Jump to content

Crank trigger tooth count


Josh T

Recommended Posts

Im going to be making up a cam trigger and crank trigger wheel and was after some input. Would an 8 tooth trigger wheel be sufficient or would a 16T be that much better? Its a 4g63 spinning to 10k max. It will be running a hallefect e7-9 cam sensor on the crank and another with one tooth off the cam.

I was planing on doing equal teeth spacing to tooth width on the crank but how big/many degrees should i run cam tooth for? Cheers

Link to comment
Share on other sites

With a missing tooth crank wheel the location of the cam tooth is less critical.  

Ideally your number of teeth should divide into 360 deg so definitely dont go for 16T.  8 teeth will be ok.

What do you mean by "modifying a standard sprocket"  What engine is this and what is the Sprocket off?

Link to comment
Share on other sites

Ahh i totally forgot tooth count had to divide into 360. Its a 4g63, i machined the teeth off the unused balance shaft sproket and will mill slots into it. 

Im assuming even 8 teeth have to be alot better than the factory two long teeth? Or does the g4+ use both rising and falling edges off each tooth like some other ecus?

Does equal tooth to gap spacing work the best on hallefect? Or smaller teeth with larger gaps?

Cheers

Link to comment
Share on other sites

Ahh i totally forgot tooth count had to divide into 360. Its a 4g63, i machined the teeth off the unused balance shaft sproket and will mill slots into it. 

Im assuming even 8 teeth have to be alot better than the factory two long teeth? Or does the g4+ use both rising and falling edges off each tooth like some other ecus?

Does equal tooth to gap spacing work the best on hallefect? Or smaller teeth with larger gaps?

Cheers

1).  First of all your target wheel needs to be a magnetically soft steel (usually low carbon steel is suitable).  Those belt sprockets are usually some kind of sintered alloy steel material so will likely give poor positional repeatability.  You might be better to start from scratch or get something laser cut.

2). For multitooth or multitooth missing trigger modes we only use one edge. Provided the 8 teeth are positioned accurately and the target is made from the correct steel then yes it should be better than two teeth.

3). Most automotive hall sensors have a bias magnet inside, you can guess the diameter of that magnet by wiping something like a paper clip over the end of the sensor. Your target wheel teeth need to be at least bigger than that bias magnet.

 

Link to comment
Share on other sites

1).  First of all your target wheel needs to be a magnetically soft steel (usually low carbon steel is suitable).  Those belt sprockets are usually some kind of sintered alloy steel material so will likely give poor positional repeatability.  You might be better to start from scratch or get something laser cut. 

 

Its magnetic and ill scope it on the lathe first. Cheers

I see my wording above may have been a little ambiguous.  I meant "magnetically soft"  not "magnetic".  The difference is a "magnetically soft" material will not permanently magnetise easily.  This is what you need for a trigger wheel.  If the metal is not "magnetically soft" then overtime the bias magnet inside the hall sensor can slowly magnetise the trigger wheel and your trigger points (base timing) will shift.  In your case if that sprocket is also sintered/powder metal like most are then even if your teeth are machined accurately the porosity will mean every tooth will trigger at a slightly different point - again not good for ignition timing accuracy.

Link to comment
Share on other sites

  • 9 months later...

Belts are due on my 4G63 soon and I was wondering about fitting a 12-1 trigger disc, is this worth doing or does it depend on other factors? Car is around 500hp and has Mivec. Are there any tangible benefits to doing this??

 

I did read on another platform that there could be an issue with controlling Mivec with an aftermarket disc- can't see why as the ECU just needs to know where the engine is and therefore what position the inlet cam needs to achieve- sure the Link will handle it :) 

Link to comment
Share on other sites

I assume we are talking Evo 9 Mivec?  If so the mivec cam position is done by a separate sensor to the trigger “sync” so you are free to change the trigger.  Note you will need to remove one tooth from the exhaust cam if you are changing to a generic multi tooth or missing tooth crank.  

Benefits of a higher tooth count is more correct/stable ignition timing during fast transient conditions so if retuned you can often run more timing before the knock limit is reached.

Link to comment
Share on other sites

Thank you for the reply and advice. 

Yes, Evo 9 Mivec. I am 99% it was to do with the ECU itself rather than a generic issue, just thought I'd ask before causing myself a problem! 

That's interesting about taking a tooth off the cam sensor trigger, without wishing to sound really silly, can the ECU not account for the 2 generated signals at the Cam sensor? 

Would it matter which of the 2 teeth is removed, one is a lot bigger than the other? 

 

Found this thread earlier, question answered.  

 

Link to comment
Share on other sites

  • 2 weeks later...

Its unable to on a generic option. either tooth can be removed.

Are you on a plugin unit or a wire in.

One thing to consider on the plugin is that the Tacho signal is driven from the crank signal so if changing to a high tooth count you will need to move the Tacho signal on to a free AUX output.

Link to comment
Share on other sites

I've just done some maths. If I calculated it right, momentary timing drift can be as big as 45deg on the standard Evo Crank wheel if rising and falling edge is used (=4 signals per Rev) 

Let's assume: n=6000rpm=100/sec,  deltaRPM=3000rpm = 50/sec 

time for 90° KW    t_90= 1/(100*4) = 0.0025sec 

delta angel   d_phi = deltaRPM * t_90 * 360deg = 45deg

That's a huge difference. Do I miss something? 

on Track datalogs I see Engine Speed ROC as big as 18000 RPM/sec !  That would mean a huge timing drift. 

 

Link to comment
Share on other sites

6 hours ago, mapper said:

That's a huge difference. Do I miss something? 

Reading on a cell phone but that doesnt look right to me. 

If your initial crank speed was 6000RPM, and you had a continuous acceleration (ROC) of 18000/sec, then after 90deg (0.0025sec) the engine speed has increased by about 45RPM.  That means the next crank tooth would arrive at the sensor about 0.00012sec earlier than the ECU had predicted it would (Based on the last two teeth).

I might be wrong and am struggling to show it clearly in words, but I can only see room for the potential of about 0.5deg error in this example.

Link to comment
Share on other sites

EDITED 

your right Adam, it was a bit late and my calculation was wrong.  It's quite simple to work the timing shift out with the equation of motio

linear motion: x(t) = 1/2 * x'' * t2 + x' * t + xo

translated to a rotative system: phi(t) = 1/2 *phi'' * t2 + phi' * t phi0

given parameter: phi'=6000 RPM/min = 100 RPM/sec phi''=18000 RPM/sec^2 = phi0 = 0

time for one revolution @ 6000RPM/min t100 = 1/100 sec 

We have 4 reference signal per RPM, so in worst case the last reference signal until ignition firing is 90deg. (Assuming the ECU processing has no timing lag) 

time to pass 90deg crank @ 6000RPM/min 
t90 = t100/4 = 1/400 sec 

phi(t) = 1/2 * (18000 RPM/sec2) *(1/400sec)^2 + (100 RPM(sec^2)* 1/400sec = 0.3063 RPM

We want the result in Crank degrees: 

phi(t)result = 0.3063 RPM *360deg = 110.25deg 

==> means we have a timing drift of 20.25deg!

The Rate of Change of RPM value is the max. I see on a flat out Lap on Tarmac on a 700hp 4cyl engine: On top of that signal processing can add another significant lag which leads to timing more timing drift.

Please let me know if someone don't agree with my calculation

Adam or Simon at what frequency is the trigger input calculated and updated. I suppose this will add a significant lag resp. time shift, isn't it? 

Link to comment
Share on other sites

On 11/01/2018 at 10:42 PM, Simon said:

Its unable to on a generic option. either tooth can be removed.

Are you on a plugin unit or a wire in.

One thing to consider on the plugin is that the Tacho signal is driven from the crank signal so if changing to a high tooth count you will need to move the Tacho signal on to a free AUX output.

Thank you for the reply. 

I'm on an EvoLinkIX PNP.

Great point on the tacho, suppose one benefit if I were to have to relocate the tacho output to an aux I could take advantage of the dial sweep on ignition! I'll have to check the wiring but a really good point. 

Completely on the fence now whether to spend the £100 odd to convert to a multi-tooth trigger wheel. 

Link to comment
Share on other sites

Hi Guys 

I just realised that I made another mistake in my calculation. I edited the post above.  Timing drift can be huge on low crank count trigger.

to get you some additional numbers, error falling in square! timing drift @ 6000rpm and Rate of change in RPM 18000RPM/sec: 

Numbers are for worst case scenario (90deg between last reference signal and spark signal)

Number of Crank Reference signal    / Timing drift

4 / 20.25 deg
8 / 5.0625 deg
12 / 2.25 deg
18 / 1.0 deg
36 / 0.25 deg
60 / 0.09 deg 

Link to comment
Share on other sites

What I can add is if you lock timing on an Evo and when watching with a good stable timing light (I use a ferret), give the throttle a hard stab you can easily see the drift.  It doesnt look as big as the numbers above but you can easily see it - maybe 4 or 5 degrees at a guess (this is after ign delay is adjusted properly).

When you do the same test with say a 2JZ VVT with 36-2 you cant detect any movement by eye.

Link to comment
Share on other sites

5 minutes ago, Adamw said:

What I can add is if you lock timing on an Evo and when watching with a good stable timing light (I use a ferret), give the throttle a hard stab you can easily see the drift.  It doesnt look as big as the numbers above but you can easily see it - maybe 4 or 5 degrees at a guess (this is after ign delay is adjusted properly).

When you do the same test with say a 2JZ VVT with 36-2 you cant detect any movement by eye.

I agree, have to check what ROC numbers we get just by throttle blips.  

Adam does the trigger decoding take ROC into account to guess when next reference signal must occur? Or does it just use actual RPM (means ECU expect steady RPM) ? 

 

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...