MicroSquirt Hybrid Alfa-n

SunsetMoto · Post by **SunsetMoto** » Sat Nov 18, 2006 5:36 am

Will it be possible to program MicroSquirt to run Hybrid Alfa-n fuel control

Bernard Fife · Post by **Bernard Fife** » Sat Nov 18, 2006 7:49 am

Will it be possible to program MicroSquirt to run Hybrid Alfa-n fuel control

SunsetMoto,

Yes. In fact, MicroSquirt can use any of alpha-N (TPS), speed density (MAP), or mass air flow (MAF). It can also 'blend' these with use settable conditions. There is more info here: http://www.microsquirt.info/uscode.htm

Lance.

SunsetMoto · Post by **SunsetMoto** » Sat Nov 18, 2006 9:42 am

Thanks for that Lance. I havn't got a μS yet I'm just trying to work out it's current limitations compared to MS2 & MSex. I intend to run it on my bike and if it's succesfull I have a few other bike projects in the pipeline. The advantages of the μS are size and waterproof case, however it would be good to have the flexibility of some of the MSex code e.g. dual map and full hybrid alfa-n. Does the blended alfa-n code use MAP as compensation to an alfa-n algorithm or does it swap from alfa-n to speed density and back again at pre-set points? I have had experience with some of the injection systems fitted to bikes and they are currently very good. However the earlier systems that changed from alfa-n to speed density usualy about 2750 RPM suffered with stumbling and hesitancy at low throttle openings, especialy V twins. Those that used speed density throughout sufferd from throttle lag just off idle. Anyway back to the point, how does the blended alfa-n use the MAP signal? I would prefer to be able to use an alfa-n algorithm with MAP as compensation, I think this would be easier to set up especialy with a pressurised airbox (Using forward motion to pressurise the airbox through front mounted air intakes: sorry for anybody that didn't need this explanation but it's a lot more common on bikes than cars)

Thanks Sean

SunsetMoto · Post by **SunsetMoto** » Wed Nov 22, 2006 1:54 pm

I've been having a closer look at the 'blend' options. If I set Alfa-n to blend setting at 3500 and set blend to Speed Density setting above my redline limit will I get alfa-n with MAP correction from 3500 to redline.

I have looked in at the Manual but there are is no information on the blend settings yet

Thanks Sean

Polux RSV · Post by **Polux RSV** » Wed Dec 13, 2006 8:04 am

From http://www.microsquirt.info/uscode.htm

The alpha-N blend logic was modified to allow alpha-N fuelling to be implemented at low rpm or at high rpm. The former is often used to maintain a more stable idle on an engine with, for example, independent runners and/or a really bad cam. The latter is often the preferred mode for bikes, since the MAP quickly saturates well before max throttle angle is achieved.

The last remark is right. Using MAP at low RPM/low TPS, and TPS at high load/RPM is far better on bikes. On the Aprilia RSV1000, the standard ECU work in this way, and the "blending" is not based on RPM, but on TPS. The two signals, MAP and TPS are exponentialy tightly linked. At low TPS, up to 10-15%, small changes have big effects on MAP. Then MAP saturates and is not usable anymore.
This give a better handling on race-track when cornering with low TPS but high RPM. Working on MAP give smooth transition

and a good feeling of what is appening with your tires

May be the code could be configured to use RPM only, TPS only or a combination of the two.

Angelo

Post by **grippo** » Wed Dec 13, 2006 8:39 pm

The microsquirt code will have the following two options:

use ve[tps][rpm] at rpm < lo_rpm,
blend ve[tps][rpm] and ve[map][rpm] between lo_rpm and hi_rpm
use ve[map][rpm] at rpm > hi_rpm

OR you can go the other way and use map, rpm below lo_rpm, and tps, rpm above hi_rpm and blend in between.

It is not directly implemented this way, because there is a table of alphaN_map as a function of tps and rpm. You can dtermine this table by driving under different throttle angles and rpms and recording the map values. Then VE always stays a function of a pseudo map and rpm. If you don't like this, then you can set the table so that it is equal to tps at all rpms and then make the array of maps indexing the VE table into tps values. This is exactly equivalent to the above implementation.

Aaron Silidker · Post by **Aaron Silidker** » Mon Oct 29, 2007 1:10 pm

Too cool, this is exactly what I was wondering about uS. My MX bike has a very very erratic map signal at and around idle because of the single cylinder and short intake, but I am sure it quickly becomes stable then saturated. Can I use Alpha N on either end of the scale (low and high) and MAP or a blend of MAP and TPS in between?

Post by **grippo** » Tue Oct 30, 2007 9:28 am

I'm afraid not. You can use alfaN at low, map at hi rpm or the reverse, but you can't go alfaN, map, back to alfaN. If the problem you are trying to solve is large pressure fluctuation at low rpm and lack of map resolution at high rpm, then you may find that with the new map sampling algorithm in v2.881 and up you get a stable enough idle that you don't need alfaN. If not, the path we are pursuing for ITBs is to sample all the cylinders, that is a sensor for each cylinder, but each sensor sampled at the same point in the engine cycle. If this doesn't work, then it's easy to add another user input rpm and allow what you want - everything is in place for it.

Matt Cramer · Post by **Matt Cramer** » Wed Oct 31, 2007 4:34 am

Of course, on a typical MX bike, the Microsquirt would have one MAP sensor per cylinder anyway.