Page 1 of 4

Water Boil Detector

Posted: Sun Oct 28, 2007 12:14 pm
by MrAl
Hi folks,

Anyone ever build one of these?
It has to detect when the water starts to boil somewhat violently
(or mildly if you can).
If you can come up with something that would be great.

After that, i would like to be able to put the detector inside the
microwave oven. The other catch is that the 'pot' has to rotate
on the platter inside the mic oven. The pot does have a clear top
but not clear sides.

Some ideas perhaps...

Infra red beam?
Standard light beam?

Posted: Sun Oct 28, 2007 12:36 pm
by dacflyer
thats a tough one...but i can tell you forget putting anything electronic in'll be fried in seconds...

Posted: Sun Oct 28, 2007 1:40 pm
by MrAl
Hi again,

Yes, that's one reason why i am posting here, to hear ideas.
I'd use a thermistor if it would survive, but then there is the
pot rotation to think about too though.

Posted: Sun Oct 28, 2007 2:29 pm
by haklesup
Since you didn't say the detector had to do anything like turn off the oven. I suggest a tea kettle whistle type attachement for the lid. If not directly integrated in the lid via an existing vent, maybe built into a plastic ring which fits between the lid and pot. If you want to get fancy (electronic), an acoustic electronic whistle detector.

Newer ovens can have a humidity sensor which along ith the program decide when something is cooked. Just bringing a cup of water to a boil is not one of the choices on my oven though. With a little experimenting I am sure you could calibrate a humidity sensor to the point you need it to trip.

A thermocouple is possible, you just need to shield and shape it correctly. Older ovens often had temp probes. A few newer models still do.

I am not aware of any microwave or conventional oven to attempt to use IR as a sensor. I am not sure why. It may have been the high cost (up until n ow) or it may be that steam and/or air hotter than the food may confuse the readings. You can probably calibrate an IR pyrometer to tell you when its done also but it may need recalibration for a different pot or food (aimed at side of pot).

Posted: Sun Oct 28, 2007 2:50 pm
by MrAl
Hi again,

Hackle, thanks for the reply.

When you said 'shield' what do you suggest for this?
I was hoping to detect the 'bubbles' caused by boiling, but i think
if i used a thermistor to detect ambient temperature inside the oven
i might be able to correlate that to the approximate boiling time,
then watch the oven for the next 60 seconds to wait until it boils.
I dont think waiting another 60 seconds would be too bad, as right
now i have to basically watch the oven for up to 6 minutes or so
to wait for the boil.

I would like to possibly hand the thermistor inside over top of the pot
to detect temperature, so what do you think would make a good
shield fot the two conductor cable and the thermistor itself too?

Posted: Sun Oct 28, 2007 3:15 pm
by Dean Huster
If you'd just go into the other room for a minute, it would be boiling when you got back. Just standing there monitoring the process visually or with an electronic apparatus will not work. Remember that "a watched pot never boils."


Posted: Sun Oct 28, 2007 3:22 pm
by jwax
Good one Dean! It boils about the time the telephone rings too.
Actually Mr. Al, I think the old standard steam whistle is simplest. Wiring in the micwave isn't going to be safe.
If you can modify an existing microwave-safe air-tight container, insert a whistle through its cover. All plastic, of course. Steam whistles can be very loud!

Posted: Sun Oct 28, 2007 4:33 pm
by pebe
We have a Panasonic microwave oven that automatically cooks vegetables. You put any quantity of vegetables in a container and switch on. A humidity sensor acts when the temp is up to boiling point by measuring the water vapour given off. The program then calculates from the elapsed time how much longer to cook the vegetables. Perfect results every time.

There are also humidity detectors on some video recorders to protect the heads.

Perhaps you could utilise one of those.

Posted: Sun Oct 28, 2007 7:17 pm
by jwax
The tricky part is keeping induced voltages from frying your humidity sensor and the wires. BUT, perhaps they could be mounted way down the airstream, out of the micfield. ??
Try mounting a humidity sensor at the air output Mr Al. See if it spikes at the time of boiling. I believe the forced air is just to keep the magnetron cool, but they may draw some air from the chamber too.

Posted: Sun Oct 28, 2007 8:00 pm
by Externet
A glass bell put to float on the surface will sound when boiling shakes it, :grin:
Won't get dizzy with the spinning platter and a nearby microphone can put the sound trough PA systems if you want,


Posted: Sun Oct 28, 2007 8:49 pm
by sparkle
As an experiment I rigged a paper humidity detector at the vent fan rear exit of my 0.6 Cu ft. microwave oven, which both vents the oven chamber and the outer shell area containing the timer/power supply/magnetron stuff. A better location would be just outside the oven chamber vent holes to sense the steam immediately exiting with less dilution from the additional room air entering for magnetron cooling. The makeshift sensor was the narrow end of a clean new $20 bill with about 1 KVDC applied from my picoampmeter/megohmmeter. I put 50 ml of tap water in a 150 ml beaker. Initial sensor resistance was 2.4e3 Megohms. About a minute after boiling commenced the 2.4e3 Megohms dropped to 2.0e3 Megohms; much later about a minute after boiling ended due to all water evaporated (judged visually thru the oven window) the resistance returned to 2.4e3 Megohms. This is with a heck of a lot of bypass air from other case vent holes diluting the steam venting the oven chamber. I would think sensitivity and response time could be considerably improved with a professional sensor and sensor location right at the inner oven chamber vent just outside the microwave power area. And if your situation has a relatively high or variable chamber humidity before boiling begins, perhaps rate of change detection would be better than threshold detection.

Posted: Mon Oct 29, 2007 3:56 am
by jwax
Excellent sparkle! Have to ask- why a clean new $20 bill?
Now to incorporate the humidity sensor from an old VCR, and closer to the source. Should work! :grin:

Posted: Mon Oct 29, 2007 9:01 am
by dyarker
"Have to ask- why a clean new $20 bill?"

WAG- No finger oil to block absorption/release of moisture.

Posted: Mon Oct 29, 2007 9:52 am
by philba
another approach is to use some sort IR temperature sensor like this one - I think you should be able to figure out how to shoot it through a hole in the inner oven liner . You'd need to figure out how to shield the device from microwave RF and avoid stray RF energy so don't try this with out knowing what you are doing. You would also need some way of tracking the rotation and/or widening the beam. I'd use a uC to manage the process.

Posted: Mon Oct 29, 2007 10:33 am
by sparkle
jwax wrote:...Have to ask- why a clean new $20 bill?...
:grin: Although I had heard some other contributed ideas boiling with potential, I was hot to try to throwing money at the problem. Since my VCR works, I was a little reluctant to disassemble it for testing with the humidity sensor! Clean ATM currrency seemed to fit the bill. And if the idea held water, I knew I could brew up a nearly equivalent cost reduced sensor material that cost 1/20 of the initial prototype, with a stack of ideas on similar materials orders of magnitude cheaper remaining to be moved from paper concept into test. :grin:

More seriously, I had investigated in possible response to a previous posted question on CRT discharge the inexpensive construction of high voltage, high resistance values with relaxed stability requirements out of common materials. I was looking for values to discharge a CRT in about a minute or less. I found that certain papers like newsprint were a bit too good an insulator, and that yellow post-it paper and manila file folders were somewhat more usable but might vary from color, mfg to mfg and date of production. Looking for a common standardized material I investigated currency, and found that it (as well as other papers) was definitely moisture sensitive just from breath. Also, finger print residue significantly brought down the resistance in a way that would vary from individual to individual, and possibly lead to sluggish responses as various ions migrated along the energized path at different rates like an electrophoresis experiment. Soiled currency is much more conductive than clean money, way more than the moisture effect alone on clean bills. Anyway I decided not to post earlier given the ongoing, more practical discussion of constructing HV resistors from LV ones in series. I intutively agree with the previous speculation that finger prints would hinder the entrance/exit of moisture and slow response. I am guessing that fast geometries would be thin, very porous films to maximize immediate moisture capture without requiring much diffusion into a material and minimize moisture trapping on the drying recovery cycle.

Paper is not a good long term solution but it sure lends itself to quick experiments. A good sensor should have some sort of liquid splatter blocking but vapor permeable splash guard so steam but not cooking mist can get to the sensor zone. Also, I'm not sure if the VCR sensor technology can withstand long term exposure to cooking vapors without steadly dropping in resistance from accumulating contamination.