-
FLUKE 8845A- residual reading for Vac when shorted (zero error)
Posted by autko on 16 Aug, 2018 12:22 -
Hello,
I recently i bought Fluke 8845A (calibrated 26.04.2018) and noticed one issue: on each and every range when measuring AC voltage (regardless of used filters) I do see residual reading of 35-40 on the two last significant digits of each range in MED accuracy mode (in high: 350-400, in LOW: 3-4). The issue is shown on on the pictures attached to the message (just three ranges but it is present on all of them).
Would like to mention that relative measurement does not help in auto ranging. It seems that the meter was not calibrated to the zero Vac value (on Vdc and all ranges it shows 0V ± 1-2 last significant digits, which is very fine). I do understand, that this is withing the accuracy of the meter (on the edge but within specs) but still this is the first time I noticed such a a behavior on any DMM- especially on the A brand product, officially high quality one... It seems that they did not calibrated the DMM for AC zero reading and no- this is not a nose case (if it was just a noise present in the measurement I should see lower numbers on higher ranges).
This situation is quite annoying especially when using 2nd function measurement: it always shows false reading and there is no simple way to reduce it in auto ranging...
Anyone had similar situation/noticed such an issue?
-
There is no such thing as "Zero AC"
- this 36uV residue (which is essentially self-noise of the meter, measured) on the 100mV AC range is already pretty good for a bench multimeter.
Cheers
Alex -
if this was just a noise it would be ok- take a look at photo on 750V range- it shows over 300mV
EDIT
ok- got your point (or at least hope so): you are referring to internal noise that is present in unit and exhibits itself as this ca. 0,035% of range. Again- this is just on the edge of tolerance that Fluke shows in its specs (0,04% of range for 100mV as we have 0V on input
) and exceeds the specs for other ranges (0,003%). Still the unit has fairly new calibration (within 1 year)
-
if this was just a noise it would be ok- take a look at photo on 750V range- it shows over 300mV
It is the noise. On higher ranges it's value is essentially multiplied as the input signal is divided so the noise relative to the input gets proportionally larger.
Cheers
Alex -
Meters typically cut off their specs on the AC ranges. You will find in the datasheets that specs only count for >x% of range. Anything below is disregarded or additional errors are added or both.
This is from the 8846A datasheet
AC Voltage Specifications
AC Voltage specifications are for ac sinewave signals >5 % of range. For inputs from 1 % to 5 % of range and <50 kHz, add an additional error of 0.1 % of range, and for 50 kHz to 100 kHz, add 0.13 % of range. -
if this was just a noise it would be ok- take a look at photo on 750V range- it shows over 300mV
EDIT
ok- got your point (or at least hope so): you are referring to internal noise that is present in unit and exhibits itself as this ca. 0,035% of range. Again- this is just on the edge of tolerance that Fluke shows in its specs (0,04% of range for 100mV as we have 0V on input ) and exceeds the specs for other ranges (0,003%). Still the unit has fairly new calibration (within 1 year)
Please read the Fluke 8845A specifications for AC voltage properly.
Cheers
Alex
P.S. ManateeMafia was quicker
-
Yes, you are right (I corrected my first reply when you was answering me On the other hand- in noise case I'd expect the readings to be a bit more chaotic (it seems that the meter do lots of statistics in the background). Still a bit strange for me...
Maye someone who has this DMM (or 8846) would confirm such a reading on his unit?
-
The noise adds some RMS background. Though noise, the amplitude is relatively constant and thus the reading can be pretty stable.
How much noise is present also depends on the bandwidth - so a cheap meter might even show a lower noise background, simply due to lower BW. A nasty property of the analog RMS chips is, that BW can depend on the amplitude. So the low level readings, below a certain percentage is usually not really accurate.
Beside noise, there can also be some offset - as an error or intentionally to somewhat compensate for the noise contribution. -
All RMS/DC converters based on computational methods obtained by analog multipliers/dividers or log/antilog circuits cannot give accurate
result at zero input signal.
The former should perform a division by zero, the latter should extract ln(0) which is clearly impossible.
Each converter has its own behavior around zero, which results from various parameters: component properties, trimming errors, noise, leakage etc.
See for example the datasheet of the LTC1966 converter, one of the most accurate in monolithic form.
http://www.analog.com/media/en/technical-documentation/data-sheets/1966fb.pdf
It is based on a different calculation method (sigma/delta) but the principle used is the same as normal log/antilog converters.
The graph named "DC Transfer Function Near Zero" (Pg. 7/38) shows the behavior of the circuit for low input voltages:
as you can see every IC has a different behavior around the zero.
Some multimeters patches the problem by zeroing the reading below a minimum limit.
The only detector that (theoretically) does not have this problem is the thermal converter.
On this behavior, see also:
http://www.gellerlabs.com/34461A%20AC%20zero.htm
-
Keysight says that on their TrueVolt series (34466A and 34470A) do direct digital sampling that is supposed do a better job, linearity wise and also able to more accurately measure signals with larger crest factors... Didn't have a chance to try it.
-
I have both FLUKE 8846A and Agilent 34410A, and did experiments with them (using FLUKE 5520A MFC and HP 3458A in ACV SYNC mode), and the 34410A with digital sampling is superior. Geller Labs did some interesting experiments too http://www.gellerlabs.com/34401A%20AC%20zero.htm
So for ACV I trust 34410A the most, and for DC 8846A
-
I have both FLUKE 8846A and Agilent 34410A, and did experiments with them (using FLUKE 5520A MFC and HP 3458A in ACV SYNC mode), and the 34410A with digital sampling is superior. Geller Labs did some interesting experiments too http://www.gellerlabs.com/34401A%20AC%20zero.htm
So for ACV I trust 34410A the most, and for DC 8846A
Yes you are correct. I forgot about 34410A and 34411A. They are TrueVolt too.
I believe it is probably based on Swerlein algorithm. -
I have the same model and I see 200 - 250 counts on all ranges with the exception of 100mV where it is ~450. HP 3478A is even worse.
-
Hi autko,
I can't explain the technical reason for it, but as opposed to DC a today's multimeters are not calibrated at zero AC voltage.
Check out the calibration instructions for a 3456A or just look up the calibration instructions for a LTC1968 chip, starting on page 22:
[...
AC-Only, 2 Point
The next most significant error for AC-coupled applications will be the effect of output offset voltage, noticeable at the bottom end of the input scale. This too can be calibrated out if two measurements are made, one with a full-scale sine wave input and a second with a sine wave input (of the same frequency) at 10% of full scale.
...]
http://www.analog.com/media/en/technical-documentation/data-sheets/1968f.pdf -
These analog RMS converters have quite some limitations at low signal. It start with the simple rectifier part - the smaller the amplitude, the slower is switching the sign. So there is an inherent reduction in BW towards low amplitude. A second point is a possible offset at the rectifier - thus a small signal might be stuck at one side and would only contribute trough it's peaks that might cross zero.
A third point is an offset due to meter internal noise. Thus even a perfect RMS converter would not show 0 but a background power. However this is not just a simple offset, as non correlated signals add up as squares. At least in theory the CPU could take care of it correctly, but it is not an easy adjustment or correction that is easily done on the analog level.
So with an analog RMS chip, one can not expect a good reading at the very low end.
So I wonder why we still see so many meters with those limited analog RMS detectors instead of digital sampling RMS, like the trueVolt ones. It does not take so much to be about on par with the analog RMS chips like AD636. Besides the accuracy, there is also the slow settling that is not that attractive.