Forum » MoTeC M1 Software Tutorial » Digital hall effect crank sensor reads position changed

Digital hall effect crank sensor reads position changed

MoTeC M1 Software Tutorial

Forum Posts

Courses

Blog

Tech Articles

Discussion and questions related to the course MoTeC M1 Software Tutorial

= Resolved threads

Page 1
Author
1649 Views

Hi guys,

Just started to tune my car the other day and the motec m1 software has a large learning curve. None the less I ran into an issue regarding my crank sensor channel changing from OK to position changed around 3000-4000 rpm and the engine sputters which makes sense, but if I let off the throttle and get back on the throttle it seems to be OK. So the first thing I checked was my crank gear trigger wheel. It appears to be fine. What is the correct way to set up your hysteresis and denounce for digital hall effect crank ref and cam sync sensors. My threshold is set to 2v and set to falling edge. Hysteresis it's set to .8 all the way across. Thank you

Hi Luke,

What Reference Mode are you using? Do you have a input capture of the signals?

Normally what is happening is that the sync signal is moving around and the edge position is changing in relation to the sync point of the crank.

Hi Black Rex,

What do you mean what reference mode? Thank you.

Hi Luke,

The Engine Speed Reference Mode that is set in 2: Initial Setup | 3, Ref Sync that informs the M1 of the Ref and sync patterns to be used to obtain cycle lock.

Hi Blackrex,

I'm using multi tooth four stroke. 12 tooth trigger wheel. Offset is 430. Blank ratio is 10% wide pitch is 150% narrow pitch is 50%. I forgot to mention my denounce for cam and crank is 0 = 100 500=80 1000=80 2000=80 10000= 80.thank you

Regards,

Luke

How close to the active Ref edge is the Sync edge? the reason that I ask is that with a multi tooth setup like that, as the Ref has not way for the M1 to recognise whereabouts in the cycle the crank is (due to no missed/additional teeth), that if the sync edge moves then the M1 can't recognise this occurring and it just re syncs on the next tooth after the sync position, so I would check using the input capture function that the Sync is not moving.

Also, with the Debounce, the higher the engine speed, the lower the Debounce values, I would start at 60 and then go down from there, try;

0 = 60

500 = 60

1000 = 40

2000 = 20

10000 = 10

If you are still getting errors try either halving or doubling the numbers in the region that you have the issue.

Blackrex,

Thank you for your help. I'll let you know how it goes. I'm also going to post a ref sync capture. Soon.

What engine are you using LukeW?

Hi all - I am having a similar issue. Here are some of the details/configuration:

2jzgte

using 15khz hall sensors for cam and crank

cam is using the 2nd cam sensor location with 36* Engine Speed Reference

Pullup = On for both engine speed and synch (wondering if how they are wired... see below... if this should actually be off)

Active Edge = Falling Edge for both

Threshold set to 4.0v (log indicates ~9.8V max for both; both are wired to battery + 12V for power and engine harness ground... not 5V ref and signal ground. This was per instructions of the sensors)

I logged the sensors V and both don't seem as consistent as I would expect. The crank trigger/wheel is a multi tooth 12 teeth configuration and the cam trigger is only a single "tooth"

Hysteresis is set at 0.5

Engine Speed Debounce = 0 RPM @ 60us; 500 RPM @ 60us; 1000 @ 40us; 2000 @ 20us; 4000 @ 20us; and 10,000 @ 10us (adjusted to above comments)

Engine Synch Debounce = 0 RPM @ 130us; 1000 @ 120us; 2500 @ 60us; 4000 @ 60us; and 5500 @ 60us (have not adjusted from a base file)

Attached Files

The M1 Logging is too slow to capture the actual voltage of the except at very slow Engine Speeds (like cranking) and with very fast sample rates. That is exactly what the Input Capture... feature is for. Use that to capture your signals and save that file and post here.

Thanks David, I will do this later tonight and post. As always, thank you for your help.

Here is a link to the picture of the Input Capture... didn't look like you can upload the file type to the forum.

[img="https://flic.kr/p/2m5N5ep[img ]https://live.staticflickr.com/65535/51250083273_7699be38f6_k.jpg" alt="" alt="" ]2JZGTE M150 Input Capture PHR Cam and Crank Hall Sensors by Romans1-16, on Flickr

Hi,

For that Engine I would use the following settings.

Engine Speed Reference Mode - Multi Tooth Four Stroke

Engine Speed Reference Tooth Count - 12

Engine Speed Reference Offset - TBC

Engine Speed Reference Blank Ratio - 10%

Engine Speed Reference Wide Pitch Threshold - 150%

Engine Speed Reference Narrow Pitch Threshold - 50%

Engine Speed Pin Pullup Control - On

Engine Speed Pin Active Edge - Falling Edge

Engine Speed Pin Threshold - 2V

Engine Speed Pin Hysteresis - 0.5V for all engine speeds

Engine Speed Pin Debounce - 4us for all engine speeds

Engine Synchronisation Pin Pullup Control - On

Engine Synchronisation Pin Active Edge - Falling Edge

Engine Synchronisation Pin Threshold - 2V

Engine Synchronisation Pin Hysteresis - 0.5V for all engine speeds

Engine Synchronisation Pin Debounce - 4us for all engine speeds

I think if it were me I would choose the rising edge on the speed pin, the falling edge is very close to clashing with either sync edge.

Possibly the case Adam, I'd really need to see the Input Capture to correctly specify it.

Thanks guys… @Stephan I posted a link to a screenshot I took of the input capture above, not sure if you saw that.

Was also curious to learn more about the 4us recommendation relative to your recommendations to the OP. I don’t know much about this setting, so thank you in advance!!!!

Hi Aaron, I saw the screenshot, for this sort of work the raw file is much better as it allows for zooming in the be able to look at the edges that are being generated to allow for the correct edge to be used, as well as allowing for the offset between the Reference and Sync signals to be measured.

As the Hall Effect sensors provide an already filtered signal to the UDig input, you do not need to have a long debounce period to handle any possible noise around the trigger point. A Mag sensor does need to have a longer debounce value at lower engine speeds that reduces in length at higher engine speeds due to the nature of the signal output from the sensor.

BTW, you can zip the file ("send to compressed folder" in Windows parlance) in order to upload it here.

@Stephen... thanks for teaching; much appreciated. Add a zip file of a data capture; special thanks to Mr. Ferguson for what should have been an obvious solution... guess double the learning for me today :)

Thanks guys!

Attached Files

Having a look at the capture, I will stick with using the falling edge for this on the Reference input, the reason for this is that the spacing on the falling edges remain the most consistent, and the edge is more cleanly defined. You could use the rising edge for the Sync as this is less critical in its exact timing as it has to fall within a window, and it is slightly further away from the Reference edge.

Attached Files

When you say "slightly further away", isnt it still only 1.5degs from disaster? If that cam edge moves just 1.5deg to the left he is suddenly going to be synced 30 deg earlier?

Okay - so let's see if I am learning here...

Seems like if I use rising edge on the Speed pin say 2.5v the line seems to be "clean" less rounded when switching from ~0V to to the rising point.

Then if I use falling edge on the synch pin 4.5V it would be 19degs past the previous rising edge of the speed pin and 12degs before the following rising edge of the speed pin.

Reading the M1 instructions it seems that it will define the reference tooth immediately before the synchronization tooth as cycle lock. Does this mean that the cycle lock will advance one tooth?

So if I look at the data capture to I measure the change in degrees from the threshold point (i.e. 2.5v).?

Thanks in advance for all the teaching :)

Okay... for those that may look back at this for reference; the setup above seems to work well providing a clean consistent synch reference. I ended up with a engine speed reference of around 62degs. I need to think through the input capture to understand the shift.