How will I set my req_fuel, I get different answers from different sources.
MegaSquirt calculator http://www.megamanual.com/v22manual/mfuel.htm (http://www.megamanual.com/v22manual/mfuel.htm) gives with inputs
Engine Displacement (cubic inches) 183
Engine Displacement (cubic inches) 4
Number of Injectors 4
Number of Squirts 2
Injector Flow (lbs/hour)
per injector 86
Injector Staging alternating
Fuel Type Gasoline
an answer of
Req_Fuel
(ms) 5.6
Downloaded
Req_Fuel
(ms) 5.6
And if I change Injector Staging = simultaneous --> Downloaded
Req_Fuel
(ms) = 2.8
OTH PhatBob manual http://www.vems.hu/wiki/index.php?page=MembersPage%2FPhatBob%2FUserGuide (http://www.vems.hu/wiki/index.php?page=MembersPage%2FPhatBob%2FUserGuide) :
req_fuel = 6.49 * (D / N / I)
I injector flowrate (cc/min)
D engine displacement (cc)
N number of injectors (eg. 4 for a 4 cyl port injection)
req_fuel = 6.49 * ( 3000 / 4 / 903 ) = 5.4
No separation between batch or sequential or which divider.
I'm using divider=1 and sequential injection.
Finally I'd like to confirm that when using Low-Z -injectors ( 2.35 ohm ) with ballast resistors and the box has 30V Flyback, then PWM will be disabled like this:
*Inj open time should be 0,
*Inj rampup_battfac must be 4080 (which diables the feature)
What are right values ( or a good starting point ) in this window:
(http://www.vems.co.uk/VEMSUserManual/InjectorSettings.png)
PS. How in h*** one marks a topic as new??
You can see threads as new if you dont post on them.
From the top...
Go with req_fuel from my manual, we have a different injector strategy to MegaSquirt
Have you really got a 3000cc four cylinder engine?
divider=1 and sequential injection is fine.
Inj open time @ 13.2V should be zero
Inj effective rampup try starting with 1000
Inj rampup_battfac should be 4080
battfac try 512
Injpwm DC should be 100
Injpwm DC added @6V can be anything as it cant add to 100%
Injpwm peak time can be turned off by using 25.5
Quote from: [email protected] on November 28, 2007, 08:50:54 PM
Have you really got a 3000cc four cylinder engine?
Yeah, those crazy Porsche engineers... ;) ::) ;D ;D
I really had no idea... can't wait to see the results of your work :)
OK, now it started like never before...but only afted I doubled the req_fuel from 5.4 to 10.8 . Although then A/FR was 10-11:1 even after warm-up enrichment didn't interfere.
I made the VE-table with tools: Generate table and used these ( hopefully someday's ) figures: 500lb-ft/4200rpm, 500hv/6200rpm, max 7200rpm @ 2,0bar = 300kpa. Top row has then numbers in excess of 160, so I can't just double the tables' values ( >>It is common practice to then half this number and to double the values in the VE table, as this means that the VE table's resolution effectively doubles and allows finer grain tuning.<< ) to get req_fuel back to 5.4 .
Highest PW I noticed was 1.3ms @ 2000rpm, @ idle it was 0.9 . So there is altogether something to do, but what?
It started like never before? Where did it fall over before? Dont forget that req_fuel has no effect on priming cranking afterstart.
If you want to double values start there, followed by the warm-up enrichments. Your engine is only in that area for a very short time, so you should really concentrate on getting the req_fuel and VE numbers right for the normal warmed-up operation of the engine.
Also bear in mind the effect of the target lambda decrease the number where you are at 160VE and you'll find that the PW increases and that means you can bring the VE value lower.
Quote from: [email protected] on November 29, 2007, 09:52:40 AM
It started like never before? Where did it fall over before? Dont forget that req_fuel has no effect on priming cranking afterstart.
How come then if I don't change anything in priming/cranking/afterstart and play just with req_fuel it either doesn't even promise, just starts but then dies, starts but dies a bit later sunning stumbling or starts and finally idles like it's going to win LeMans 24hrs... Req_fuel amounts are then ( about ) 5.4-> 7 -> 8 -> 9 -> 10.8
Quote from: [email protected] on November 29, 2007, 09:52:40 AMIf you want to double values start there, followed by the warm-up enrichments. Your engine is only in that area for a very short time, so you should really concentrate on getting the req_fuel and VE numbers right for the normal warmed-up operation of the engine.
Yes, this is the goal. I just tried with doubling the VE-table and halfing the req_fuel. When warm, I had to subtract 15 units from VE-table to get A/FR near stoichiometric. Still PW remained below 1.0. OTH Inj effective rampup was still 1000. I haven't yet tried to play with that, it's next thing to do.
Quote from: [email protected] on November 29, 2007, 09:52:40 AMAlso bear in mind the effect of the target lambda decrease the number where you are at 160VE and you'll find that the PW increases and that means you can bring the VE value lower.
Quote from: pete95zhn on November 29, 2007, 12:40:49 PM
Quote from: [email protected] on November 29, 2007, 09:52:40 AM
It started like never before? Where did it fall over before? Dont forget that req_fuel has no effect on priming cranking afterstart.
How come then if I don't change anything in priming/cranking/afterstart and play just with req_fuel it either doesn't even promise, just starts but then dies, starts but dies a bit later sunning stumbling or starts and finally idles like it's going to win LeMans 24hrs... Req_fuel amounts are then ( about ) 5.4-> 7 -> 8 -> 9 -> 10.8
Then you must have cleared the afterstart phase then! What sort of warm-up enrichment value does it settle on?
Quote from: [email protected] on November 29, 2007, 01:34:01 PM
Quote from: pete95zhn on November 29, 2007, 12:40:49 PM
Quote from: [email protected] on November 29, 2007, 09:52:40 AM
It started like never before? Where did it fall over before? Dont forget that req_fuel has no effect on priming cranking afterstart.
How come then if I don't change anything in priming/cranking/afterstart and play just with req_fuel it either doesn't even promise, just starts but then dies, starts but dies a bit later sunning stumbling or starts and finally idles like it's going to win LeMans 24hrs... Req_fuel amounts are then ( about ) 5.4-> 7 -> 8 -> 9 -> 10.8
Then you must have cleared the afterstart phase then! What sort of warm-up enrichment value does it settle on?
WUE is flat 140% from -40C to +5C, then linear to 0 at +70C. A bit rich though, but engine runs. I'll adjust it when this PW issue is solved. If I multiply VE-table by two ( ignoring boost area ) and divide req_fuel by two, PW drops to 0.3, increasing to 0.7 @ ~1600rpm. I played a little with Inj effective rampup, it's now 400. I'll get it lower, but I ran out of gas just when I rememberd that I can datalog all this...
IAC ( Bosch pwm -style ) is another question. It seems that I have to use pretty low % to control idle. 30% gives 1300rpm idle, 20% ~1000rpm...TB ( 65mm ) doesn't apparently close fully. Engine idles now @ 57kPa, which is better than before ( ~70 ).
OK, I'm getting now pissed and I have to admit that I have looked for fully commericial EMS's...
*Whatever I do, I can't get the PW higher I've earlier written. I've tried req_fuel from 5.4 to 22.0 with matching VE-table, engine idles and can be tuned close to 14.7:1 A/FR, but PW stays well below 1.
*It also doesn't rev above 2k, dies just below. Acceleration speed has no effect. I have checked all settings but haven't found yet any bugs, because it looks like ther's some kind of limiter on...or the trigger just can't handle higher rpm. Acceleration enrichments have little effect.
*One issue, probably related to previous ones is that when trying to open the thottle very slowly ( and if it doesn't die ), A/FR goes to very rich ( 10:1 with black smoke ), rpm down to 800 and intake pressure to 90 kPa.... ie it idles stumbling and shaking. Even after closing the throttle again this continues for 3-5 seconds, then is suddenly jump back to 1300 rpm/65 kPa/14:1 A/FR. Only good part with this is that EGT goes down too...
Is it possible that all this is related to some hardware issue? Can I check eg PowerFlyback's funtionality? My ballast resistors are 6R8's ( 6.8 ohm ) and injectors 2.35Ohm.
Its clearly chucking more fuel in, else you wouldnt get black smoke out...
Please download your configs and tables and send them to me. What Firmware are you running?
Rob
You haven't got your overrun fuel cut set to something like 2000rpm have you ?????
Hilly
Failing the full config and tables can you post a picture of the the warm-up enrichments.
We were looking at this and one thing sticks out. You wrote: WUE is flat 140% from -40C to +5C, then linear to 0 at +70C.
We argue about what this means but I get the impression that you have 0 in the 70C cell. That means that you remove all fuel and that the engine only runs on some end of scale rounding error.
You should have 100% in the 70C bin. That means that the warm up enrichment is turned off.
Jörgen
Quote from: hilly on November 30, 2007, 06:24:41 PM
You haven't got your overrun fuel cut set to something like 2000rpm have you ?????
Hilly
IIRC no, but I'll check.
Quote from: Jorgen on November 30, 2007, 06:51:43 PM
We were looking at this and one thing sticks out. You wrote: WUE is flat 140% from -40C to +5C, then linear to 0 at +70C.
We argue about what this means but I get the impression that you have 0 in the 70C cell. That means that you remove all fuel and that the engine only runs on some end of scale rounding error.
You should have 100% in the 70C bin. That means that the warm up enrichment is turned off.
Jörgen
Sorry, misprint, it should have read 100%, that's what's in that cell.
Quote from: [email protected] on November 30, 2007, 05:26:13 PM
Its clearly chucking more fuel in, else you wouldnt get black smoke out...
Please download your configs and tables and send them to me. What Firmware are you running?
Rob
Firmware is 1.1.24. Config & tables ( although for batch fire ) can be found at my Wiki memeber's page http://www.vems.hu/wiki/index.php?page=MembersPage%2FPeteKrgr (http://www.vems.hu/wiki/index.php?page=MembersPage%2FPeteKrgr). Latest ones I'll send tomorrow.
Quote from: [email protected] on November 30, 2007, 06:49:15 PM
Failing the full config and tables can you post a picture of the the warm-up enrichments.
E-mail sent.
http://www.vems.hu/wiki/index.php?page=MembersPage%2FPeteKrgr%2FConfig
Areas i would be looking at
tps_low=22
tps_high=FF Both these need calibrating for the TPS sensor, is your sensor really 255 at 100% WOT??
Oh and if you are running batch fire the table generator will be miles out. The numbers it seems to produce are huge in relation to what is actualy required.
Knock back the VE values where you are having the problems.
You could try the VE learn mode and set it agressively so that it pulls the VE values back quickly.
Quote from: Sprocket on December 01, 2007, 04:30:49 PM
Oh and if you are running batch fire the table generator will be miles out. The numbers it seems to produce are huge in relation to what is actualy required.
Knock back the VE values where you are having the problems.
You could try the VE learn mode and set it agressively so that it pulls the VE values back quickly.
No, I'm not running batch fire anymore. I will correct tables to Wiki later. I had to go and re-calibrate the TPS again, that 255 wasn't right. I remember I had something like 224 and it now gave 215 at WOT and still 34 at closed position, TPS clicks when closing. Should it be 0 when fully closed?
EDIT: There seem to be one real and big mistake, ie. duration of acceleration enrichment. I have set it to 25
ms but now I was told that there's misprint in MegaTune and it's really 25
s, so no wonder why it runs very rich while trying to move the throttle...!!
Even if your not running batch fire, the table generator still returns huge numbers
Looking at the VE table on your Wiki page your VE at idle is around 40 and in the part throttle mid RPM range its 128, that is a very large hike, if you are over fueling, this i feel is the area you should be looking at.
Also, low ohm injectors with the resistors require a reasonably lagre ramp up time. I recon this is playing a major part in increasing the flow rate of the injectors with them being sluggish open and close. I suffered similar problems when trying to get my engine running and the VE table was miles out!!! almost double!! The injectors and the resistors work and work well, you just have to tune for it.
This is the current VE table:
(http://www.9xxfin.com/forum/files/ve_table_pieni.jpg)
I have to correct these too to Wiki too.
EDIT: I have to correct these too to Wiki, but only after they're more correct. This table gives 12:1-14:1 idle
Your engine idles at 50 kpa?
Quote from: Sprocket on December 01, 2007, 06:17:41 PM
Your engine idles at 50 kpa?
Yes, unfortunately...big valved head ( 48mm in /45mm ex ) etc. No found leaks, though. Before modifications idle kPa was around 35-40.
Personally I'd have a line below idle at 40kpa.
I've got your setup and will do some tests later.
Rob
I noticed one peculiarity when looking throuhg one datalog with MegaLogViewer. There reads " Fuel Pressure 5 bar". Where does it get that figure? I'm using 3 bar and if the box really thinks that used pressure is 5 bar, it's really significant difference and IMO causes short PW ( thus requirement to use double req_fuel ) with other given numbers being accurate.
I haven't noticed either any field with aquestion of used fuel pressue or requirement to use fuel pessure sender. Is this 5 bar pressure some kind of basic setting in all firmwares or can it be changed? Or just MegaogViewer error?
How about divider then? I have all the time used Divider = 1 with both batch & sequential fire, but here's a 2.3l turbo Saab with almost similar VEtable, injectors and boost but he's running batch with Divider = 2 and req_fuel 4.1 ...
With batch fire and a divider of 1, all the injectors fire every engine event, 4 times per cycle. setting the divider to 2 will halve the injection events resulting in the injector firing only 2 times per cycle, but the injector squirts twice as much (required fuel value is doubled).
One thing i am unsure of is what the required fuel calculation is calculating a figure for. I think it is a figure for sequential injection where the injector fires once per cycle, if you then decide to run batch fire with a divider of two, i would think the required fuel value should be halved, or batch fire with a divider of 1 the required fuel value should be halved and then halved again. This however, depending on how you have sized your injectors will determine how big or small the pulse width is. If you sized your injector to be big enough to provide all the fuel in one squirt (sequencial), the same injector will be massively oversize when running batch fire with a divider of 1. If your calculated required fuel pulse width is already small, running batch fire with a divider of 1, the pulse width will be quarter of that, and idling will be problematic????
There is no reason not to run sequential injection, as long as you have wired the injectors to individual output channels on the ECU.
The required fuel value is the figure the algorythms use to determine the injector size, thus it can determine the size of the pulse width to provide the required fuel for the current operating condition.
This theory is a little unclear and i cannot find anything in the Wiki about it. Needles to say, i was untill recently running a divider of 1 on what would be concidered batch fire, i had to halve the calculated required fuel value. after some thought and a little experimentation, I now run a divider of 2 and i now use the calculated required fuel value, with the VE values the same, still achieving the same fueling
I hope i havent confused you too much, LOL :D if i have, just ignore what i said, lol, im just brain dumping :D
Ignore the fuel pressure value.
If you have no fuel pressure transducer, then the ECU will not be reading anything.
Hi,
It sounds like your setup is messed up. I would normally have sent you a config that would get you started but as you use the auditrigger setup in an unsupported application I don't have a setup for you.
Could you post config.txt, tables.txt and the msq from the megatune you are running?
We all prefer to look at the setup in different ways and if you prepare a fresh package with all of them you will be more likely to get someone to review and find the problem of your setup.
It's not unlikely that your problems are related to the Porsche trigger.
Jörgen
Sami pointed out that if you use a 1.1.23 Megatune with 1.1.24 and up firmware your PW will not show the right value. The value was 8bits in 1.1.23 and it's 16 bits in 1.1.24.
I have also seen that you use 16X14 tables, that is not recommended. It's only used in Finland and getting help outside finland will be hard. 12X12 is more then enough with the type of fuelcalculation we use. It's different with more primitive fuel calculations that is used by many other systems, or when you use Alpha-N which is always equally primitive.
Jörgen
Quote from: Jorgen on December 02, 2007, 07:44:33 PM
Hi,
It sounds like your setup is messed up. I would normally have sent you a config that would get you started but as you use the auditrigger setup in an unsupported application I don't have a setup for you.
Could you post config.txt, tables.txt and the msq from the megatune you are running?
Well, it could be very much so that I'm the first ( or second ) one trying to use original triggers with v3.3 and there's no setup ready. Altogether the original triggers setup is very close to Audi's, the significant difference is no cam sensor and different tooth wheel count at flywheel. Both are originally run with Motronic.
Here's a lot of this and that: http://www.vems.hu/files/MembersPage/PeteKrgr/VEMS+check+files.zip
Quote from: Jorgen on December 02, 2007, 07:44:33 PMWe all prefer to look at the setup in different ways and if you prepare a fresh package with all of them you will be more likely to get someone to review and find the problem of your setup.
It's not unlikely that your problems are related to the Porsche trigger.
Jörgen
Quote from: Jorgen on December 02, 2007, 08:40:22 PM
Sami pointed out that if you use a 1.1.23 Megatune with 1.1.24 and up firmware your PW will not show the right value. The value was 8bits in 1.1.23 and it's 16 bits in 1.1.24.
I have also seen that you use 16X14 tables, that is not recommended. It's only used in Finland and getting help outside finland will be hard. 12X12 is more then enough with the type of fuelcalculation we use. It's different with more primitive fuel calculations that is used by many other systems, or when you use Alpha-N which is always equally primitive.
Jörgen
I got info about modifications from Sami, I'll do those and report findings. What's the problem with 16x14?
I know that there are few guys waiting for my results before they decide which EMS they'll run. IMHO this setup is worth supporting and developing. Although I'm not a genious with programming, the work I'm done ( the questions and very valuable answers from the VEMS community ) will help others in the future. Market in the US is much bigger and they're really in the need of affordable EMS, if teething problems are solved.
Pete
Solving teething problems is what we do :)
OK, at least the PW -problem is now solved. I changed Vemsv3.ini-files in 1.1.23 Megatune mtcfg for same ones from 1.1.24 and PW @ idle is around 3.0, without any other modifications. It's now easier to start solving remaining questions.
So your next post will be a picture of you smiling and holding a dyno sheet ;)