vaf question

[Rocketman]

I do have adapters available, $35 shipped or available at Ohiostiva for $30

[sketchman]

Yeah, I have a shroom filter that came with the car on it, but it's still a VAF, and it annoys me.

That basic MS kit is still a good ways from a full install.

I'm not sure how they do the fuel pump thing.

I've just ordered an Arduino nano($12 shipped), and I'll be playing with this to see what I can do.

EDIT: Just went back and looked at the site again and saw the fuel pump switch thing.

[sketchman]

So here's what I have so far.

The Arduino reads the MAF signal and converts it to a float value which will be a fractional representation of the max voltage that the MAF can output. Then it takes that and multiply by the max voltage that the VAF can output to send to the ECU. IF the Arduino can do all that fast enough, that is. The fuel pump switch can be done easily enough, I think. A simple relay triggered again by the Arduino when appropriate. The IAT sensor would probably be easier to adapt from the same car the MAF comes from because it's already separate, or I guess if you wanted to you could cut it out of the VAF.

I'm just not sure yet how fast the Arduino can do multiplication and division on floats, which would have to happen REALLY fast. It can read 10,000/sec so there shouldn't be any issue there, but it's the time in between reading and writing that will make or break it.

It may have to leave the math out and use predefined values to look up and use. I just don't know yet.

EDIT: Did a bit more reading. It can do 30k float div/sec and multiplication is even faster. @ 8k RPM you're spinning 133.33333 rev/sec. I'm thinking it'll be fast enough.

[Chobobulous]

From what i got out of tinkering with the vaf it isn't a 1-100 kind of thing. Its 1-10 1-10 1-10. The fuel pump switch is just that a switch it is simply on and off. I bet ot could ne wired to the ignition or to a kill switch.


Correct me if im wrong please

[sketchman]

^I'm not sure what you're saying there. The 1-10 1-10 1-10 bit, I mean.

Even if the fuel pump is just a switch in stock form it may be too much current for the Arduino to output and still may need a relay to do it. I've never done any investigating into this before, though, so this is all new territory.

[Chobobulous]

The switch is mechanical. A metal bar attached to the vane closes it when the vane moves open and OPENS it when the vane is all the way shut.

What i ment with the 1-10 thing there, it seems that the maf uses a straight line likethis/ the vaf voltage goes like this //// moving up to 10 dropping back to one and moving back to10 over and over.

1-10 are example voltages and i hope i made myself a little clearer

[sketchman]

I just went out and grabbed a spare VAF I had and played with it a bit. I have it powered with an old ATX PSU. This gives you both voltages to power the VAF and test it. It seems to be a smooth sweep from 0-11.70v from what I'm seeing on the multimeter. Also I got a reading on the VAT sensor at room temp so I have a baseline.

Anyway, this is more involved than it should be, because instead of using a 5v max signal for the VAF and making things simple, it uses 12v. Why that was done when they had a 5v ref going into the VAF anyway, I have no idea. Anyway, this is annoying because electronics don't like to measure and output such high voltage, which I have found out by searching for components.

What this means simply is in addition to the Arduino to control what's going on, you also need components to supply variable voltage higher than the output pins on the Arduino can do. At the moment I have found 2 different ways to do it. Not sure which one I want to go with yet.

[sketchman]

The switch is mechanical. A metal bar attached to the vane closes it when the vane moves open and OPENS it when the vane is all the way shut.

What i ment with the 1-10 thing there, it seems that the maf uses a straight line likethis/ the vaf voltage goes like this //// moving up to 10 dropping back to one and moving back to10 over and over.

1-10 are example voltages and i hope i made myself a little clearer

OK, I understand now. But see above post. What I'm seeing is a gradual transition from 0 at closed flapper to full voltage at wide open.

The fuel pump switch I can't quite figure out. There are 7 wires coming out of the VAF plug. All of them I have accounted for by playing with the flapper while monitoring with the MM. None of them are a plane on/off controlled by the VAF flapper from what I've observed.

[Chobobulous]

With the pins looking at you and the vaf sitting properly the far left 2 pins are for the cutoff switch., i have one ripped apart at the house ill send photos in like 20 minutes

Maybe mine is broke that's why its not a clean sweep.

[Chobobulous]

Here it is. I have taken the temp sensor out.

[Chobobulous]

Here it is. I have taken the temp sensor out.

[sketchman]

With the pins looking at you and the vaf sitting properly the far left 2 pins are for the cutoff switch., i have one ripped apart at the house ill send photos in like 20 minutes

Maybe mine is broke that's why its not a clean sweep.

OK, now I see. The Haynes manual didn't explain that in the pinout. It just showed 12v going into those 2 pins and I assumed it was only used as a 12v ref, so I didn't even bother testing them. So I'll have to test the current draw with the car running to see whether or not the Arduino can sink enough to hold it closed.

EDIT: Just had another thought. Mattdickmeyer posted saying the VAF is more of an on/off switch even on a very mildly modded Festiva. He said that at 4.5K RPM, it just flips completely open. IF I go wire up my meter and observe that happening, it should be a lot simpler to get this done. Instead of having to build a circuit to out 0-12v, I could just have the Arduino output 0- 5v(the max it can do), and then when it pegs it can close a transistor?/relay? to the full 12v.

More investigations needed.

[Chobobulous]

How are you powering it again? With out using the car i mean. This weekend i will hook this vaf. To the car and observe it. But it looks like once its open it is open.

Im assuming that this is a lot more difficult than what we are thinking. Otherwise it would have been done before. It took those Australian guys forever to do it lol.

Edit. Another thing i noticed was the wear pattern of the electron electronic piece. Only about2/3rds of it is used. That tells me that the vane is never 100% open. This is confirmed by the oil stains in this vaf housing. When the vaf is opened to as far as it was used it only made about a3/4 to a 1 in gap.

[sketchman]

Haynes shows 2 powered pins. One 12v and one 5v. Just so happens a molex connector on a PC power supply has both of those and 2 grounds side by side. Power it up and hook up the MM to read volts from the signal pin. Then I just moved the flapper by hand. I also read from the VAT(IAT) sensor and verified I had the proper pins by slowly breathing hot air onto the VAT and then fast moving air to cool it down.

Those Auzzies also have an RPM input and software to "tune" the curve, and are using a lookup table instead of a straight interpretation from MAF to VAF signal. What I want to do is just translate the signal and keep it as simple as possible. I THINK the Arduino is fast enough to do the calcs and pass on the signal on the fly, but I don't know that yet. It also may take some conditioning instead of a straight conversion, but I don't know that yet either. There is another guy who has done this DIY style on the cheap. I'll see if I can dig up the post.

[sketchman]

Found it. Unfortunately the guy neither give details about how to do it, nor does he provide a part to purchase.

BUT, he did it like I want to do it and with a microcontroller identical in speed to the Arduino's.

http://homepage.ntlworld.com/mx6-uk/massair/massair.htm