Jaguar A/C Vent and Fan Control

[jagx09]

I am thinking of designing a simple more practical solution to replacing the over priced, hard to obtain, failure prone A/C computer in an older Jaguar XJ-S.

The blower fan speed selector would simply need to be a pot with resistors possibly to control lo - med - high speeds on the fans.

The other task would be a circuit or simple solution for controlling the 'blend flap motors' These two stepper motors turn half moon shape flaps or doors inside the A/C unit box that control air flow to the different ducts for the vents on floor, center / side vents and defrost vent.

These servo motors are controlled by the existing computer design. The computer controls the current flowing to these and stops sending the supply when the flap door reaches it's end travel and can no longer turn when placing it in it's needed position to direct the air. Anything short of this would mean the motor continuing to turn and the gears inside the servo casing the motor is in being stripped. So the design I need to control these motors basically has to send the supply current only when needed and to stop the supply current flow when the flap is in it's final position and needs to stop along with the motor.

I have a link to show the cutaway diagram of what the blend flaps look like inside the A/C unit if anyone is interested in this project.

 

[crutschow]

How does the computer know when the door flaps reach the end travel? Does it monitor the motor current?

 

[jagx09]

How does the computer know when the door flaps reach the end travel? Does it monitor the motor current?

I am not sure, I imagine it would have to. My best guess would be the position of the feedback potentiometer at the opposite (shaft) end of the flap. Flap shaft into motor on one side, opposite end flap shaft into feedback potentiometer on opposing end of flap. The fb pot end flap shaft is a half moon shape that fits into the (matching) center hole of the fb potentiometer. Because the matching male to female flap shaft into the fb pot only fitting one way, my thought I this is how the computer knows the position of the shaft (and fb pot) at all times, it is likely factory calibrated by design this way.

However from what I understand about this Delanair MK III system, the feedback potentiometer is there to receive voltages from the 3 temp sensors throughout the cabin that measure various readings and communicate with the computer.

https://www.skjagtech.co.uk/support-pages/generic-support/delanair-mk-3-support.html

 

[crutschow]

If there's a pot on the flap shaft, then that's how the computer knows the flap position. The pot output is proportional to flap position.

The temp sensors' outputs likely go directly to the computer. The computer then controls the flap position based upon the sensors' readings. For example, the higher the temperature delta between the temp setting and the sensor setting, the further open the flaps would be set. It would make little sense to route the signals to the pots.

 

[jagx09]

If there's a pot on the flap shaft, then that's how the computer knows the flap position. The pot output is proportional to flap position.

The temp sensors' outputs likely go directly to the computer. The computer then controls the flap position based upon the sensors' readings. For example, the higher the temperature delta between the temp setting and the sensor setting, the further open the flaps would be set. It would make little sense to route the signals to the pots.

Yes agreed, makes much sense. Thank You. So would you have any ideas on a simpler design to achieve similar results? my thought is to replace the temperature pot with a multi position switch replacement that supplies the motors with the voltage they need to mainly move the flaps to position for each basic setting - defrost, upper vents for A/C and lower vents for heat.

 

[crutschow]

You need to measure the pot output voltage at each desired flap position. Then you could use a number of comparators, such as an LM339 to control the motor voltage on and off in response to your switch settings and the pot position voltage. You may need a little logic circuitry to get it all to work properly. Do a state table of all the switch positions you need and what result (flap position, AC on/off, etc.) you want for each position. Then go from there.

Post the state table in this thread if you need further help.

 

[jagx09]

You need to measure the pot output voltage at each desired flap position. Then you could use a number of comparators, such as an LM339 to control the motor voltage on and off in response to your switch settings and the pot position voltage. You may need a little logic circuitry to get it all to work properly. Do a state table of all the switch positions you need and what result (flap position, AC on/off, etc.) you want for each position. Then go from there.

Post the state table in this thread if you need further help.

Thanks Carl, appreciate your taking the time, I think I can take it from here, working with the feedback pots simulating the same basic functions of the computer then should suffice, will let you know if I need any more help. Thanks again.

 

[holabr]

jagx09,

I would be very interested in your progress since I also have an old XJ-S in need of a HVAC control. I really don't want to pay what they are asking for a rebuilt unit that is based on 30 yo technology. Although you seem much more knowlegable concerning the electronics than I am, I am really interested in learning. So far I have been afraid to tear into the beast.

 

[marcbarker]

An interesting posting - redoing the Delainair Mk3 ECU. I think these cars are a piece of English heritage and should be preserved for as long as possible. I've repaired these ECU's a few times, it is possible. I too also had considered re-doing the ECU from scratch, but had decided against it at the time.

About 10 years ago I published some webpages about the Delanair Mk3 including some details of testing and servicing, I'd also drew a functional block diagram of the ECU and loosely described how it works, but leaving out details of the D to A converters etc. Although I'd terminated the site (far-out.demon.co.uk/cardiy) earlier this year, the information from it has already passed into the public domain (including some still-prevailing myths and idiosyncracies that I'm to blame for, sorry!). However, on the whole I'd say what's been copied from my pages had been improved on, indeed some websites have done a really great job on it, and it's a great feeling to see people developing what I'd done.

But back to the ECU, I'd say used units are in reasonably good supply, because of the rate that xj40's and xjs's are still being scrapped. Also don't forget that the blower units themselves fail more often than the ECU. Here's an example page about it: Air conditioning blower control

What I'd say about redoing the ECU, it's more complex in functionality than I think believed. I think the work involved in reproducing, possibly even equalling (or surpassing?) the performance of the original ECU shouldn't be underestimated. I think a good place to start is to study the technical service manual for the Delanair Mk2 (with cam operated microswitches) to see things like how they have Cabin Demand SetPoint offset, depending on the amount of temperature difference between Ambient and In Cabin sensors. You often find previous versions have more in depth discussions of the design, that's later omitted in the next update. The D-MK3 appears to be the same functionality as the Mk2, but has a microprocessor, A to D and D to A's.

A new "Mk 4" nowadays would likely have a PIC on board, using on-chip A to D's for the 4 temperature sensors, PWM outputs to drive the 2 fans, and appropriate bi-level outputs to drive the blend flap servos.