Crank sensor offset - getting it right

Give input on tuning as well as any tips and tricks you may have. Also feel free to share base mapping files for various engine types.
SMR
Posts: 137
Joined: Tue Jan 16, 2018 9:38 am
ECU Model: S100 Pro
Distributor: DTA
Location: Aberdeen

Re: Crank sensor offset - getting it right

Post by SMR »

Here's the cranking scope trace I took, looking for the cam sensor. I took about half a dozen - all showed a trigger at 14 teeth but not sure if that's anything.
Reassembly test f cyl1 200712 single spark.jpg
Alex DTA
Posts: 1622
Joined: Fri Feb 28, 2014 1:03 pm
ECU Model: S40 Pro
Distributor: DTA
Firmware Version: 79

Re: Crank sensor offset - getting it right

Post by Alex DTA »

SMR wrote: I suspect the one tooth per revolution isn't enough at cranking speed....
One tooth per cam rotation is fine for sequential, and all that is needed. More teeth are only required for VVT (and quick sync in later vehicles).
SMR wrote:Here's the cranking scope trace I took, looking for the cam sensor. I took about half a dozen - all showed a trigger at 14 teeth but not sure if that's anything.
Reassembly test f cyl1 200712 single spark.jpg
Still not picking up the cam pulse reliably.
katana
Posts: 455
Joined: Sun Jan 27, 2013 6:08 pm
ECU Model: S60 Pro
Firmware Version: 29

Re: Crank sensor offset - getting it right

Post by katana »

Try tightening the gap between tooth and sensor or making tooth wider / longer? I DIY'd a cam trigger in a GSXR - single M6 steel bolt and magnetic sensor detected at cranking speed, likewise std Zetec arrangement ok at sub 400rpm cranking. Alternatively, turn off sequential below idle rpm - it'll just run wasted / batch fire until phase sync is established?
Rob Stevens
Posts: 1247
Joined: Thu Jul 28, 2011 4:10 am
ECU Model: S100 Pro
Distributor: None-Coil On Plug :)

Re: Crank sensor offset - getting it right

Post by Rob Stevens »

Or dont worry until it's running and just switch to sequential >1000rpm where it will probably be fine. Once running sequential I highly recommend the knobs box to swing the injection point around, I found that it runs best when its richest as you swing the knob
SMR
Posts: 137
Joined: Tue Jan 16, 2018 9:38 am
ECU Model: S100 Pro
Distributor: DTA
Location: Aberdeen

Re: Crank sensor offset - getting it right

Post by SMR »

Thanks Rob - was planning this, although it cranks at more than 1000rpm, so will try at the max of 2000rpm (car idles around 1800/1900 or so).

I've bought a second had knobs box so will give this a go.
Katana - I've already made a larger diameter wheel with deeper & wider tooth and narrowed the gap, although i could try an even smaller gap if necessary.
stevieturbo
Posts: 3577
Joined: Tue Aug 02, 2011 12:08 pm
ECU Model: No ECU
Location: Norn Iron

Re: Crank sensor offset - getting it right

Post by stevieturbo »

Holy F....That must be a starter from hell !!! cranking at 1000rpm
SMR
Posts: 137
Joined: Tue Jan 16, 2018 9:38 am
ECU Model: S100 Pro
Distributor: DTA
Location: Aberdeen

Re: Crank sensor offset - getting it right

Post by SMR »

It's a 3kw external stater motor off a diesel truck or something - drives directly onto the end of the crank as there's no space for an onboard starter, hence the fast cranking speed.

Ok, first bit of testing done - injector deadtime table compiled using a variable PSU to power the car; managed to swing between 11.5 and 13v, as well as the 13.8v from my other PSU. Got enough numbers to plot a curve and enter into the software (Considerably longer deadtimes than I had found online).

Adjusting the crank sensor offset from 72 to 76 didn't make much difference, so will leave it at the correct 76 and sort the ignition map on the dyno. The cranking angle swapped between 10 and 0 degrees BTDC didn't make any noticeable difference as it pretty much starts immediately, so let at zero.

Last test is to see if I can get it to run sequentially! It's a real pain as I can only run it for 10 - 15 minutes or so before it gets too hot, and then need to leave it to cool down for an hour. (I'll maybe pop over to GWR tomorrow and borrow one of the Raptor plug-in coolers).
SMR
Posts: 137
Joined: Tue Jan 16, 2018 9:38 am
ECU Model: S100 Pro
Distributor: DTA
Location: Aberdeen

Re: Crank sensor offset - getting it right

Post by SMR »

I tried sequential; I got a enough cam signals for the diagnostics page to show a cam position of 68 degrees which I entered into the software.

If I slowly raised the revs to the 2000rpm change-over point, it stalled. If I rev'd it quicker, I'd get a few "blips" on the engine before it died.
On checking the diagnostics page, I've had 500+ cam errors so clearly I need to check out the sensor clearance to the tooth. Of course, with the engine installed, I can't really get to the sensor location to acurately measure the clearance.... I'll keep trying!

I did notice that quite often during blips of the throttle (in batch or attempted-sequential) the diagnostics page would flash up "Engine over speed" - any idea what this is?

And a further question! SWIN wouldn't allow me to tick "One coil per plug" and a 2000rpm changeover, if I still had "twin spark" ticked in the engine config - I guess that until I have a reliable cam signal I can't untick the twin spark?
stevieturbo
Posts: 3577
Joined: Tue Aug 02, 2011 12:08 pm
ECU Model: No ECU
Location: Norn Iron

Re: Crank sensor offset - getting it right

Post by stevieturbo »

coil per plug sequential would negate the twin spark option ( assuming one plug per cylinder of course ).

So can't see how the two would live together. Although I've a funny feeling I recall Allan saying years ago that on the S, even if you do not tick twin spark, the outputs are mirrored on the higher numbered outputs anyway.

So just wire as normal, 1 output per cylinder and all should work ok.

In short..untick it lol.

But it does sound like you're having crank and cam issues here. Are there errors on both ? rejected pulses etc ?
SMR
Posts: 137
Joined: Tue Jan 16, 2018 9:38 am
ECU Model: S100 Pro
Distributor: DTA
Location: Aberdeen

Re: Crank sensor offset - getting it right

Post by SMR »

I had another go yesterday; raising the cam sensor gave some sporadic cam sensor readings along with a couple of hundred errors, but I need to bodge up a "loose" sensor mount so I can try moving the sensor while the engine is running. Moving the sensor closer (from first-installed position) didn't seem to make much difference but produced around 500 errors in the same timeframe.

StevieT - I phrased my comment badly - what I was getting at is my confusion over sequential injection and sequential ignition; the S100 appears to allow one or the other or both? the car has been up and running in batch for 2 seasons now, but when I built the loom, I ensured all coils and injectors were wired in firing order, and fitted in a hall (GT1) cam sensor - just never got the opportunity previously to play with it.

I did play with the injector deadtimes and quite interesting results! The chunky 12v PSU I have has a PCB-mouted pot to adjust output voltage so I could set the car running voltage from 11.5v up to 13.5v (It's a total-loss system onboard). Managed about 6 different voltage settings and using the deadtime test came up with a set of times.

I plotted these out on a graph (old fashioned using paper!!!) along with the numbers i found on the net for my IWP043 injectors; The spec was for 3 bar and I'm running 4.5bar (a fixed reg) - the deadtimes I measured mirror the standard numbers but consistently offset by around .3ms hich seems to tally with what I could find online about the relationship between deadtime and pressure. I picked off the relevant numbers from the graph and entered into the deadtime table - a good educational afternoon!

Alex - perhaps a couple of busg in DTASwin (I'm using v90.22);

1. Sometimes when I went to enter numbers in the deadtime table, the decimal places dissapeared and I was left with a cloumn of single digit numbers. I deleted that map and used a previous map and all was well. this happened a few times?

2. When saving maps, the comments box sometimes appears to lose most of the text info?
Post Reply