A/C interrupt with timer for car - help needed

[mojozoom]

I'd like to make a circuit that shuts off the vehicle A/C system for about 5 seconds when you start the car moving from a stop.

Both the clutch and brake pedals have switches on them, so any time they are both depressed then the brake is released, I'd like the circuit to trigger a relay that'll interrupt the power to the A/C while the timer runs. I'm guessing this would require some sort of AND gate, a latch, a 555, some sort of power transistor and relay?

I just have no idea where to start. This is really my first grass roots electronics project. Thanks!

 

[panic mode]

why would you want that?

 

[cowboybob]

I'm going to assume that you want the load of the compressor removed when trying to accelerate from a stop.

If I understand your scenario correctly, consider this for a moment; every time you hit the brake (however slightly) and fully disengage the clutch, simultaneously, your system would disengage the compressor.

Under stop/go traffic conditions (with subsequent cabin heating issues) that might be bothersome.

I admire your thinking, but practically speaking, it might be easier to just manually disable the A/C at stops.

As an afterthought, I suppose you could tap the speed pick-up signal (which would give you a "vehicle is not moving" indication), to accomplish your goal.

But, again, the stop/go issue would rear its ugly head...

 

[Nigel Goodwin]

I'm going to assume that you want the load of the compressor removed when trying to accelerate from a stop.

Does it make the slightest difference? - I can't say I've ever noticed any such problems, and the AC load is only a tiny fraction of the engine power.

 

[mojozoom]

Cowboy, I'm beginning to agree with you. Probaby best to stick with the tried and true method of monitoring engine load via manifold vacuum and switching it off based on the MAP sensor voltage.
That way I don't have to deal with switch debouncing either I guess.

 

[cowboybob]

Mojo, here again, as manifold vacuum is max (and airflow is minimum) with the throttle closed, any time you remove your foot from the accelerator, your compressor would disengage (with whatever system you use to do that).

Just a thought, but there are times when, at a dead stop, I shift into neutral to reduce the transmission (actually, torque converter) idle load (to save some gas. Maybe not much, but every bit helps [especially for the '74 vette]). The neutral position switch could be tapped for that condition.

Might be a pain to implement a connection, but that action would for sure be an indicator of a dead stop (555 timer OFF, relay OFF, compressor OFF) and, then, after a return to drive (timer ON with a 555 driven 5 sec relay ON delay, then compressor ON again). This would also work from a "cold iron" (car off) situation.

I'm sure there are others ways to skin this cat (my insincere apologies to PETA).

 

[mojozoom]

Manifold vacuum decreases with engine load, so it's a workable methodology. The MAP (manifold absolute pressure) sensor typically outputs a voltage signal scaled over the pressure range of the sensor, so I just need to build a switch that compares the MAP sensor signal to an adjustable reference voltage and use that to trigger a relay.

The concept of using manifold vacuum to cut off the A/C is nothing new - here's an EPA evaluation of a device that used the concept back in 1980:
https://www.epa.gov/oms/consumer/devices/pb82178534.pdf

The EPA concluded that a fuel economy improvement of up to 4% could be had depending on A/C system type, driving conditions, and the amount that the A/C was used. It's been from 95F - 106F here in KC for the past few weeks, so I'm using the A/C a lot right now.

Here's a circuit that essentially does what I'm looking to do, so I'll probably take this path:
https://www.autospeed.com.au/cms/A_110955/article.html

 

[cowboybob]

Simple enough solution.

I didn't realize that you were primarily interested in ALL hard acceleration situations.

The system is, essentially, an OpAmp voltage comparator (with the volyage being compared coming from the MAF) whose output is then used to trigger a 555 timer attached to a relay. Voltage comparator hysteresis and Timer delay adjustments will need to be included in its design.

Is this what you were looking for?

 

[mojozoom]

That's the ticket. I wonder if a timer really would be needed at all if the hysteresis is set up right. I really want to avoid excessive cycling of the compressor clutch.

 

[cowboybob]

...I wonder if a timer really would be needed at all if the hysteresis is set up right...

That's a good idea. Controlling compressor OFF timing by the OpAmp comparator window values (and the rise times by the RC component of the input) would be more in sync with hard acceleration duration timing.

 

[mojozoom]

I also want to implement a two stage shift light on the car. Maybe a microcontroller would male sense then.
I know the Picaxe is a no-no for auto apps, but I liked the cheapness and easy programming. Is there a microcontroller that's good for this type of thing?

 

[cowboybob]

I also want to implement a two stage shift light on the car.

(My italics)

Not sure what you mean... Something like a light indicator of staged RPM values for a particular gear (just a SWAG on my part)?

 

[mojozoom]

I'd like a green led at say 3500 rpm as my normal driving shift point then a yellow at say 6400 rpm for best performance shifts.

Typically this would be an LM2917 freq to voltage converter and an external op amp. If you used an AVR though you could set a different shift point for each gear, as the differing gear ratios in the trans really require different shift points to maximize acceleration.

If you pick up the vehicle speed signal (VSS) from the trans and divide it with the engine rpm you can figure out what gear the car is in, based on the known gear ratios for that car.

Without a microcontroller thats out of reach though.

Did I mention that I want one more feature on this circuit? I really want some indicator of when the car has engaged the VVT. Again it should just be a matter of grabbing the signal, tweaking it, and comparing it to a reference to light an led.

 

[crutschow]

Does it make the slightest difference? - I can't say I've ever noticed any such problems, and the AC load is only a tiny fraction of the engine power.

It does depending on the size of the engine. My car has a 2.5L engine and it definitely has a noticeable reduction in acceleration if the AC is running, even for just normal acceleration from a stop.

Some time ago I installed a vacuum switch designed for just that purpose, to cut off the A/C when the manifold vacuum dropped below a certain point, indicating acceleration. The problem was that it was a manual transmission and so the A/C clutch cycled on every shift. To avoid undue wear on the clutch I added a delay circuit to increase the turn-back-on time to give me time to shift. I also added a bypass switch so If I was climbing a long grade I could keep the A/C on, even if the manifold pressure dropped. It all worked quite well. Unfortunately it broke and I never got around to fixing it.

 

[mojozoom]

If we compare to the MAP sensor voltage we should be able to adjust the vacuum setting for each car. That should help with the hill issue.

Maybe we use two timers so there's a minimum time that the A/C is off, and a maximum time as well to cover the long hill issue.

 

[crutschow]

...........................
Maybe we use two timers so there's a minimum time that the A/C is off, and a maximum time as well to cover the long hill issue.

Sounds like a good plan to me.

 

[mojozoom]

So to get this done I need a comparator and a 556 timer. The shift light would need a LM2917 and two more comparators, and the VVT light would need yet another comparator. Plus all the ancillary stuff to condition signals.

The AVR is looking better all the time...

Maybe a Teensy?

 

[cowboybob]

"AVR" means what in this instance? I'm not making a connection...

 

[mojozoom]

AVR's are Atmel microcontrollers. The Teensy is neat because it just takes a usb cable to program it and its similar to an Arduino. Plus its only $16.