philba

I have a project that I'm thinking about doing. It's a programmable food dehydrator. One of the key ideas is to sense the amount of moisture being removed and use that to track a drying profile. The advantage would be that the dryer would shut down before the food is overdried. I would compare input air humidity vs output air humidity so I'd need 2 sensors.

Unfortunately the humidity sensors I've looked at seem to be very slow (measured in minutes) and quite expensive. Most seem to be capacitive. Anyway, I've thought about building capacitive based sensors myself but don't know if that's the best approach - is the different so small as to be a problem. Are there better approaches? Maybe ultrasonic transmission time? RF absorbtion?

BeeBop

I don't know if this will be much faster, but for sure it's cheaper:
If you remove the outside shell of a ceramic covered resistor, say a 1 W - ceramic can easily be chipped off - the resistance changes with the humidity. I can't remember just how I verified this, but if I recall, said resistor was put into the feedback loop of an op amp.

Hero999

Capacitive seems good, cheap and I imagine it's instantaneous.

How does your dehydrator work?

Are you using heat or are you freeze drying food?

__________________
I also post at the following sites:
http://www.stop-microsoft.org http://www.heated-debates.com
Screen name: Aloone_Jonez

philba

I was thinking of heat for the first effort but I have been looking at vacuum approaches as well. Basically, it's a box that can accept screened racks which hold the food. Air is brought in at the bottom, heated and distributed via a fan. Exhaust is at the top. Humidity would be sensed at the inlet and outlet. air temp would be controlled based on a differential humidity profile. The lower the differential humidity, the lower the heating. As the food dries, the differential would drop. A given food would have a specific profile. This way, it would prevent over drying.

The only question I have about capacitive is how much of a difference between 30% humidity and 70% (for example) there would be. I guess I'll need to do some tests.

Leftyretro

Humidity can be a complex measurement problem. The most common measurement is relative humidity measuring the amount of water vapor relative to the maximum water vapor that can be held at that specific temperature, hense the term relative. Absolute humidity is a different measurement and would involve the actual amount of water vapor present. if I recall it's a more difficult measurement to make.

Not sure which is the desired measurement for your application as you are using the measurement to control the temperature, so a possible catch 22, or maybe not. Research or copying a proven design should prove constructive.

Lefty

philba

I think RH should be just fine since I'm looking at the differential between intake and exhaust RHs. I doubt it needs to be that precise as what I really care about is when the exhaust RH starts getting close to the inlet RH (factoring in temperature). I just don't want to spend a lot of money on sensors if I can make them myself.

Humirel has several sensors that are about $10 so that might be ok.
I'm playing with the idea of using ultrasonic time of flight to measure RH however, that may not be much cheaper than the humirel sensors.

wmerrin

Take a look at how newer clothes dryers know when they are "done" - they must be using some sort of humidity sensing of the exhaust air.

Wally

speakerguy79

philba

gee thanks for pointing out the web site for the sensors that I mentioned I have looked at. Any specific reason for posting the link?

speakerguy79

Sorry, only read the OP. Humirel is what we use at work to maintain proper humidity in our fuel cell systems. If there is something better/cheaper, let me know

krazatchu

I thought it would be nice to have my bathroom fan on a humidity switch...

Also need it for something else too...

I've seen it mentioned here a few times, that a resistor with the paint scraped off will vary with the humidity level...

I haven't seen it implented so I thought I would give it a try...

I tried it with a TL072 since I had one handy, might try it with a 741 as well since I have some around here..

The first circuit shown works great... except that the TL072 doesnt swing to GND... it's low position is 1.6v... hence the funny output series of invertors...

The second circuit I did not test, it should swing to GND ...

So,.... how well does it work?

It seems to be able to sense my finger from about 5 mm distance... or if I blow on it from about 30cm...

Thought it might be heat... so I tried wet paper near it, that works too...

It turns on and off very quickly... on time is too short to count and off time is perhaps a few seconds at the most...

R3 and R8 I used to model the naked resistor...

Attached Images

humidity switch.png (48.5 KB, 37 views)

bryan1

Hiya Phliba,
Eh mate you might want to checkout the dallas I-buttons. There is one that does humidity and temp on the same package. They can be set for any timing period and just thrown in and read later or you can connect it up and do a force read. Here is a link to the one I reckon might suit your needs

http://www.maxim-ic.com/quick_view2.cfm/qv_pk/4379

Cheers Bryan

__________________
" The only way to avoid human error is to avoid the use of humans"

krazatchu

I'm curious as too how you are progressing on your dehydrator...

I have a bit of experience in this area... So I'm interested in the approach your taking... Here are a few tips I will pass along...

1)Theramal dehydration takes alot of energy... due to hydrogen bonding between the molecules... It takes approx. 2.25 kJ of energy to convert 1 gram of liquid water into steam...

2)Once converted to vapor... the volume is massive... 1 L of liquid water becomes abour 1160 L of vapor... this requires significant airflow to remove...
Dust and airborn particulates become a problem with increased airspeed..

3)Heated air is very inefficient... especially without using a heat exchanger / heat recycle system... using a heat exchanger takes a large space due to required surface area...

4)Vacuum systems are a more effiecient approach... but the volume of vapor gets larger as the pressure drops... As well, the vacuum pumps are quite expensive.. especially the ones designed to handle water vapor... Vaccum systems are normally called MVC (mechanical vapor compression)...

5)If the sensors take a few minutes to react.. that's probably ok... since the transpiration rate decreases as the product becomes dryer... drying will slow down significantly near the end of the batch...

There are some other options available within thermal drying and quite a few in the non-thermal catagory... It all depends on the requirements of what your drying and the final product...

I hope this info helps... feel free to inquire if you ahve any questions..

And sorry to all for the non-electronics related post...

Michael

Conradb

One can only imagine this is a meat-drying box? Infrared lamps similar to those used for raising chicks work well, and a small fan (space-heaters set to minimum work). More interested in how calibrating the 'resistor' sensor worked out, any photos of your resistor might encourage others to try building such a cheap sensor if it did work.

krazatchu

hey~~

I was thinking some might be intresting in the resistor humidity sensor as well... as I have seen it mentioned in a few threads but have not seen it in application...

I will be going out to pick up some parts later today if I make the time, otherwise Wednesday... I'll pick up an LM393 and test it ...
My goal was to have a set humidity point at which it triggers... not so much a sensor as a switch...

Perhaps I should start a new thread for this instead hijacking this one

Thanks,
Michael