36-1 wheel decode problem

All Megasquirt 1 support questions. See also the Documentation

Moderators: jsmcortina, muythaibxr

Post Reply
milosch01
Helpful MS/Extra'er
Posts: 90
Joined: Sun Aug 28, 2005 1:20 pm
Contact:

36-1 wheel decode problem

Post by milosch01 »

I am running 029i on v2.2 hardware. Using a 36-1 wheel and the Honeywell GT1 sensor. Suspecting my umbilical to the relay board, I pulled them and set everything up on the bench. I verified that I am getting some kind of trigger from the GT1, readings from CLT and TPS, voltages are there, etc.

The GT1 test was pretty crude, using a screwdriver passing over the sensor rapidly.

Back on the car, I get a pretty steady though incorrect RPM reading if in MSNS mode at idle of about 300. The car is running it's own ignition and fuel, currently. Switching back to generic wheel with MSNS off I get no rpm reading at all.

I hope to have a scope soon but for now all I can say is that I am relatively sure that the sensor is ok. The mount is steady and the wheel is centered with a 1-2mm gap from the edge to the sensor.

It's a straight 6, and I have to admit I have no clue what to set for the wheel trigger points using 3 spark outputs. I have tried variations on 3 trigger/return pairs, e.g. 1-6, 13-16, 25-28.
jsmcortina
Site Admin
Posts: 39585
Joined: Mon May 03, 2004 1:34 am
Location: Birmingham, UK
Contact:

Re: 36-1 wheel decode problem

Post by jsmcortina »

milosch01 wrote: Back on the car, I get a pretty steady though incorrect RPM reading if in MSNS mode at idle of about 300. The car is running it's own ignition and fuel, currently. Switching back to generic wheel with MSNS off I get no rpm reading at all.
Something isn't reading right - 300rpm in MSnS mode means the sensor is giving pulses equivalent to 300/12 i.e. 25rpm which is not right at all.

You should be seeing say 9600rpm in MSnS mode (assumes 800rpm idle)

If your multimeter does frequency, probe the sensor output and see what it says 800rpm = 480Hz on 36-1

James
I can repair or upgrade Megasquirts in UK. http://www.jamesmurrayengineering.co.uk

My Success story: http://www.msextra.com/forums/viewtopic ... 04&t=34277
MSEXTRA documentation at: http://www.msextra.com/doc/index.html
New users, please read the "Forum Help Page".
milosch01
Helpful MS/Extra'er
Posts: 90
Joined: Sun Aug 28, 2005 1:20 pm
Contact:

Post by milosch01 »

I had the input circuit wired for points, which I used successfully with a pertronix ignitor II. I altered the input circuit as recommended but am still getting similar readings. Since the pertronix popped, it could be that my U4 is bad. Anyway, I removed D5, R10, and D8. Installed a 1.5k resistor in place of D8. Ran a jumper from XG1 to spot where D5 anode was. Ran another jumper from the bottom of D8 (now 1.5K) to 5V. I guess I still need an external pullup resistor and tried another 1K wired from TACH to 12V. With or without that, still nothing but the same 300rpm reading. The alternative to schematic 2 for 5V sensor was what I used. I have tried also wiring the GT1 to 5v VREF pin as well for its supply.
milosch01
Helpful MS/Extra'er
Posts: 90
Joined: Sun Aug 28, 2005 1:20 pm
Contact:

30-2

Post by milosch01 »

I had a little better success using my 60-2 on the camshaft. Readings were relatively stable and around 10000 rpm reading. My thought is to remove 2 more teeth opposite the two already missing and set it up as a 30-2 wheel. I still won't know accurate cam position for TDC, but was thinking this would be the same as a 30-2 on the crank. All I can say is that trigger wheel and sensor mounting have far exceeded any other difficulties I have had with MSNSe so far ;)
milosch01
Helpful MS/Extra'er
Posts: 90
Joined: Sun Aug 28, 2005 1:20 pm
Contact:

Post by milosch01 »

Works! Will drive it around today to make sure the mount is stable before I try to use it. But, I finally have a stable RPM reading with the Allegro sensor. It's glued onto a piece of aluminum with some JB Kwik. I know that's not ideal, but at least it is exposed and not cooped up under a heat shield or cover. The 60-2 is hacked to be a 30-2 wheel and MSNSe seems to be ok with that using triggers of 3-8,13-18,23-28.
Image
milosch01
Helpful MS/Extra'er
Posts: 90
Joined: Sun Aug 28, 2005 1:20 pm
Contact:

30-2 wheel problem during cranking

Post by milosch01 »

I have verified that I am getting an excellent square wave from the sensor and also at the IRQ signal to the processor. I am using 029q on a v2.2 board with an allegro hall sensor and a virtual 30-2 wheel (electromotive 60-2 with opposite teeth cut out) and 3 coils ABC from 3 VB921s. However, 9 times out of 10 it won't sync.

I wonder now if there is something not right with the wheel itself. Maybe I need to cleanup the missing tooth sections, which were intended to be ground down to the level of the valley of the remaining teeth?

My triggers are setup currently as 17-22, 27-2, and 7-12. Is it ok to cross over the gap between trigger/return? I would assume so as long as the ECU knows where the wheel is (synced).
milosch01
Helpful MS/Extra'er
Posts: 90
Joined: Sun Aug 28, 2005 1:20 pm
Contact:

Bad plots

Post by milosch01 »

I did a couple of plots from a spreadsheet and rarely does it spot the missing teeth. Maybe it was a bad idea to try to do this on the cam, since I am basically showing it two different wheels in a row. This must be confusing the decoder. Either that, or my aluminum mount is becoming a speedometer with the cam sprocket face spinning about a 1/4" away, or the same thing is affecting the sensor magnets. I wonder if I would have any better luck using the GT1 pointed at the wheel radially?

http://www.groupwhere.org/megasquirt

What's the most frustrating is that this worked with the sensor and wheel on the frontside of the cam with the exact same spacing to the sprocket face, etc.

FWIW, the sensor is an Allegro 665 with the input circuit modified for 5V operation. I will hopefully have some 625 and 674 samples from them to try, the latter being specifically designed for automotive cam sensors according to their docs.
milosch01
Helpful MS/Extra'er
Posts: 90
Joined: Sun Aug 28, 2005 1:20 pm
Contact:

Post by milosch01 »

Come on, guys ;) I know it's a weird setup, but maybe it's a problem with the input circuit cap or something. Like I said I am getting a very nice square wave out of the allegro. Any ideas?
Keithg
Super MS/Extra'er
Posts: 2413
Joined: Sun Mar 06, 2005 9:15 am
Location: Chicago, IL, USA
Contact:

Post by Keithg »

I have been watching, but do not know what to say. The only thing I can comment on is that I do not think you can use tooth 2 on a 30-2 wheel. The missing teeth are 1 and 2, so the first tooth is 3. I used a 665 sensor and got a great sync from a 60-2 electromotive wheel on the crank. I did not run the car on it, but it was idling on its points ignition and I got a great sync in MT that matched engine rpm. The electromotive teeth are too small for a GT1 sensor, but the Allegro seems to sense them just fine.

(edit) your 36-1 wheel looks like ity would work with that GT1 as well. The problem with the elctromotive was that the teeth are too shallow.

KeithG
milosch01
Helpful MS/Extra'er
Posts: 90
Joined: Sun Aug 28, 2005 1:20 pm
Contact:

Post by milosch01 »

Man, not being able to use tooth #2 is huge point ;)
(edit) OK, so I should probably use 18-23,28-3,8-13. Will let you know. Might even need to bump that another tooth as well, but I think I tried all of these in the process.
Keithg
Super MS/Extra'er
Posts: 2413
Joined: Sun Mar 06, 2005 9:15 am
Location: Chicago, IL, USA
Contact:

Post by Keithg »

Well, you have made your cam wheel to have available teeth 3 through 30, you cannot use teeth 1 and 2. as far as I understand the wheel decoder stuff.

Personally, I would think you would be better off with a crank wheel, though. 36-1 is well documented. If that GT1 cannot read those big teeth you show on your web page, the 665 surely can. IMHO, YMMV...

KeithG
milosch01
Helpful MS/Extra'er
Posts: 90
Joined: Sun Aug 28, 2005 1:20 pm
Contact:

Post by milosch01 »

All I had read showed that 0 and 1 were unusable. Does this depend on the tooth count or did I just read it wrong?
Keithg
Super MS/Extra'er
Posts: 2413
Joined: Sun Mar 06, 2005 9:15 am
Location: Chicago, IL, USA
Contact:

36-1 wheel decode problem

Post by Keithg »

Ok, I am mistaken, again... I set up mine a while ago and have not looked back at it since (sorry for the mis-information). Are you dead set on using the cam sensor?

Upon re-reading, you are right. tooth 30 and tooth 1 are out of bounds. I just ran my spreadsheet and got:

90 deg missing tooth (first one)
60 deg trigger angle
18 dbtdc crank timing
trigger, return
2,6
12,16
22,28

HTH,

KeithG

On 6/2/06, milosch01 <milos@groupwhere.org (milos@groupwhere.org)> wrote:

All I had read showed that 0 and 1 were unusable. Does this depend on the tooth count or did I just read it wrong?



This post is at: http://www.msefi.com/viewtopic.php?p=123527#123527



Posted by email.
milosch01
Helpful MS/Extra'er
Posts: 90
Joined: Sun Aug 28, 2005 1:20 pm
Contact:

Post by milosch01 »

Thanks for clearing that up. I am using the cam sensor due to mounting issues at the crank. I have little room to spare up front with the current old-school flex fan on the water pump. This engine is longer than a 350, leaving me slicing up room to fit the wheel, fan, and maintain clearance to the radiator. No room to mount it on the backside of the balancer from what I can tell, and it would throw the spacing off. Also, the balancer assembly inserts into a front engine oil seal. Perhaps I could use electric fans, at which point I should have better clearance on the crank.

It's not impossible, and it would definitely make it easier to work on the crank re: wheel setup. Using the cam-mounted wheel, I am able to hide the sensor and wheel on the backside of the cam sprocket and under the rear sprocket cover.

There is a 3rd possibility, and that is to mount the wheel on the backside of the auxiliary drive sprocket. This 3rd sprocket is the same size as the crank sprocket, so it would be like a standard crank setup. It would involve modifying covers which mount underneath the crank sprocket, so I would have to pull all of that off the engine to get the wheel mounted. Again, not impossible. Just a pain.

Maybe this method is a little too slick to get past the code. But, it has worked on a couple of occasions - engine started and idled smoothly. So, something simple is getting in my way now...
milosch01
Helpful MS/Extra'er
Posts: 90
Joined: Sun Aug 28, 2005 1:20 pm
Contact:

Post by milosch01 »

James, I am not sure if you got my email, but that code did the trick. The car is now running smoother than ever, carb or FI! I found some fuel issues with the filter fittings and went ahead and rebuilt the pressure regulator just in case.
Post Reply