fabricating a VSS?

[dcb]

just did some bench testing of device in post 25:

+ connected to 12v battery
- connected to battery ground
o connected to 12v via a 5k resistor.

output at o with no metal near tip, 12 volts

output at o with metal at tip, 0 volts

device current draw @ 12v output, 8.25 milliamps

device current draw @ 0v output, 11.6 milliamps

notes:
the steel had to be pretty close to the tip to set the output low, like less than 1/4 inch.
There were no intermediate voltages on the output.
the voltage output at o should work fine as a speed input to the guino.

[usergone]

Looks like I aughtta go buy me one of them.

[dcb]

If you like to buy one, it isn't horribly expensive, I will be testing it on the car within a couple weeks too. I've been wanting a generic vss option for a while, and this seems like about the easiest option. The injectors are already pretty generic since they are easy to find on the car, and there only two wires to choose from (guess or measure). So with a "bolt on" vss you wouldn't have to even look at a schematic, regardless of the car, but you would have to locate an appropriate spot for it and bracket it up, which is probably right up a lot of modders allies anyway.

My only outstanding concern is in the frequency of pulses being outside what has been tried. We had to put some code "hacks" in to make the reed switch vss signals look acceptable (i.e. debouncing periods) that limit the max speed the guino can register. There was an old metro needed it, but my saturn (which uses a hall effect) doesn't, so that delay ( VSS Delay ms) can probably be set to 0 for this device, but I haven't tested it yet.

So if I manage to tuck it inside the wheel, and my tires get 850 revolutions per mile, and I have 10 holes in the steel wheel, that will be 17000 interrupts per mile for the guino vss setting (850*10*2). The "typical" setting is 8000-10000

edit: if a rotating part that doesn't have any protrusions or ferric material, where it is reasonable to mount the sensor, can't be found, then drilling and taping a couple opposing machine screws onto a better situated rotating surface would likely suffice.

[usergone]

That is the last setup I tried. You can see the micro switch with a roller, and that grey blob is JB-weld, making a bead that pushed on the switch. This is the last one that I tried, which did not work.


If you look at the driveshaft, there is a perfect spot right after the weld (between the round part and the bearing connection) which should give me 2 passes/revolution of the driveshaft. Fabricating a mount would not be very difficult at all, I would simply make one very similar to the one you can see in the picture.

[dcb]

ahh, good old chunky ujoint . There I'm thinking of pointing it at the inner corners of the yoke, i.e. lust left (forward) of the jbweld, pointing at the centerline of the yoke.

also putting longer ujoint bolts in and clamping some two sheet metal "paddles" on would be easy, and point the sensor forward so it only sees the paddles as they go by would make for a well balanced signal with more even on and off time.


Little plumbers tape, tightened with a machine screw, to hold a machine screw threads out, might be all you need for bracketing on that diff.

[usergone]

Actually, the bracket you see with the wire wrapped around it is on there very tight. The wire was just to hold it while the JB cured. I have not tried to break it off yet, but I suspect I will need a big hammer

And in post 34, I meant on the shaft side of the U-joint, to the left of the bearings. I don't think the shape is very good just to the left of the bead of JB-weld, it has an odd profile (like a rectangle that has protuberants on the middle of the long sides)

[usergone]

I have ordered that sensor, and am now awaiting its arrival

[usergone]

The sensor arrived today, so I got busy installing it. I used some sheetmetal with a rib to reinforce it and hopefully keep it from slamming down into the driveshaft.

The initial tests were good, but I kept the speed down very low, because the JB-weld I used to attach it was not cured.

[usergone]

Update: I took it for a better test drive, but at 30 MPH, I went over a bump and the sensor (I can only speculate) hit the drive shaft, ripped itself off of the mount, and was dangling beneath the car by one wire.

I still have the sensor, thankfully

[usergone]

I finished mounting the sensor again. This time I put it right up by the transmission, where I had the magnetic reed switch a few tries ago. I welded bold heads onto the driveshaft, and put the sensor so that they would pass by as it spun.

In the first test (down the street very slowly) it worked great. Then I pulled onto a bigger street and got up to about 45. When I looked down, I saw that the guino was reading half the actual speed. I figured that it might be the sensor not being able to pick up the bolt head flying past, so it might be sending just one signal out.

Then I realized that the datasheet said that it could switch up into the kilohertz range (if I remember correctly) so I looked in the config part of the guino, and... voila. There was a VSS microseconds setting, which was on 2.

I set it to 0, took another fast test drive, and it read correctly. Now I'll need to take a look at the speed when I am up to 60, to see if the guino can register that fast.