Digital Temp Control for the A/C

[ganun]

Not leaving well enough alone and just because I'd like experiments.

Rather than the rotary temp switch (thermostat) control of the compressor, (ancient tech) why not a digital controller where the ON and OFF temp for the compressor operation can be easily set.
Say 38 and 34F, with the sensor in good contact with the evap outlet pipe with the intention of maintaining the evap as close to freezing as possible but not below.
Yeah I know this is the function of the thermostat, but with digital control one can really play with it.
A concern is causing too much cycling on/off of the compressor because the evap is getting too cold in which case I would think increasing the amount of 134 in the system to raise the low side temp a bit.
So what's the consensus, + or -

This controller retains memory when off.


Minimal mod really, just put it here, leave the the old knob in place just cut the wires in a convenient place.

[RNajarian]

Wow! Innovative idea! I bet this catches on

[OldSpool87]

Neat concept. Sub'd if not for the awesome AC donnybrook to ensue.

[Ronnie's.930]

I've got wood (for your idea George, not for YOU, bro!).

I use a similar device for my bedroom window a/c - a digital temp controller that overrides the original thermostat and tricks the unit into thinking the incoming air is warmer than it actually is. The controller has a room temp sensor, a heating element that attaches to the original temp probe, and a frost sensor so that the compressor will shut off briefly when the evaporator coil begins to freeze. I use this in order to make my room very cold for sleeping (much colder than the window a/c thermostat will allow).

It looks like you have a similar idea in play, but without the frost sensor? Regardless, very interesting!

[ganun]

Cool, you've done it already.....Not an original idea, it comes from several people, looks like you're doing it, same thing, basically a frost sensor for the evap.
So you are a transplanted Eskimo living in TX, ....feel sorry for you!

[Ronnie's.930]

Originally Posted by ganun

So you are a transplanted Eskimo living in TX, ....feel sorry for you!

Exactly! I am as misplaced, climate preference wise, as anyone could be.

One example of how a temp controller like that is worth its weight in gold is when the outside air is not warm enough to make a non-modified a/c unit work long enough to remove interior humidity. Several months of the year around here, the outside air is "cool" enough that a good sized window unit will quickly knock the room temp down to the unit's minimum setting (typically about 64 degrees), but it does not run long enough to clear humidity, so with the controller, the a/c runs more and dehumidifies the room much better (makes the interior air much colder than 64, too, buttofcourse).

[spyerx]

A PID that controls a digital relay could easily do this. You can set a curve with a range to keep it cycling too fast.

http://www.ebay.com/itm/DC-12V-24V-PID-Temperature-Controller-K-sensor-25A-SSR-/320843501848?hash=item4ab3c35118

[ganun]

Spy, don't think PID is needed or wanted here, the typical solid state relays and fast switching is not what the compressor likes.

[ganun]

A quick update,
With a 12v battery, some wires, a cup of ice water and a multimeter my son and I played around with the controller some to test its functions. The hardest part was trying to understand the translated instructions, but once we figured out the logic then it was a piece of cake. We got two, one in C and one in F, at a cost of 6 bucks each why not.
They both work the same,.... a quick press of the SET button, then scroll up or down with the arrows to the desired set-point. A 5 sec press of the set button lets one scroll thru settings such as DELAY and DIFFERENTIAL. Diff is the key setting, it is simply how many degrees of temp rise before the compressor reengages after it reached the set point and shut off. A time delay can be added also but we will see if we need that function.
OK, with the controller working we would put the sensor in a cup of ice water and see the relay click open when it reached our setpoint of 34F then take the sensor out of the ice, let it slowly warm up and see that at 39F it would click close. We will start with these settings i.e. SETPOINT 34F, DIFF 5F .

[ganun]

Finally this afternoon it was cool enough (80s) after some massive storms to get in the garage and try to install the controller.
Intent is to make this plug and play with no mods to the original system so that its just 2 wires that are unplugged from one controller to the other and the original is undisturbed except for being unplugged.
The work:
First, pull the 2 wires off the thermostat. G/W (power) and G (to comp).
The original thermostat gets power from the fan switch right next to it so unfortunately that means having to deal with the fun console to reach the 2 wires and pull them off the thermostat.







The wires are just short so we will about 4 inches of 14 gauge with a spade terminal so that the original wire terminals left intact.
Also....Wondering why the oe wires are ~12 to 14 gauge we measured the amp draw of the clutch and relay in the circuit. Got 1.86A , so the controller rating of 10A is fine.

[ganun]

The controller gets power from a small jumper wire soldered to the G/W (recall that the G/W gets power when the fan switch is turned on). I'll update the schematic on the first post with the complete wiring later. Its really simple!
So....time to make the 2, four inch extensions, one with a jumper attached, and then thread them thru the back of the console and under the carpet. No fun here. But years ago I had opened up that miniscule slot to something more reasonable so as not to force the wires to go up and over the back side of the console in a most convoluted manner, .......what were they thinking!



I forgot to mention that I made a simple plastic panel with my sons hobby 3d printer to support the controller.

Time to test, so temporarily find a ground for the system at the bolt that holds down the pass side floor board, and temp sensor will be sent forward tomorrow to attach to the evap outlet with some heat transfer paste and a tiny hose clamp..... key to on, rotate the fan to turn on the ac and it looks good so far.


AMBIENT TEMP

SET POINT




Hopefully we will finish tomorrow, so far its taken about 4 or 5 hours going slowly.
Gee, even my old Peugeot has a Bosch computer and input from 4 sensors ( outside air, inside air, evap and heater core temps ) to run the climate control

[ganun]

Today we attached the sensor to the evap outlet with some heat transfer paste and a small hose clamp.



Well it was time to see if it worked........Started the car with a big fan blowing under the car to the back.
Temp was set at 33F + 5 diff>>>>>>after about 5 to 10 minutes the sensor got to about 36+F and the cabin air temp about the same BUT checking the evap coil temp showed 28F ....so clearly the set point had to moved up.
Picture shows after about 5+ minutes, it was getting real cold.
Sensor coincides with air temp unintentionally, I really want to monitor and control the temp inside the evap coil, which is colder, to prevent icing.






To be continued...

[Ronnie's.930]

George, love your project and detailed posts. Like I said before, very sweet idea.

[tirwin]

Love this project.

I have been thinking about this for a while. Since I deleted the center console I have temporarily relocated the A/C controls to where the ash tray used to be. I have been thinking about trying to come up with something that would replace the fresh air controls and integrate the A/C controls. The hard part is making it look retro and not uber-modern.

[Duc Hunter]

I have wanted to find a way to temporarily monitor the EVAP temp, so I can determine where not eh story dial is "cold enough". My plan was to then note/mark that point and remove the eval temp sensor. I am eager to see how your adventure goes.

Side now, my expansion valve was not working very well until I covered its sensor on the same eval line you are using was covered in cork tape. Just an idea.

[ganun]

Thanks guys!...appreciate the encouragement, still have to actually drive the car to fully test it and see IF the combination of set point and diff value will give the desired results. ie always ice cold AC
BTW still waiting for comments from Kuehl and the gang

[kuehl]

We started a project a few years back on a complete AC control, let's call it a 'head unit', alike a 964/993 however limited to AC and not heat. Function wise the circuit is simple and logical, controlling compressor cycle and evap fan speed in one unit, microprocessor based.

From a marketing point it had limits. 2 models were required (bolt in wise), pre 86 and 86 post center consoles require 2 different fascia mounts. Then there is the users preference for originality in look (2 oem style knobs to control settings however digital or graphic display; the quirk). Looking at the number of models on the road in light of those whom either want something 'better' or those that have issues in maintaining optimal evap core temps, that weighed with the 'factors of production' (minimum production run), margins, and price point .... simply put: the economy of scale to bring a product to the market that works and is profitable to maintain. Hence, we put the project on the side for the moment.

But, we have in the works a simple digital based thermostat designed to be a 'drop in' replacement for the OEM and will have a user adjustment feature; we have been testing it for the past 3 months (busy golf season).

Ganun, my thoughts: You did a clean job with the install thus far. The look of the unit in console will be a personal choice for the car owner. I'm sure there are many fellow PP members that will take on your concept here. My only suggestion to you is to move the digital thermostat's probe from the evaporator outlet and simply put it in the right place in the evaporator core.

Most 911/930 owners frustrations with the operation of a good system is evaporator freezing. This is simply an issue of not getting the thermostat probe properly located in the evaporator core. An oem mechanical thermostat has no issues turning on and off the compressor when the core reaches the freezing point provided their thermostat works (test it) and the probe is in the right place.
If you are using a system that attempts to adjust compressor on/off cycle "after the fact" evap core outlet temperature, you are making things too complicated.
Try moving your digital thermostat's sensor into the evaporator core: as the refrigerant travels through the evaporator the temperature of the refrigerant rises as it takes on more heat from the core mass. By the time the refrigerant gas exits the evaporator core the heat value is higher than the average core temperature.

Have fun with the project.

[ganun]

Originally Posted by kuehl

Ganun, my thoughts: ...... My only suggestion to you is to move the digital thermostat's probe from the evaporator outlet and simply put it in the right place in the evaporator core.

Most 911/930 owners frustrations with the operation of a good system is evaporator freezing. This is simply an issue of not getting the thermostat probe properly located in the evaporator core. An oem mechanical thermostat has no issues turning on and off the compressor when the core reaches the freezing point provided their thermostat works (test it) and the probe is in the right place.
If you are using a system that attempts to adjust compressor on/off cycle "after the fact" evap core outlet temperature, you are making things too complicated.
Try moving your digital thermostat's sensor into the evaporator core: as the refrigerant travels through the evaporator the temperature of the refrigerant rises as it takes on more heat from the core mass. By the time the refrigerant gas exits the evaporator core the heat value is higher than the average core temperature.

Have fun with the project.

Thanks Kuehl, ..... I think you're absolutely right, you confirmed what I was thinking about....I didn't want to make the hole to get the sensor into the evap but it's going to happen, I think I'll just pull out the original sensor, put it safely aside and enlarge the hole enough for the digital one. You know from the first run I saw that what you said about "after the fact", .....i.e. a temp difference between the outlet and the evap core of maybe 8 to 10 F.
Thanks

[GH85Carrera]

I can appreciate all your work and it is cool to see digital controls.

It is just my personal opinion and I could be wrong but, I see it as a complex solution to a problem that does not really exist.

Not that there is anything at all wrong with that!

My simple analog OEM temperature control knob works just perfect with my Dual Kuehl condensers and evaporator. No doubt I spent more doing that than your digital display.

Good work and excellent documentation!

[ganun]

Thanks Glen,
Like I said I couldn't leave well enough alone, besides in a matter of less than 10 minutes the original thermostat can be reconnected, it's only 2 wires to reconnect, but I really want to give this a good push forward.
BTW So far I only have an incredible 6 bucks invested and I am learning. For instance it's interesting to see the expansion valve function as it cycles open more or less and how it reflects on the evap temp. Thinking about, if I have to go back to the original thermostat it it might be cool to leave the controller installed, just passive reading of the evap temp, just to monitor the function of the original thermostat.
The joke is that P called its ac system "CLIMATE CONTROL" nothing could be further from the truth. It could not be a more basic system.