Motogadget Motoscope Tiny speedometer with Motometer (BMW) board

[diesis]

I'm installing an aftermarket speedo from Motogadget (Motoscope Tiny) and using the stock BMW board to convert the factory hall sensor signal to a 6 pulse / revolution signal for the motoscope.  I've wired in the BMW board (motometer) as specified by a great thread here (http://www.motobrick.com/index.php/topic,6814.0.html).  The motoscope was configured for 6 pulses and a wheel circumference that was reasonable (just wanted to see the needle move).

I turned everything on and let it spin in 1st, nothing on the gauge.  Adjusted rpm & gear, still nothing on the gauge.

Generally I would assume this is my wiring ignorance; however, connectivity to the hall, and supply voltage to the motometer board check out with a multimeter as they should.

Hooked an oscilloscope up to the output from the motometer board and here is what is getting fed to the Motoscope gauge:



Red arrow shows ground, green is the peak voltage (which at ~2V/div is about 7.5V).  The square wave frequency varies with the engine and wheel rpm.  This seems to indicate to me that I've got the BMW board installed correctly (surprise) and that it is working (also a surprise given my track record). 

So what am I missing?  This signal is being fed to the motoscope tiny (orange wire) and I've verified it to that connection point. 

The motoscope tiny test for the speedo (connect speed sensor wire to ground and look for needle movement) is successful, so I assume the speedo isn't broken.

Is the BMW board providing too high of a voltage (Motoscope specifies 5V)?  Have I been inhaling gas and exhaust fumes for too long and I've lost any reasoning capability?  (All of the above can be correct).

Any help appreciated, as always, thanks for the amazing support.

[diesis]

Update -

I measured the signal from the Motogadget provided wheel sensor and it looks like a nice spike, very short in duration.    That isn't sufficient for design requirements, so I fed the Motogadget with a function generator.  Here is what it takes to get a consistent speed reading:

Vpeak-peak >= 10V  (Couldn't test the top end as the function generator didn't go above 10V)
Pulse width: 1-19% duty cycle

So, it looks like I need to convert the stock bmw motometer board from a 5-7Vpp, 50% duty cycle signal to something that is ~12Vpp, 5-10% duty cycle.

Any ideas?  A 5-12V level shifter might work well (seeing as I have a 12V source hanging about on the bike) to get half way there.  Are there any ICs that do duty cycle adjustment?  Otherwise I might be able to get an Arduino to do that portion.

[Laitch]

Will the information in this blog do you any good?

[diesis]

Thanks Laitch - I'm pretty sure that's rbm's project.  He ended up using the reed switch (the one provided by Motogadget).  Knowing rbm's EE ability, I should probably take that path as well (and likely will for the short term).  However, I'm still curious to figure this one out.

[rbm]

The BMW board uses the following schematic.

V2 is a 7.5V Zener which is setting your peak output voltage.  Possibly changing that component for a 10V zener might solve the problem, however I'm not sure if that will also change the threshold voltage for the second stage of the LM2904, which might change the pulse duty cycle.

Alternatively, you could build a current amplifier using an open-collector NPN transistor that pulls its output to ground. 

This circuit comes from the Brotbox by Maru Labs.  Apparently, it is supposed to work with most Motogadget gauges.  I think a MosFET might be even better because I found the Motogadget to be sensitive to the sender that gets connected to it.  Early on, I prototyped two boards using an NPN transistor and found my Motogadget Classic to show the wrong speed.  This was during my build and I got impatient to get the bike on the road, so I gave in and used the reed switch.  I'm playing now with a new design of my TGPI board to include a speedo amp based on the design above.

[diesis]

Allright - got this one figured out (it appears).  This isn't as nice of a solution as RBM's (which would be a single board), but works in conjunction with the motometer board salvaged from the bike.

I spoke with a colleague who is pretty quick with this stuff and he suggested a resettable one shot timer based on the 555 IC.   The circuit diagram is as follows: 



Vin (left side) is the output from the motometer board (the 50% duty cycle square wave).  Vout will be connected to the motogadget.  I added a Zener diode between the 12V supply (Vcc) and ground for a little bit of protection.

I put this together on a protoboard and am getting a readout on the speedo that is consistent (still need to do some calibration in the speedometer itself, as right now 25 is about 40 - I may just be in metric).  However, looks good.

That said, if RBM puts something cleaner together, I'll be buying it.  My solution is functional, but certainly a hack.

Cheer$,

Dave

[rbm]

I guess the Motogadget wants a clean square wave input and the 555 supplies that pulse form.  You might want to put a 12V zener circuit on Vcc along with a dropping resistor.  The Ne555 is rated for supply voltages between 5V -16V with absolute max voltage being 18V.  Sometime motorcycle and automotive electrical systems can output spikes that exceed 20V.  The zener will ensure that the max voltage to Vcc is 12V and that R1/C2 time constant is maintained since it is voltage sensitive.

The open collector circuit I presented in a previous post is a level shifting circuit.  The output of the circuit is connected to a pull-up resistor of 2.2K - 10K Ohms, one side of which is connected to +12V and the other side of which is connected to the output transistor.  The output signal will then swing between 0V and 12V (instead of 0V and 5V which is the logic level of the opamp, V-Reg).

[rbroen07]

How did this ever end up? Did you get the motoscope to work with the OEM sensor?


Rasmus

[diesis]

I've done a poor job checking in.  The Motogadget Motoscope Tiny works great using the circuit described a few years back and the OEM wheel sensor.  Somehow the circuit has survived a few years (all be it with minor riding) in a really janky PVC tube stuffed up underneath the tank.