KenF27
|
| posted on 24/3/08 at 09:11 PM |
|
|
data log issue
Could anyone explain why the o2 drops off drastically after full throttle. Could it be to rich and fouling the lamda. I have hooked up a wide band
lamda with readout and when accelerating gently it runs about 12 to 13 to 1 which I would think ok. even upto high revs, but if the throttle pedal is
open fully then it goes to 17 to 1. I have attached the datalog. Any comments would be greatly received either on the main issue or anything that
looks odd. Ps It accellerates ok but it is still is not smooth and the megasquirt is running on a narrow band lamda.
|
|
|
|
|
CairB
|
| posted on 25/3/08 at 08:11 AM |
|
|
Your MAP signal doesn't change much with throttle position, but from the O2 sensor(s) it looks like your air flow and fueling requirement does.
As shown in the plot of MAP v TP:
 
KenF27TPvMAP20080325
At full throttle I would be expecting to see higher MAP value than 93kPa unless your high up.
This sort of points to a problem with your MAP signal.
Is it being limited in some way?
From your previous posts I assume that you are using individual throttle bodies then I suggest that you look at using Aplha_n or hybrid Alpha_n as it
can overcome the problem of a steep change in fuel requirements over the 90-100kpa region.
I'm slightly puzzled by your MAP signal though as it appears very smooth for ITB's but still responsive the throttle over a limited range.
How is your MAP signal configured?
The coolant temperature in the log only just gets over 70C so it's mostly in the warmup mode.
Cheers,
Colin
|
|
|
paulf
|
| posted on 25/3/08 at 03:47 PM |
|
|
It probably is going lean at full throttle but the narrow band will not show how much it is lean by , the AFR reading will not be accurate apart from
at 14.7 therfore the 17/1 may be any amount that is lean.
I have set up with a narrow band but always tried to keep it on the rich side and used plug readings as well to confirm it was not running lean.A wide
band is much better and will give an accurate reading that you can use .
Paul.
|
|
|
KenF27
|
| posted on 25/3/08 at 08:49 PM |
|
|
Thanks for the replies gents. I am running on Alpha N but noticed tonight that the map sensor was set at 250kpa. Would this be the reason my maximum
map reading is in the low 90s.I have also put in a small accumulator in line as well. Its the type you get on carbs but I have just found the fuel
filter I was going to use first. I am using GSXR600 bodies on a 1.8 Zetec with a cut down zetec inlet manifold for the minute.Each throttle body has 2
small take off pipes. I have connected one take off from each body together then to the map sensor on the squirt and the other take off pipes are
connected together then onto the idle valve. As the wideband is only connected to a digital readout and not to the squirt. If I want to check the AFR
I have to take out the narrowband and screw in the wideband, thus not allowing the narrowband to make any adjustments. Am I doing this wrong or should
I be turning the ego correction off in the settings, also When adjusting the fuelling on the ve table using Alpha N should i be looking along the revs
and up the map or should I be looking for the tps value somewear else. Thanks Ken
|
|
|
CairB
|
| posted on 25/3/08 at 09:36 PM |
|
|
If your on pure Alpha_n, TPS based ignition and TPS based accel enrichment than I don't believe that MAP will be needed.
I am using a modified version of extra 021u so some aspects may be different for your version but...
I think that alpha_n uses table 1.
The axis labels I think are set by megatune configuration and don't matter.
The values on the vertical used to be TPS raw ADC values, i.e 0-255 for 0-5v.
This can lead to confusion as the datalog defaults to throttle %.
Check this by using megatune - tuning - table 1 option and operate the throttle to see where it moves in the table.
I know it caught me out when I first switched to alpha_n.
If you disconnect the narrowband when using the wideband then I would turn off ego correction.
Is there anyway you can connect in the wideband as you will make life so much simpler. Otherwise it may well be worth investing in an LC1 or
equivalent.
Cheers,
Colin
|
|
|
KenF27
|
| posted on 25/3/08 at 11:17 PM |
|
|
I have set Alpha N on the constants page 1.EGO is set to be active above 1300rpm, also open loop mode is set to go open loop @raw 192 on the TPS.
Could this be leaning out the mixture on WOT. I am still confused with the VE table. If Alpha N is selected why does the VEtable 1 show kPa and RPM.
Should it not show TPS and RPM. Re the wideband, it does not have the controller to work with the Squirt. Thanks Ken
|
|
|
CairB
|
| posted on 26/3/08 at 09:49 AM |
|
|
Looking at you log it looks like once the coolant is over about 70 the ego correction is swinging + to - 15%.
It looks like your rich in some areas and lean in others, obviously (;-).
When the VE table is optimised there should not need to be as much EGO correction.
I don't think that the open loop is causing your problem it is the table itself, it needs to go through more tweaking.
Have you tried processing the datalogs through any of the VE analysis tools from the msefi site?
The VE table for Alpha_n looks different to that for speed density.
It tends to be triangular in form.
An easy way to visualise this is to consider that to get to 90Kpa the throttle has to be open more as the rpm increases.
Here is my current table - Please note that this is for hybrid alpha_n so will be distorted slightly, but should give an idea of the general form.
 
Hybrid_Alpha_n
Note that mine displays Kpa but is actually raw TPS ADC.
It doesn't bother me enough to alter megatune.
Its that way beacause it is Megatune displaying the data from the MS and it is configured to label that axis as Kpa.
I think that it can be altered in one of the ini files for Megatune.
HTH, Colin
|
|
|
KenF27
|
| posted on 27/3/08 at 01:33 AM |
|
|
Thanks for the advice Colin and Ian. I will have a go at altering the alpha n tps values when I get time , busy at present for the next few days. The
only ve analysis I have used is MLV. I have looked at the msefi site but not sure what you meant. Cheers
ken
|
|
|
.jpg)
