Having been through the exercise at work, I can tell you that thermocouples are
not linear! There is actually a website you can visit to obtain calibration
information on different types of thermocouples (www.temperatures.com,
http://www.temperatures.com/tctables.html). Links here give you all the
information you were ever afraid to ask about all kinds of temperature sensors!
Regards
Chris Cameron
Quoting Nigel Charles <72016.3721@compuserve.com>:
> Message text written by Fred Fillinger
> >In noting discussion re temps, I am wondering if anyone is testing
> their temp gauges for accuracy, before drawing conclusion and doing
> mods.From experience in designing digital readouts from the type sensors
> on
> the Rotax, the issues aren't trivial.<
>
> With the uMonitor you calibrate the senders as part of the setting up
> procedure. I agree most senders (except thermocouples) have a non
> linear
> resistance/parameter characteristic. RMI supply their own senders with
> the
> unit rather than use the Rotax ones which overcomes the design problem.
> Obviously the calibration needs to be optimised at the critical points
> in
> the scale. For instance fuel gauges are usually calibrated at full and
> empty and the non linearity in between is not so critical. For CHT the
> high
> end of the range is more critical and therefore the upper setting can
> be
> calibrated using boiling water. For the low end melting ice in water
> gives
> approx zero deg C. Although thermocouples are ideal for temperature it
> is
> important that their cold junctions (ends of the leads) are in a stable
> cool environment. The thermocouple is measuring the difference between
> the
> hot and cold junctions. With that in mind my thermocouple connectors
> are
> aft of the firewall where the temperatures are more stable giving
> greater
> accuracy. Calibration with digital instrumentation is much easier than
> conventional analogue gauges as the calibration points are set using
> digital software. In use, although analogue is easier to see at a
> glance
> accurate readings and slower trends are better observed using digital.
> An
> rpm gauge is the only engine gauge which I feel needs analogue display.
> Ironically these are often the ones which need the most setting up.
> Mine
> overeads by 200rpm at 5800rpm. Until I had checked the accuracy of my
> digital rpm I was only using 5600rpm for take-off losing significant
> performance.
>
> Nigel Charles
> _______________________________________________________________ _______
>
|