Cam Sync - AEM Infinity - Starting Issues - Cam Sync - Page 2

mikedsilva · 05-09-2021, 07:31 AM

Quote:

Originally Posted by icarp

Andrew and Rasant /or Infinity? were having trouble with the sensor for a while .
This could be the same gremlin. It had to do with the 60-2 and the sensor specs .
Ian

Still hoping to hear from Andrew... no reply yet...

mikedsilva · 05-09-2021, 09:40 PM

So, it appears to be "fixed".....

I don't quite understand why tho... there was another thread on Pelican that I forgot about, where the original poster had an almost identical problem to me... pulled engine, swapped cams, reinstalled and no run. He ended up ditching the AEM and going with Link or Haltech (I forget which one). Another poster in the same thread ended up upgrading his firmware and seems to be running OK. I was hesitant to upgrade firmware and at this point, still have not.

I borrowed a scope tool that plugs into your laptop.. and managed to scope the cam position sensor and the crank position sensor. First time I used a scope.. took me hours to figure it out and the scaling and stuff.. still not right but just wanted a visual.
I sent vids and pics of the results to a friend to confirm that the sensors were indeed sending the correct signals. And they were. So it wasn't a failed sensor. Now this is the part that I dont understand....

It was working before fine..
Then it began to start only half the time...
Then it wouldnt start EVER.

It was suggested to remove the cam sensor, rotate the pinion, and try to get it to sync on a different part of the cycle. I wasnt exactly sure what I was aiming for.. after rotating it 180 degrees I still got a no start event. Roating 90 from there, and still no start. Rotated again and ended up syncing at 56deg according to the Diagnostic tab in Infinity Software. Looking at the scope, it shows the cam signal on top, coinciding with the missing teeth signal of the crank pulley.

But now it starts every time and I have no more Cam Sync Errors.

Is this considered fixed?
If it is, then why did it work before, and then work less and finally not work at all?

No sensors have been changed.

Here's my vid.. please don't laugh, I don't understand electronics at all.

'78 SC · '78 SC's Garage

Congratulations, Mike. Good news, though getting something to work but not understanding why it works (and why it stopped working) can be frustrating.

Is this one of the threads you are referring to? Hall sensor EFI madness

At this point, do you believe the basic problem was the "phasing" of the cam synch signal relative to TDC? Even though you did not rotate the pinion on the EDM initially, you could have effectively rotated it 180* by installing the cams on the "other" TDC. The fact that it started first time for the cam break in but would not start reliably after that is confusing, though.

When the problems started, you resynced the cam signal without moving the EDM, though it did not seem to be stable and you were getting sync values >360*. Only when you eventually pulled the EDM out and physically rotated it various amounts (180*, 90*, ??*) could you get to no Cam Sync Errors. Don't ECUs have restrictions on where the cam signal needs to be relative to TDC? Like within x degrees of TDC? Rising edge versus falling edge?

Glad you're up and running, looking forward to the comments on this.

Steve

mysocal911 · 03:53 AM

Quote:

Originally Posted by '78 SC

Don't ECUs have restrictions on where the cam signal needs to be relative to TDC? Like within x degrees of TDC? Rising edge versus falling edge?

That an absolute necessity (no distributor), especially when you have such a wide cam pulse width signal.
Typically, a cam position pulse signal width is about the same as the crank signal pulse width,
and not multiple crank pulses as seen on the scope waveforms, i.e. one cam pulse for every two crank rotations.

Dr J · 07:46 AM

Congratulations on solving Mike. In hindsight I looked at the MSX3 manual because it talks about cam tooth vs crank. . I know you are using the Infinity as I am, but I still use the MSX3 manual as a good source of info for EFI. Maybe before the cam edge was close to the missing tooth.

mikedsilva · 05-10-2021, 12:55 PM

Quote:

Originally Posted by '78 SC

Congratulations, Mike. Good news, though getting something to work but not understanding why it works (and why it stopped working) can be frustrating.

Is this one of the threads you are referring to? Hall sensor EFI madness

At this point, do you believe the basic problem was the "phasing" of the cam synch signal relative to TDC? Even though you did not rotate the pinion on the EDM initially, you could have effectively rotated it 180* by installing the cams on the "other" TDC. The fact that it started first time for the cam break in but would not start reliably after that is confusing, though.

When the problems started, you resynced the cam signal without moving the EDM, though it did not seem to be stable and you were getting sync values >360*. Only when you eventually pulled the EDM out and physically rotated it various amounts (180*, 90*, ??*) could you get to no Cam Sync Errors. Don't ECUs have restrictions on where the cam signal needs to be relative to TDC? Like within x degrees of TDC? Rising edge versus falling edge?

Glad you're up and running, looking forward to the comments on this.

Steve

Hi Steve
I wish I was smarter.
I DON'T understand exactly why it is working now. Your comment about installing the cams on "the other TDC" sounded good, except when I rotated the cam sensor 180, that should have effectively put me back on "the other tDC" and it still didn't fire. In fact it carried on and backfired exactly the same. Here's a lesson, don't stand in front of the exhaust pipe when cranking... my shirt melted, and all the hair on my stomach and left leg got singed!!

I feel it has something to do with picking up the cam signal first, before the TDC signal during rotation. The document posted by DrJ kinda explains it, but it also says the cam trigger can be anywhere during the revolution EXCEPT for in the middle of the missing teeth on the crank.. but if I am reading the scope right, isn't that exactly where my triggers are now aligned?

mysocal911 · 05-10-2021, 06:39 PM

Quote:

Originally Posted by mikedsilva

I feel it has something to do with picking up the cam signal first, before the TDC signal during rotation. The document posted by DrJ kinda explains it, but it also says the cam trigger can be anywhere during the revolution EXCEPT for in the middle of the missing teeth on the crank.. but if I am reading the scope right, isn't that exactly where my triggers are now aligned?

For proper syncing with the crank sensor, the timing of the cam sensor is based on the state change (plus to minus, or minus to plus)
of the signal and not the voltage level. If the ECU used the negative going edge of the cam signal relative to the crank TDC, then the
cam signal (negative edge) occurs about 70 degrees before TDC, based on the scope image.

Mr Beau · 05-10-2021, 06:57 PM

If crank sensor is 60-2 then you should be able to run batch/waste spark and not worry about the cam sensor.

jpnovak · 05-10-2021, 07:05 PM

MIke,

the 360 out is likely due to the noise on the cam inputs signal. This occurs about 36 teeth before the TDC indicator.

I suspect that the noise is crossing the threshold for negative edge trigger on the cam signal causing a false angle calculation. If this angle is different than the cam offset values it will increment the error counter and not start or misfire.

You have to find the source of that noise and eliminate it. First step here is to disassemble the top of The Rasant cam sync mechanism and thoroughly clean out. Check for metal shavings on the underside of the VR sensor and that the rotating bolt and surrounding surfaces are clean.

I have had problems where these threshold value settings in software do not specifically do anything to counteract noise. The hardware inputs on the ECU are also not filtered to reduce noise. You may end up building a noise filtering circuit on the cam input line. The most simple is done with a capacitor between signal and ground in combination with a resistor in series with the signal line. Higher order filters work better but are more complicated to install.

Mr Beau · 05-10-2021, 07:35 PM

Posted before seeing it was fixed. Probably what's happening is the ECU is looking for the cam signal at a particular location. You were close, but not close enough initially.

mikedsilva · 05-11-2021, 03:36 PM

thanks for all the responses.
I did remove the cam sensor from the Rasant housing to inspect.. clean as a whistle inside. Remember, this was working fine till I swapped cams and reinstalled. As some have said, it must have been something I had altered to cause the no-start situation.

I don't fully understand what Jamie was saying above but I suspect "threshold" and "false angle calculation" are def what I was experiencing.

My suspicion is that '78SC (Steve) was right that when I installed my new cams and retimed my engine, I might have set them so that the cam sensor was synced 180 out from when I pulled the engine. How would I have done this? Simple - by rotating the crankshaft 360 before starting my cam timing process as this would have rotated the Rasant sensor 180 degrees..... Of course I did not keep a track of how many times I rotated then engine when I was timing the cams.

The next thing I did, was pull the Rasant adapter out of the motor and re-aligned it to be more central over the crank trigger wheel and re-adjusted my gap (it was perfect anyway). When I re-installed it, I must have inadvertently rotated the pinion on the Rasant Sensor 1 tooth as it was now syncing at a different spot and the Cam0Timing Degrees was different to earlier that day.

Then I tried to rotate the pinion 180 but was still having issues. So from here it was pretty much guesswork to arrive where I am at now. I'm having zero CamTiming Errors. Took it for a drive, and seems completely fine. If anything, I think the starting is now 'better' than it has ever been.

In hindsight, when I first assembled the engine, it must have been a complete fluke that I got it running by aligning the cam sensor at the right spot, or one of the right spots. I don't remember the install manuals being so specific about where the signal for the cam needed to be in relation to the crank trigger. (apart from being 360 out and needing to offset).

So, going back to before I removed the Rasant sensor housing and rotated the pinion: could I have simply input another 360 into my Cam timing offset, to achieve sync on the other cylce of the motor?

jpnovak · 05-11-2021, 04:00 PM

Mike,

The only way to truly get this right is to use a timing light and set the reference angle so that a fixed timing "0" matches TDC on the light. Adjust the cam offset until they align.

If the engine does not start then you are 360 out.

None of the AEM documentation mentions anything about where the cam signal can be relative to the crank TDC Angle. Most systems are quite specific about allowable offset windows.

Glad its working well for you now.