My project is a 1987 Mercedes 2.3-16v, motor rebuilt with 11:1 compression and a few other goodies. Here are the specs of the Vems install:
-380cc Bosch injectors
-Modified Saab fuel rail with stock Bosch FPR
-BMW E36 325i throttle body with TPS
-BMW coil
-GM 1 bar MAP sensor
-Webshop MAT and CLT sensors
-Custom intake plenum and intake
-Custom wiring harness
-GT101 picking up 2 bolts on the dampener, stock distributor
I'm using 1.0.73 firmware. Everything is working well after some issues getting firmware loaded. I tried uploading the newer firmware 1.1.70, but the ecu froze in bootloader mode again and again with few different versions of Vemstune I tried (through 4-28). So I'm going to stick with 1.0.73 for now.
I've tried getting the car started but it's progressing slowly. I've figured out the req fuel at 9.82 and I'm working with the cranking/afterstart settings now. If I set the cranking rpm threshold to 2000 (cranking advance set to 12), the car starts right up and idles fairly smoothly around 1100 with a pw of about 5.8. The firing while cranking setting is set to alternate banks.When I bring the cranking threshold down to 300, the car starts right up then stumbles badly right away around 500-700 rpm. Ive tried raising the ve until the pulsewith is around 6, just like during cranking, but it still stumbles. I have the injectors set up sequentially:
0 to 4
1 to 1
2 to 2
3 to 8
divider set to 1, alternate bans from h[0] to 3..0
I have the spark table in all the idle areas set to 12, just like cranking, and the ve is 55 in all the idle areas, so it should run just like when the cranking threshold is set to 2000 rpm. Everything is disabled as far as ALS, spark idle control etc. What's the best way to get the car running on the normal tables? Can I disable the afterstart settings by putting in zeros for the settings (don't really know what to put in there)? I have the afterstart settings set to 50 engine cycles at the moment.
If the car stumbles after cranking then its your afterstart settings that you need to work on.
Watch the injector PW as the engine catches and see how quickly it drops down to the idling PW, then experiment with the afterstart settings.
It helps if you can get the engine warm enough to idle on its own - then you'll know the target PW that you're aiming for, and will then be able to work on the afterstart settings in order to smooth the transistion from cranking to idle, I'm betting the transition is abrupt.
Thanks Rob! Got the car idling reasonably! The calculated req fuel is 9.8, but I bumped it up to 13, with 72 in the VE and .98 in the lambda around idle. The pw is around 4.2-4.5, and the map is in the high 40's with a 950rpm. Does that seem reasonable? Haven't hooked up the wideband yet as I didn't want to foul it while I got the car going.
I take it I should stick with the proper req fuel (especially for 1.1.70), so then should I bump the ve higher? 72 already seems high to me for idle...
I tried the 6-3 nightly build of VT and was finally able to upload 1.1.70!! Previous versions of VT always gave me a bootloader error, so I have been using 1.0.73, which was also acting up a bit (like 2 cells of the ve table kept changing to smaller values by themselves repeatedly). I haven't ran the car on 1.1.70 yet, and the wideband calibration crashes the 6-3 nightly version, but it looks promising. A strange thing is that earlier versions of VT don't recognize the ecu... (I tried installing/using 3-8 and 4-28, in different folders)
Tried to start the car today on 1.1.70, it wasn't successful. Everything seems in order except the coil/ignition. The car idled fine under 1.0.73, but 1.1.70 is blowing the ignition coil fuse within a few seconds. I changed the "coil charge time map scale" from 2 to 0 and it seemed to help a bit although I can't be sure as I haven't cranked too much since changing that. The coil charge time at 13.2v is 2.43 (it was 3 under 1.0.73).
I checked that TDC after trigger is good with a timing light, and it is. The car seems to be backfiring when it does spark/tries to start. Trigger signal is good, and around 200 rpm is shown when cranking. I think it has something to do with the trigger/ignition settings, but everything seems set up right to me.
Once again it's a single coil triggering off pin 35 with the stock distributor, and 2 bolts on the crank with a GT101. Have a look at the screenshots please.
(http://i281.photobucket.com/albums/kk209/bartxb/Vems/primtrig.jpg)
(http://i281.photobucket.com/albums/kk209/bartxb/Vems/ignset.jpg)
(http://i281.photobucket.com/albums/kk209/bartxb/Vems/reftooth.jpg)
(http://i281.photobucket.com/albums/kk209/bartxb/Vems/ignout.jpg)
(http://i281.photobucket.com/albums/kk209/bartxb/Vems/ignout2.jpg)
Use 4 ignition outputs in the table, just define the same one in all the 4 positions.
Also, "Number of teeth on wheel = 4".
The trigger tooth table should read, from top to bottom : 0, 3, 2, 1
Hint: Don't give fuel to the injectors until you know for certain that the ignition hits it's marks on all cylinders, my advice should make this work but it's so not fun to soak the engine when the ignition is screwed up.
Try 1.1.74. 1.1.70 has a low-rpm cranking bug, we have similar blowing fuse at start or stop with bmw m50.
Did you get it running with the settings Mattias wrote?
is it still suposed to be 180degrees toothwith, ?
/Skassa
I've been busy the last few days, but yes, the settings worked great! 180 degree tooth width. Thanks Mattias!!!
I haven't tried the newest VT yet, but hopefully the wideband calibration won't crash so I can start tuning! :-)
The ignition fuse hasn't blown since the new settings, what are the major advantages of 1.1.74 over 1.1.70 other than the low rpm bug fix? Should I upgrade even though the fuse is okay? I've had such bad luck upgrading FW until the nightly versions of VT came out that I'm a bit hesitant to do it unless there is a compelling reason.
Okay,
Great news,
It´s very eacy to upgrade from 1.1.70 to 1.1.74, nothing like upgrading from the old 1.1.44,
tools , firmware, firmware upload, then pick 1.1.74, upload.
After that you upload your own 1.1.70 config, and its like the old setup, it says found 2new table use default, press yes.
Job done...its a 3min thing :)
/Skassa
I've had a bit of a break from getting my car running due to life :), but now I'm back on it! I'm using 1.1.70 FW. I downloaded the latest Vemstune (10-19-10) and tried to calibrate the WB02, no luck. At least Vemstune doesn't crash when I open the calibration box, but it doesn't do anything either. I must be doing something wrong...
-I've double checked that WB02 is wired correctly to the EC18 per Phatbob's guide.
-I have 12v going to the WB02 when the Vems is on verified with a DVM (key in run position, engine not running).
-I have the sensor connected to the harness and outside the car (not in exhaust).
-I have the sensor body grounded to the chassis.
-When I turn the car on, after a short while I can feel that the WB02 sensor body is getting hot (more like really warm, I can still hold it fairly comfortably in my hand).
I turn the ignition to the run position, Vemstune starts reading my sensors etc after a few seconds and then I open the WB02 calibration screen. I hit start calibration and nothing happens. I wait a few minutes and the 02 gauge still reads 0.0 . The calibration constant in the window is 190.
Shouldn't the 02 guage read something close to the ideal, or at least something other than 0.0 a few minutes after starting the calibration? Maybe the various values in the WB settings window under general settings are wrong... Help doesn't seem to have any suggested values for them, I believe I'm using what Mattias has in his config as that's the one I loaded as default.
Do you mean the sensor gets hot while the engine is running and the sensor is hanging loose outside the exhaust ?
That is correct. With the engine off, the sensor should not get hot until you start the calibration or the engine is running (rpm > 0), that's when the heater starts.
You can use the default settings in the dropdown box at the bottom of the wideband settings dialog.
The sensor gets hot after I start the calibration. I double checked the various wideband settings in the base settings drop down and they are the same as in your 1.1.70 config file. Default is selected at the wideband settings drop down as well.
The WB02 calibration gauge still shows 0.0% after a few minutes. It doesn't seem to matter what number I put in the calibration (190 is there to start with). Could the WB02 sensor be bad? Maybe the heater works but the sensor is bad. Is there any way to test it?
Someone must have a separate wideband controller that you can test your sensor with (free air calibration).
I went ahead and bought another LSU4 at a local parts store, Bosch 17014, only $42.
I plugged it in, grounded it, turned the ignition on (car not running), all my gauges/values came up in VT. I went to WB02 calibration in the menu, hit "start calibration", and... nothing but 0.0 on the calibration screen after a few minutes of waiting. The sensor felt warm after a short while so I guess the heater is working. So I guess my original sensor was ok! Well I have a spare now.
What is the next step in getting the WB02 calibrated? I've checked the wiring a few times, all correct per Phatbobs instructions. I've checked continuity on the 4 wires between the EC-18 and the WB02 plug, good. I've checked that there is voltage to pin 3 on the WB02 plug (12.7v).
(http://i281.photobucket.com/albums/kk209/bartxb/Vems/WB02settings.jpg)
i had similar problem...whatever calibration number i entered it couldnt get it to read more then 13-15 i think,but after few times trying it just went to ~21 with same numbers?!
Here are the WB02 specs as supplied when I purchased VEMS:
Board_version=v3.3
pump-= 4.0V
wbo2_pump_pw_zero=0x64
wbo2_nernstdc_target=0x86
pump+= +2.1mV/200 Ohm
AREF=4.9V
(C103=1nF, no D100)
I've updated the firmware a few times and got stuck in bootloader mode a few times as well requiring a hard restart, would these settings have survived? Is there some way to check what the current settings are through VT? Are they needed anymore, or is it handled differently in VT now?
Are Phatbob's WB02 calibration and setup instructions still valid? Can I calibrate it with that method running 1.1.70 FW?
Getting the WB02 calibrated/working is the last hurtle to cross!
In the ECU calibrations dialog :
wbo2_pump_pw_zero=0x64 = 100
wbo2_nernstdc_target=0x86 = 134
http://www.vems.hu/wiki/index.php?page=GenBoard%2FManual%2FWideBandHardwareTest
Try this, cheers andré
Ps do this at sensor connector not the ecu,
Because you know also your wiring is ok.
Thanks guys!! Turns out that 255 was in both the pump zero and nernst figure windows! Silly I'm sure to the more experienced! So, for others who may read this at some point, all the WB02 settings are not in the WB02 window, the two critical ones are in the ecu calibration window.
Now on to getting it running well! I've managed to set up the idle pretty well at this point, its steady, MAP values in the low 30's, and the lambda is good. What a difference having the WB02 makes!
I've set the req fuel (9.8 : D=2,300cc, N= 4 cyl, I=380cc/min), the lambda target table, and the spark table. Both of these tables were adapted from someone with the same car/similar motor running VEMS, but using 1.0.73. Should be a pretty good starting point.
As I understand it, having the above items set, now I start tuning to the desired lambda only increasing the fuel via the values in the VE table. At idle the values are around 80, and at 2k rpm with no load it seems to need around 110 in the fields to have around the proper lambda (injector pulse width is around 3.5 or so I think). Is this the proper way to go? So if I need 160 or 220 in the fields for the proper lambda it doesn't matter as that's what's required, or should I change something so that the VE values stay under 100?
The one main reference I'm using is Mattias' config, and his VE table doesn't go above 94, and it's mostly in the 60-80 range. Just making sure I'm not going about things incorrectly.
Also in the injector settings dialog, what is recommended for the "effective rampup" and "voltage compensation" during tuning and after? How are they adjusted and what is the procedure? I can hear the sound of my motor change slightly at idle if I change the rampup values, but it doesn't seem to effect the lambda or rpm noticeably.
The WBO2 settings dialog is just for the software PID control, and is general settings that work with any ECU and sensor.
All three WBO2 calibration values, not just the two mentioned but for the sensor too, are in the ECU calibrations dialog.
The engine in my config is a 2 valve/cyl engine with poor head flow and aggressive cams compared to your "formula 1" engine - this is why you see much higher VE values in your config, and a much flatter map as well since you don't loose as much VE at low rpm and idle. You also have to consider that I might have underrated my injector flow, and you might have overrated yours. Fuel pressure and actual measured flow will be factors that affect the value of the calculated req fuel.
Use the values from my config for the injector settings, you might use 0 for the effective rampup actually. It's a lot more complicated than it seems but really works good enough with these settings, read the "F1 for help" text in the dialog. I'm not the only one that wish for a better scheme for the injector opening/closing, it might be one of the firmware updates the coming year. With larger injectors, this stuff gets more critical if you want to get the idle and low load sorted out.
Nice project!
Do you have photos of the modified Saab fuel rail and intake plenum? A couple friends are also interested in converting their 2.3-16 to EFI.