cam and crank trigger combo inside a distributor (megasquirt

[Paul H]

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 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 commercial sensor, home brew packaging)

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

I'm sure it does but I just wanted you to explain the math behind the decode

[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 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 commercial sensor, home brew packaging)

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

I'm sure it does but I just wanted you to explain the math behind the decode

Sadly I suck at math, or I would have become an EE...
...for example, I still consider "management" to be a 4 letter word.

If you are really curious how it works in detail, the MS2-extra and older versions (1 rev back) of the ms3 code are downloadable for all to see... std vanilla C code.

[Paul H]

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 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 commercial sensor, home brew packaging)

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

I'm sure it does but I just wanted you to explain the math behind the decode

Sadly I suck at math, or I would have become an EE...
...for example, I still consider "management" to be a 4 letter word.

If you are really curious how it works in detail, the MS2-extra and older versions (1 rev back) of the ms3 code are downloadable for all to see... std vanilla C code.

I'll give the download a miss and I'll go and hang some sky hooks and mix up some tartan paint

[Redline Weber]

MATHWORKS : Another SIMPLE statement, a RULE, is what even divider can be used for a Crankshaft Target.
THINK : A "gear down" TW = 2X of the crankshaft tooth count.

Thus we ALSO make, sold for many years, a CAS TW of 48-2 (24 crankshaft) count.
The MATH states fifteen degrees for a V-8 engine each tooth.

NOW when I READ about a 36-1/2 "gear down" I am amused GREATLY. (18@crank)
MATH=4 divided by 18 is NOT even close to being an even count for a VW-Type.

The next GREAT ERROR is about cam belt, oil drive, valve train harmonics deflection as being intense.
TRUE, the crankshaft is still the best location, in the center is BEST for TW location.
THOUGH the truth when observer, not GUESSED ABOUT, IS that the "high res" CAS will AVERAGE its general position filtering out these vibrations.

The Gear Down single engine position encoder is a viable method when understood.

Lance

[Paul H]

Very good Lance but do you realise that a lot of people getting into efi don't know what a lambda sensor is or the difference between a Hall and VR sensor

[Piledriver]

While that (24-1 or 48-2 gear down as a requirement) may be a rule for your system, 36-1 (or -2) gear down (and even a 16-1 gear down) works splendidly on MS2-extra and MS3 on a /4.
..As Barbie once said, "math is hard" but not for a decent ECU.

The next setup I plan to try for comparison is 60-1, as the 72 tooth 32 pitch gear "wheel" won't quite fit in a 009.
(and only to better test for the theoretical errors I have only seen to date using a ~minimum resolution wheel setup, and even then ~only noticeable near cranking speed)

(Reiterating) On MS2-extra or MS3, there is no requirement for tooth count to be divisible by the number of cylinders, beyond having a minimum valid tooth count,,, on a 4 cylinder the minimum is 4 teeth (4-1 on crank, 8-1 at cam speed is valid)
More teeth provides better resolution. Result must be an even # when divided by two.

I will not disagree that a cam speed wheel works well on an ACVW.
I will say I'm not a big fan of VR sensors, but most problems folks have with them are malpractice on the installers part.
Hall sensors tend to be much more forgiving.

[Redline Weber]

The "rule" is NOT a "my system" rule, this is Binary Rule.

The processor will NEED TO DO MUCH LESS WORK when the count is even AND does NOT require "endless" computation as a Repeating Decimal could require.

Thus processor "overhead" is REDUCED, a "touch of class" in the code.

Lance

[Piledriver]

Understood, but rounding was invented long ago, and the incurred error and overhead is minimal if implemented properly.
For any reasonably modern ECU it should not really be an issue.

[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.

Just an update... As of today this is running with a new LED. I ended up not using the voltage divider and just put a 330 ohm resister in front of the black wire, power to the led, and used a 1k resistor to pull up 5v to the gray wire. nice clean signal ~0-5v. won't be able to drive till the salt is off the road but it idles nice in the garage

[juki48]

I got my little optical sensors from china today. these are really neat and SUPER cheap, ~$3 shipped. seems like a very good option for an optical trigger. they should be able to handle very high tooth counts too, higher than anyone will want to make unless you have access to a laser cutter. the led would also be helpful for initial setup and troubleshooting. in this photo I am running it off 5v and it outputs 3.56. wish I had found these before I spent $30 on a crane cams sensor.

[Devastator]

Neat!

[Redline Weber]

OK Juki, A GOOD LESSON taught from NASCAR before my 60-2 TW was fitted to the GN engine.

The under-hood temps reached over 300F melting the magnetic flux inside the REQUIRED MSD Nascar approved distributor.
The "fix" was to go OPTICAL.
THIS method also allowed for NO dirt on the sensor.

Method : Your optical board could be located remotely, away from the engine.
Then fit a light pipe to communicate the required signal from the shutter
wheel to your sensor and light source.

Worth a try ?

Lance

[juki48]

melting the magnetic flux

Are you saying they were using magnets that were degrading from temperature? if so, why not use higher temp rated magnets? if your not saying that, I don't know WTF you are talking about... I design solenoids that work at 220°C and my flux doesn't melt

[Piledriver]

I think he said they used fiber optics to get the electronics away from engine heat.
(Good plan)
Some of the engineering plastic fiber will take quite a bit of heat, and glass multimode will take ... whatever abuse glass fiber and kevlar will tolerate.

Probably not an issue for most applications.

[Paul H]

So does it all work then or do we have a barrel of broken biscuits ( English for nonsense )