[time-nuts] Temp/Humidity control systems?

Neville Michie namichie at gmail.com
Wed Oct 26 15:54:36 EDT 2016


Hi John,

such a project is quite practical.
I spent a working lifetime in textile physics research.
Air conditioned rooms were in great demand and access was limited.
So I did what I could with what I could get.
In spite of what conventional wisdom (or lack of wisdom ) says you 
can get good performance from a reverse cycle window air conditioner unit.
For humidity control, you should be able to make do with unilateral control downwards,
set the room to a humidity that is lower than the lowest ambient humidity. You may even compromise
by setting humidity at say 40% and tolerate the rare occasion that it drops below. It may be worth it 
so avoid the mess of bilateral control.
Now, so far you have the recipe for a system that produces great swings in temperature and humidity,
not what you want.
The important step is the controller and a small fan or two that gently swirls the air around the room,
at a velocity about 0.5m/s, which is only just on the threshold of perception. The swirling of the air 
means that within a time frame of 30 seconds the air is mixed into one thermal mass.
The control has several requirements.
The temperature must have a very rapid response time one or two seconds. A very small glass 
encapsulated thermistor in the general air flow are used. Thin film humidity sensors can be fast 
enough for humidity control.
The controller is directed at motor control. 
The motors are switched with substantial zero voltage switching solid state relays, - no QRM.
The rules are for both motors, the motor can be switched 
on within a second of the control temperature being reached, but only if a “decompression time” has elapsed.
No noisy switching is possible because any on signal starts the decompression timer.
This would be about 2 minutes for a window unit. This is necessary to avoid harm to the compressor through
over frequent cycling, and it does not have to start under load.
This may seem catastrophic for control, but the 
result is benign.
The fan in the window unit and dehumidifier are left running, probably on “low”.
Changeover from heating to cooling on the reverse cycle unit occurs automatically when a 6 minute timer 
detects no motor operation. This may seem a bit slow, but it happens at a time when heating or 
cooling are hardly needed.
How it works: Your 2 * 3 * 3 metre room contains 18 Kg of air, effectively in a single mass.
When the window unit cuts in the temperature rises of falls at a constant rate of about 2 degrees per minute. When the 
set point is crossed this ramp will stop within a few seconds. The temperature then slowly falls or rises
until after 2 minutes IF it is on the wrong side of the set point the motor starts up again.
Typically the motor runs 15 seconds and stays off for 200.
My rooms had an average temperature that did not shift more than a tenth of a C. They only used the minimum power 
needed, when required, peak variation was about 0.5C. Remember you have to keep the air volume mixed.
The thermometer was mounted in the general air stream about half way around the loop from the AC unit around the room.
One of my rooms had 30% windows and double brick walls but not other insulation.
The window units did not have shortened lives because of the “frequent cycling” because they were started under the right conditions.
I did publish this 30 or 40 years ago in an Australian electronics magazine.

cheers,
Neville Michie



> On 27 Oct 2016, at 2:59 AM, John Ackermann N8UR <jra at febo.com> wrote:
> 
> I may have the opportunity to build a small "clock room" and am considering whether I could make it an environmentally controlled space.  I'd like to learn about the options for doing this.
> 
> The space would probably be 6x8 feet or so, in a basement with one outside wall.
> 
> Can anyone point me to purveyors of the hardware to do something like this?  Because I'll have a limited time to build this, I'm looking for something that uses more-or-less off the shelf gear, and not a whole lot of custom engineering.
> 
> Thanks!
> John
> _______________________________________________
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.



More information about the time-nuts mailing list