# Fluke 5100B

## Ask a question - edaboard.com

Robert T.
Guest

Tue Jul 29, 2014 7:33 pm

Hi all,

I am looking to calibrate 16 bit A/D converters used in several
configurations (DC voltage, AC voltage, and RTD) and was wondering if a
Fluke 5100B 4.5 digit calibrator would be a good choice accuracy-wise?

According to blurbs on the web, these are the specs.

The 5100B AC/DC Calibrator calibrates precision meters that measure AC or
DC Voltage (0-1100V), AC or DC current (200 uA-2A), and Resistance (1 ohm
to 10 Mohms). Accuracies are: DC voltage and resistance (0.005%), DC
current (0.025%) AC Volts (0.05%) and AC current (0.07%).

Thanks.

Guest

Tue Jul 29, 2014 7:33 pm

In sci.electronics.basics Robert T. <nospam_at_dev.null> wrote:
Quote:
I am looking to calibrate 16 bit A/D converters used in several
configurations (DC voltage, AC voltage, and RTD) and was wondering if
a Fluke 5100B 4.5 digit calibrator would be a good choice accuracy-
wise?

A 16-bit A/D converter has 65,536 steps over its full range. If you
want to have resolution down to one step, you need something that's
accurate to at least (1/65536)*100 or 0.00153%, and preferably more than
that.

If you're willing to accept being three or four steps off, then the
0.005% is OK. It depends a lot on what you're doing with the digitized
data.

Matt Roberds

Robert T.
Guest

Wed Jul 30, 2014 3:50 pm

On Tue, 29 Jul 2014 14:31:01 +0000 (UTC) mroberds_at_att.net wrote in Message
id: <lr8b72\$5m9\$1_at_dont-email.me>:

Quote:
In sci.electronics.basics Robert T. <nospam_at_dev.null> wrote:
I am looking to calibrate 16 bit A/D converters used in several
configurations (DC voltage, AC voltage, and RTD) and was wondering if
a Fluke 5100B 4.5 digit calibrator would be a good choice accuracy-
wise?

A 16-bit A/D converter has 65,536 steps over its full range. If you
want to have resolution down to one step, you need something that's
accurate to at least (1/65536)*100 or 0.00153%, and preferably more than
that.

If you're willing to accept being three or four steps off, then the
0.005% is OK. It depends a lot on what you're doing with the digitized
data.

Thanks Matt. I think that I can live with a few counts off.