Saabscene Saab Forum - Saab Technical Information Resource banner
1 - 20 of 47 Posts

·
Registered
Joined
·
1,448 Posts
Discussion Starter · #1 ·
I had the recent pleasure of getting to pour over the WIS for both the 9-5 and 9-3 range (as well as a glimpse into the EPC) ... and I think I may be able to get a good explanation going as to why T7 tends to adapt downwards.

Firstly according to the WIS both T5 AND T7 control boost by air-mass/combustion and this is not unique to the T7 engine management. Knock control on the T5 calculated permissible load via converting air-pressure in KPa to air-mass/combust. T7 does exactly the same thing, and when at WOT it does it using the MAP sensor and NOT the Mass AirFlow sensor. (The MAF sensor can't respond quickly enough at sudden load changes, though it works well when cruising. It also determines the low pressure airflow into the turbo so that the ECU can keep from overspeeding it at high altitude.)

T7 DOES use the MAF sensor for general boost control. This was a logical step since it sits right next to the turbocharger. But here is where the problem arises.

Say you boot the
pedal while stock ... everything seems fine, the MAF sensor controls the boost to a certain air-mass/combust and the MAP sensor verifies that result via KPa and a known Volumetric Efficiency. Now say you go and put a 3" turbo back exhaust on it ... now the Volumetric Efficiency has changed. Now at that given level of pressure in KPa there is actually more air going into the engine than it is fuelling for.

On T5 the result is a slight increase in power coupled with a slightly leaner mixture. On T7 you get a totally different response. Here's why:

The MAP sensor which controls fuelling on T7 cannot see the change in airflow, but the Mass AirFlow sensor CAN see the change. Now the two sensors are in conflict.

The Mass AirFlow sensor is now reaching the same air-mass/combust at a lower boost pressure. But because Knock Control bases it's air-mass/combust readings of the MAP sensor just like in T5, it never reaches its "Target" air-mass/combust. So it signals the ECU to raise the boost to that given air-mass/combust level. However the MAF sensor is already AT that level. Now the two sensors are constantly in conflict with eachother.

Hopefully that makes some sense. It's really just a glitch in the programming. They changed boost control to the MAF sensor, but still kept knock control on the MAP sensor. Knock control has control over timing, fueling, and boost, and simply wasn't updated to the MAF sensor in it's boost control third of its programming. Had it been, there would have been no adaptation downwards. Though still probably none upwards either.

If the difference between airflow readinds from the MAF and MAP sensors is great enough it throws a code that indicates a leak in the system and runs off the MAP. (Though at a conservative "limp home" setting.)

Those with T5 should also be weary of upgrading parts. It still has the potential to make the engine run lean. Should be ok though for small changes.

Make any sense?

Dubbya~
 

·
Registered
Joined
·
1,448 Posts
Discussion Starter · #2 ·
After re-reading my own post, I think it could be summarized as follows:

Knock control on T7 gets it's READINGS from the MAP sensor (Manifold Pressure), but ALL of the boost control is done via the Mass AirFlow sensor.

When you increase the Volumetric Efficiency with an exhaust the ECU is reading from the MAP sensor that it needs to boost higher to reach its "target", but when it goes to apply those readings to the MAF sensor it finds that it is already AT the boost target.

What it really needs is an adaptive Volumetric Efficiency map. That would be a relatively simple change to the ECU, and would allow modifications to keep from causing downward adaptation. It would even give it an edge over T5 in that mods would not change the air/fuel ratio either.

Unfortunately the factory engineers were not expecting it to be modified.

Dubbya~
 

·
Registered
Joined
·
25 Posts
Dear Adrian

I'm sorry to disappoint you, but you're wrong. The T5 system is ONLY MAP sensor based and the T7 system is ONLY MAF sensor based.
So the mbar boost level on T7 is not used at all for internal calculations of the ECM, it's just a result. The only value used internally is the mg/comb, means the airmassflow/time divided by the rev's.
Adaptions take place as soon as the calibrated values of the PID doesn't match 100% with the actual. This can happen in case of mechanical failure or hardware modification. Every part of the engine has been calibrated (cosidering some tolerance) during the development, from the airfilter to the endpipe.
Anyway, boost adaptions are "less" powereffecting than fuel adaptions, at least on highway driving, but give the bigger "feeling" impact cause they affect the turboresponse.
There's nothing you can do to change this behaviour, I'm sorry.

Trionic
 

·
Registered
Joined
·
1,448 Posts
Discussion Starter · #4 ·
I will quote from the WIS.

From the Description for the Manifold Absolute Pressure sensor. WIS section engine ---> T7 ---> MAP sensor

"During rapid load changes, the engine load or calculated air mass/combustion based on mass airflow sensor value can differ from the actual mass being taken into the cyllinder during the induction stroke. This is due to the inertia of the large volume of air present in the intake system. The intake pressure is therefore used to correct the air mass/combustion used in the basic calculation of fuel quantity."

Obviously if the sensor is not correcting properly due to changed VE, there will be conflict.

Also, knock control is most active during spool up, when it is uncertain how much boost will generate knock. This is when the MAP sensor is used to correct the air-mass/combustion value.

I am sorry if that cam across differently in my previous post. But it should be made clear that T7 DOES have a MAP sensor. In fact, you can disconnect the MAF sensor and run on the MAP sensor alone. I have a video of myself doing that.

Dubbya~

p.s. We may BOTH be wrong. It's just a theory on my part, which happens to fit the information in hand quite well. Very well in fact. The basis of this as a knock control related problem fits in well with why it doesn't ALWAYS adapt downwards. Nick T was able to get good results on stock software because there was no knock. So the discrepency was ignored and there was no adaptation in either direction.

At least ... that makes some amount of sense.
 

·
Registered
Joined
·
25 Posts
Hi Adrian

you are right when you say that the T7 has a MAP sensor, in fact it has two MAP sensors. Anyway these are used only as backup function in case of a MAF sensor failure and the fact that you can drive a T7 car without MAF is only cause it has been foreseen for this use (limphome function) and the poweroutput is reduced.
And I'm surely not wrong, mine are not theories.
 

·
Registered
Joined
·
1,448 Posts
Discussion Starter · #7 ·
Originally posted by Trionic:
[qb]and....

the knock has NO effect on the boost or fuel adaption. [/qb][/b]
Well for that one, I will need to quote the WIS again.

Two parts, here's the first:

"From the combustion signals, the control module will be aware of which cyllinder has fired and if a knock signal above a certain level is registered at the same time by the control module, it will gradually retard the ignition timing for the cyllinder in question until the knocking ceases. Timing correction then slowly returns to Zero.

If the mean value of the ignition retardation on all cyllinders exceeds a certain level, fuel enrichment will take place. If the mean value of the ignition retardation of all cyllinders further despite fuel enrichenment, the maximum permissible air mass/combustion will be limited."

I think that pretty well disproves what you just said? Unless you are referring to the long term fuel adaption. (There is no long term boost adaption.)

Then there are two kinds of long term adaptation in Trionic 7. Additive and multipicative. Then there is maximum permissible air-mass/combust.

Here I will show that both "adaptations" have no affect on the car at full throttle.

Quotes for both kinds of Adaptation:

Multiplicative Adaptation- "The fuel quantity is always multiplied by a multipicative adaptation factor that is 1.0 when the control module is new or has been de-energized. The Multiplicative adaptation takes place during patial load. Purging is interrupted during adaptation and no other sources affecting the fuel may be present."

Obviously that isn't related to this argument.

Additive Adaptation- "The amount of fuel that is added or subtracted until the closed loop value fluctuates around 1.00. The reason that additive adaptation is necessary is that air leakage when idling leads to a major fault, which cannot be adapted using multiplicative adaptation since the amount of fuel would then be too great as soon as load increased....

... The following conditions must take place for Additative Adaptation to occur:

- Closed loop active

- No purging in progress

- Coolant temperature exceeds 78C

- Engine speed 800-1000 rpm

- engine load 85-235 mg/c

- Car stationary"


Obviously that does not apply to full throttle either.

The maximum permissible air mass/combust IS a form of short terma adatption. As stated in previous quotes it is VERY dependant on knock.

If the engine were to knock during spool up with an inaccurate MAP reading caused by inaccurate VE, it would get the wrong mg/c airmass limitation sent to knock control. This would either cause the knock to get worse, and cause significant retardation, or it would cause a lower than necessary restriction on air-mass/combust ... which would show up on the dyno (as Phillip HS saw) as a power output which got lower with each dyno run. Much more so than heat soak would account for.

I'm sorry if you consider this an argument. I'm trying to keep as few oppinions in this as possible. However you are arguing with Saabs own Workshop Instruction System program here (not arguing with me) ... and I am not sure that's an argument in which you would continue to benefit from perpetuating.

If you have any concrete information, please do share it! Or if you work for Hirsch, and would like to give me a free stage III upgrade for my troubles!

Dubbya~
 

·
Registered
Joined
·
1,448 Posts
Discussion Starter · #8 ·
Also, the WIS only lists ONE MAP sensor. This is because the other Absolute Pressure sensor is on the charge intake pipe, and not the intake manifold. Thus it cannot be called a "Manifold Absolute Pressure" sensor.

Again, not arguing with me here. This is the WIS program. Take it up with Saab.

Dubbya~
 

·
Registered
Joined
·
25 Posts
Fuel enrichment has nothing to do with fuel adaptions, so...
The 2 MAP sensors work against each other, so both are used.
The problem is that you misunderstand the WIS cause you dont know the function of the whole system. Knocking has NO EFFECT on fuel adaptions or boost adaptions. It effects the actual output during knocking conditions on every load, not only full throttle. The knockings are "listened" from very low loads and than you get a "knock adaption" based on fuel quality and engine behaviour (This you cant find in the WIS) that tells the ECM how much "knock resistant" the engine is (learning function).
And I'm not disputing with SAAB about it (or with the WIS) but with you, cause you could create confusion here in the board spreading "not correct" informations.
But, seen from another perspective, you're right, I should not be here and discuss with you about this.
 

·
Registered
Joined
·
1,448 Posts
Discussion Starter · #10 ·
Well if you know so much, I'd be happy to be enlightened. I think we're both misunderstanding eachother since what you say doesn't seem to fit the facts for me, and what I say doesn't seem to fit the facts for you.

Firstly, why aren't there two MAP sensors listed in the WIS? Are you referring to the charge pressure sensor in the intake pipe? Would seem odd for a program designed to instruct on the workings of the motor to leave out that there are two MAP sensors. Possible? Maybe. But when I look under the hood, I see one.

Secondly, there shouldn't be any knockings under low load. So I'm not sure I understand what you mean by that. Knock, while varying in degree and form, generally does no occur in low loads. If it did, it would not be wise for an ECU to increase the load.

The knock resiliency of the engine could theoretically be observed by extrapolating the temperature curves inside the engine.

However the last SAE paper I read from Linkoping UNiversity on exactly that subject did not include a conclusion or well tested method. Since that paper was published well after the release of T7, I find it hard to believe that Trionic, which is designed to be a reliable engine control system, would be ahead of the current theoretical systems at the time.

I'm very curious where you're getting this information. Some of it seems ... stretched. While I'm more than willing to quote WIS, or SAE papers, you continue to debate based on personal experiences which you do not share.

I'm more than willing to be proven wrong, but I'd like to see more than just "I say this, so that's how it is. You are wrong and don't understand what you're talking about." Which is very much how you're coming across right now.

Maybe just cultural and language differences?

Actually. I think we ARE misunderstanding eachother here. But I do not think I am incorrect. You will need to give more evidence to make your claims credible!

Dubbya~
 

·
Registered
Joined
·
14,305 Posts
OK I'll attempt a bit of intellectual oil pouring here...

I suspect we do have some language problems that are not helping.

I'll deal first with the points that seem easy...

a) knock vs boost adaptation
I tend to agree that knock should have no effect on boost adaptation. The way I see it is that knock detection is designed as a "safety measure" in case of temporary adverse conditions. You would not want a temporary condition to have an effect on long-term adpatation.

B) use of MAP/MAF sensor.
I think what Trionic is saying is that for adaptation purposes T7 only uses information from the MAF sensor

c) interpretation of WIS
I would suggest that the WIS was written in a way to describe the operating principles of stock engines for fault finding purposes. It was not written as background information to be used for tuning/performance upgrades.

Overall, I suspect that if we were all sitting in the same room, the discussion would be resolved in a matter of minutes. Meantime, for those that may want to contribute further (and please do if you feel you have something relevant to add) might I, without wishing to sound rude, suggest that you read very carefully and try to understand the points that folk are trying to make
.

I'll try to spend some more time considering the various points later and find some more commonality .
 

·
Registered
Joined
·
1,086 Posts
I agree that there is some misunderstanding here. If Trionic was able to talk to Adrian in person, this would probably be finshed quite quickly. I know Trionic personally and I can say his knowledge of the T7 is probably beyond what WIS states.

Shouldn't you be working, Trionic?
 

·
Registered
Joined
·
25 Posts
You're right, Eric, I should (I am at) work.
But this story about adaptions it's a longtime discussion here, so I just wanted to help a little....
But apparently it's not needed, so it's better I go back to work.
 

·
Registered
Joined
·
3,892 Posts
Originally posted by Eric (9-3 Hirsch):
[qb]I know Trionic personally and I can say his knowledge of the T7 is probably beyond what WIS states.  
[/qb][/b]
I'm guessing that this is somewhat of an understatement Eric, hoping I've put the subtle hints together correctly and added 2+2=4......
 

·
Registered
Joined
·
6,531 Posts
Originally posted by Trionic:
[qb]You're right, Eric, I should (I am at) work.
But this story about adaptions it's a longtime discussion here, so I just wanted to help a little....
But apparently it's not needed, so it's better I go back to work. [/qb][/b]
I think we could all gain a huge amount from your posts, so please feel free to set the record straight.
I agree, WIS is used as a troubleshooting tool, rather than a definitive guide for the hugely complex beast that T7 is, let alone T8
 

·
Registered
Joined
·
563 Posts
Originally posted by Trionic:
[qb]You're right, Eric, I should (I am at) work.
But this story about adaptions it's a longtime discussion here, so I just wanted to help a little....
But apparently it's not needed, so it's better I go back to work. [/qb][/b]
Your help is needed and very much appreciated. Without it, it keeps us guessing and speculating, using the limited sources of information we have.
 

·
Registered
Joined
·
1,448 Posts
Discussion Starter · #20 ·
Actually this would probably be resolved quickly if I could speak fluent Swiss German.

I'm also aware there is a LOT not in the WIS program ... but for it to be outright wrong?

I hope Trionic sits down and makes a good long explanation post. Or ... if someone who speaks fluent Swiss French and is good with technical bits would like to translate.

Technical issues always make language barriers worse because all the subtleties in speach when being pedantically specific become very apparent.

As for adaption, I think some of the discussion is missing the point. If the ECU were confused over target boost values, and actual boost values. And were the modified VE to confuse it further, during a knocking situation where it must react quickly it may choose the safer route and adapt excessivly down.

Which is what I was trying to say in the first place. Doesn't that make some sense?

As Trionic said, everything from the air filter to muffler has been programmed. (It all effects the VE, or Volumetric Efficiency.) So if you go and change the VE, things get screwy. Since the Mass AirFlow sensor reads a measured value which is not affected by VE, while the MAP sensors read a value that IS affected by VE, having the two sensors give different readings has to have some effect on power on the dyno.

Maybe the word "adaption" isn't the best to use. Might be the problem with the language barrier. Long term fuel adaptions (Additative and Multiplicative) are explained in a previous post. The boost should always be some mg/c value that only changes when the maximum permissible by knock changes.

Makes sense to me? Seems to go along kinda with what Trionic is saying ... though how he's saying it is very different.

Dubbya~
 
1 - 20 of 47 Posts
Top