| Products > Test Equipment |
| What's the cheapest 0.02% accuracy handheld meter |
| << < (3/16) > >> |
| bdunham7:
--- Quote from: miegapele on July 12, 2023, 06:58:47 pm ---Now I wanted more accuracy and started looking for 0.02%/0.03% accuracy multimeter. --- End quote --- The devil is in the details. A stated 0.02% 'basic DC accuracy' is not a complete spec. You need to look at the total specification for each range, typically expressed as (% of reading + counts) or (ppm of readin + ppm of range) or something like that. So on your RM303, according to the specs if you are measuring 1V, the specified error is 0.05% of reading (5 counts) plus 3 counts for a total of 8 counts. And in reality, it may or may not meet those specs, IMO that figure seems a bit aspirational, especially since it is claimed for all ranges. The Fluke 289, for example, is 0.05% + 20 counts in the lowest mVDC range, whereas your RM303 claims the same accuracy in the 20mVDC range as it does for all the others--0.05% +3 counts. |
| J-R:
500,000 count mode is only available for DCV and DCmV and I don't see where the accuracy specifications are defined in that case. |
| David Hess:
When looking for higher accuracy, I also expect a fixed 10 megohm input resistance or a gigohm input, which is going to mean a bench meter. |
| J-R:
Some thoughts on the accuracy specifications. It is true that some test equipment greatly exceeds published accuracy specifications. I could imagine some reasons for this, such as the manufacturer wanting to maintain the accuracy over a wide range of operating conditions or life of the product, extra margin for internal component variations from suppliers, shared manufacturing lines, etc. However, the main issue is HOW do you leverage that accuracy performance? Some of my methods include obtaining calibration plus data, having a collection of references, and comparing against other test equipment where I have calibration data already available. Without this information, you are going to be somewhat limited to using the published numbers and the reputation of the manufacturer. It's a bit tricky to leverage the 500,000 count mode of the Brymen DMMs, so if you are planning to bank on that I would suggest at least a high-quality 5V reference to check it against (or even calibrate it if desired). At lower points in the ranges it can be particularly trust-worthy, so for example 1V. I calibrated my BM869s a year or so ago and it's still basically perfect in this case. Although it's more of an intellectual exercise. Just get a decent bench DMM if you're bumping into this requirement level. |
| bdunham7:
--- Quote from: J-R on July 13, 2023, 03:13:13 am ---However, the main issue is HOW do you leverage that accuracy performance? --- End quote --- If you are a hobbyist that doesn't need to prove anything to anyone, then your imagination is the limit. For example, you can use calibration certificates in reverse to imagine you know what a meter actually reads to 10X its spec if you like. Or dial them in to the last count and ignore the tempco that you can't control for. If you need certifiable results, then I think the whole point is that you don't leverage the difference between specified and actual performance to achieve a tighter spec, but rather to get a greatly increased probability--hopfully a near certainty--of meeting the spec. The two parameters (reading uncertainty vs the probability of being in spec) are intertwined and inseparable. |
| Navigation |
| Message Index |
| Next page |
| Previous page |