Automated heater box project

[Reinsch]

Hello members,

I thought I will share my project in regards of automation of heater box. As my project car #4454 is undergoing resto-modification, all parts are reviewed and upgraded if possible.

I explain most of the details in video attached.

if you want to know prior to watching - I did standard renovation of box, but also lined it inside with silver tape to increase efficiency. On top of that I've added electric stepper motor to control air mixing flap (no more mechanical cable). Direction flaps are still based on vacuum actuators, but those are connected to air solenoids (no more leaky rotation switch). Heart of this project is Arduino microprocessor that will be also gathering data from few temperature sensors, to achieve demanded cabin temperature (this will be set by dial knob).

For testing purposes I'm using touch screen, but vehicle control panel will have knob. More in video:


https://www.youtube.com/watch?v=U54ae7XVkqA

regards,
Marek

weekly updates of my project on facebook group (Delorean vin 4454)

[jangell]

Very nice! I’ll be following this closely.

I’m working on a similar system, but you’re a bit further along than I am. I’m going to use a 7” touch screen to replace the climate control knobs, driven by a Mac laptop (which will eventually do more than just this). An Arduino will be the Mac’s interface to the car, and will be used to control a similar relay board (also for heated seat control) by means of a transistor board (I’m worried about overdrawing the Arduino without it). That board will also control the vacuum solenoids for the vent/seal/vac, just like you’re doing. And I’m using a variable speed motor controller for the blower motor.

I hadn’t figured out where to put the temp sensors yet — I’ll just do what you did, probably. putting one where the fan resistor lives is a good idea. Since my evaporator is in the car, I don’t know if I’ll be able to mount a sensor on it. I’m also considering a sunlight sensor, which is apparently a thing that modern cars do to amp up the A/C a bit when it’s really sunny out. The logic is that the passengers can feel the sun on their skin, but temp sensors elsewhere in the car might not be able to read that.

I’m keeping the cable for the hot/cold split flap, but I’m going to drive it which a high-torque servo motor. I admit that I still need to test it to make sure it can actually move the flap, but I think it should. My box still being in the car makes it difficult to replace the cable with a stepper/servo directly

My plan is to do full manual and full auto modes, including fans speed, using my 2015 Subaru Outback’s behavior as a reference. The part I’m trying to figure out is how to adjust everything when it’s close to the desired temp. Should the fan ramp down before the hot/cold flap, for example. I’m probably just going play with the Outback’s controls and see what it does in different scenarios, but I’m also curious with what you’ve come up with for that.

I have the Mac software coded (mostly, with a UI aesthetic similar tin Apple CarPlay, as that is what the stereo immediately above it will show most of the time) and the Arduino communication working, as well as tests with the relays, servo and vacuum solenoids. I really just need to test the blower motor variable speed control and then I can start installing it.

For the final install, I’ll have a 12v to USB-C power supply for the Mac laptop (it will automatically sleep when the car is off), a 12v to USB-A power supply for the Arduino and touchscreen, and I’ll be relocating the headlight and hazard switches to a pod near the right driver’s side knee pad area, at least for now. I decided that it would be a bad idea to trust my homebrew software to the headlights and then lose them in the middle of the night when the app crashes. the Mac will sit in a sleeve mounted in the cabin, although I haven’t figured out the details yet. I may install a second battery dedicated to these additions so that I don’t risk draining the starter battery when the car is off.

I’ll be very interested to see your system when it’s all done — good luck!

— Joe

[81dmc]

Hello members,

I thought I will share my project in regards of automation of heater box. As my project car #4454 is undergoing resto-modification, all parts are reviewed and upgraded if possible.

I explain most of the details in video attached.

if you want to know prior to watching - I did standard renovation of box, but also lined it inside with silver tape to increase efficiency. On top of that I've added electric stepper motor to control air mixing flap (no more mechanical cable). Direction flaps are still based on vacuum actuators, but those are connected to air solenoids (no more leaky rotation switch). Heart of this project is Arduino microprocessor that will be also gathering data from few temperature sensors, to achieve demanded cabin temperature (this will be set by dial knob).

For testing purposes I'm using touch screen, but vehicle control panel will have knob. More in video:


https://www.youtube.com/watch?v=U54ae7XVkqA

regards,
Marek

weekly updates of my project on facebook group (Delorean vin 4454)

Very cool!

Why don't you swap all the vacuum pods for electric actuators? Then you could blend modes if the design of the box allows.

[DMC-81]

Wow, that's amazing! Great job!

[vwdmc16]

This is a fantastic update idea. It would be super great if you could select defrost and face vents at the same time, I really wish that was a common option on cars.

[Reinsch]

@jangell

You are right, Arduino power output is limited. Manual says : max per pin 20mA, recommended 10mA. Power sum for each group should not exceed 100mA. Max for entire package 200mA. This is for Mega board.
I use relay board with galvanic separation to arduino (via optocouplers) and separate power supply (step down buck 12V-5V) to avoid any interference from solenoids/relays. I had this issue on early stages of my project. Thanks to this solution, problem is gone.

PWM control module for blower in my case is a bit too weak. I currently use 15A unit, but it's getting really hot despite only using 10A out of it (limit on my test power supply). I would suggest 30A unit and some good cooling for heatsink. In modern heater boxes, amp for blower motor sits in box intake, just after blower, so you get good cooling. I will be cutting hole opposite to stock resistor location (top of the box). It is better location then stock resistor hole as not exposed to rain water.

Algorithm for controlling fan speed and flap position is quite complex. I have opportunity at work to consult my work with automotive heater box designer and he explained me how it all works. You need variables such as evaporator temp (unless you assume its always ~1*C when A/C is on), coolant temp (unless you assume its always 85*C), outside/box intake temp and cabin temp. Fan speed should depend on difference between actual and set temperature - greater the difference, higher the speed of fan. However, once you set it to +30*C for cabin and it's -10*C outside, you don't want to blow cold air into your face, so you need to consider engine coolant temp. Some cars also have 3 modes for auto fan (soft, normal, fast).
There are also constants that will be specific for vehicle and those are "tunable factors".

Sun sensor is good idea, especially if you can get one with datasheet to understand relation between light and resistance (they are usually photoresistors).

I was also thinking about servo motor, but it was hard for me to determine required torque. Big advantage is known position of your servo.
After trying some smaller stepper motors (too weak, despite gearbox) I decided to apply confirmed solution - a stepper from modern heater box. It has a lot of power and is able to move a mixing flap easily at highest blower speeds. Only disadvantage I found is that I need to redesign bottom air guide, as it is clashing with my motor.

I don't think I mentioned in my video, but I had to add Teflon bushes for air mixings flap. Otherwise it would be quite hard to rotate it, as shaft is held by 2 metal clips. Those clips are pushing shaft against plastic housing. I guess you need that in stock design as you want to keep your flap in set position. However, solution with stepper motor does not require that friction. For holding flap in position, you apply small current to motor winding (so called holding current) and flap won't move.

I will share my further development. Currently working on putting new engine to frame and frame-body marriage after that. I hope to have this setup in vehicle by end of year.

Please share your project too. Maybe it will inspire others!


@81dmc
swapping vacuum actuators for electric ones, would require quite a redesign. Direction flaps usually operate in OPEN or CLOSE manner, so it would be a major overkill. However, it is possible.

@DMC-81, vwdmc16
Thanks!

regards,
Marek

[Bitsyncmaster]

If money is no object, these servos should have plenty of power and they hold without power.

https://www.aircraftspruce.com/catal...SABEgLcUfD_BwE

[DMCMW Dave]

If money is no object, these servos should have plenty of power and they hold without power.

https://www.aircraftspruce.com/catal...SABEgLcUfD_BwE

Before building anything new (especially displays and processors) into a car you should verify that the temperature range (stored and operating) is appropriate to an automotive environment. -40C to +90C seems to be a common automotive interior spec. Although for use in a "classic" car, -40C is probably more extreme than you need to worry about. -15C is probably good enough for a garaged car.

[jangell]

You are right, Arduino power output is limited. Manual says : max per pin 20mA, recommended 10mA. Power sum for each group should not exceed 100mA. Max for entire package 200mA. This is for Mega board. I use relay board with galvanic separation to arduino (via optocouplers) and separate power supply (step down buck 12V-5V) to avoid any interference from solenoids/relays. I had this issue on early stages of my project. Thanks to this solution, problem is gone.

I'm using the Mega as well. I planned on using optoisolators for the eventual digital inputs from the 12v systems (for an upcoming custom instrument cluster project), but for some reason I didn't think of them for the relay board. I will be using a 12v to 5v buck power supply (I said USB A before; my mistake) for all the 5v devices, including the Arduino, although that could also run off of USB power from the Mac.

PWM control module for blower in my case is a bit too weak. I currently use 15A unit, but it's getting really hot despite only using 10A out of it (limit on my test power supply). I would suggest 30A unit and some good cooling for heatsink. In modern heater boxes, amp for blower motor sits in box intake, just after blower, so you get good cooling. I will be cutting hole opposite to stock resistor location (top of the box). It is better location then stock resistor hole as not exposed to rain water.

I like the idea of putting the motor controller in the path of the blower motor airflow to keep it cool. Avoiding the resistor area is a good idea as well (for water), but since mine is in the car I'll probably have to face of the box near the glove box rather than the top. For motor control, I'm using a Polous dual motor controller configured to run one high-amp motor (max load is 24A with 60A spikes when set up this way). I have an old motor I'll be using to test this. I should probably mount heat sinks on the chips just to be sure I don't overheat them, though.

Algorithm for controlling fan speed and flap position is quite complex. I have opportunity at work to consult my work with automotive heater box designer and he explained me how it all works. You need variables such as evaporator temp (unless you assume its always ~1*C when A/C is on), coolant temp (unless you assume its always 85*C), outside/box intake temp and cabin temp. Fan speed should depend on difference between actual and set temperature - greater the difference, higher the speed of fan. However, once you set it to +30*C for cabin and it's -10*C outside, you don't want to blow cold air into your face, so you need to consider engine coolant temp. Some cars also have 3 modes for auto fan (soft, normal, fast).
There are also constants that will be specific for vehicle and those are "tunable factors".

The complexity is part of my concern. I figured that there's a few states that the system can be in:
- Cold engine: Keep the fan low until engine temperature his high enough to actually heat the cabin, then go into normal automatic mode.
- Warm A/C system: Same, but waiting until the evaporator cools down.
- Cabin Too Hot, Normal Operation: Hot/cold split to "cold", max fan. Fan and hot/cold split reduce as you near the desired temperature, but it's unclear to me in what ratio to adjust the fan speed and hot/cold sprit relative to each other. Maybe turn off the A/C part if it's not humid in the car or outside and the exterior temperature is close to the desired interior temperature.
- Cabin Too Cold, Normal Operation: Hot/cold split to "hot", max fan. Adjust both as you near desired temperature as with A/C.
- Recirculation flap position would be open if the exterior temperature is closer to the desired cabin temperature than the current cabin temperature.

I also figured there'd be a 1-2 degree cabin temperature hysteresis where the system would sit idle so that it isn't flipping back and forth all the time.

Some googling found that cabin temp sensors are usually mounted on a tube attached to one of the vent ducts, which I didn't expect. The idea seems to be that the moving air in the duct pulls cabin air into the tube and past the sensor. That way you should be getting a constant flow of cabin air that should better represent the current cabin temperature, rather than trying to put a series of sensors around the cabin, so I figure I'll give that a try.

I also noticed that my Subaru's fan is always on, I assume at a minimum so that it can more accurately monitor the temperatures and to cool the blower motor controller (now that I know that it's probably mounted in the path the blower).

Sun sensor is good idea, especially if you can get one with datasheet to understand relation between light and resistance (they are usually photoresistors).

I think the main trick will be finding one that you can buy off the shelf. I think they just put a hemispherical lens over the photoresistor, but I'm not totally sure. I also figured I could make a table easily enough myself if I did get one from a junkyard, but a data sheet would be a lot easier.

I'm also thinking of a humidity sensor to decide if the A/C should be on vs. just using the vent. Also, a humidity sensor near the rear view mirror seems to be used in some cars to detect if the window may be fogged up and automatically turn on the defroster when starting the car, so I may try that out as well.

I was also thinking about servo motor, but it was hard for me to determine required torque. Big advantage is known position of your servo.
After trying some smaller stepper motors (too weak, despite gearbox) I decided to apply confirmed solution - a stepper from modern heater box. It has a lot of power and is able to move a mixing flap easily at highest blower speeds. Only disadvantage I found is that I need to redesign bottom air guide, as it is clashing with my motor.

I don't think I mentioned in my video, but I had to add Teflon bushes for air mixings flap. Otherwise it would be quite hard to rotate it, as shaft is held by 2 metal clips. Those clips are pushing shaft against plastic housing. I guess you need that in stock design as you want to keep your flap in set position. However, solution with stepper motor does not require that friction. For holding flap in position, you apply small current to motor winding (so called holding current) and flap won't move.

It is really, really hard to turn my hot/cold dial by hand, especially at the extremes. I was hoping this was the gearing mechanism in the dial, but now I'm worried that it's the flap hinges that are hard to move. I'm definitely not taking the box back out to replace the hinges. I had it out a year ago when I refurbished the A/C system, but didn't think about the hinges or coating the box with silver tape.

I didn't think of the fact that the flap has to also push against the air from the blower as well. The 20 kg/cm servos I got feel pretty strong, but I don't know if they're THAT strong.

I will share my further development. Currently working on putting new engine to frame and frame-body marriage after that. I hope to have this setup in vehicle by end of year.

Good luck!

Please share your project too. Maybe it will inspire others!

I definitely will, once I have something more interesting to show.

@81dmc
swapping vacuum actuators for electric ones, would require quite a redesign. Direction flaps usually operate in OPEN or CLOSE manner, so it would be a major overkill. However, it is possible.

I had wondered about this as well. It was my original plan until I realized I could electrically control the vacuum. I sort of wanted to upgrade to electric just to get rid of the vacuum actuators, but they work, and it didn't seem like it was worth the time investment.

Thanks for the ideas!

-- Joe

[jangell]

If money is no object, these servos should have plenty of power and they hold without power.

https://www.aircraftspruce.com/catal...SABEgLcUfD_BwE

While the installation of a Mac in my car does imply that money is no object, it's an older Mac. Those look cool, but I really hope I can find something more affordable if need be. It's nice to know they exist if all else fails.

Thanks!

-- Joe