VEMS Community Forum
Technical => Triggering => Topic started by: DButters on October 16, 2015, 11:02:33 pm
-
Hi all,
Before I burn too much more time getting this thing going, does anyone know why a rotary engine trigger pattern would fail on the latest firmware?
I copied the settings from my already running rotary to a new one, but used the latest firmware.
I get spark but the timing appears to sync to a different spot each time I start cranking. Seems consistent once I get going.
The trick with rotary triggers is the sectrig pulse is every 360 degrees.
Trigger log is nice and stable, it is here: https://drive.google.com/file/d/0B8c8BE2zEwc1cVduQnN0ei1tSFk/view?usp=sharing (https://drive.google.com/file/d/0B8c8BE2zEwc1cVduQnN0ei1tSFk/view?usp=sharing).
Will also rollback to same firmware as on my other car and see how that goes.
-
Hello Dbutters,
I reviewed your triggerlog, the spark linking up exactly with the sectrig pulse seems a bit odd to me, more likely some crosstalk/capacitive or inductive coupling than anything else.
Could you take a triggerlog of engine cranking 8-10 seconds without ignition and injection fuse and provide a short vemslog of the same sequence ?
After receiving both files i'll investigate your case.
Best regards, Dave.
-
Hi Dave,
Crosstalk where exactly? The ignition events in that triggerlog are those calculated by the ECU, they cannot be crosstalk. Unless I am missing your point.
I have chucked the oscilloscope on and the triggerlog matches what I am seeing on the scope.
I have rolled back to 1.2.11 firmware and things are trying to start, trigger log looks good, I just have fuelling issues now.
So I would say either Rotary trigger is no longer working, or the config needs to be massively different for the triggers and will not work straight after upgrade.
-
Update: I have confirmed their is either
- firmware incompatibility
- breaking config change
on upgrade from 1.2.11 to 1.2.31 with my setup.
1.2.11
-- injectors firing OK at configured pulsewidth
-- ignition firing OK
1.2.31
-- no injection at all, test mode shows operation OK though
-- ignition (only leading coils verified) OK
Can anybody throw any ideas into the ring at this stage? The 1.2.31 firmware seems to reconnect much quicker to the software, and the ignition works so I would like to go with the latest, but no injection is a show stopper.
I hope its just something in the config after upgrade?
https://mega.nz/#!nUcBRLyC!b2LnFM-658owgi0R0LCBlG8skqDewFgraL2n-gL2oYY
-
Hello DButters,
I have not received, please take both with 1.2.31 firmware:
Could you take a triggerlog of engine cranking 8-10 seconds without ignition and injection fuse and provide a short vemslog of the same sequence ?
After receiving both files i'll investigate your case.
I did however review your 1.2.31 vemscfg (v3.3_u004837-A-2015.10.18-20.50.08beforebackto1.2.11.vemscfg), perhaps you forgot to review the output of the config validator ? There are many trigger related config erros pointed out there that need correcting. I was in a good mood today so decided to jumpstart your 1.2.11 to 1.2.31 transition and have corrected them for you, corrected config:
http://vems.hu/dave/dbutters/v3.3_u004837-2015.10.18-11.24.11_dbutters_rx7_corrected.vemscfg
Please take a triggerlog and vemslog (as suggested above 8-10sec without ign and inj fuse after uploading that config and restarting ecu) and upload it here i'll review; Thanks!
Best regards, Dave
-
Thanks Dave - I appreciate your good mood :) Apologies I skipped uploading the trigger, I did take one and it was all clean, will upload later to satisfy your curiosity.
You are right there are many trigger related config validation fails.
But be aware that these are all put in by the 'wankel 2 rotor' defaults in VemsTune, not by me.
And the same config works fine in 1.2.11.
I will upgrade again and try your config out in around 10 hours time. FYI I tried the 'released' 1.2.30 and had the same problem (ignition only no injection) also.
Thank you hugely for your offer of assistance, always great to see the VEMS community active.
-
Hello DButters,
I will upgrade again and try your config out in around 10 hours time. FYI I tried the 'released' 1.2.30 and had the same problem (ignition only no injection) also.
Thank you hugely for your offer of assistance, always great to see the VEMS community active.
Okay, perfect; One thing i might have missed, it is recommended on a setup which has has injectors wired phase sequential to disable injector angle control (put all sites to 720 degree).
Best regards,
Dave
-
Thanks Dave,
RE: injector angle control, I have found it makes a massive difference to idle quality to set this as per Mazda's recommendations, particularly on a bridgeported (huge overlap) rotary. FYI I have spent over 10 years installing VEMS, and a couple of years rotary engine tuning on the 1.2.11 firmware.
Sadly I did not get time last night to test the config changes, though I did get the car running for the first time (both ignition and injection) on the 1.2.11 firmware! Always a good day when an engine first runs, so I decided to leave it at that until tomorrow.
-
Hello DButters,
Always a good day when an engine first runs, so I decided to leave it at that until tomorrow.
Understandable , indeed its nice to end the day on a high note. Will review when the files (or questions) present itself.
Best regards, Dave
-
After a bit more running on 1.2.11, upgraded to 1.2.30
I also found I was running an out of date VEMStune, the newer one was installed but the desktop link was to the 2014 version.
So my pick was the old vemstune not downloading correctly to the newer firmware, which is reasonable.
So now with newer VEMSTune, got it running on 1.2.30 for a couple of short tests.
triggerlog
https://mega.nz/#!CEc3haxB!76jwpibrcnvnL7KH7dKhb-OBm6djmMejIrfM5MdCaeo (https://mega.nz/#!CEc3haxB!76jwpibrcnvnL7KH7dKhb-OBm6djmMejIrfM5MdCaeo)
config
https://mega.nz/#!mBcxRIiJ!P9efc8tEt4ZPhvdQMvh5dc__R2Y0Y022F8C0YFYD-CU (https://mega.nz/#!mBcxRIiJ!P9efc8tEt4ZPhvdQMvh5dc__R2Y0Y022F8C0YFYD-CU)
running log
https://mega.nz/#!XE9TCBKK!EGHnt8I_UVAccmjOA2Z_9WzXjOLANiuaGmr_O1-bsrs (https://mega.nz/#!XE9TCBKK!EGHnt8I_UVAccmjOA2Z_9WzXjOLANiuaGmr_O1-bsrs)
However I am terrified to use it, the two times I did the ECU locked up at some point and froze all outputs on, LCD filled with garbage, filled my engine/exhaust/floor with fuel. Never saw this in a couple of days of testing on 1.2.11.
I see others in the forum talking about the same problem... seems to be when it goes from running to non running state.
Any suggestions anyone?
-
Hello Dbutters,
However I am terrified to use it, the two times I did the ECU locked up at some point and froze all outputs on, LCD filled with garbage, filled my engine/exhaust/floor with fuel. Never saw this in a couple of days of testing on 1.2.11.
I see others in the forum talking about the same problem... seems to be when it goes from running to non running state.
The outputs freezing is is something that should never happen, that is very strange! I have reviewed your configuration (from vemslog) and there a lot off strange settings still present, i have corrected these for you and uploaded a corrected config for you here:
http://vems.hu/dave/dbutters/v3.3_u004837-2015.10.22-10.34.01_dbutters_1_2_30_corrected.vemscfg
Please try this on car to see if it behaves, there should not be any issues. Could you upload a short vemslog of the car running with this corrected config after ? Thanks!
Best regards, Dave
-
After a saga of failed firmware updates, config uploads and flooded engines, I am back on the 1.2.31 firmware.
Dave, the config you sent me gave me all outputs OR'ed, so I recreated as best I could and see no config reason for:
Same issue - car runs nice but LOCKS fuel pump AND injectors on when it stops running!
It does not seem to do it if the car does not fire off cranking, only if it goes from running to non-running.
This is exacerbated by the fact that I am outside the car with a remote start switch so it gets a good 5-10 seconds full on before I can turn the key off.
I wonder if nobody else notices this just because they are turning off the key to stop the engine rather than it dying off while tuning? Or if there is still a gremlin in the config, which I have gone through page by page multiple times to no avail.
Special favours to the person that can help me figure this one out.
Config:
https://mega.nz/#!SEFUEQpZ!poshbtJfhGFZISje85hAIPqcB7eoHehpjaQJ7EKtoGI
Log:
https://mega.nz/#!PYcSzQaT!aHHyKz1LyNgbAPbYxP9P6P6M9eoFN3qcd9B44VEE8QU
-
Hello Dbutters,
I have reviewed the changes you made compared to my provided 1.2.30 config (yours is 1.2.31) not sure what happened there, but there are many configuration problems.
To rule out some things lets reduce it to the minimal function set, so please disable everything that is not needed for basic running or has bad configuration now:
* prime pulse (set to o)
*staging (tps = 100, map = 510, rpm = 25500)
* both wheelspeeds (enter speed sensor calibration = 0 for both)
* injector angle control (set all curve angles to 720 degree, apply configlet inj_curve_no_cam_sync )
* gear dependent retard, set all sites to 0 degree
* fill clt/map ign retard table properly, fill with full 0
Could you make these changes listed above (or apply my config to your ecu), than save the configuration (no need to start the car) and upload it so i can review if the offending sections are correctly configured after ? Thanks!
Best regards, Dave
-
Config modified, burned to ECU and checked using 'config validate' in VEMSTune (I forgot the cmdlet to download the config from the device)
https://mega.nz/#!eIVClQBb!KZnXpP0g6JHxBFtS2ehI0nSbvqn1zXTIw-uH8EAsJgQ
Will try running this in an hour or so and post results.
-
Hello Dbutters,
I have reviewed your config: v3.3_u004837-A-2015.10.27-14.44.04_1_2_30_cleaning_2.vemscfg, its moving in the right direction, just a few leftovers that need correcting:
* accel enrichment bins dTPS, adjust the lowest bin from 2 to 6. second 10->13.
* shiftcut disable cut spark / cut fuel; activates below adjust from 2.16->1.2; activates above adjust from 2.49->5, min tps adjust to 100 and min rpm adjust to 25500 (never engages)
* launch rpm curve adjust left bins so highest speed is on top and monotonous decreasing towards lowest bin.
Apart from that you should be good to run some tests on car. Definitely adjust accel enrich related before trying on car, preferably fix all 3 mentioned above.
Best regards, Dave
-
Thanks Dave - back in the country now so will try this out. My last tests (1 loooong hour ago) still showed the same flooding and fuel pump on issue.
-
OK, one of those last changes did the trick - no more horrendous flooding! Not sure how they caused such a symptom, but I'm just happy I can start working on the other problems with the install now.
Cheers Dave!