27-06-2018
Just a little note. Why does EgoCorrection work on an 4.2 sensor for years, and when I convert to a 4.9 sensor EgoCorrection does not work on full throttle. Sensor position is not changed and engine is not changed. Same ECU same FW.
regards
Per Andersen
Copy of report feeback:
Hello Per,
I understand your case, from control and hardware side there are however some differences between LSU 4.2 and 4.9; LSU 4.9 has a higher heat target (780degC vs 750degC for Lsu 4.2) and wider openings at the front (these allow for faster reaction, but also cause more effect on heating and cooling of the sensor element) as you know on FULL throttle the exhaust flow is the highest, thats why its effects (influence on sensor temperature are highest there).
Without the control loops on target the lambda measurement can be significantly off (inaccurate) thats why the controller disables ego control automatic mixture feedback.
Between 1.2.30 (your current firmware) and 1.2.38 (latest stable release) there have been some changes to enhance the wbo2 measurement in borderling conditions, you might benefit from upgrading it.
I hope this helps; if you have any further questions, feel free to ask.
Best regards,
Dave
=============================================================
26-06-2018
I now have changed the LSU 4.9 to a new one. I have gone through the wires and found nothing wrong.
EgoCorrection still turns off at full throttle, and the Lambda value still seems to be okay, as before with the old sensor. EgoCorrection works fine on part throttle.
I converted from 4.2 to 4.9 about 1½ years ago, as described on the WIKI. With the 4.2 sensor and same FW 1.2.30 EcoCorrection worked fine.
I have another ECU in a Volvo 940 also with FW 1.2.30 where I converted to LSU 4.9, at the same time as the one above, and here EgoCorrection works fine all the time.(see powerrun14 logfile)
http://www.vems.hu/files/perboddum/powerrun14.vemslog
=============================================================
18-06-2018
For no apparent reason my EGO correction turns off at WOT.
No accell enrichment is active and throttle pot is stable at 100%.
At part throttle EGO correction works fine.
Please look at the logfile http://www.vems.hu/files/perboddum/ego1.vemslog
EGO correction turns ON after accell is done and then EGO correction turns suddenly OFF
I have been through my config file and cannot see anything that should stop the EGO correction from working.
I have attached 2 logfiles.
http://www.vems.hu/files/perboddum/ego1.vemslog
http://www.vems.hu/files/perboddum/ego2.vemslog
best regards
per andersen
==============================================================
With this 12-1 trigger signals and high compression, the missing tooth is sometimes shorter (downto 1.55) than the normal tooth (upto 1.62) so this will not work.
As a bandaid 2 holes can be shut with tape as a test (conversion to 12-2 wheel). Or install a suitable wheel (this wheel is, though the most unfortunate choice for missing-tooth pattern - reason below -, would work with a normal engine of lower compression, or higher flywheel mass; but not in this super-engine).
Primary and camsync trigger is hall.
We couldn't set up detect trigger sequense from Nissan 300ZX HALL module with 12-1 triggerwheel
- and it might actually be impossible if the time of a normal pulse sometimes exceeds appr 175% time of previous normal pulse.
It's been known since long, that 12-1 is the most prone to sudden pulse-time variation at cranking (especially with high compression), so it takes more work (eg. analyzing triggerlog - the numerical output is more descriptive in this case - and setting advanced filtering threshold) to set it up right (and best to avoid it). Better options:
- 12-2
- 24-1 or 24-2
- 36-1 or 36-2
- 60-2
- even 8-1
- actually, anything is better choice than 12-1
When turning distributor by hand with ocsilloscope connected there is no doubt where missing tooth is. Allways 11 tooths between the missing tooths and the camsync signal comes on triggertooth 0 (first tooth after missing tooth)
I have also tried disabling camsync all together, but it makes no difference.
When turning the distributor by hand (approx 80..100 rpm in vemstune) sometimes the sparkplugs ignite.
The rpm signal in vemstune (1.5.24) seems to be very stable. No flickering rpm signal.
Triggerlog when engine is cranking. (with wrong ignition output settings, but trigger setup is correct)
http://www.vems.hu/files/perboddum/cranking.triggerlog
Triggerlog when distributor is driven by hand
http://www.vems.hu/files/perboddum/turnbyhand1.triggerlog
http://www.vems.hu/files/perboddum/turnbyhand2.triggerlog
Vems config
http://www.vems.hu/files/perboddum/setup6.vemscfg
Picture of triggerwheel 12-1
If you don't change to a better wheel (eg. to 12-2) at least check position
- if the missing tooth is also enlarged (at cranking, high compression) that makes it more likely that you can set proper threshold
- obviously more than 150% needed (150% is the default threshold for N-1 wheel, when "missing tooth advanced filter" is disabled
- if the normal tooth is max 165% or so, than it will work with good setting...
- If slowing due to compression sometimes causes normal tooth time to be more than appr 175-178% of previous tooth, than out of luck: then impossible to set it correctly and changing to suitable pattern is unavoidable.
- actual max % of normal tooth (when cranking with high compression) is ... ?
- the min % of missing tooth (usually 190%-195%) in this case is ... ?
Per i have reviewed your triggerlogs and indeed the periods between teeth seem very irregular.
1) Lets start out with with following, could you make a triggerlog of engine cranking 8-10 seconds without injector and ignition fuse to verify if the irregular periods are caused by some noise issue.
2) Secondly could you make another triggerlog (again 8-10 seconds) but then spinning the trigger not by hand but by some something with stable rpm (perhaps use a battery powered drill), that way we can verify, if with stable RPM the trigger periods are even.
If you upload both these triggerlogs to your memberspage i will review and provide hints on how to proceed.
Best regards, Dave
Thanks Dave
Scope of triggerwheel 12-1 with power drill at approx 200 rpm
I have done some more test with power drill (50...120 rpm) and the distributor connected to Vems.
Vemstune displays the rpm but as soon as rpm drops and then gains rpm again I often get PRI.trig error in Vemstune and the sparkplug stops sparking.
It seems to me that it is a combination of low rpm AND varying rpm at the same time that confuses Vems.
I can have steady 60 rpm and there are no trigger errors and the sparkplug ignites
Tomorrow morning I will try install the distributor and make the logs
Hello Per,
When reviewing your scope shots it seems there also is a very slow rise rate on the primary trigger pulses (red signal, ch1) could you set both probes to the same amplification, set the scope Y range the same for both channels, trigger on yellow signal (ch2) rising edge and make a much smaller timebase shot so i can accurately see:
1) falling secondary triggger (yellow, ch2) in center of screen and one or two primary edges (red, ch1) left and right.
2) additonally please supply the VEMS triggerlogs (.triggerlog files) as requested in previous query.
3) make sure the secondary trigger is enabled and connected to the VEMS when making these.
Best regards, Dave
Distributor connected to Vems and drill operated.
http://www.vems.hu/files/perboddum/drill1.triggerlog
http://www.vems.hu/files/perboddum/drill2.triggerlog
I cannot make a triggerlog or normal log, because Vems constantly looses the serial connection (physical serial COM1, not USB adapter) during cranking. Vems supply voltage is around 9.35 volt during cranking.
Note:Battery is in the back of the car.
Hello Per,
If the VEMS looses serial connection during cranking this means the ecu resets (most likely because of voltage dip during initial cranking engage) but if the ecu does not reconnect during cranking there might be some problem with the integrity of your GND wires from battery to chassis, GND from chassis to engine or VBATT cable from battery in the back to the engine. While performing the other checks it makes sense to verify these also (for good connection, corrosions, sufficient size etc.)
Best regards, Dave
Thanks Dave.
The car is rebuild from scratch with new wires in 2009. No rust or corrosion what so ever. Have only run some 4000km.
Connected 2 batteries and the engine fires up, but still looses sync at 1500 rpm. No camsync.
http://www.vems.hu/files/perboddum/running1.triggerlog
http://www.vems.hu/files/perboddum/running1.vemslog
With camsync.
http://www.vems.hu/files/perboddum/running2.triggerlog
http://www.vems.hu/files/perboddum/running2.vemslog
Hi Per,
Nice to see it running, but in your triggerlog with CAM sync i do not see secondary trigger showing up (perhaps you did not power cycle the ecu after making the changes in configuration ?). Before attempting anything else we should verify that all triggers are in working order _without fuse to INJ and IGN_ at cranking.
Very important in your running1/2.vemslogs, CLT/MAP IGN adjust table is not filled reasonably, please correct before attempting any further tests
Best regards, Dave
Hi Dave
Camsync is connected to EC36-13.
With scope connected to pin13 on Vems printboard I get a fine camsync signal. Why is it missing in the log.
I am pretty sure the previous triggerlog with camsync is recorded after a reboot.
This triggerlog is for sure after reboot.
http://www.vems.hu/files/perboddum/running3.triggerlog
Hello Per,
I have consolidated my request to single block perhaps this clears its up.
When reviewing your scope shots it seems there also is a very slow rise rate on the primary trigger pulses (red signal, ch1) could you set both probes to the same amplification, set the scope Y range the same for both channels, trigger on yellow signal (ch2) rising edge and make a much smaller timebase shot so i can accurately see:
1) falling secondary triggger (yellow, ch2) in center of screen and one or two primary edges (red, ch1) left and right.
2) Lets start out with with following, could you make a triggerlog of engine cranking 8-10 seconds without injector and ignition fuse to verify if the irregular periods are caused by some noise issue.
3) make sure the secondary trigger is enabled and connected to the VEMS when making these.
To summarize, before we have verified correct operation of BOTH triggers, no starting attempts (with fuel and spark) are required (even cordless drill tests are suitable for these scope tests, perhaps easier to perform than cranking the engine all the time).
Best regards, Dave
Hi Dave
Cranking engine triggerlog
http://www.vems.hu/files/perboddum/cranking2.triggerlog
My Vems config
http://www.vems.hu/files/perboddum/setup12.vemscfg
Scope with distributor (connected with 12volt and 1K pullup on trigger+camsync) drill operated.
Hi Per,
The scopeshots look perfect, i still suspect scope amplification difference (1X/10X) resulting in the X 5v/50v range to get same height on screen but lets not worry about that now.
I tried to review cranking2.triggerlog but the file seems to be empty. Could you re-upload or verify ? Thanks.
Best regards, Dave
Hi Dave
You are right. One probe set to x10 and the other x100.
Cranking triggerlog uploaded again
Have updated vemstune ini files from web. Was missing CLT/MAP ign settings. This solved loosing sync at 1500 rpm.
But it seems I still don't get camsync into triggerlog and vemstune signals Less sec.trig.
Trigger with camsync falling.
http://www.vems.hu/files/perboddum/cranking3.triggerlog
Trigger with camsync rising
http://www.vems.hu/files/perboddum/cranking4.triggerlog
scope of both signals during cranking. Watch the camsync.
Why does it behave like that when connected to vems. Both signals comes from the distributor Nissan hall unit.
Hi Per,
Great to hear updating ini file manual update provided you with the correct ini files to match your firmware (as recommended on GenBoard/UnderDevelopment/FirmwareChanges).
I have no idea whats going on with your secondary trigger signal, it seems some kind of noise superimposed on signal. Secondary trigger also does not show up in the triggerlog(s). My suggestion would be to disable secondary trigger for now, you are running distributer ignition anyway.
Make sure to check if your tdc calibration is set up correctly (with timing light) before attempting any further tuning.
Best regards, Dave
I still have problems with enabling camsync.
We have changed from 18K to 2.7K pullup and the scope signal is now good on pin13, but no signal into Vemstune.
We have tried firmware 1.2.23 and 1.2.11 but no signal in Vemstune.
It seems that some components for camsync (hall) is missing according to Peter Jensen.
Help needed please
16/3-2014
Hi Per
It seams you need a resister on your board.
This resister is a 0 Ohm so you can solder a thin jumper wire and sec. Trigger will work
16/3-2014
Camsync now gets into vemstune after soldering the jumper wire
16/3-2014
Trigger sync request for 12-1
Issue using pri trigger only:
When cranking vems read the pre. Trigger signals nicely but when it start one or two cylinders vems loose sync, see log files her:
http://www.vems.hu/files/PerVolvo/v3.3_u008135-2014-03-15-16.31.04.triggerlog
- sometimes normal tooth is up to 1.62 from previous normal
- most of the time the ratio for missing is 1.55 to 2.40
- THIS MEANS the missing tooth is sometimes shorter (1.55) than the normal tooth (1.62) so this will not work.
- the wheel cannot be rotated so the missing tooth is aligned with TDC compression. So another wheel is needed.
2 batteries have been connected doing all the test on this site so far, but we have to use only one battery at race events.
Also this is a high comp engine and maybe it rotates too slow/quick when it fires up. (see top of page)
Cranking is 70..113 rpm and varies much due to high compression)
Do we need a new firmware low tooth count or what could we try.
Thanks for your time
Peter
Hello Peter & Per,
Could you elaborate to exactly what the problem is you are seeing ? During previous logs which were submitted by Per (admittedly using dual batteries) the engine was (reportedly) running fine.
Cranking such low (and also unstable) rpm is not recommended and very hard to sync to. Try to get cranking rpm up to at least 120-150 rpm stable (and i suspect any issues you might be seeing will be solved).
One possible solution (we actually recommend this for 12-1 in help): http://vems.hu/vt/help/v3/v3_primary_trigger_settings.html, more specific section: "For 12-1 missing tooth wheel" is enabling advanced filtering with recommended thresholds.
Please report if this benefits you.
Best regards, Dave
17-03-2014
Hi Dave, thanks for your time
Note. Camsync is disabled.
There is no way to get the cranking speed higher than 113 rpm, even with dual batteries. Around 113 rpm seems to be startermotor max. (motor is new). Same startermotor is on a Rover SD1 V8 and it is also running at around 113rpm and it is starting fine (has other brand of EFI controller).
When we sometimes gets the engine started (after 10..15 sec of cranking because of sync loss) it runs fine.
The problem is that the sync loss happens just when the engine starts to ignite a couple of cylinders (and rpm rises fast) and then the ignition is cut and Vems starts to resync and the same thing starts all over again.
I have tried many sorts of "Advanced filtering", but it does not seem to have any major impact. I can, however, adjust the filtering so much that I don't get ignition pulses.
Per
Hello Peter & Per,
I will run some tests locally to verify operation during fast rev up from cranking, using 12-1.
Best regards, Dave
17-03-2014
Thanks dave.
Here is the latest log where the engine starts.
http://www.vems.hu/files/perboddum/v3.3_u008135-2014.03.15-13.23.22.vemslog
If I drill operate the distributor and mount a sparkplug to the coil and vary the rpm, I can see that sometimes the sparkplug stops igniting. Vemstune also reports Less Pri trig.
It seems that it is every time the drill changes a little in speed.
Per
18-03-2014
Hi Per,
Your trigger reference tooth table has this sequence:
07654321
03690369
Sorry if I have understood this wrong, but since 0 shows h10 on mouseover, and 1 shows h11, it might be worth a shot to change the table to:
07654321
09630963
\Alex
18-03-2014
The engine runs fine when it finally starts after some 10..15 seconds of cranking
Per
19-03-2014
Thanks for the suggestion for 12-2
I have now tested a 12-2 distributor on the car with a powerdrill and a
sparkplug connected to the coil.
The 12-1 wheel I have taped one hole next to missing tooth so now it is 12-
2.
Vems now seems much happier. It is now very difficult to provoke trigger
errors at (60...200 rpm), even at varying rpm.
Sparks rarely drops out,as it did before (during resync)
12-2 have now been tested in the car. It sometime still have trigger errors when cranking. Each time I start the engine there is 40/60 change of trigger error. When cranking with no trigger error the engine fires up after 3 sec.
The problem seems to be the width of missing tooth. Will I have similar problems if I use a 24-1. 24-1 missing tooth width in relation to trigger tooth is the same as 12-1 ??
In my case I think 24-2 might be better ??
Per
2014-04-01
Changed the triggerwheel to 24-2.
Now the engine fires up within 3 seconds (remember slow cranking) and only on rare occations it looses sync when cranking. I think that now it only looses sync in some 15% of all crank attempts.
Do you have triggerlog when that happens ?
- The actual circumstances would be important to see.
- When last time we tested this pattern (and other related N-1 and N-2 with and without camsync), with bench signals (quartz-precision signal and also with some intentional variations) ignition was very predictable and reasonably fast.
- It would be nice to see the actual pulse-train that results in that behavior (maybe it can be improved at the trigger HW level, eg bigger capacitor, stronger pullup, or HALL-dirac filter).
4/4-2014
We are investigating all kind of trigger wheel we did make and run 24-2 now and seams promising using only pre trig. If we try sec trig it is hard to start and losses serial comm so we scooped the 12v supply near ecu and it make a short drop to 6.5-7 v and I think ecu restarts and therefore it´s also impossible to make a trigger log.
Question: Can we add a capacitor for backing up this voltage drop: size and where to fit it please.
Tnx Peter
Just when the starter motor engages we have a short downwards voltage spike to 6.0..6.5 volts and it then raises to 9 volt after 100ms. This causes the ECU to reset. However the the engine starts after 3 sec of cranking (no camsync)
With camsync the engine will not start.
With camsync and 2 batteries the engine starts fine (no trigger errors)
(Allow sync without camsync=Enabled)
MembersPage/PerBoddum/SierraCosworth - Sierra Cosworth
MembersPage/PerVolvo - Volvo 960 VEMS