For cars with ITBs, or extreme cams the Manifold Air Pressure (MAP) cannot be relied upon to give a good signal - the resolution is poor and the pressures fluctate wildly. So its common practice to run setups like this as AlphaN where the Throttle Postion Sensor (TPS) is used instead of the MAP sensor, this gives very sharp throttle response and smooth running.
AlphaN is a pain to setup, so for this reason VEMS ecus tend to be supplied using Speed Density MAP sensor to make starting the car and initial tests easier. So once you have everything setup and the car idling you can change to AlphaN using the following steps:
(Remember to save your configuration before you start so that you can return to working settings).
In MegaTune select Alpha-N/Speed-density blending:
(http://www.vems.co.uk/VEMSTech/MTExtrasAlphaNMenu.png)
And you will get the following screen on early releases
(http://www.vems.co.uk/VEMSTech/MTAlphaNSpeedDensityEarly.png)
Of this screen on the later releases
(http://www.vems.co.uk/VEMSTech/MTAlphaNSpeedDensityLate.png)
for max RPM for Alpha-N (rpm) enter 25500
and min RPM for Speed Density (rpm) entere 25500
If you have MAP(load) multiplication ensure that it is set to Disabled
Once you've set these values take a look at the main screen - you will see that the MAP gauge reads 0.0 kPa.
Press the throttle down and MAP will rise to 255 as Throttle Position goes from 0 to 100%.
If you do not get the full range on the TPS then you will need to check the TPS calibration.
Now we need to change the Maps go to: Settings->VE Table and fill the kPa settings with the following values starting from the bottom cell: 0, 2, 5, 12, 23, 35, 53, 75, 115, 165, 225, 255
Make sure that you burn the change to VEMS by using File->Burn To ECU In the VE Table window.
To help you keep the car running (as ITBs tend to have terrible IACV control) its not a bad idea to put a "kick" up in the Ignition advance down at the lowest Load and RPM cell. In Settings->Spark Table Put a value of about 30 down on 0kPa and the lowest RPM cell.
When your engine goes to stall it will get a sudden advance which will catch the stall and cause the engine to rev.
Once your VEMS is better calibrated it is possible to remove the Ignition Kick, and use the Ignition based advance which gives a similar function but is better controlled.
Okay,
Great info, as i have newer seen anyone run alphaN,
Looking forward to info regarding boosted itbÃ,´s, alphaN with Map :)
/Skassa
Paz runs it with his 16V 4A-GE ;D
Along with a couple of other ITB'ed 4A-GEs and a few Hondas.
do you need to disable map load multiplication in the constant screen too????
settings->constants
Yes its the same setting, and should be the same value as set in the AlphaN window.
This is a right pain to do you know... Hope it's worth while ;)
I am having the odd difficulty with lambda not being stable, but I'm wondering if this is because the map is a long way wrong in places.
We found with the Corollas that its a load more to setup, but once its running nicely the response is so much nicer.
i gave up in the end and went back to kpa/rpm,
thats not good news, I think mine will have to run AlphaN, and would like to get it running before going on the rollers... Oh well, we will at least try, maybe I am lucky!
It takes time to get AlphaN setup nicely, you will have to completely remap everything including changing the breakpoints, one trick for an even idle is to put an RPM point 200rpm eitherside of your idle rpm - so if your target idle speed is 950 your points go: 750, 950, 1150.
Another trick to keep the engine alive when you're tuning it is to put an advance kick in the ignition table at the lowest RPM cells, we did this on Paz's car so that I could tune the idle to allow him to break in his cams on a completely blank AlphaN map. Paz has a 45degree advance spike at ~600rpm on 0% throttle now as he's not too bothered about tuning part of the map he seldom uses.
Good point about cam bedding. That is a point I have thought about. My Twin-cam head has new reground cams in it and I would like to keep the engine around 2000 rpm for 15-20 minutes to bed them in properly. I have a feeling this could be very difficult as I am sure to have some troubles getting it to run that long the first time it starts....
well one way I've had of getting an engine to run without load, easily is to mess around with the cranking and afterstart values, by putting the Cranking threshold at 2500 the ECU never considers the engine to be running and holds advance at 10deg and scales the fuel based on the Crank Cold and Crank Warm PW values. I'm sure we can rustle up some fair guesses for what the PWs would be.
Sounds doable then! Good. Will come back to this when everything is ready. (May be a while yet)
I had another go at alphaN last night. It was considerably better, but still nothing like as smooth as SD. Lambda control seems to be unable to cope properly with the large AFR excursions presumably caused when the idle controller changes the amount of air entering the engine.
I didn't want to risk this setup on the road without a co-pilot, so I'll leave it for the moment until I get some time spare on some rollers somewhere...
The speed density map is now quite well refined, and almost allows the car to do 30mph without emulating a kangaroo...
BTW are there any software updates about negative acceleration in the pipeline - like fuel cut and resume that don't result in lurches? (Is that just my fault for having too many CCs?)
It takes time to smooth and refine it, SD does so much in the way of correction and smoothing.
I think it would have helped me if the IAC position could be made to affect the AlphaN TPS/MAP axis when the IAC is active. This would make the map 3d for idle instead of 2d.
Sounds like you're falling into the race car / road car gap here....
TPS based systems are generally used on cars that can't use a MAP system because the signal is too noisy. That generaly means lumpy cams etc. For these engines idle quality isn't an issue, so they don't use idle bypass valves.
You want a road car based setup, but are trying to get around a noisy/aggressive MAP signal using TPS, which limits your options.
Things to try
1) Use the ignition based idle controller. The IAC should only really be used for quite slow corrections, or to copensate for additional loads (like friction when cold or alternator loads). The ignition path should then do the fast work to stabilise the idle and react to changes in load (before the air path has a chance to). Hopefully the idle igntion will become 'proper' PID controlled at some point, with error compensated PID terms tables (are you listening Rob? ;))
2) Try fitting a damper in the signal line to the MAP sensor - something like a 0.5mm restrictor hole. This might calm the signal down enough to use it (or it mihth not)
Regarding you other point about shunt on fuel cut/reinstatement, check that you have blended out the VE table to the minimum pulsewidth (you'll need to investigate what that is) and also blend away the ignition at the low load sites (you may need to play with the load sites to get this right). This will smooth the torque away as you tip out and enter fuel cut, hence reducing the torque step between fuel on and fuel cut. You will still get some step since you don't have control of the throttle an ignition in the same way as you do on a modern torque based system, but you should be able to improve things.
Oh, and what's SD?
i would guess sd is speed-density?
Already doing 1 - it's the only way I've found to tame the idle. (It works very well too, even without PID control) I'm not using the idle stepper other than to set the base idle point. (I have no AC etc and don't care at the moment if the idle speed creeps around slowly)
I am not finding the MAP signal all that noisy having spent another night tweaking things (I used the thinnest hose I could find) - I just wanted to try alphaN to see what it was like.
About the fuel cut: It sounds like I need to drop the ignition away more than I currently have! I hadn't blended out the VE table so I can do that too, and it now makes sense. (Amazing how it's obvious now...) I was a bit worried about it all going lean and wrong on me...
SD = speed density. Laziness on my part ;)
Big improvement with deceleration now :) Nic will be pleased at the return of engine breaking.
Still a little shunty at low speeds, but then that's wild cams and letting someone like me doing the cal for you...
I really must try using more than 50% throttle one day. I haven't felt the need yet, even when a new BMW 6 series tried to get away from me at some traffic lights ;)
Do you still have to use acceleration enrichment when using AlaphN...?
Yes, but you tend to use only very small amounts, the fuel map tends to provide the fueling adjustments.
Rob
do you use barometric compensation when using AlaphN or am I imagening things ....?
You have a choice to use barometric correction or not.
Quote from: [email protected] on August 15, 2007, 08:30:46 AM
You have a choice to use barometric correction or not.
How does one check this?
If this setting si turned off, it could explain the issue we discussed on the telephone.
If it IS turned on, can you do a quick explanation of using the blending option.
Alpha N to 4000rpm/SD from 4-9000rpm
Setting a new RPM bin, and setting the new throttle/KPA bin combo.
When blending, does the KPA axis value mean KPA or throttle constantly?
Barometric correction is enabled/disabled using Settings->Constants then selecting On or Off in the Active drop down below the Barometric correction label.
AlphaN blending is found in Extras->AlphaN / Speed Density Blending
Set MAP Multiplication to: Enabled
Set max RPM for Alpha-N (rpm) to the RPM point just above where the MAP signal starts to read cleanly.
Set max RPM for Speed Density (rpm) to a value just below the Alpha-N value.
For alpha N, I would assume barometric correction should be left on to help compensate for
Hot/cold weather, the ramair effect when driving?
Is it safe to say it should only be done BEFORE mapping begins?
When going blended, is it correct to say that at the point you know what your stable vacuum will be,
you should change the nearest TPS axis point to match this?
I'll paint brush some illustrations.
Quote from: [email protected] on November 03, 2008, 06:33:58 PM
Barometric correction is enabled/disabled using Settings->Constants then selecting On or Off in the Active drop down below the Barometric correction label.
AlphaN blending is found in Extras->AlphaN / Speed Density Blending
Set MAP Multiplication to: Enabled
Set max RPM for Alpha-N (rpm) to the RPM point just above where the MAP signal starts to read cleanly.
Set max RPM for Speed Density (rpm) to a value just below the Alpha-N value.
Quote from: [email protected] on November 03, 2008, 06:33:58 PM
Barometric correction is enabled/disabled using Settings->Constants then selecting On or Off in the Active drop down below the Barometric correction label.
AlphaN blending is found in Extras->AlphaN / Speed Density Blending
Set MAP Multiplication to: Enabled
Set max RPM for Alpha-N (rpm) to the RPM point just above where the MAP signal starts to read cleanly.
Set max RPM for Speed Density (rpm) to a value just below the Alpha-N value.
does the barometric correction be set to on or off when mapping?
Quote from: irishtwincam on November 03, 2008, 11:25:05 PM
For alpha N, I would assume barometric correction should be left on to help compensate for
Hot/cold weather, the ramair effect when driving?
The only way you'll see the compensation for ram air is with the MAP sensor connected to a suitable feed on the inlet manifolds.
It should be enabled before mapping, what state is it currently in?
Its on.
Need to get the speed density thing set up over christmas!
My ITBs are joined using pneumatic fittings:
(http://www.martinbruwer.com/pic6.jpg)
They're perfect for the job, cheap and available from a few places in Cambridge as well as on-line:
http://www.technobots.co.uk/acatalog/Online_Catalogue_Push_Fits_134.html
All you'd need is One stud elbow:
http://www.technobots.co.uk/acatalog/Online_Catalogue_BSPT_Stud_Elbow_267.html
Three stud tees:
http://www.technobots.co.uk/acatalog/Online_Catalogue_BSPT_Stud_Tee_271.html
And the pipe to join them up, you just need to decide on the right tube size.
i still have a small problem i have to set my idle every now and again as the car decides to shift the idle around (yes the idle screw is secure) mapped the car idle was grand drove home and car was parked for two days then when i started it i had to set the idle again as it wouldn't idle correctly it kept dying.is this due to the atmosphere?mapped the car at night as that was the only time the private road was available :-) also it was quiet cold when the car was set to idle.also have to judge what rpms to set it to as when i turn on my lights (rpms drop under load) it drops about 150rpms.so is there anything i can do to or check without adjusting my idle before i take it in for the nct (mot)?also the barometric adjustment is set to on
Getting a clean idle from cold is very tricky - you really need an idle control valve.
What are your ignition advance idle settings like?
Quote from: [email protected] on November 05, 2008, 01:15:55 PM
Getting a clean idle from cold is very tricky - you really need an idle control valve.
What are your ignition advance idle settings like?
Car was fully warmed up.i will check all setting when i get home in about 2 hours have a map here of how the car ran on the dyno after being road mapped will post up that aswell
Quote from: [email protected] on August 14, 2007, 09:16:43 PM
Yes, but you tend to use only very small amounts, the fuel map tends to provide the fueling adjustments.
Rob
What acceleration enrichment base settings should be used with AlaphN? I'm using the following:
(http://www.locostbuilders.co.uk/upload/Screen%20shot%202010-10-18%20at%2010.50.39.jpg)
They should be about the same as with Speed-Density, except you probably need less since the throttle input quickly responds to give you the correct fueling.
Start with something like 100/40/10/0 for rpm based enrichment, and 5/10/20/40 for TPS. Then work from there on the TPS table alone.
Lean blips in the lambda graph are caused by either rich mis-fires or simple too little enrichment. Watch tpsDot gauge and lambda graph. Datalog gear changes and transient behaviour, this stuff takes time, but it's easy to at least avoid the too lean condition.
Thanks again Mattias.
Does Alpha-N/SD blending work the same way in VEMSTune? Sorry if its a silly question but most of the info here is for the older software regarding settings to do blending. Obviously the procedure is still the same :)