Airmass sensor usage on 993tt ECU
11-27-2005, 11:07 AM
#16
Intermediate
Thread Starter
Join Date: Jul 2005
Location: Stockholm,Sweden
Posts: 49
Likes: 0
Received 0 Likes
on
0 Posts
So the ECU on the 993tt uses the MAF sensor as load signal and the boost control function in the ECU uses that same signal or?
I read somewhere that the boost control was an open loop regulator!?
Anyone knowing if this is correct?
In that case there are much room for imporvement when it comes to creating higher boostlevels on low revs.
I read somewhere that the boost control was an open loop regulator!?
Anyone knowing if this is correct?
In that case there are much room for imporvement when it comes to creating higher boostlevels on low revs.
11-27-2005, 12:22 PM
#17
Banned
Rennlist Member
Rennlist Member
Join Date: Jul 2004
Location: California, Bay Area
Posts: 577
Likes: 0
Received 0 Likes
on
0 Posts
The turbos are controlled by a frequency valve connected to the wastegates. The ECU cycles the valve based on throttle postion, MAF readings, RPM's etc. The ECU has no idea what actual boost levels are.
I stand by my previous suggestions. They both have major advantages over the stock setup.
I stand by my previous suggestions. They both have major advantages over the stock setup.
11-27-2005, 02:29 PM
#18
Addict
Rennlist Member
Rennlist Member
Thomas
I can't help you in much of a technical way, but feel I should balance some of the negativity about the stock MAF system.
There are VERY few people in the world who can program the factory 993tt Motronic properly, getting it to produce more boost in a reliable way seems to be the main stumbling block and is why Todd @ Protomotive goes via the fixed boost controller. The Motronic, when mapped right is ABSOLUTELY the best system IMO, it allows the boost to reach a high level low down, giving some hefty torque numbers (I see up to 1.45bar at peak torque on a MAF motor) and then can follow the knock curve quite precisely allowing the maximum power to be obtained from a given set of components.
The other big benefit lost with fixed boost is the overboost function which when done right makes the car very powerful on the road - punching the throttle at for example 5000rpm in 3rd gear will give me an "overboost" of 1.2+ bar for a couple of seconds before the boost is brought back to the new knock curve by the ECU. This translates to an incredibly torquey motor.
The ECU effectively varies the boost all the time according to the inputs it recieves.
About the MAP sensing - this is the next step for a Motronic engine, the GT2 EVOs ran MAP with the Secan intercooler and all diffrerent intake set up. The factory used TAG (because according to Paul Frere TAG were a small efficient company which could work more easily with the needs of the race program)
RS Tuning converted the Motronic to run MAP with all the features described above to give reliable 640hp 3.8 engines which were usually the front runners and faster than other claimed ~700hp engines with inferior torque charactaristics.
I can't help you in much of a technical way, but feel I should balance some of the negativity about the stock MAF system.
There are VERY few people in the world who can program the factory 993tt Motronic properly, getting it to produce more boost in a reliable way seems to be the main stumbling block and is why Todd @ Protomotive goes via the fixed boost controller. The Motronic, when mapped right is ABSOLUTELY the best system IMO, it allows the boost to reach a high level low down, giving some hefty torque numbers (I see up to 1.45bar at peak torque on a MAF motor) and then can follow the knock curve quite precisely allowing the maximum power to be obtained from a given set of components.
The other big benefit lost with fixed boost is the overboost function which when done right makes the car very powerful on the road - punching the throttle at for example 5000rpm in 3rd gear will give me an "overboost" of 1.2+ bar for a couple of seconds before the boost is brought back to the new knock curve by the ECU. This translates to an incredibly torquey motor.
The ECU effectively varies the boost all the time according to the inputs it recieves.
About the MAP sensing - this is the next step for a Motronic engine, the GT2 EVOs ran MAP with the Secan intercooler and all diffrerent intake set up. The factory used TAG (because according to Paul Frere TAG were a small efficient company which could work more easily with the needs of the race program)
RS Tuning converted the Motronic to run MAP with all the features described above to give reliable 640hp 3.8 engines which were usually the front runners and faster than other claimed ~700hp engines with inferior torque charactaristics.
11-27-2005, 05:26 PM
#19
Burning Brakes
Join Date: Jul 2004
Location: Official Jack off extinguisher
Posts: 1,173
Likes: 0
Received 1 Like
on
1 Post
Originally Posted by K24madness
Thomas,
Lots of ways to address the poorly designed stock setup. Glad to see you thinking outside the box. One look at the stock setup and even a fool would see it far from ideal. You would be hard pressed to find another turbo car trying to control boost without a MAP sensor.
I am not sure if you are looking to achive better results of you like the challange of the project.
From a results POV Protomotive has achive what I consider to be one of the better balanced approaches. Todd installs a EBC to controll boost levels. The EBC is second to none when trying to control stable boost levels. There are many many great products on the market. Greddy is one fine example. He then fine tunes the setup from there. Since he has done lots of cars this way he has really got the setup dialed in quite well.
Motec is altogether the best option. The time involved to get the car setup correctly makes it less desireable from my POV.
If you got the time it would be interesting to see the results from your approach.
Lots of ways to address the poorly designed stock setup. Glad to see you thinking outside the box. One look at the stock setup and even a fool would see it far from ideal. You would be hard pressed to find another turbo car trying to control boost without a MAP sensor.
I am not sure if you are looking to achive better results of you like the challange of the project.
From a results POV Protomotive has achive what I consider to be one of the better balanced approaches. Todd installs a EBC to controll boost levels. The EBC is second to none when trying to control stable boost levels. There are many many great products on the market. Greddy is one fine example. He then fine tunes the setup from there. Since he has done lots of cars this way he has really got the setup dialed in quite well.
Motec is altogether the best option. The time involved to get the car setup correctly makes it less desireable from my POV.
If you got the time it would be interesting to see the results from your approach.
I would like to hear your explaination why Bosch did not use a "fuzzy logic " boost controller on that car.
11-27-2005, 05:41 PM
#20
Banned
Rennlist Member
Rennlist Member
Join Date: Jul 2004
Location: California, Bay Area
Posts: 577
Likes: 0
Received 0 Likes
on
0 Posts
Pole Position
The 996tt uses MAP & MAF sensors. This is the ideal setup. No need for a EBC there. This is the same as MoTec.
TB
Running boost spikes past the intended levels is not ideal. While you may like the way it feels there are many downsides to it. Stable boost levels control heat and detonation thresholds. While RS may have made the best of the worst it is far from ideal. I am sure that if you asked RS to give you a certain boost level and keep it flat across the RPM band they would tell you they would not be able to do so with the current setup. Boost can't be controlled that well with the stock setup.
The 996tt uses MAP & MAF sensors. This is the ideal setup. No need for a EBC there. This is the same as MoTec.
TB
Running boost spikes past the intended levels is not ideal. While you may like the way it feels there are many downsides to it. Stable boost levels control heat and detonation thresholds. While RS may have made the best of the worst it is far from ideal. I am sure that if you asked RS to give you a certain boost level and keep it flat across the RPM band they would tell you they would not be able to do so with the current setup. Boost can't be controlled that well with the stock setup.
Last edited by K24madness; 11-27-2005 at 05:57 PM.
11-27-2005, 06:29 PM
#21
Originally Posted by pole position
The Alzen 996tt (now retired) was/is the fastest race 996tt on the planet and survived 24h races at the Ring setting lap records. IMO those are real world results and not shoestring mag shootouts or 1000whp cars on 6speedonline.
I would like to hear your explaination why Bosch did not use a "fuzzy logic " boost controller on that car.
I would like to hear your explaination why Bosch did not use a "fuzzy logic " boost controller on that car.
11-27-2005, 06:57 PM
#22
Addict
Rennlist Member
Rennlist
Site Sponsor
Rennlist Member
Rennlist
Site Sponsor
Boost can be controlled at the code level in the Motronics.. I think what TB has tried to mention was that the overboost that occurs with the Motronics is a function of the ECU.. It definately isn't boost creep. Case in point, my Stage 2 Z/C turbochargers have given some tuners head-aches to tune with the Motronic, thus the addition of the EBC. Guenter could place the boost level where we wanted it. Higher in some area's and lower in others. Infact, the stock ECU untouched dialed in rock solid boost levels and control.
11-27-2005, 07:13 PM
#23
Banned
Rennlist Member
Rennlist Member
Join Date: Jul 2004
Location: California, Bay Area
Posts: 577
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by Kevin
Guenter could place the boost level where we wanted it. Higher in some area's and lower in others. Infact, the stock ECU untouched dialed in rock solid boost levels and control.
11-27-2005, 07:56 PM
#24
Addict
Rennlist Member
Rennlist Member
Originally Posted by K24madness
On the dyno maybe in 4th gear only
Of course the boost varies in each gear, that is the benefit. In a high gear on the Autobahn in 6th the boost can be higher as there is more airflow through the I/C and the Motronic does its stuff - a fixed controller will limit you massively compared to full Motronic control.
Anyway, we'll be testing the theories out soon. my 993tt 540hp/736NM (RS engine dyno) Motronic controlled motor v's 993RSRs 540hp/750NM (Chassis dyno) 993GT2 fixed boost motor.
11-27-2005, 07:58 PM
#25
Burning Brakes
Join Date: Jul 2004
Location: Official Jack off extinguisher
Posts: 1,173
Likes: 0
Received 1 Like
on
1 Post
K24 madness ,
Your exact words " the EBC is second to none when trying to control stable boost levels" indicated that you included all modern motronic cars in your evaluation and for that reason alone I brought out the Alzen example on which Bosch was in charge of the ECU.
EBC's and eliminating built in safety features in the motronic are a indication that the programmer is unable to get complete control of the ECU.
Your exact words " the EBC is second to none when trying to control stable boost levels" indicated that you included all modern motronic cars in your evaluation and for that reason alone I brought out the Alzen example on which Bosch was in charge of the ECU.
EBC's and eliminating built in safety features in the motronic are a indication that the programmer is unable to get complete control of the ECU.
11-27-2005, 08:13 PM
#26
Banned
Rennlist Member
Rennlist Member
Join Date: Jul 2004
Location: California, Bay Area
Posts: 577
Likes: 0
Received 0 Likes
on
0 Posts
Pole
You really took that out of context. Guess you failed to read the statement above it then. "you would be hard pressed to find another turbo car trying to control boost without a MAP sensor"
Yes the 996 has a MAP sensor and controls boost very very well. So does MoTec. If you want to retain stock motronic (993tt) the EBC is second to none.
You really took that out of context. Guess you failed to read the statement above it then. "you would be hard pressed to find another turbo car trying to control boost without a MAP sensor"
Yes the 996 has a MAP sensor and controls boost very very well. So does MoTec. If you want to retain stock motronic (993tt) the EBC is second to none.
11-27-2005, 09:36 PM
#27
Nordschleife Master
Thomas,
Back to your original question, the MAF is used as the load value for metering fuel and controling ignition timing. MAF is not the ideal method for engine load measurement and is most commonly used by OEMs to comply with emissions specifications for 100k miles. Since engines wear over time, the volumetric efficiency changes and therefore, the mass of the air through the engine. Since a MAF measures the mass directly, it can continue to meter fuel properly as the engine wears. I believe it is a limitation in a performance environment. Further, in a turbocharged application, it is difficult to use MAF as an indication of load since there may be the same mass of air at different load sites, depending on where the turbocharger is spooling. I think that MAP is a much truer indication of engine load, further, if you have an ECU that does barometric compensation, then MAP/BAP would be the ideal measurement of engine load.
As for your question on the system you are familiar with on the Saab. If you have and know the software, and the ECU can receive information from a crank position sensor (60-2 mag sensor), cam position sensor (hall sensor), TPS, MAP, AT, ET, and you can set tables up for MAP or TPS as load, then you are all set. You'll need 1 ignition output, 6 injector outputs at a minimum. Idle air control and Boost control are nice as well since you can integrate turbo boost control and cold warmup idle control.
Back to your original question, the MAF is used as the load value for metering fuel and controling ignition timing. MAF is not the ideal method for engine load measurement and is most commonly used by OEMs to comply with emissions specifications for 100k miles. Since engines wear over time, the volumetric efficiency changes and therefore, the mass of the air through the engine. Since a MAF measures the mass directly, it can continue to meter fuel properly as the engine wears. I believe it is a limitation in a performance environment. Further, in a turbocharged application, it is difficult to use MAF as an indication of load since there may be the same mass of air at different load sites, depending on where the turbocharger is spooling. I think that MAP is a much truer indication of engine load, further, if you have an ECU that does barometric compensation, then MAP/BAP would be the ideal measurement of engine load.
As for your question on the system you are familiar with on the Saab. If you have and know the software, and the ECU can receive information from a crank position sensor (60-2 mag sensor), cam position sensor (hall sensor), TPS, MAP, AT, ET, and you can set tables up for MAP or TPS as load, then you are all set. You'll need 1 ignition output, 6 injector outputs at a minimum. Idle air control and Boost control are nice as well since you can integrate turbo boost control and cold warmup idle control.
11-29-2005, 04:20 PM
#28
Intermediate
Thread Starter
Join Date: Jul 2005
Location: Stockholm,Sweden
Posts: 49
Likes: 0
Received 0 Likes
on
0 Posts
Thanks everyone for good input!
I think I have a fairly good understanding of how things work then.
With the SAAB trionic system I think the crank position sensing is the same as on the 993 (58x) but the idle control and the knock detection would be something to think about. On SAAB the knocking is detected by ion signal measured over the spark plug and no acoustic knocking sensor is used. So one possibililty would be to use the SAAB direct ignition system as well. I don't know if it would meet the ignition demands of the Porsche but I guess with direct ignition things would not be worse than on the original setup? Then the idle control must be taken care of since the SAAB uses e-throttle for that. But a PWM output to control the idle valve with a fairly simple control algorithm would probably do.
I think I have a fairly good understanding of how things work then.
With the SAAB trionic system I think the crank position sensing is the same as on the 993 (58x) but the idle control and the knock detection would be something to think about. On SAAB the knocking is detected by ion signal measured over the spark plug and no acoustic knocking sensor is used. So one possibililty would be to use the SAAB direct ignition system as well. I don't know if it would meet the ignition demands of the Porsche but I guess with direct ignition things would not be worse than on the original setup? Then the idle control must be taken care of since the SAAB uses e-throttle for that. But a PWM output to control the idle valve with a fairly simple control algorithm would probably do.
