MegaJolt and MPH

spetom · Post by **spetom** » Thu Oct 16, 2008 7:43 pm

Theoretically, could the MegaJolt calculate the speed of the car?
Using data from the sensor, and a user-defined value for the size of the wheel, could a fairly accurate speedometer be created?

(Of course, I'm ignoring the fact that there isn't really a way to output this, but a speed would be interesting to have when you're data logging a journey, alongside RPM, etc...)

Or is data from the sensor not directly accessible by the MegaJolt, since it is intercepted by the EDIS? Would this instead be a job for MegaJolt 2?

Alexander_Monday · Post by **Alexander_Monday** » Thu Oct 16, 2008 7:55 pm

The MJ would not know what gear you are in to be able to do that.

spetom · Post by **spetom** » Thu Oct 16, 2008 8:09 pm

Hmm... good point.

Well, it was a nice dream while it lasted

nickj · Post by **nickj** » Sat Oct 18, 2008 1:25 pm

No need to worry about what gear it's in, just log the rpm of the driveshaft. Together with the distance
travelled per driveshaft revolution, you have a loggable speed of vehicle.

Nick

Captain Midnight · Post by **Captain Midnight** » Thu Dec 04, 2008 5:40 am

If I used a sensor on my motorcycle wheel, how accurate could it be?

Post by **brentp** » Thu Dec 04, 2008 8:48 am

Likely not very. The polling interval in the configuration software for runtime data is about 200ms; at higher speeds signals from the wheel sensor may be missed, which would throw off the speed measurements.

Captain Midnight · Post by **Captain Midnight** » Fri Dec 05, 2008 5:33 am

What could be done to increase the accuracy? I am running a flat track motorcycle and data logging RPM and MPH would be very helpfull.

Dean924s · Post by **Dean924s** » Fri Dec 05, 2008 3:37 pm

Captain Midnight wrote:What could be done to increase the accuracy? I am running a flat track motorcycle and data logging RPM and MPH would be very helpfull.

Some sort of hall sensor that was reading the chain gear (I assume that you are chain driven) The hall sensor signal would then have to be processed in to a a 0-5 volt signal. This could then be interpolated into speed. You would have to calibrate it some how initially but I don't see this is all that hard a problem to solve. The hallefect sensor is the easy part. The unit from a 2000 Ford expedition dif would work. The hard part / unknown is the hardware to convert the signal form the hall sensor to a 0-5 volt output that the MJ unit could read. I am sure it is out there you would just have to dig around for it. The resolution of the MJ unit is the only question here but it would be fun to play with.

Post by **brentp** » Fri Dec 05, 2008 6:54 pm

Yes, An external frequency to voltage converter would work; it could be calibrated in the firmware to show actual MPH. Good idea!

The solution with fewer external parts would be to log a pulse per wheel resolution, but would require calculations after data logging to convert to MPH.

For the motorcycle wheel, what will be the RPM of the wheel at the maximum speed you will see?

Dean924s · Post by **Dean924s** » Fri Dec 05, 2008 10:27 pm

This looks interesting

http://www.api-usa.com/pdfs/api7580g.pdf

Alexander_Monday · Post by **Alexander_Monday** » Sat Dec 06, 2008 6:57 pm

If you are an electronics experimenter and can adapt and build from datasheets, I have used the TC9400 for this type of application before. It is a very stable and accurate frequency to voltage or voltage to frequency chip and is cheap. Make sure you filter the output good, because is does have a sawtooth ripple superimposed on the DC output. Also, implement a simple power on reset with a cap from pin 11 to VDD, or after you have built the circuit you can pull your hair out trying to figure out why every once and a while at random it decides not to work.

A 555 can also be configured for frequency to voltage, but is not as easy to implement unless you have a square wave input to start with.

You could also make a simple hi-pass or low-pass filter with filtered output with a couple of op-amps, however the TC9400 would be simpler to implement.

With any of these methods, you will of course have to calibrate the AUX input to give you a reading that will correspond to the speed you are going, make the mod for 0 to 5V input by disabling R6 as shown in the vehicle installation guide, and I would highly recommend that you parallel a 5V or less ( I have a 4.7V ) zener diode with the AUX input and ground, and a 1k or so resistor in series before the zener to prevent an external circuit over voltage from wiping out the processor in the MJ. Brent, you might want to do this in future versions on user inputs?

Unfortunately, I am very busy at work and home, so I do not have time to play with a circuit to give you the specific schematic and parts values you would need.

http://www.microchip.com/wwwproducts/De ... e=en010486

Post by **brentp** » Sun Dec 07, 2008 12:49 am

Thanks for the tip- I'll take a look at that datasheet.

For V4, The Aux and TPS inputs on the MJLJ do have input protection- but we avoided using zener diodes as they typically begin conducting near the zener voltage, affecting the top end of the ADC range.

Alexander_Monday · Post by **Alexander_Monday** » Mon Dec 08, 2008 12:41 pm

Yes, a crowbar of some sort would be better than a zener, but the zener is quick and easy protection, and I am interfacing with a thermocouple multiplexer I built that in normal operating range should never peak above 4V, and I had a 4.7V zener laying around, so---