Maker Pro
Maker Pro

AC excitation of RTDs and other resistive sensors


Bill Sloman

Jan 1, 1970
There's a good book on the subject - H.K.P.Neubert's "Instrument
Transducers" .

It's been out of print for a long time - the second edition was
published in 1975 - but you can still buy it second hand

The first edition was enthusiastic about AC excitation of strain
gauges and resistance thermometer, in part because the Blumlein bridge
gives more sensitivity than the Wheatstone bridge bridge&f=false

As Rayner and Kibble point out in "Coaxial AC Bridges"

- also out of print - the impedances of the arms of centre-tapped 1:1
transformer winding made with twisted pair can - with care - be equal
to about one part per billion, and dissipate quite a lot less heat
than their resistive equivalent. Ratio transformers - made with
twisted bundles of wire, rather than just twisted pair - only get to
about one part in ten million, but standards laboratories love them.

AC excited bridges are more complicated than their DC equivalents, but
the complication buys you quite significant performance advantages,
and you could pack it all into a couple of square inches of board
space, even with through hole parts.

Bill Sloman

Jan 1, 1970
2012 - 1975 = 37.

That's the second edition. I read the first edition as a graduate
student around 1968 - I may have a reference to it in my Ph.D. thesis,
but that's been in Australia for more than a year now, along with my
second-hand copy of the book (sold off by a US academic library).

Most of the references in the book were to work done in the 1940's at
Farnborough on wind-tunnel measurements. Neubert may have known
Blumlein personally. Classic work doesn't have a shelf life.

Spehro Pefhany

Jan 1, 1970
Moreover, if you have the processor bandwidth to sample many times within
an AC cycle, you also get to average out a bunch of quantization noise
and ADC nonlinearities. It doesn't make a silk purse out of a sow's ear
-- but it can gain you amazing improvements.

A very useful technique for getting down to the limits. I'm using an
FPGA to avoid any question of processor bandwidth.

Best regards,
Spehro Pefhany

Bill Sloman

Jan 1, 1970
A very useful technique for getting down to the limits. I'm using an
FPGA to avoid any question of processor bandwidth.

My 1996 paper

Sloman A.W., Buggs P., Molloy J., and Stewart D. "A microcontroller-
based driver to stabilise the temperature of an optical stage to 1mK
in the range 4C to 38C, using a Peltier heat pump and a thermistor
sensor" Measurement Science and Technology, 7 1653-64 (1996)

describes a system where we used 20-bit sigma delta ADC to measure the
bridge output and it mentions that we could have used "reversing DC"
AC excitation with another 20-bit ADC from the same manufacturer, had
we had a -5V power supply.

"Crystal Semiconductors do sell the CS5520 20-
bit bridge transducer A/D converter which supports low frequency,
'reversing DC', AC-bridge excitation, but it
requires both +5 V and -5 V supplies, which ruled it
out in our application)."

The only programmable logic device we used was the ICT PA7024 which
was an antique even back then, but it was - just - good enough to do
what we required and we already had the programming tools for it.

The technology has moved on a bit since then. You can get much faster
sigma-delta A/D converters, which does open up other possibilities,
and FPGA's can be programmed to act as pretty much any kind of multi-
threaded signal processor you feel like (and can afford to buy the IP