cam and crank trigger combo inside a distributor (megasquirt

[Paul H]

When I mentioned my device to an ECU engineer regarding trigger patterns and the twin sensor the first thing that came out of his mouth was the issues of the Honda sensor obviously they'd found problems during testing maybe Honda fixed this with better wiring and filters but the problem reappears when the engine is rewired with aftermarket ecu

[Piledriver]

The 36 tooth wheel I'm using in a 009 (rc car spur gear) and 3/8" od $7 home built Hall sensor assy. is giving me the cleanest signal I have ever had. The gear could easily be 2X the current diameter if I could source one off the shelf.

[Piledriver]

I think that this may just fit inside a 009 body, with a sensor attached to the OD of the case.
I suspect 72 tooth will be sufficient, and the price is right.
(equivalent to 36 teeth on the crank, w/32 pitch gears as I am running now)
...well, it almost fits, would work in a mallory.

http://www.ebay.com/itm/191005403107?_t ... EBIDX%3AIT

There is also a ~similar steel 60 tooth CNC machined wheel for a touch more, guaranteed to fit.

[Paul H]

So how do you get that working as a cam sync ?

[Piledriver]

Tig a spot of fill between 2 teeth for the missing tooth.
(I actually overdid it and made a 36-2 with a touch too much heat, but MS is fine with that too.)
For actual production you would have something made, but for a quick test I don't mind spending a few seconds with the welder and file.

At distributor speed the missing tooth IS the cam sync.

Works for sequential, only one sensor/input required, 24-1 with a vr is what Lance sells, and although I'm not a huge fan of vr sensors, the gear down missing tooth wheel works fine... tried it with a 16-1 using a Mallory Unilite with extra slots in the shutter, and with the 36-2 I fabbed up in a 009.

I discovered the sensor I was using (based in an Allegro `667) would not work well with a tooth ground out, but worked great with a tooth filled in. (missing vs. filled likely causes shift in the electronics automatic gain etc)

The sensor/filter caps and pullup resistor all slide inside short bit of 3/8" SS tubing, which gets filled with 5 minute epoxy.

[juki48]

amazon has some gears, not as cheap though
http://www.amazon.com/Boston-Gear-H3224 ... gear+steel
http://www.amazon.com/dp/B004N631VC/ref ... t_asn[url][/url]

[Steve Arndt]

I have some pictures to share of the Redline Weber 24-1 drop in trigger wheel setup that is in the works.
These pics show the unit mounted in a lathe chuck for testing up to 6000 RPMs. A billet top cover is in the works to finish off the package when used for triggering only. It is also possible to use a rotor and single coil. The fitted sensor is an OEM Honda unit.








Here is a link to a datalog from a Redline ECU that was used during testing which demonstrate a clean wheel pattern decode and RPM trace up to ~6000 RPMs. (CSV spreadsheet file open in excel, plain text):

https://docs.google.com/file/d/0B12q4Qf ... VdpcWg4YTg


I am now an authorized Redline supplier. PM me for details. I will be able to provide the product and support (ECU packages, sensors, wiring, IGNx coils)

[Piledriver]

Cool, congratulations!

[juki48]

how much? and how does the signal look at cranking speed?

[Redline Weber]

I also fit the LM-1815M as the VR amplifier.
At cranking RPM of (???) amount.
I "see" a clear sync with my ECU at 160-165 RPM.
This my be due to my ability of ONLY moving the Lath Spindle by force of hand.

Your question may be better asked by demanding output voltage measurement at a specific RPM.

Tooth timer frequency could affect low RPM sync, this case is that faster is NOT better.

ALSO, the immunity of EMI is good with our sensor though we KEEP the wire harness fitted with my Conductor/Cable/Drain "close" to the sensor, not as pictured for the Lath Test.

Lance

[juki48]

FYI, I'm pretty sure 5v borked the infrared led in the crane 700-0020 sensor. mine is dead after testing it out on the bench. I took it apart and there is nothing to limit current in there. I looked up replacement led's online and noticed there are none rated for more than 2 volts. also, the crane sensor will need a 5v pullup with 10mA current max. I think I'll make a little adapter with a voltage divider and 1k pullup resistor so I can power the led with ~1.6v and provide the 5v for the photo transistor to pull down. like this

I can easily put those 3 resistors in the harness connected to the sensor.
I also ordered some of these to play with at ~$4 each

[Paul H]

Tig a spot of fill between 2 teeth for the missing tooth.
(I actually overdid it and made a 36-2 with a touch too much heat, but MS is fine with that too.)
For actual production you would have something made, but for a quick test I don't mind spending a few seconds with the welder and file.

At distributor speed the missing tooth IS the cam sync.

Works for sequential, only one sensor/input required, 24-1 with a vr is what Lance sells, and although I'm not a huge fan of vr sensors, the gear down missing tooth wheel works fine... tried it with a 16-1 using a Mallory Unilite with extra slots in the shutter, and with the 36-2 I fabbed up in a 009.

I discovered the sensor I was using (based in an Allegro `667) would not work well with a tooth ground out, but worked great with a tooth filled in. (missing vs. filled likely causes shift in the electronics automatic gain etc)

The sensor/filter caps and pullup resistor all slide inside short bit of 3/8" SS tubing, which gets filled with 5 minute epoxy.

I understand the missing tooth for cam sync but the crank signal will be getting a missing tooth on rev one and then no missing tooth on rev two

[Piledriver]

Since 24-1 at cam speed was stock on very common Hondas, I would be surprised if any modern ECU didn't support at least that particular wheel.

It works well, MS 2-3 (maybe even ms1), Redline EMS support it (at least)
Megasquirt 2-extra/3 generic wheel code is very flexible as to tooth count.

My current setup is 36-2. My initial setup was 16-1 with a modded Mallory Unilite trigger wheel.
The next version will probably be a 60-1.

[Paul H]

I'm no expert on Hondas but I think they just had 24 tooth with separate cam sync equivalent to 12 on the crank.
If you put a missing tooth on the cam wheel it doesn't equate to 12 on the crank. Most the ECU's these days can be programmed with pretty much any pattern

[Piledriver]

Paul, I assure you it does work, and for fully sequential operation, one sensor, one wheel at cam speed.

Redline Weber has been selling that 24-1 with VR setup for awhile, for a variety of engines.

It wouldn't work well on many V6/V8s or any motor where the distributor gets driven via the oil pump via a long sequence of drive chains etc, but on an ACVW is all direct drive gears, unless you have slop in the distributor or a dorked up brass gear it seems very stable.

To try it, I made up a 16-1 using a modified mallory unilite rotor/shutter wheel setup, then later made a 36-1 in a 009 using a home brew sensor setup ($7 Allegro Hall sensor, home brew packaging)

It saves an input as well, (I/O is in short supply on MS2-extra compared to possible features)