Saabscene Saab Forum - Saab Technical Information Resource banner

1 - 18 of 18 Posts

·
Registered
Joined
·
14,305 Posts
Discussion Starter #1
I'm after some collective tomes of wisdom here, 'cos I'm trying to get my head round what's going wrong with my turbo set up. This gets a bit heavy and involved, so please bear with me!

The background:

I was running an Abbott ECU with MBC that saw me get up to 290hp. I'd decided that I wanted to "comfortably" push through 300hp, and that the route to go had to be software. About the same time, my turbo was starting to smoke, so I upgraded it to a T3/T28 roller bearing hybrid rated for 350hp. I first fitted this and noticed an increase in lag at the top end- not surprising though.

I then ordered new software, which arrived and was installed. There was a problem with it, so it went back and the ECU was apparently checked out and set up on another car (in Sweden) then returned to me.

Things were looking good, lots of
but the peak boost crept up to about 2 bar. I was also getting the occasional instance of it holding base boost for a while then suddenly rising up, so I suspected perhaps I had a faulty APC solenoid.

On the rolling road, peak power and torque were down on the expected by 30 or so horses & lbft. I sent details of the session off to the software folk who said that there was a terrible boost spiek (true- the max should only be 1.65 bar apparently) and thaty it was not holding up boost high enough at the top end to reach the maximum power.

Last night I fitetd a new APC solenoid, and briefly saw a peak of nearly 2.3 bar
before the pressure blew the hose of the intake manifold.

The tuners say this is happening because the spool up time of the roller bearing turbo is too quick once it gets going. I'm struggling to see why this should be the case and why the APC doesn't stop it.

I poped my old Abbott ECU back on last night for comparison and that performed as expected.

Now, apparently the standard MAP sensor only maps to about 1.6 bar. In the equivalent kit that the tuners supply for my level of tuning, they include a 3 bar MAP sensor. They said I didn't need it, just the software and bigger injectors. When I queried it, they said the 3 bar MAP sensor is only there to enable them to map a fuel cutoff at 1.95 bar. However, I thought information from the MAP sensor was critical to APC- without it, how can it know how much to open/close the solenoid?? Or is the operation of the APC vakve controlled by other parameters like info from the DI cartridge , knock sensor etc.

Which I suppose is the nub of my query: How does Trionic APC work?

Anyone got any (sensible
)thoughts?

Thanks

Mark
 

·
Registered
Joined
·
8,216 Posts
mark, I can understand Maptun's difficulty here in that they don't have a car exactly like yours to develop the software on. However, your turbo is tiny compared to Fredrik's on his 450hp monster and I thought he still used Trionic.

The Saab workshop manual shows MAP sensor voltage outputs up to 1.75 bar, but that is absolute pressure, so only 0.75 bar above atmospheric. That may be simply the highest check-point for testing and not the highest reading it can manage. I thought I heard somewhere that the standard sensor is 2 bar (which would be 1 bar atmospheric if it iss 2 bar absolute). With the standard sensor, my boost hits 1.6 bar maximum in normal operation, and occasionally overboosts briefly (say in very cold weather) and the fuel cut-out (pressure guard) cuts the engine. So my overboost protection works.

If the turbo is really too big and is causing overboost, surely you should be hitting the cutoff. At the very least, I would ask Maptun to supply a 3 bar MAP sensor and remapped software if necessary, just to protect your engine while you figure out the real problem.

As I mentioned before, I was also getting the situation where sometimes I would floor it and the boost would stick at base (usually in 2nd gear), then suddenly "let go" and rise to where it should be. I only ever noticed this after fitting the JT exhaust system and it did this even with no other modifications (standard Aero ECU and the heat plate still in place). Dr. Boost on Saabnet said this was normal, but I'd never seen it before fitting the JT. It still did it after I fitted the Maptun ECU, but I have learned to avoid it. I'm sure I could still make it happen, though.

I suspect the rate of rise in boost in 2nd gear is too much for Trionic to cope with and it gives up, failing safely with base boost.

I can see how the turbo could "take off" once it got going. If it reaches a point where it really starts to turn and the boost pressure rises rapidly, this will naturally produce more exhaust gas and the turbo will spin up faster. Perhaps while this is happening even the wastegate can't cope with the rate of rise in pressure. How does it behave with the APC disconnected? Perhaps just running on base boost won't let the turbo reach the operating region where this situation arises?

Anyway, regarding the use of the MAP sensor, this is from the Saab manual:

The pressure sensor contains a pressure-sensitive ceramic sensing device, an amplifier and a circuit for temperature compensation. The pressure sensor is connected by a short special hose to the engine intake manifold after the throttle and is supplied with 5 V from ECM pin 43 and grounded from ECM pin 67.
Depending on the pressure in the intake manifold, the pressure sensor supplies a proportional voltage to ECM pin 22.

The ECM uses pressure and temperature in the intake manifold to calculate the engine load, i.e. how great an air mass each cylinder draws in. The fuel injection time is proportional to the air mass drawn in. At atmospheric pressure (100 kPa), the voltage from the pressure sensor is approx. 1.9 V . Information from the pressure sensor is also used to calculate ignition timing and for boost pressure control.

If the hose comes loose or the sensor fails, and in the event of a break in the circuit, the engine control module will use the throttle position as a default value.[/b]
Looks to me like feedback from the MAP sensor is used to control the boost pressure. Given that Trionic seems to have a pre-programmed boost curve, I can't see any other weay it would work.

Any use?
 

·
Registered
Joined
·
14,305 Posts
Discussion Starter #4
Yes, Bill, thanks.

The info from the Saab manual confirms what I thought, namely that there is a problem with the MAP info- be it because the sensor is either a) out of range or B) (as may be the case) faulty as if there is a cut out at 1.95 bar, why isn't it operating?(it did cut out at about 1.6bar with the Abbott ECU)

I will try to get hold of another MAP sensor to try although if you remember we did try one from your car a little while back when I driveability problems and that didn't have any effect. I asked Maptun to send me one but they said I didn't need it...

As far as the turbo spooling up too quickly goes, the overboost is more likely to happen in 3rd or 4th gear than 2nd, when the rate of boost rise under full acceleration is highest. The "sticking" I occasionally encountered would happen in 3rd or 4th as well for me. I've tried it with the APC solenoid disconnected and it just sits there at 0.55bar base. The actuator and wastegate are appropriately rated for the turbo.
 

·
Registered
Joined
·
8,216 Posts
I would expect that if we were both having problems with excessive rate of rise of boost, it would happen in different gears and under different loading given the differences in the turbos. I've already e-mailed you an offer to lend you the sensor off one of my cars for a quick test.
 

·
Registered
Joined
·
82 Posts
Mark,

You're not the only one to have problems with the spoolup speed of roller bearing turbos, scooby mappers have been having the same problems. The situation seems to arise that the even with a highish boost threshold (a sympton of the large turbine required to allow the exhaust gases from 350+ to pass through) the turbo spools up very quickly once it gets going. The lag inherant in the wastegate control sytem (pipe lengths, switching times etc) cannot catch up with it. The trionic is probably able to dump in more fuel, pull the timing far more quickly than the boost system can pull the boost.

This isn't so much of a problem if the car is being mapped with the turbo as they can slug the rate of boost rise by being deliberately conservative with the turbo control parameters.
 

·
Registered
Joined
·
136 Posts
I’ve done some research on stock and 3.0 bar MAP sensors and as I understood it the voltage output scales were different, requiring new software to switch sensors.
Boost Cruncher


BillJ in your quote you list some of the ECM pin numbers. Do you know here I can find a list of all the pins? Preferably for a GM900 but the 9000 should also be helpful.
Thanks
Eric
 

·
Registered
Joined
·
8,216 Posts
Originally posted by Mark E:
[qb]As far as the turbo spooling up too quickly goes, the overboost is more likely to happen in 3rd or 4th gear than 2nd, when the rate of boost rise under full acceleration is highest.[/qb][/b]
On that subject, I just thought of another possible difference between your car and mine. I have no boost limit in 2nd gear. Unless you specifically ask them to remove it, Maptun software will have this limit.
 

·
Registered
Joined
·
4,215 Posts
it really has to be the map sensor
it is the device which defines what everything else does ie fuelling,opening wastgate etc
now if you are operating out of that sensor range then you would have to interpolate from what the reading was doing before it got to that point ie rate of rise and use some assumptions on where it was going
now i guess to be able to do this it would have to call for some clever programming
but another problem do these kind of sensors suffer from hysteresis ?
again this would screwup your interpolation..
I would guess you are in the twilight zone of boost pressure and your trionic doesn't know which way is up..
however with your boost properly measured you are back in business
or Am I talking [email protected]
 

·
Registered
Joined
·
8,216 Posts
Sounds like a sound argument to me. Although I believe these sensors don't exhibit hysteresis, I agree completely with the rest of the reasoning and would think that a direct reading of boost pressure is bound to make things more straightforward and reliable.

I think it may have been a mistake for Maptun to agree to supply an ECU for a configuration they haven't been able to test.

Any more thoughts from Johnny, Mark?
 

·
Registered
Joined
·
14,305 Posts
Discussion Starter #12
Good info guys, pretty much confirming what I thought- I need to concentrate on the MAP sensor.

I've emailed Johnny again yesterday saying that I can't see how APC can function without accurate info from the MAP sensor, and clearly something is wrong in that area because I'm not getting overboost fuel cutoff. I asked him to semd me a 3 bar MAP sensor. I also reminded him that I did tell him my full configuration (incl the turbo spec) when I asked whether their Stage 5 would be suitable. I haven't heard anything back yet.

I also spoke to Giles at Abbott who was very helpful- they will do some diagnostics for me when the car goes there in a few weeks to have it's head done if I haven't sussed it by then.

I'm not convinced yet about this rate of rise thing- my last setup with a hybrid turbo and Abbott software actually spooled up a lot quicker towards it's peak of 1.6 bar.

I'm gonna get to the bottom of this one sooner or (more probably
) later...
 

·
Registered
Joined
·
14,305 Posts
Discussion Starter #13
Update-

I had a useful session yesterday with Tech 2 strapped to my car. One of the most useful functions it has is the ability to record and display various parameters in real time on the display. Using this function, I was able to monitor the output of the MAP sensor. It's analogue output is quantised and displayed as a level between 0 and 255. A value of about 20 is recorded on full overrun (-0.8 bar) and 255 corresponds to about 1.6 bar. Thus when I plant my foot on the floor and the pressure whizzes up past 1.6 bar, the sensor is out of range and not delivering accurate info to the system. This is shown in simple terms on the display as a trace that rises up to max and then has a flat top!

Clearly beyond this point the system doesn't know what's going on and therefore doesn't really have a chance of accurately controlling what's going on. It also confirmed that my software is mapped for a standard 2 bar MAP sensor.

Armed with this information, I had a useful chat with Maptun and although in an ideal world I'd take the car to them, we agreed that the best starting point would be to use a 3 bar MAP sensor and re-map the software. That way if the boost does rise very high, the system will at least have a fighting chance of being able to control it. Beyond that, it may well take a trip over there to sort it
.

I also spoke to my turbo specialist who confirmed that roller bearing turbos have given problems to other tuners (notably Cosworths) because their spool up speed post 3000rpm is so quick.

He also said that the cause of failure on my last hybrid turbo was the sustained high boost I was running, even though it was a 360 degree thrust bearing. So, whilst such bearings are better than standard ones for higher performance, they are still not as robust as roller bearings- which is why I spent the extra on it in the first place. I'm convinced this is absolutely the right turbo to use for reliability; I just have to make it work now.

(Useful info point- he reckons that the most Mitsubishi units have 360 degree thrust bearings as standard, which means that 9000 Aeros are well engineered for high boost to begin with)

Anyway, a bit of thinking about how to cure the spike problem has led me to a possible solution using a Dawes Device in parallel with the W and C ports of the APC solenoid, which I think should prevent the spikes. I've posted separately about it here

I'll let you know...
 

·
Registered
Joined
·
563 Posts
Why didn't Maptun advice a 3 bar MAP sensor from the beginning? Once the poweroutput on their SAAB 9000 tuningkits exceeds 300 bhp, there's a 3 bar MAP sensor in the package
 

·
Registered
Joined
·
82 Posts
The robustness of the Mitsubishi turbo depends on which series is being used. The TD turbos (04 on Aero, 05 on the early Impreza) are very robust and will even withstand being run without a dump valve.

The TE-05 as fitted to the later C900s and I believe the 2ltr 9000 Carlsson in the UK are not.
 

·
Registered
Joined
·
8,948 Posts
On that subject, I just thought of another possible difference between your car and mine. I have no boost limit in 2nd gear. Unless you specifically ask them to remove it, Maptun software will have this limit.  [/b]
And I wonder why they do that
 
1 - 18 of 18 Posts
Top