I was looking for the concept of "Full-scale error" (related with laboratory measurement terminology) on the web without success. I would appreciate your help.
-
I just Googled "full scale error", got some 85 million hits, and just a cursory glance at the first page showed some good stuff. – EM Fields Oct 10 '15 at 12:44
3 Answers
An error expressed as a percentage of full scale means that any measurement made will fall between the limits given by that quantity.
For example, an analog voltmeter with a range of zero to 100 volts specified as having an accuracy of +/-1% of full scale can be off by plus or minus 1 volt when there's between zero and 100 volts across the meter.
So, with the meter reading 1 volt, the actual input could be anything between zero and two volts, and with the meter reading 100 volts the actual input could be anywhere between 99 and 101 volts.
- 17,527
- 2
- 19
- 23
-
-
I don't believe it even came up, let alone being touted as "better". – EM Fields Oct 09 '15 at 16:22
-
i am looking for the reason why companies may choose to state the error in one way or the other... – nass Oct 10 '15 at 11:39
-
Probably because that's the way it's done, and everyone - Well, almost everyone - understands what it means. – EM Fields Oct 10 '15 at 12:36
-
-
The definition varies a bit, but often it must be added to a 'zero' error, and unless otherwise specified, applies only under reference conditions- warmed up for 1 hour, ambient temperature 20°C +/-1°C or whatever. Temperature changes to scale and zero, and aging often must be added on top of that. In the case of thermocouple instruments, the cold-junction compensation is added on top (and may be specified as a ratio such as 20:1, meaning a 20°C change in ambient temperature could cause an additional 1°C error).
An error specification of 1% F.S. means that if the scale is 4mA to 20mA, that error will not exceed +/-160uA (1% of the full scale deflection, which is 20mA-4mA = 16mA). If they're complete d*cks it could mean +/-2uA error, as I said, definitions vary.
So if the measurement is at mid-scale (12mA) then that error could be +/-2% of reading.
- 397,265
- 22
- 337
- 893
As an example, an error of 10 mV in a full-scale deflection of 2V would be a "full-scale error" of 0.5%. In other words you compare the error with the limits of measurement of the device and express that as a percentage.
- 456,226
- 28
- 367
- 807