are your wiper intermitant speed linked to your cars road speed?

mine aren't, the intermitant speed is controlled on the end of the wiper control stalk, there's a control that is rotated to change the wiper operation on it, it stays at what you set it, despite changes in road speed

I am sure that the speed sensor would a straight pulse generator varing with speed and would be linear as its speed is directly dependant on the rpm of the gearbox output shaft. It makes no sense to pulse width modulate this signal. The signal output from the cluster however would be serial data.

Cheers
TW.

some fairmonts and fairlanes etc had intermittant wipers that changed speeds depending on road speed

EL Futuras and above (for EF's it was Fairmont and up) got speed sensitive intermittant, as well as the stalk adjustment.

Try a test, set your stalk to the slowest speed, and while stationery, time how long between wipes (for me IIRC it's 8 secs), then take off to about 60kM/H, still with the stalk set to the slowest, time the time between wipes again, for me this drops to about 5 secs.

cool, i'll try it out, never noticed any diff before

what i'd really love is speed sensitive power steering, they had that in all the ELs and models after didn't they? ... wonder if it could be adapted to an EF, inexpensively that is

so @ 10kM/H you should see 17.3pulses/sec with a inter pulse delay of 57.8ms
@ 40kM/H = 69.4ps = 14.4ms
@ 50kM/H = 86.8ps = 11.5ms
@ 60kM/H = 104.1ps = 9.6ms
@ 70kM/H = 121.5ps = 8.2ms
@ 80kM/H = 138.8ps = 7.2ms
@ 90kM/H = 156.2ps = 6.4ms
@ 100kM/H = 173.6ps = 5.7ms
@ 110kM/H = 190.9ps = 5.2ms
------------------------------------
@ 220kM/H = 381.9ps = 2.6ms

If your uP can time the pulses down to 2.6ms you can have LIVE (ie. instantaneous at each pulse) speed displayed. If your uP isn't that fast, do a .1sec count (divide the pulses/sec number above by 10) and update the speed every 1/10th sec.

Wow,
I didn't make quite that many examples for myself.
Only that the number of pulses p/sec x 60, then divided by 6250,
then x 60 again = Km per hour.

ie. 17.3611 pulses per sec x 60 = 1041.666
1041.666 / 6250 = 0.1666
0.1666 x 60 = 10 Km per hour.

since my mcu don't like decimal, I have to multiply first, then divide,
so it's more like:

pulses per sec x 100, x 60, /6250, x 60, x 100 = Km per hour



Or I could look at the cheats way out..
Does anyone know where the right hand side instrument cluster
leads end up? I would like to find the wire from pin 14 (blue),
if I can do it from under the dash..
Cheers, Art.

pin 14 if I'm not mistaken is the data line from the cluster to the trip computer (if any), so won't be any good to you.

Yeah, that's right, that data line carries the speed on it in serial format
so the calcs are already done. look here:
http://www.fordmods.com/forums/viewtopi ... light=baud

I just pulled my dash apart to tap the speedo out line which is supposed
to be data according to the docs, but it's just another pulsing line
even at 1000 baud it would be easy to catch the speed often enough for a very fast updating display, and while I've got that line, it might be
possible to do something else with it later on.
Art.

True, but if you've used the diagnostics mode, you'll see that the data is only refreshed every second, ok when converting to analogue dials, but seeing it all the time will get frustrating as it won't be smooth.

If I were you I'd stick to decoding the speedo pulse input, would allow better output.

I'll find out how often my Mum's Toyota Echo updates.
I was only going to do it every half a second the other way.

I made a tacho for a motorise dskateboard that I used to
have that updated every pulse as you were saying before,
but the results were undesireable, the last two digits couldn't
even be seen because they were changing so much..
Art.

that can be fixed with some programming (a hysteresis function), for example.

if before_rpm is +- 10 of now_rpm then don't_bother_updating

@rt, good to hear that your making progress.

So what data line are you using ? is it 'speed in' to the cluster or 'speed out' from the cluster to ecu, bem & cruise ? or are you using the trip comp data line ?

Cheers.

I pulled my dash apart and connected a lead to the speedo out line,
then put it all back together only to find out it's not a data line..
just another pulsing output the same as from the BEM.
So I'm going to pull it apart again and connect to the trip comp data line (14).

BTW, I've had a talk to dang, and it appears that the speed data is
available four times per second. Have another look at the digital speedo in the odometer screen.. it's updating at about this rate, not once per second.

I have proven that the pulses do vary in length, but strictly speaking,
it's not a PWM signal. A real PWM signal has a shorter off time as the on
time increases. You can bring the car to a complete stop with the speedo
out line either live, or off.
Art.

The signal from the speedo (the actual geabox sensor) is just a square wave with a varying frequency. The faster the speed the higher pulses per sec the shorter the pulse width.

Instead of counting pulses per second measure the pulsewidth of the signal. It's much faster and if you dont want it to jump around just keep averaging the input untill it is updating as fast as you want.

What sort of micro are you using. I did this with a picaxe (see vehicle datalogger) but with no hardware access to the timers it was frustrating. My new micro can count these pulses in the background.

I wouldn't bother trying to read and decode serial comms from the car. Too hard and model specific. Just use the direct signal from the gearbox and it will work the best.

On another note the Vehicle datalogger V.2.0 is coming along nicely. It's not mounted in the car yet but bench testing and developing of software is moving along. I have all the timers set up and the analogue measurements (8 in total) are working. Including calibration tables to read coolant and temp (or whatever you want) sensors.

Currently working on the digital inputs/outputs. Hope to be able to measure: pulsewidth, Km/hr, duty cycle, RPM, frequency, digial MAP. with a very high accuracy.

It also has the 4x20 line backlit display, 38400 serial comms, and eeprom storage of 5min of data. .

Still a prototype but have located suppliers for the circuit boards and 90% of components. Should be very useful for ppl upgrading to programmable management. eg. can datalog injector Pulsewidth against RPM and load to determine the factory fuel mapping.

Will keep ppl informed.

Sorry for thread hijack!