Hybrid Car – More Fun with Less Gas

simple stirling hot air engine made at home with simple tools ,2 cans a baloon - Page 8

register ::  Login Password  :: Lost Password?
Posted by Morris Dovey on January 3, 2012, 11:01 pm
On 1/3/12 4:22 PM, Curbie wrote:

Not really, since the gas is being consumed in the most literal sense of
the word. I had originally thought I could apply the Ideal Gas Law to
derive pressure from temperature - but that depends on knowing the
number of moles of gas involved, and I don't see any way to pin that down.


All good stuff - I'm working with an Arduino Mega 2560 and like it
fairly well. The Mega cost a bit more but I opted for higher speed and
more I/O capability just to be on the safe side.

The headache here is that I want to push the reactor to 250C right from
the beginning, and have an end goal of 380-400C operation.

I really do like that $.75 price tag!

Morris Dovey

Posted by Curbie on January 4, 2012, 12:16 am

Well, for distilling, the project this was design for, temperature and
pressure are directly related and I choose to measure temperature, the
project has LCD, button, output switches, and an command/response OS
that might be of some use for you, but for this project, the command
sequence is NOT important so there's no run queue.


I also considered the Mega but the LCD is the bottleneck for loop
timing and I solved that by breaking up the display outputs with
timers. The loop time is under 1ms.

I think you're in the land of multiple thermocouples, they're are not
very accurate, but a way to get more accurate readings, is to read the
thermocouples many times, then average their readings by time slice.

I know I saw pressure sensors, but I don't recall what their ranges

Good luck,


Posted by Jim Wilkins on January 4, 2012, 12:49 am

The hydrogen-rated pressure sensors I found cost $00 each, and had warnings
about long-term hydrogen embrittlement of their sensing disphragms.

I checked three type K thermocouples from different sources in boiling water
and saw less than half a degree error. The stainless clad probe was best by
only a little over the 24 and 20 gauge welded wires.

Pure lead melts at 327.5C. If it's pure it will melt all at once, if not it
will solidify over a temperature range. It's used for black powder cast
bullets. Lead for modern bullets is a lower-melting alloy.

The inaccuracy of thermocouples comes from variations in their alloy
composition, and mainly poor contact with the object being measured.
Statistical averaging won't help, it's a systemic error, in my experience
always low. The ones wired to my wood stove rise or fall in steady 0.1C
steps as it heats or cools with almost no noise bounce.


Posted by Curbie on January 4, 2012, 1:47 am
 Hi Jim,

Good stuff as always.

I did tried multiple thermocouples to even out the variations in their
alloy compositions and statistical averaging to average out noise on
the wires, then calibrated with temperature sensors, and was never
able to get the thermal couples to maintain <1C in relation to the
temperature sensors. Probably just me, through I got 2 batches a 3
DS18B20 temperature sensors from 2 two different sources and they read
<.5C between all senses between both batches, just cheapo
thermocouples I guess.


On Tue, 3 Jan 2012 19:49:06 -0500, "Jim Wilkins"

Posted by Ralph Mowery on January 4, 2012, 2:22 am

The TC is not very accurate.  Where I work we have instruments tracable to
the NIST (think that is the right leters).  We have in usage several
thousand TCs.  Mostly J types.  The place we get most of them from specify
the accuarcy of  plus/minus 3 deg C at 300 deg C for us.  They ordered a big
roll of TC wire for us and most of them are cut from the same roll.  This is
done so that hopefully they will all be off by about the same ammount.  Some
are dual and triple elements in the same tube that is less than 1/4 of an
inch in diameter.  Sometimes they are very close, but I have measured them
to be about 2 deg C differant at close to 300 deg C.

We do use some RTDs.( resistance temperature detector).  They are suspose to
be more accurate,but do not seem to be as reliable and you must make sure
the connections are very good.

This Thread
Bookmark this thread:
  • Subject
  • Author
  • Date
please rate this thread