Autosport Labs

Mounting a trigger wheel, in some circumstances, may not be an option. Therefore Megajolt is loosing potential sales. Future versions of MegaJolt (Megajolt2) may incorporate different trigger options. Searching on the Internet it seems possible to emulate a 36-1 trigger wheel electronically using a PC's parallel port or a handful of components and a microprocessor programmed to trigger an EDIS module via a squarewave.

Researching lead me to a few options to emulate the trigger wheel. Such as rotating magnets paired with a hall effect IC, 360 slot opto trigger disks as used in the Nissian distributors and modify the slots to 36-1. As well as all the ingenius trigger options members of this forum have produced.

One option, that caught my attention, was the possible use of a simple 2 pole magnet and a 360 degree magnetic rotary encoder IC such as AS5145. Specs say resolution is 0.0879 degrees and good to 30,000rpm.( See http://www.austriamicrosystems.com/eng/ ... ers/AS5145 ) The output from the encoder IC would then feed into a microprocessor to produce the required 36-1 squarewave output. It could be mounted on the crankshaft or distributor shaft and the microprocessor would adjust to the different situation via a switch or dip switches. This could be in kit form for the electronics and maybe use the new Autosport Labs CPU Module for the microprocessor. As with a normal trigger wheel mounting, the mounting of the magnet and encoder IC would be a challenge. However, the size of these components are very small which may make mounting easier. Taking the idea further, the separate microprocessor could be dropped and its function done by the microprocessor in Megajolt if that was possible.

A feature of the AS5145 is that it can tell the angular position of the magnet, hence the crankshaft, even when stationary. This could be the basis of an totally different ignition system reducing the programming required in the microprocessor.

Hello,

Thank you for the reference to the AS5145. It's a fascinating device! However, it seems to be best suited for low speed positioning applications, so it doesn't seem appropriate to manage timing requirements for high speed crank position applications.

From their datasheet:

The device is ideal for industrial applications like
contactless rotary position sensing and robotics;
automotive applications like steering wheel position
sensing, transmission gearbox encoder, head light
position control, torque sensing, valve position sensing
and replacement of high end potentiometers.

While it may be able to *track* at high speeds, the data interface from the chip is much slower- it offers a serial interface as well as a low frequency PWM output.

I think this might make a great contact-less TPS sensor or steering angle sensor, something good for RaceCapture.

Back to ignition systems: Future Megajolt X hardware will not be restricted to 36-1 wheels nor the Ford EDIS module. Megajolt2 will initially support 36-1 trigger wheels, but have the flexibility in software to handle a variety of trigger wheel configurations- even distributor mounted encoders as Dean924s, alexander and others have convincingly demonstrated.

Megajolt/D will handle distributor based ignition systems, and accommodate VR, hall-effect and optical sensors.

Ultimately we will strive to balance ease of use with flexibility- I think the future will capture 90% of what people will need.

Hope this helps!

Thanks for response! I am glad I did not waste any effort trying to get this to work!

Did a little more research- found something more appropriate:

http://www.austriamicrosystems.com/eng/ ... ers/AS5040

This particular unit has a quadrature output mode, meaning that it outputs a pulse every 0.35 degrees- the kind of signal output needed for an ignition application.

A really cool device- I'm quite tickled by it.

The next question is how to mechanically adapt it to the engine. On a crankshaft, a magnet might be glued to the crankshaft bolt, and the sensor would have to be suspended over it.

A distributor would be the better application- the magnet would be attached to the top of the distributor shaft, and then the sensor could be suspended above it.

You would gain the contact-less benefits of this setup; the drawbacks are that every distributor installation would be custom. So- while it's appears to be technically superior, the 'old-style' distributor with a VR sensor and a 4-pole encoder are the easiest way to get 'acceptable' performance. Still - 1024 poles in the magnetic encoder vs. 4 poles in the traditional setup is compelling!

Not sure what kind of sensing we'll end up using this for, but I'm sure we'll think of something.

Thanks again!

Has anyone tried to use a software sound synthesizer to playback a "sample" of a 36-1 EDIS wave from the sound card output of a computer for EDIS testing ?

I was thinking something like this http://www.audiomulch.com/info.htm which can be downloaded for free and used without restriction for 90 days.. Even after that, it will still work and playback, you just cant save files and so on which wouldnt affect this use of it.

It would be a simple matter to record an EDIS wave, load it into a sample-loop-play module with variable-speed playback and twiddle a knob on the screen to vary the playback speed and thus the "RPM".

This would save having to reboot from floppies in dos mode, which the other PC simulation of a 36-1 sigal I saw required.

The only drawback I can think of is having a wire connected to your computers sound card in close proximity to 20Kv sparks


Personally, I just got a little metal disc, chopped slots in it with an angle grinder and spun that in a drill.. works great and much safer for you computer, but was just a thought

http://www.heggs.co.uk/vrssim/ would seem a possible solution

brentp wrote:...

Megajolt/D will handle distributor based ignition systems, and accommodate VR, hall-effect and optical sensors.

Ultimately we will strive to balance ease of use with flexibility- I think the future will capture 90% of what people will need.

Hope this helps!

looking forward to it. woohoo!

What's the specification for the VR signal? I've started playing with PIC processors (again) and a VR emulator sounds like an ideal application.

I have built circuit to emulate the trigger wheel. However it was intended to allow bench operation of a system; a variable osc sets the equivalent rpm while a hand vacuum pump feeds the MAP input. It just uses several common CMOS counter & decoder chips, seems to drive the VR input OK. The EDIS cuts off at 8000 rpm; don't know if its the quality of my waveform, or an EDIS limit.

THE VR input seems to be pretty high impedance, would like a nominal average voltage level of about 1.54V. Sine wave is good, a somewhat filtered square wave seems to be OK. At the missing tooth, the blue lead goes from negative peak to a center voltage, holds there for a cycle, then resumes going toward the positive peak. The grey wire is inverted.

Bruce Roe

I am looking into alternative options for sensing crank position for an engine where a regular 36-1 trigger wheel is not an option. What does the MJ look for in the VR signal? Is it amplitude? Zero-crossing? Edge? Appreciate a hint in the right direction.

- Rob

The amplitude of the signal will vary greatly with rpm. These things pretty much focus on the crossing points. Get a scope & check it out. The MJ is hard wired to receive a 36-1 signal, what is your approach? Bruce Roe

Thanks Bruce.

My approach is either an aluminum wheel with 36-1 slots or a white drum with 36-1 black stripes, and an reflective optical sensor. That will give me a TTL block signal that looks a like what a VR produces. If a zero crossing is required then a series capacitor can convert the TTL signal into pulses going positive and negative on each edge.

- Rob

The MJ expects a differential input; you might try transformer coupling. I suspect the amplitude needs to increase for higher rpm operation. Magnetic coupling is common, because it doesn't suffer so much from dirt. good luck, Bruce Roe

The MAX9924 we are using in Megajolt/C apparently works with TTL signals as well as variable amplitude VRS signal. Impressive device!

On that note, we're working on a small breakout board for the MAX9924 to interface VR sensors to TTL. Thought it would be a useful board for experimenters.

Actually, to be pedantic, the Megajolt box itself knows nothing about the 36-1 wheel at all..

Its the EDIS module that wants the 36-1 Zero crossing signal and it converts it into the PIP signal which is what the MJ box listens to.

There are a number of 36-1 wheel simulator widgets available on the net.. Often called "Stimulators" and used for Megasquirt Fuel Injection Systems. But they work equally as well with the Megajolt, since both can talk to EDIS modules.

The one I use is the "JimStim" - http://www.jbperf.com/JimStim/ - It has a number of analog knobs on it that can be cross wired to whatever input for MAP/Temp/Aux input simulation as well

Very useful for bench testing

- and yes, the EDIS modules are where the 7000-8000 Rpm Rev limit comes from. I tested a few different EDIS modules (different colored stickers) on the bench and found they stopped firing at different points with nothing else changed in the setup.. I forget the details, but its in one of my previous posts if you look them up. I think it was 7400 and 8200 Rpm from a Green and Orange Label EDIS..