erratic AFR, surging

[_jim_jones]

i have a 170cc single cylinder 4 stroke FI motor running ms2 extra 3.4.2 . im finally to a ok place in my tune to ask for some help. My AFR are all over the place and the idle and low cruising speed the engine is surging alot with the AFR. sometimes off idle it will stumble and die when throttled. I have done a few runs with VEAL and it fixed the higher rpms of the map but lower is not helping it. Please if anyone sees anything out of place in the MSQ settings, timing or ve table. the cylinders and injectors are set to 4 in order for the tach to be accurate.

[TimoL]

Have you tried running with the EGO correction disabled (just set the Controller Auth to 0%)?

According to the logs the EGO correction seems to be oscillating quite a bit. If you don't want to disable it entirely try running it in basic mode.

Lean spike (due to the oscillation) timed with a quick throttle opening will surely stall the engine.

Edit:

You have set the idle target AFR of 14.3 to 1 which might be too lean for you engine. At least I've been having problem with getting single cyl engines to idle near stoich. Around 12:80 - 13.50 to 1 has been more stable usually.

[_jim_jones]

i have only ran it without ego correction while tuning with VEAL, i have been disabling it by setting "dont include afr table" in general settings is this the wrong way to do that? should i leave ego correction on during VEAL sessions?
I will test run and log with ego correction disabled, thank you for the input.

[rickb794]

I don't like your hack to make the tach work.

Screw the tach, make it run right first then worry about the extras.

There were 10 reported errors when I opened your MSQ.

What are you doing for TPS?

I would rescale your VE table, wasted cells at the bottom the motor will never make 0kpa.
On that note the cells in the lower left corner change a lot (I'm a SD guy I don't know how this relates to AN)

[TimoL]

i have only ran it without ego correction while tuning with VEAL, i have been disabling it by setting "dont include afr table" in general settings is this the wrong way to do that? should i leave ego correction on during VEAL sessions?
I will test run and log with ego correction disabled, thank you for the input.

Having the "dont include AFRtarget" selected is OK. The difference is that if you make changes to AFR table you have to re-tune the VE table. If you include the AFR target it's a lot easier to make changes to target AFR's in the future because the VE table does not need re-tune. Please note that changing the option now requires re-tuning the VE table.

If you have stable EGO correction you can leave it on while tuning with VEAL but personally I like to disable it.

The reason I suggested that you turn of your EGO correction is because it's oscillating heavily. You can see the EGO correction jumping between auth limits in the pic below (taken from your data log) and AFR following.

ego_corr.PNG

You could try lowering the P-term (Proportial Gain(%)) from the WB controller settings or just run the basic algorithm.

Do you know the dead time for your injector? Setting it correctly will make the tuning a lot easier, especially around the idle where the pulse width is low.
I assume you are using the default settings for now?

[TimoL]

I don't like your hack to make the tach work.

Screw the tach, make it run right first then worry about the extras.

There were 10 reported errors when I opened your MSQ.

What are you doing for TPS?

I would rescale your VE table, wasted cells at the bottom the motor will never make 0kpa.
On that note the cells in the lower left corner change a lot (I'm a SD guy I don't know how this relates to AN)

He is running pure Alpha-N, map signal is not used at all. I don't think he even does have a MAP sensor.

In Alpha-N the TPS is used as load, therefore the VE table must have a 0 load row.
The table could use some re-binning in the low load area. I would go with something like this:

ve.PNG

I don't see how the number of cyls would cause any trouble since he is not reading MAP signal.
I too am running pure AN with the number cyls and squirts set to 4 on single cylinder engine and it's the best tune I've had since this day.

[_jim_jones]

Sorry i should have tested the msq attached was from msdroid and i always get errors when i bring it back to TS. I updated it with a clean msq.

I have tested without ego correction and for the most part the symptoms are gone and its running little cooler, latest log is too big guess ill have to make another . i will try setting the WB controller to basic algorithm, i might have a problem in my innovate WB, controller went into error mode a few times today so i will have to get that squared away first before i try to get ego correction working again.

i do not know my injector dead time and used the default. i will start working on figuring that also.

[_jim_jones]

re-binning the fuel table as suggested has worked wonders, idle is much more stable, afr much more stable and bike is alot more tamed and easier to ride at lower speeds.I imagine tuning will be alot easier now thanx so much for all the suggestions.

[TimoL]

re-binning the fuel table as suggested has worked wonders, idle is much more stable, afr much more stable and bike is alot more tamed and easier to ride at lower speeds.I imagine tuning will be alot easier now thanx so much for all the suggestions.

Glad I could help.

As the tune progresses, you could still re-bin the table if you notice big jumps between rows (or columns as well).

With such a small pulse width as yours the injector dead time will play a big part in the tuning consistency.
If possible, you should measure the injector dead time.
All the extra hardware you need is a some kind of measuring cup with a good resolution and an extra battery if you want to easily measure the dead time at different voltage (which needed for dead time / voltage correction).

I measured mine so that lower voltage dead time was tested when the fuel pump was running of the motorcycle's battery (the current draw drops the voltage quite a bit).
And for the higher voltage measurement the fuel pump was powered by an external battery.

You have to ofcourse remove the injector for the testing so that it can be placed above the measuring cup.