nom

A quick (and rather large - sorry!) graph... [img]
http://www.kuba.to/images/Scoobynet/APSvOEM.gif
[/img]
Not too sure what a 0.5 AFR point is - I have more detailed numbers if anyone interested this was a run in identical conditions on a MY00 with a standard Tek2 (so a lot richer low down, marginally leaner up top).
The honeycomb mesh jobbie was left in place as usual. The MAF was also checked & OK.
So I'm not at all convinced by Dave (APS) points 3 (a) & (b). From the data that we have here so far (UK), they don't look right.

john banks

Conservatively to me it looks like 2 AFR points. I suspect your ECU is hitting areas of the map which are completely inappropriate and there is a real risk of det.

The Link manual has these as a guide for lambda voltages, % CO and AFR.

0.84V 5% CO 13:1 AFR
0.86V 6% CO 12:1 AFR
0.88V 8%+ CO 11:1 AFR

Now we have all had the debate on narrowband lambda sensors. The car in the other thread with the under-reading MAF setup (not the sensor oddly) was detting when it was at 0.85 V. Before the rolling road we tried it again with a standard map and it was audibly detting on a hot day when pushed and the ECU was pulling out massive amounts of timing. I think if we left it much longer the ECU would have started dropping boost, but clearly you back off when you hear detonation.

The standard car runs often over 0.9V >10% CO <10:1 AFR. For safety I am only mapping most cars to about 10.5:1 AFR on the standard turbo, yet Nom's car looks like >12.5:1 AFR - at least as lean as the theoretical point for peak power and possibly a bit leaner still.

This is Andy Forrest mega power territory (but he has done a lot to ensure safety at this level in terms of thermal management).

Nom it would be useful if you could post the MAF voltages with similar graphs too?

[Edited by john banks - 8/27/2002 1:14:38 PM]

nom

I shall dig about & get the info.
And yes, there was det - not seriously bad stuff except when on the rollers (around 28C outside & not very nice inside!) when we were playing .
I just have to find the stuff on the laptop... and the laptop... hmmm

Bob Rawle

Pavlo, I know that map based systems use tps and "think" that the jecs does, certainly the boost control maps are excusively tps driven (with rpm), My interest was nothing to do with air filters but merely to allow me to inter relate my datalogs with the actual map load lines. I only have to touch my throttle of load and the load point moves instantly but only up the rpm range not across, you need to load up noticeably to get that to happen, tps could be a minor partner in the point selection but I believe it is one.

As far as afr points are concerned then I have datalogs showing far more than the 0.5 mentioned here ... my own car has shown some what over that and the P1 significantly more, the graph posted shows the effect quite clearly i think and some cars will be more affected than others, histrionic ... JB we are in complete agreement here.

I would again emphasise that there is only appears to be a problem with this on cars that are using the std ecu with maf, if remapped or if using a map based system then the filter "should" not cause any issues.

Having done some more work on this it's my view that the maf output may be being "jittered" or "dithered" by the open filter system allowing air pulsation to come into play, the maf will not be able to react to fast air pulses during the flow (positive and negative) and will then only generate a LOWER average. As mentioned the factory induction is designed to smooth and damp these pulses, since takng off the APS and installing the HKS DD my maf readings are considerably more stable. Also the car has become much smoother on closed loop with lower and less frequent excursions of afr correction.




[Edited by Bob Rawle - 8/27/2002 8:13:11 PM]

APS

There is something else I should have covered in my previous post.

Compared to most markets, the U.K. Subaru distribution is quite confused. As most of you would know, in the U.K. you find:-

1) U.K. delivered models.
2) U.K. delivered models, importer modified.
3) Japanese domestic models, grey imported.

The models you receive which are exactly the same as ours, are the first category. Truthfully, we have done no engineering work on the second two categories.

However, from field reports we have received worldwide, easily the biggest issue with usage of our HFA’s on 97/98, 99/00 models has resulted from the installer deleting the Honeycomb mesh at the entrance to the MAF meter. Please do not do this under any circumstances.

Apart from the proceeding, the situation is exactly as outlined in my previous post.

All the best,

Dave,
A.P.S.

nom

...but there's about 2 AFR points removed as shown on that graph up there, and that's on your type (1) (U.K. type not modified by any importer - whatever they would be doing!) with the 'honeycomb mesh' still installed. There is quite a difference between half and two AFR points.

Pavlo

Bob,

I know the boost is completely TPS driven, I even had S Done telling me it wasn't though!

I know what you mean about the load and blipping the throttle, but I did see an identical rise in airflow when doing so. I will remove the TPS, crack open the throttle (so idle switch is off) and adjust TPS without opening the throttle. This should say one way or the other, even if my load measurement is suspect, I would expect to see a timing/fuelling change if load was TPS sensitive.

Paul

Pavlo

Bob, (et all)

I did some further testing tonight:

I removed the TPS and monitored airflow and load seperately.

Leaving the throttle spindle alone and just cranking up the sensor gave an increase in load, and a very slight increase in airflow due to the engine speed going up (moved to richer part of map i guess).

Conversly, leaving the Sensor just open and revving to about 4000rpm for a second or two gave a massive increase in airflow but only a moderate increase in load.

And to finish I compared it against normal, where the load and airflow are the same at LOW revs, but airflow takes off at higher RPM whilst load increases less.

Max load at about 4-5k rpm was something like 17 (hex) during the acceration, and dropped once holding the revs.

Max airflow was about 41(hex) I think and that was at the highest speed I revved to just as it finished accelerating.

so there you go.

Paul

Moles Dad

Thanks APS for your input, doesnt mean a lot to me and I admit that

Thanks to all for the replies even though I was left behind after about the 4th post

I will leave my 'cold air intake' 'till another chapter unfolds in the Mole Lightweight....in other words when I get to the ECU and no MAF part of the saga!

Thanks again all

Bob Rawle

Pavlo well done that man, so that says that moving the sensor does change the load point without changing the airflow (clearly) and that leaving the tps alone but opening the throttle changed airflow at the same load point (roughly) so tps is a factor, if so makes sense due to the need to allow for the changing system (throttle butterfly variation etc)

Its odd that the Subaru parts book lists the same maf and related parts for all of their MY cars wether Au/Euro or Jap spec ... eveluate one and you have efectively covered them all and this applies to the MY01/02 ranges as well.

No one I know of bar Harvey has removed the bee filter on their mafs when installing the APS CAK, certainly the P1 I looked at was completely std in that area apart from the filter.

On my own car removing the mesh made roughly three fuel points difference (not afr points I should add just map amplitude.)