Date: Fri, 21 Apr 2000 08:40:01 -0400
Reply-To: Ezra Hall <ehall@TOGETHER.NET>
Sender: Vanagon Mailing List <vanagon@gerry.vanagon.com>
From: Ezra Hall <ehall@TOGETHER.NET>
Subject: LP tank level detection proposal, its Friday..
Content-Type: text/plain; charset="us-ascii"
Given that it is Friday now, I will start another thread..
I searched the archives for info on detecting Liquid Propane level in the
Manchester tank. It seems there are no elegant solutions. Pressure is not
an indication of propane level until there is no liquefied propane left in
the tank, too late unless you happen to be near a filling station, plus
pressure is temperature dependant. A mechanical float would require welding
and re-certification of the tank, I am not interested plus I couldn't find
a sensor to weld on! That leaves thermal methods. Detection of difference
in temperature along the outer surface of the tank requires high flow
rates to cool the liquid through adiabatic expansion, we don't flow a lot
of propane with Westy appliances, so this doesn't work well. The only
practical method remaining (sure you could use acoustical waves or RF, but
that gets expensive and complicated) is detection of thermal resistance and
mass. Presence of LP on the other side of the canister wall decreases
thermal resistance and increases thermal mass to absorb heat whereas
portions of the tank with only gas have a relatively higher thermal
resistance and smaller thermal mass. Liquid crystal level gauges take
advantage of this by measuring temperature at the surface of the tank when
heat is applied to the surface. This elegant solution is described in the
archives as the "fling the boiling water" method. Ok, as you can guess by
now, I want some sort of electronic warning that tells me the tank is low.
Unlike my recent thread on towing a wheelbarrow sized motor to power you're
A/C system while camping, I think this concept is actually fairly straight
forward to implement, the question is, how much interest would there be for
such a system? Well, here goes:
- To apply heat to the tank, an electrical heating element will be used (a
flat thin film strip)
- To detect temperature, inexpensive thermistors will be used
- A minimum of two sensors is required to detect that LP is below a
threshold, more sensors could be used for finer granularity
- In the two sensor scenario, one thermistor is placed at the desired
liquid level threshold, the second would be used to set the compare
temperature of a gaseous portion of the tank (near the top)
- A simple op-amp window comparetor would be used to compare voltage from
the level sense to the voltage from the reference sense. The window will be
set large enough to prevent false readings, ie. A certain temp delta will
be required. Of course, measurements are only taken during or after heat
application via the heating strip.
- Additional level sensors and comparetors could be added for additional
levels of granularity
This really is a simple solution. The only difficult part will be affixing
the sensors to the tank, perhaps a stainless steel band around the tank
could hold them in place vs. an adhesive that could fail. The electronics
and LED(s) could be placed in the same housing with the batt and water
level sense. When you flip the switch, you would also get an LP reading. If
there is enough interest, I can develop this into a product and sell to
list members at a decent price. The above should be enough info for anyone
with the skills to fabricate and build such a device. Anyone interested?
Thanks,
Ezra (going camping this weekend, with full LP tank)
'88 Westy, 94K
'86 944, 260K
|