Saabscene Saab Forum - Saab Technical Information Resource banner

1 - 20 of 28 Posts

·
Registered
Joined
·
14,305 Posts
Discussion Starter #1
This is a problem that I know has troubled several of us for various reasons; myself most recently trying to match new software to a mega turbo
.

The problem that we generally suffer is in trying to achieve sustained higher boost levels, during mid range spool up, the boost pressure can rise too high triggering fuel cutoff. If we back it off to avoid fuel cutoff, we lose out on top end power. I think I'm right in saying this applies to pretty much all models- be it those with separate APC boxes and pots to adjust or Trionic with only base boost and MBC to play with.

My solution is relatively simple but is designed to work with the APC valve still operational- even if a bleed valve is being used.

To understand how it works, we need to consider the way the APC valve operates. Under electrical control, it in effect acts as a variable pressure diversion valve to determine how much of the charge pressure (port C) is applied to the actuator (port W). Basically, it will stay "shut" up to a certain pre-determined point (max boost), not allowing charge pressure to be applied to the actuator but instead diverting it back into the system (port R) and after then, will start opening thus allowing pressure to be applied to the actuator to avoid any further increase in boost pressure.

The spike is caused when, for any one of a number of reasons (software, hardware or both), the APC valve is unable to react in sufficient time to prevent the boost pressure exceeding the maximum allowed before fuel cutoff.

So what is needed is something to make sure that the pressure differential across the W and C ports of the APC does not exceed a certain value- in other words, a pressure relief valve that opens very quickly at a preset pressure.

We already know of such a device- commonly sold as a Dawes Device. My solution is to tee it in to the W and C pipes and adjust to suit. This can also work if you are using a bleed valve in the W line as long as you tee it in after the bleed.

I haven't tried it yet as although I have a Dawes Device, I don't have and tees, but will try to find some soon.

In the meantime what do you all reckon? Have I found a "miracle cure" or have I lost the plot?
 

·
Registered
Joined
·
563 Posts
Very good thinking, Mark. After reading your theory, it seems like the solution to boost spike problems is very simple (hindsight
) I hope this really works.

However I wonder why a Dawes would be faster than a selenoid valve. The only reason I can come up with is the vacuumlines leading to the selenoid being to long and causing delay. In that case the dawes should be connected as close as possible to the beginning of the vacuumlines leading to the W and C ports on the APC valve.

Keep us informed about the progress of your experiment.
 

·
Registered
Joined
·
14,305 Posts
Discussion Starter #3
Maarten,

The problem with spikes comes from the inertia of the complete system- including the software preventing it from reacting quickly enough to control the rapid rise in boost.

The Dawes I have is one of the higher spec ones which is reckoned to operate very quickly, so we'll see...
 

·
Registered
Joined
·
1,676 Posts
For reference when determining which part of the spike is mechanically (inertia) related and which stems from the APC/Trionic system (including solenoid valve): I replaced the APC solenoid valve on Saabine with a three-way manual boost control valve (same as Abbott's). The result, measured with Saab's own big workshop boost pressure gauge: boost pressure initially rises to 1.05 bar, then settles at 0.95. Dont know whether the spike would be proportionally or progressively bigger with higher boost pressures, though.
 

·
Registered
Joined
·
1,676 Posts
Just occured to me that the boost spike might be bigger with a faster spooling turbo bacause it is on a steeper incline in the boost curve when the wastegate opens. Then again, it might be smaller because it reacts faster to he wastegate opening, as a product of the same lack of inertia.

Oh well, what do I know...
 

·
Registered
Joined
·
4,895 Posts
The problem with spikes comes from the inertia of the complete system- including the software preventing it from reacting quickly enough to control the rapid rise in boost.  [/b]
Indeed, there's "lag" inherent in any digital processing, especially on older hardware, I wouldn't expect a reprogrammed Trionic 1 ECU to react as quickly as a Trionic 8 for example. I don't know what hardware Saab use in the trionic boxes so I don't want to speculate too much, maybe all trionic units use the same chipsets for all I know.

The Dawes device is mechanical so it eliminates the processing lag.
 

·
Registered
Joined
·
8,216 Posts
Originally posted by Mark_A:
[qb]I'm led to believe that Trionic uses a version of the Motorola 68000 cpu ... but this could be an urban legend. [/qb][/b]
Don't know about T7 and T8, but T5 uses a 68332. This is part of the 68300 family, which is a part of the 68000 range aimed at embedded applications (as opposed to general-purpose computing such as in a PC or MAC).
 

·
Registered
Joined
·
8,948 Posts
Don't know about T7 and T8, but T5 uses a 68332. This is part of the 68300 family, which is a part of the 68000 range aimed at embedded applications (as opposed to general-purpose computing such as in a PC or MAC).[/b]
And exclusivly licienced to GM for automotive use.

Andrew
 

·
Registered
Joined
·
16,524 Posts
The Apple Mac used the 68000 series Motorola in the earlier computers. Last used in the series was 68040 which ran at 66MHz and was superseded in 1996. Perhaps it's time for a faster chip? Or have GM done this already?

Mac are still using later Motorola chips and they seem to be lagging behind the PC world in clock speed at the moment.

Perhaps a faster chip would allow extra sophistication like a rolling average on the boost calc, so that the fuel cutout wouldn't be frightened by a very short duration bit of overboost while the wastegate reacts.
 

·
Registered
Joined
·
8,948 Posts
quote:
--------------------------------------------------------------------------------
Originally posted by Scaero owner:
And exclusivly licienced to GM for automotive use.
--------------------------------------------------------------------------------

Surely not?  
I'm pretty sure we used it in an Ethernet switch 7 or 8 years ago.  [/b]
An Automotive Ethernet switch? I've not seen that.

For Automotive use of the 68000 chips, GM had/has exclusive controll, for other uses, other people had rights.
No one else could use a 68000 series chip in an automotive application.

Andrew
 

·
Registered
Joined
·
403 Posts
Mark E,

I'd be very interested in your findings on this one, as I am contemplating doing a very similar experiment with my dawes device.

Would I be correct in saying that the 3 port apc system is purely a control device for the wategate and has no actual feed-back to Trionic ?
And can it be removed/replaced completly?
Cheers

Damo
 

·
Registered
Joined
·
3,618 Posts
The Apple Mac used the 68000 series Motorola in the earlier computers.  Last used in the series was 68040 which ran at 66MHz and was superseded in 1996.  Perhaps it's time for a faster chip?  Or have GM done this already?

Mac are still using later Motorola chips and they seem to be lagging behind the PC world in clock speed at the moment.
 [/b]
Ahh.. you have fallen foul of Intel's megahertz myth.

The architecture of the Motorola-IBM G4 processors that reside in today's Macs are chalk and cheese compaired with the Intel P4 series systems.

Even AMD don't play the megahertz game anymore after their recent renumbering of the product line to indicate relative performance to an intel chip rather than megahertz (because a 1.5Ghz clocked AMD processor is capabile of pushing as many calculations through as a 1.8Ghz Pentium 4)

Same goes with the relative performance capabilities of the Ultra Sparc processor in Sun systems against higher clocked Intel systems.

The G4 architecture includes two very significant features that allow the chip to run at lower clock speeds to give similar or improved real-world performance as a system whole -

The Velocity Engine
Behind the PowerPC G4?s performance is its aptly named Velocity Engine. The Velocity Engine processes data in huge 128-bit chunks, instead of the smaller 32-bit or 64-bit chunks used in traditional processors (it?s the 128-bit vector processing technology used in scientific supercomputers ? except that IBM/Motorola added 162 new instructions to speed up computations). In addition, the PowerPC G4 can perform four (in some cases eight) 32-bit floating-point calculations in a single cycle ? two to four times faster than processors found in PCs.

The PowerPC G4 with Velocity Engine works with the overall system architecture to accelerate the data-intensive processing required by video, audio and graphics applications. Among the key G4 features is a vector permute function capable of rearranging data in the registers ? a real benefit when converting data from one format to another (often necessary with audio, video and graphics apps, which typically need to save data in a number of different formats). These vector processing advantages give the PowerPC G4 a significant edge when it comes to visualization. Making the PowerPC G4 perfect for everything from digital video, graphics and 3D games to astronomy, the biosciences and predictive modeling.

And at less clock cycles, which equals less power requirements, which equals less heat.. which equals better enclosure designs, with less fans which equals better battery life (in the case of a mobile application) and less noise.

Also please remember that overall speed of a system is not dictated by processor design alone, but by RAM type, system Bus type, symetrical processing ability, speed of Hard disks, architecture of the operating system and it's ability to exploit the hardware in the best manner.

Therefore I would counter what people are always finding out, that when you compare Oranges to Oranges - the same application running on similar spec'd systems (as long as the application is using the capabilities of each system) - timing the same task to complete - that the Apple architecture still is faster than the equivilant Intel based machine.

I'm not biased - it's just the very reason why I moved to using a Mac for work purposes many years ago.. and every time I trade up.. it still proves true.
 

·
Registered
Joined
·
16,524 Posts
Originally posted by Mark_A:
[qb]Ahh.. you have fallen foul of Intel's megahertz myth.
[/qb][/b]
Not true!! I was just speculating that a chip that was superseded by Apple at 66MHz in 1996 may be in need of retirement. My current year old Apple is "only" 933MHz but that's a lot more than 66MHz. And it seems fast enough for me.

Don't forget that I new Lisa

However, there have been reports that some multimedia stuff is benchmarking faster on Dell than on Apple at the moment. I take it with a pinch of salt 'cos the report came from Dell


But then, what do I know about computers, I'm only an engineer. I just use them and only tinker with them when the IT man can't fix them.....and then I find I can't either


Oh the joy of a son with a computer science degree
 

·
Registered
Joined
·
14,305 Posts
Discussion Starter #16
Flippin 'eck! I thought this was meant to be a serious car tuning topic!!
Now I find I'm wading my way through all this computer hardware tosh!!

Anyway, back on track:

DaMob- Yes, you're right. The APC valve has no feedback as such other than via other sensors feeding back to the ECU what is going on in terms of boost, knock etc.

It can indeed be removed and relpaced with a mechanical control- Abbott and others sell just such a device.
 

·
Registered
Joined
·
3,618 Posts
However, there have been reports that some multimedia stuff is benchmarking faster on Dell than on Apple at the moment.  I take it with a pinch of salt 'cos the report came from Dell  [/b]
Well, actually the report was posted by Adobe, using Dell hardware, however the problem is that the particular software in question being benchmarked was not multi-processor aware (in otherwords the software runs [expletive deleted] on the current breed of multi-processor power macs), so the benchmarks posted by Adobe just show how [expletive deleted] their software runs.. not the true processing power of the machine.


Sorry Mark_E.. back to your massive turbo woes.
 

·
Registered
Joined
·
8,948 Posts
Mark,

Are you saying you are going to attemp to fit the daws to allow a max boost of 1.6 or there abouts (depending on the max you want), while still allowing the apc to knock the boost down if it wants?

Andrew
 

·
Registered
Joined
·
136 Posts
It is an interesting idea and worth looking into. I have seen another solution to this problem, called a pop-off valve. It was attached near the throttle body and functioned like a giant MBC. It was set just under the fuel cut-off pressure and vented the excess air pressure to the atmosphere.

But I feel that there are more factors involved governing max boost other than just APC control. One problem that I recently discovered was that my wastegate actuator was not stiff enough to stay shut. The exhaust gases leaving the engine were forcing it open, bypassing the turbine and causing a decrease in boost. I can watch my boost oscillate at higher rpms as it plays out. The obvious solution is to tighten the waste gate by shortening it, which I did. This gave me a couple extra psi at high rpms but some experimentation caused unsteady ignition at high rpms forcing me to back off the wastegate. I know you have upgraded your turbo Mark E and I’m assuming the wastegate actuator with it. So you may or may not be facing the same problem with the wastegate being forced open.
 

·
Registered
Joined
·
14,305 Posts
Discussion Starter #20
Andrew,

Yes, that's it in a nutshell- I want to use the Dawes as boost limiter.

Eric,

I've also heard of the pop off valve and whilst possibly slightly more of a guarantee that it will limit boost pressure, personally I'm not a fan of dumping air out of the system. Although it's not a problem for me 'cos I have an MAP sensor, for any car that meters the air this [iwill cause problems as it will be overfuelling.

I touched on the issue of uprating wastegate actuators while back here . Basically, you should use a higher base pressure spring rather than shorten the actuator rod in order to resist opening under exhaust gas pressure. My actuator has a 7 psi spring which is just under my base boost.
 
1 - 20 of 28 Posts
Top