Supra Forums banner

1 - 16 of 16 Posts

·
A zone of danger.
Joined
·
1,175 Posts
Discussion Starter #1
It pains me to write this since I've always done pretty well with research and manuals on my own, but I've hit a roadblock. I've applied every piece of information I could find with zero change in status - the cam and crank count, but just won't sync. So before I have to find a means of wrangling an o-scope, perhaps the more experienced than I will spot an oversight on my part.

Basics:
AEMTuner v3.2, 02v02 firmware

Mechanical timing:
Per 1) the manual 2) past experience 3) verification from another VVTi GTE-sporting member and 3) This post on myIS, the timing is correct. I used the "crank @ 0, dashes = cam cover" method, just like a regular GTE.

Advanced pickups:
- I used the VVTi setting from the Cam/Crank wizard for for the Crank Tooth Control table, Sync MX, teeth count, teeth missing and any other changes it provides.

- I used this post on Clublexus to adjust the CKP sensitivity since the 36-2 has smaller teeth and likely a shorter waveform. I also upped the CMP sensitivity, though no real change there. I saw a slight improvement at 6000rpm from 2000rpm and both better than 500rpm, but the Tooth counts were still sloppy and inconsistent. Not sure if this is a logging issue or I'm actually getting truncated counts and missed teeth:

Is this still a noise issue? Sensitivity? A setting I missed? I never looked that closely at my non-VVTi logs since I never had a sync issue, and I started with that as a basemap to make changes.

I also tried flipping the rising/falling edge settings per dre99gsx's findings in troubleshooting in case of a change due to new triggers...no luck.

Ultimately, this is my advanced pickups page for the stock VVTi CKP & CMP connected to my stock non-VVTi harness:



If someone has any ideas, please let me know. I'm at the end of my rope after assembling this thing all summer and just want to get tuning.

TIA,
Mike
 

·
Registered
Joined
·
2,183 Posts
I had use 0.06 for Crank H Sense Level. It would start fine on a cold/hot engine during the winter at 0.10v, but during summer on hot restarts it would not see the crank signal at cranking speed. Hot sensor = lower voltage.

It looks like the AEM is seeing cam and crank for you, at least. You'll need to take an internal log at the fastest rate to see each tooth during cranking. Get these channels at the fastest rate (Fast 1-7 and Slow 1-7 are all 250Hz with the new firmware). Then we can see exactly what's going on.

 

·
A zone of danger.
Joined
·
1,175 Posts
Discussion Starter #4
Well, Kurt to the rescue once again. The only thing I changed was Crank H Sens to 0.06 and it Sync'd in about 750ms. Traces are obviously much cleaner as well, though internally logging may have helped:



I had one good crank left in the battery after the log, so I quickly turned on the injectors & coils and tried again. It gave me a quick cough so its close now, but it's never run 93oct on the 1450cc's (for this purpose, easier to start than E85) so the start tables, base duty, etc need cleaning...not to mention timing validated/synced.

Thanks a ton Kurt. Hopefully this will be the only hiccup, but VVTi setup is also uncharted territory so we'll see if that's correct down the road. :)

Mike
 

·
A zone of danger.
Joined
·
1,175 Posts
Discussion Starter #6
Well, the pickup woes continue... =/

- Ign Timing was resync'd; couldn't find the timing mark at all so I pulled the plugs and adjusted Ignition Phase tooth until the mark was 10deg advanced, matching Crank Advance. 12 tooth GTE was 6.00 teeth, 36-2 VVTi is 22.58 teeth.

- To ensure I wasn't spraying closed valves, I resync'd the Injector Phase as well. Final is 20 teeth, putting Fuel Inj #1 at 11.5 teeth. Best I figure its 15deg ATDC, give or take 15deg...enough to make the car run.

I setup a spylist at 250hz, so at 4ms I shouldn't be missing events in the logs. But according to what I see I'm not consistently firing the coils and injectors, suggesting my pickups are to blame yet again. The greatest observation, however, is my Fuel Tooth inconsistency. I frequently get a reset at 12 instead of 24:




I'm baffled. The Cam Tooth is very inconsistent and I'm getting Sync Errors, but I was under the impression the Fuel Tooth reset only occurs when you reach the two missing teeth twice (one full cycle/12 synthetic teeth per rev), hence the 36-2.

Furthermore, when the AEM does "know" where it is, I only get one 24 tooth cycle where every injector and coil fire on time. At that point its likely not a pickup concern but sensitivity calibration (or so logic would say.. :dunno:).

I'd appreciate any thoughts. The car "starts" quickly just like it always has, just won't stay running.

Thanks
 

·
A zone of danger.
Joined
·
1,175 Posts
Discussion Starter #7
The plot thickens: A Tooth appears to be incrementing properly, but Fuel Tooth isn't tracking.



According to the A Tooth Control table, the function values are:

0 Do nothing.
1 Process significant edge.
2 Process alternative edge, used for crank firing.
3 Force reset of A_Tooth, if loss of sync.
4 Test for synchronisation
5 Process significant edge and Test for synchronisation.

6 Process alternative edge and Test for synchronisation.
7 Reserved

I don't get an A Tooth reset, so apparently it's not a sync problem. So the question now is why doesn't Fuel Tooth follow A Tooth. Hmmm...
 

·
Registered
Joined
·
2,183 Posts
I setup a spylist at 250hz, so at 4ms I shouldn't be missing events in the logs. But according to what I see I'm not consistently firing the coils and injectors, suggesting my pickups are to blame yet again. The greatest observation, however, is my Fuel Tooth inconsistency. I frequently get a reset at 12 instead of 24:

I'm baffled. The Cam Tooth is very inconsistent and I'm getting Sync Errors, but I was under the impression the Fuel Tooth reset only occurs when you reach the two missing teeth twice (one full cycle/12 synthetic teeth per rev), hence the 36-2.
The only time you will reliably capture every coil/injector event is when dwell or PW are greater than 4ms. It's missing those due to the timing of the sampling and it's normal.

Your fuel tooth resetting at 12 is definitely a problem and is due to the cam pickup. Look at A Tooth, Fuel Tooth, Crank Count, and Cam Count and you'll see exactly where you're missing cam teeth. You can see the missing tooth section of the crank in A Tooth and Cam Count for reference. One rev will have two cam triggers and the other will have one. Cam teeth will all be fairly evenly spaced so you will easily see any that are missed. In addition, Battery Raw will indicate each compression stroke. You might try Cam H Sense Level at 0.10v.
 

·
A zone of danger.
Joined
·
1,175 Posts
Discussion Starter #9
Great info, thanks Kurt.

I've been tinkering all day, and not 5 minutes before you replied it finally tried to idle. I was hunting horribly since I changed the Idle% vs Target table, and the Fuel Tooth still resets every 2nd or 3rd cycle, so maybe I just got lucky. I rewired the crank pickup connector, but you're absolutely right - it caught right after I made the cam pickup changes. A bad decision on my part, but if you look at the Advanced Pickup page I posted above I made the same changes to the cam pickup that I did to the crank...when the cam teeth aren't any different, at least not on the scale of the 12T crank vs 36-2. I put in the basemap values and boom..fire.

I'm hoping it starts again with some better timing and Idle%/Target, then try to find a cam setting that works but the guesswork is a bit tiring.
 

·
A zone of danger.
Joined
·
1,175 Posts
Discussion Starter #10
For those of you watching (or searching in the future), I'm almost certain I have a cam noise issue caused by hardware. No adjustments to the Advanced Pickups removes the extra cam count causing my sync to drop. Tonight I started the car with the settings below and it jumped to a solid idle so the fuel/timing/IAC values are fine. I tried tipping in, and immediately the timing jumped all over as soon as the revvs hit 1100. Also, it will idle just fine, then without any changes from me it will start hunting wildly once warm (even with all error tables zero'd).

The above points to intermittent connection, variability with temperature/failing sensor. I had no issues with my original harness, but this is the original VVTi sensor that I clearly have no history on (sold my USDM sensors already..doh). Going to try a different sensor, add 2.2k ohm to signal/ground to reduce any noise, and see if I can strengthen the grounds.

One very strange issue: As soon as the engine starts the internal log data immediately begins logging implausible data on every channel (Eng Spd > 17000, etc etc). This concerns me a bit as to the reliability of the AEM...
 

·
Registered
Joined
·
2,183 Posts
One very strange issue: As soon as the engine starts the internal log data immediately begins logging implausible data on every channel (Eng Spd > 17000, etc etc). This concerns me a bit as to the reliability of the AEM...

I think it will do this if you have it set to loop logging. The new firmware isn't happy with it set to this. I like using a switch for internal logging, personally. I hope I'm wrong, but I believe that this issue will not be fixed as the S2 platform is now dead.

The pull-down resistors on the 6100 box are 2.2k for cam and crank. You can change these to lower 4.7k and use higher voltage pickup settings, but this will require some skill with a soldering iron. For reference, I'm using IS300 mag pickup sensors with my original 6100 box with no issues. These sensors age, so it will probably be different for others.
 

·
A zone of danger.
Joined
·
1,175 Posts
Discussion Starter #12 (Edited)
Well I checked and I do indeed have loop logging enabled. Next time I try starting I'll turn it off and have a look at the data.

With regard to the IS300 sensor - is here a benefit vs the MKIV, or just a matter of circumstance that you had one (ie - still PNP)? I intended to order a new sensor from Curt today, but if there's merit in running a different part number, I'm willing to try.

Thanks

:Edit:
I called Curt today and he ran the P/N's for the IS300, non-VVTi & VVTi GTE, and they're all the same. New sensor will be here tomorrow.
 

·
A zone of danger.
Joined
·
1,175 Posts
Discussion Starter #13
Quick update:

- New cam sensor did not resolve the issue.
- I tried various resistors and anything weaker than 4.7k ohm resistance to ground won't allow sync. 4.7k did create a more solid idle, and Sync Tooth moves less frequently, but it still moves. No good.

Next steps:
- Overlay cam & crank wires with shielded twisted pair with a dedicated ground for the shielding

- Considering a pull-up resistor on at least the crank signal to increase the wave amplitude so I can decrease the sensitivity. This may be an act of desperation. Have to see what AEM says but they always give me some bullshit runaround about VVTi being an unsupported platform rather than telling me what might help their hardware perform better.

- If still failing its back to the 12 tooth wheel. Theoretically this should still work since AEM always converts to 24 events anyway, and it's just not worth the headache anymore.
 

·
Registered
Joined
·
2,183 Posts
If you were using the stock cam (1 tooth pickup) and the 36-2 wheel and a fixed cam angle then you would see Sync Tooth stay at one value once the engine was running. With the VVTi cam (3 tooth pickup), it is normal for that parameter to jump around like you're seeing. At each angle, it should be moving between three different values.

Channels to watch for pickup issues while running (at high speed with internal log) is Crank Tooth Period, T2PER, Cam(T2) - Crank, Timing Errors, and Sync Errors. Timing Errors should not increment, Sync Errors will swing from 0-1 the whole time and should not ever go higher than 1. At 1,000rpm (60ms/rev), Crank Tooth Period should be around 5ms, T2PER 40ms, and Cam(T2) - Crank will hold steady around a certain value. Any sudden jumps in any of these channels will indicate erroneous extra or missing teeth.

If there were excessive noise induced over the pickup wires, you'd see extra teeth being processed and sync/timing errors incrementing. This would call for steps to eliminate the noise, or try the higher pull-down resistor to try and deal with it. You can determine if this is happening by looking at a scope reading of the pickups at the ecu plug or simply by examining your fast internal logs. I think you should determine that erroneous extra teeth are actually causing you a problem before adding a pull-down resistor. What you're doing is making the signal weaker, which would necessitate different pickup settings for each different resistor. If it turns out that you're dropping/missing teeth due to a weak signal, then the fix would be to change the pull-down resistors inside the box. Since you're consistently getting sync and starting the engine, I highly doubt that this is an issue for you. Signal will be at it's weakest on a hot sensor during cranking.

From your description above, it sounds like you may be looking at the wrong channels to determine what your pickup issues are and applying fixes that may or may not be adding to the problems.
 

·
A zone of danger.
Joined
·
1,175 Posts
Discussion Starter #15
If you were using the stock cam (1 tooth pickup) and the 36-2 wheel and a fixed cam angle then you would see Sync Tooth stay at one value once the engine was running. With the VVTi cam (3 tooth pickup), it is normal for that parameter to jump around like you're seeing. At each angle, it should be moving between three different values.
This was in the back of my mind, as to how the AEM would be able to track three teeth for sync instead of just one. I was able to see visually 2-10-18-2-10-18 repeating though I haven't looked at the log. Good to know.


Channels to watch for pickup issues while running (at high speed with internal log) is Crank Tooth Period, T2PER, Cam(T2) - Crank, Timing Errors, and Sync Errors. Timing Errors should not increment, Sync Errors will swing from 0-1 the whole time and should not ever go higher than 1. At 1,000rpm (60ms/rev), Crank Tooth Period should be around 5ms, T2PER 40ms, and Cam(T2) - Crank will hold steady around a certain value. Any sudden jumps in any of these channels will indicate erroneous extra or missing teeth.
Sync Errors did indeed toggle 0-1-0, but I also accumulated Timing Errors. The rest I didn't look as closely at in the log, only A Tooth vs Fuel Tooth to look for the dropout.


If there were excessive noise induced over the pickup wires, you'd see extra teeth being processed and sync/timing errors incrementing.
Again, Timing Errors, but the only extra tooth I see is on the Cam Tooth.


This would call for steps to eliminate the noise, or try the higher pull-down resistor to try and deal with it. You can determine if this is happening by looking at a scope reading of the pickups at the ecu plug or simply by examining your fast internal logs.
I've been trying to get my hands on the scope at work, but it's in use so I'm stuck with the log I'm afraid...


I think you should determine that erroneous extra teeth are actually causing you a problem before adding a pull-down resistor. What you're doing is making the signal weaker, which would necessitate different pickup settings for each different resistor.
Agreed, and that's what I did...sort of. Once down to 0.1 on Crank H Sens (maximum for sync) there's only 4 points of resolution left before running back into zero and defaulting to 0.2V.


If it turns out that you're dropping/missing teeth due to a weak signal, then the fix would be to change the pull-down resistors inside the box.
I'm decent with a soldering iron, but this option makes me nervous for fear of damage. If this was my only remaining course of action, I'd go back to the 12 tooth wheel and known good USDM sensor.


Since you're consistently getting sync and starting the engine, I highly doubt that this is an issue for you. Signal will be at it's weakest on a hot sensor during cranking.
Also an issue. Cold start is good, but after 3-4min of idle and a 5min soak, it fails to sync (or at least well enough) upon next crank event.


From your description above, it sounds like you may be looking at the wrong channels to determine what your pickup issues are and applying fixes that may or may not be adding to the problems.
The only thing I've changed so far was attempting to use the resistors in the event noise was present, but those were not permanent and have already been removed. Everything else as just been calibraiton changes per our discussion, or cannot make matters worse (ie - new overlays with beefier wiring, etc).

I'll send you an email shortly. I'm sure that means will be more efficient than screen shots and trying to explain what it is I see.

Thanks
 
1 - 16 of 16 Posts
Top