MXL Pista speed input sensor/speedo?

[deathbyspoon]

Hey guys -
I have a Mxl Pista that's going in my '83 930 and am trying to figure out what to do about the speedometer. I'm using my factory speedo with the trans to keep track of miles on the car.

Is anybody on here using a magnetic speed sensor? If so, can you please post links for the pulse generator, sensor, and harness?

If not, what are you using for your speed input?

Thanks!

[Walt Fricke]

I wouldn't know a Mxi Pista from a pizza.

But why can't you use the regular speedometer signal to feed your new system also?

Your transmission has a disk with eight bar magnets on it, arranged radially. It is attached to the diff, and positioned so that only a thin piece of the case is between it and the stock speedo sensor.

The stock sensor is nothing but a reed switch. Magnet passes by, it closes. Magnet moves on, it opens (or it is the other way, doesn't really matter). Eight pulses per wheel revolution.

When the reed switch is closed, it grounds the sensor wire from the speedometer. For some reason, Porsche runs a pair of wires all the way from the speedo to the sensor, but one of them at the speedo is just grounded there.

So your speedo picks up the pulse the grounding creates, and uses it. My understanding is that the speedos of that era then convert the digital (really, binary) pulse into a voltage, which varies with pulse rate. That varied voltage then drives a motor to turn the odometer over. And that voltage also tries to rotate the speedo needle against the spiral spring. More pulses, higher voltage, odo motor turns faster, speedo rotates higher.

I think I have seen what the voltages involved are, but can't remember what they are. No more than 12, anyway. And I bet the input impedence of the D to A converter in the speedo is very high, which I think should mean you could just tie the speedo signal wire and ground to your gizmo's signal input and signal ground. It, too, probably has a high impedence, and in any case it is not the input voltage things are using, it is the frequency. And a frequency generator as simple as a reed switch isn't going to be affected by a little more voltage draw.

In fact, the factory manual simple test for speedo function versus bad sender is to attach a wire to the speedo signal terminal, and manually ground and unground that wire rapidly. The needle should rise some when you do this. Hardly precise, but you can trace wire faults this way from the connector inside the tunnel by the shift coupler. And if the speedo seems to work in the test, but not when you drive the car or rotate the right rear wheel, then the sensor is probably the culprit. But, in any case, nothing you do is apt to mess up your speedo.

If you needed something more isolated to get the right voltage into your gizmo, you could easily rig up an optoisolator or the like, so the frequency would come out the same, but not over whatever the specs for the pizza might be.

[ONQRACIng]

We just added an MXL to our 911 Club Racer. We are adding an "L" bracket off the trans near one of the CVs. The magnet will be added to a small bracket off one CV bolt, and the sensor will be mounted of the "L" on the trans. Harness will be lengthened via a Pegasus supplied extension. Pretty straight forward.. Ibet one could goggle an image of such brackets..

Andrew

[deathbyspoon]

Originally Posted by Walt Fricke

I wouldn't know a Mxi Pista from a pizza.

But why can't you use the regular speedometer signal to feed your new system also?

Well, here's the Pista. If i had a choice, I'd rather eat a pizza.



Anyhow, here's the specs on the sensor:
Supply voltage
4.5-24 V DC
Supply current
10 mA (typ.) 20 mA (max)
Output signal type
Pulse 0-50 Volts
Maximum output current
20 mA
Maximum operating frequency
100 kHz

That being said, can i split the signal from the trans to run both the factory speedo and this other speed input?

ONQRACIng - You're sort of inline with what i had in mind. Something like this:



From pacific customs. Do you have a link to the sensor/pickup you purchased?

[ONQRACIng]

My Pista came with all the need base sensors. Temp and speed were in the box.

Same idea, minus the chopper wheel on the shaft, that is like a crank sensor. Ours will run only one magnet, and the speed setup will do the math from wheel Diameter, to when it see the magnet.

here is where I saw the sensor we have.


https://www.pegasusautoracing.com/bigpicture.asp?RecID=5456
Andrew

[Quinlan]

The way ONQRacing describes is exactly how I have mine mounted. On the inboard cv joint on the driver's side.

[deathbyspoon]

Great info. Is just what i imagined, though i was overthinking that i had to use something like the pulse wheel pictured.

Just so im clear: 1 magnet, 1 "pulse" per revolution, and use the software included to calibrate the reading based on the tire diameter.

Got it!

[ONQRACIng]

Roger that.

Pretty easy stuff.

[Matt Romanowski]

Go GPS for the actual speed. It will give you real speed, better track maps, and get rid of a beacon. If you want to use a wheel for speed, it should be a front, not a rear so you don't get spin/slip involved.

[ONQRACIng]

Matt is absolutely correct.. GPS is the way to roll, if you have that add on in the budget..

[Matt Romanowski]

Talk to Jerry Austin (JANDSAUSTIN1@VERIZON.NET) and tell him I sent you for a GPS reciever. He'll take care of you. The budget isn't too big and it's worth it's weight in gold. You'll be able to do far more with the GPS map than just the track map.

[TimT]

I welded a little tab to the cv keeper and mounted a magnet, and made a bracket that mounts on the diff cover for the sensor, 1 pulse/wheel rev. works well enough for speed...somewhere I have a pick from underneath the car showing the set up installed, cant locate it now though

You need to know the rolling diameter, not the overall diameter of the wheel for better accuracy.

Keeper with magnet:



You don't need to use the cage type pulse ring. You can mount a sensor and orient it such that the heads of the cv bolts are triggers (in some case 6 pulse/rev). If you have 4 bolt cvs you need to come up with a different solution

GPS + wheel speeds can provide some useful info regarding slip etc.. not sure if you need all 4 wheels to have speed input though

[Matt Romanowski]

I figured I would give you an idea of what you miss out on with only having the wheel speed. With just a wheel speed, you can build a track map like this and it will allow you to get sector times.


When you add the GPS, it will allow you to "overlay" any information you want on top of the map, with colors to delineate what you are looking at. The first one is longitudinal G force indicating braking forces leading up to a corner.


This one shows speeds around the track


This one shows lateral G forces. Their are a few other items I find it easier to look at the map and see rather than just in a graph.


So, you can see that the GPS adds a lot to your analysis. On a regular day, it also elimanates needing a beacon and makes your day much easier. It will also allow you to compare line differences from lap to lap.

[Mahler9th]

Is you application for a street car or a race car?

If you need more AIM stuff and/or help, I suggest Dave and Ellen at Veracity Racing Data. They are friends. They used to race a Porsche in our area (they lived in Sunnyvale if I recall correctly) and have quite a few Porsche customers.

They know their stuff for sure.

[Walt Fricke]

I am seeing two kinds of sensors. One would be a coil, and when a magnet passes by it creates a voltage. The other (Hall effect?) is a magnet with a coil around it, which creates a field. When a metal object passes by (like a CV bolt head), the field is altered, and again a voltage is created or at least altered. Both produce a pulse, but one needs a moving magnet, and the other does not.

The Hall effect setup (if I have that right) is perhaps more flexible. It is what triggers my Electromotive EFI and spark setup from a toothed wheel. I think the sensors on a 3.2 flywheel are the same.

Some applications requaire, or at least work better, with a higher frequency, which gives finer resolution. Tires rovolve about 800 times per mile, which is a very low frequency when looked at in terms of cycles per second, even at high speeds. Which is why you see things like the add-on for the CV shown. Or think of what you see on ABS axle hubs.

If you are reading CV bolts (seems a convenient thing to do to me, and you get six per revolution), you can also easily make a bracket which will allow your sensor to read the outer CV bolts. Easy to drill and tap two small holes into the aluminum banana arm to hold it in a suitable place so the sensor can be adjusted to let the bolt heads pass by at the needed distance.

This kind of thing is what has had me thinking of just using the stock Porsche speedo signal. I think the makers of these nifty electronic dashes and such are not making them specifically for Porsches, so that would not particularly occur to them. The VDO speedometer is not set up to use either a Hall or a magnetic sensor, but it would be interesting to know if it would work with any of the devices which are set up that way. I'm pretty sure that using it for dual duty would not interfere with the speedometer function, but that's not based on experience or even a strong electronics background. It would simplify things, though, as no brackets, no new wires, and 8 pulses per revolution to boot.

On a much newer car, perhaps one could pick up something from an ABS sensor, again assuming all these circuits are high input impedence - like your digital voltmeter, which does not drag down the voltage it measures thanks to solid state circuitry these days. Lots has changed from when a VTVM was the hot ticket.

[Matt Romanowski]

Originally Posted by Walt Fricke

Some applications requaire, or at least work better, with a higher frequency, which gives finer resolution. Tires rovolve about 800 times per mile, which is a very low frequency when looked at in terms of cycles per second, even at high speeds. Which is why you see things like the add-on for the CV shown. Or think of what you see on ABS axle hubs.

If you are reading CV bolts (seems a convenient thing to do to me, and you get six per revolution), you can also easily make a bracket which will allow your sensor to read the outer CV bolts. Easy to drill and tap two small holes into the aluminum banana arm to hold it in a suitable place so the sensor can be adjusted to let the bolt heads pass by at the needed distance.

A lot of ABS reluctor wheels are 30+ teeth, so the more the better as far as teeth go. Most people say 6 is really a minimum. You also want to measure it around 20 hz to get the best data.

[Walt Fricke]

I would think that for wheel speed or some similar low RPS rotating part, a design could use period rather than frequency to increase resolution, if having a lot of teeth presented problems.

[TimT]

I just got out my laptop with Aim Racestudio config software, and apparently you could use the speedo sensor from a 930 box..

In the config for for speed, input the rolling dia of the tire, and then 8 pulse/rev.

I'm still driving cars from the stone-ages with the 901 tranny