:16 pm | #40 | |
Octode
Join Date: May 2016
Location: Oxfordshire/Bucks borders, UK.
Posts: 1,604
|
Re: Meter Suggestions for Basic Alignment/Calibration
Quote:
With our Avo's we strive for perfection, accuracy, resolution and above all, the agility, adapability and capability to be our best friend on the bench.
__________________
Avometer, vintage Fluke and Marconi collector. Also interested in vintage Yaesu and KW. |
|