Menu toggle

944 DME Project

List updated, new I would miss something! But yeh they are all generic really, the board is designed to be quite modular so its all down to the software. I'm going to love being able to play with the software got loads of ideas to test out :D.

I'm goint to make something a bit more special for the turbo as I don't think this would be the best way to achieve what I want. Its a little bit more ambitious but its a natural progression from this project and will make use of the core components.
 
I have made some more progress on this project.

AugTronic.jpg


We are very close now to producing the first boards for testing. Cant wait!
 
Fantastic work Tom.
If I understand correctly this will be a completely new ECU reusing the original harness and boxes, but everything else being upgraded as no "part" of the original Bosch Motronic programming will be reused?
 
Thom, I will be keeping the harness, box and the analogue board in the DME. You have probably seen the two boards in the DME, well the one that doesn't have the EEPROM chip and has the loom connector on it (the analogue board) is basically just a driver board, it doesn't have any logic. The entire digital board with the processor the EEPROM and all the other associated circuitry is scrapped. The new board and its processor will have brand new code but the maps will be based on the original factory data.

I will also be reverse engineering the analogue board so I will be able to replace that too. It will probably be upgraded from the original design, to support two injection drives and various other enhancements.

Then I want to take all the design from the two boards and create a single PCB which will take both the connectors in the 951 into one PCB. This will provide a properly integrated EMS that would plug straight into the 951. This is what Porsche should have done rather than just slapping the KLR into the mix...
 
Have you looked into active knock control like with the original set up and a possibility to run a full sequential injection?
 
This board can take the knock input from the KLR so I can do active knock control rather than use the 4deg retard that the KLR resorts too. Its just a matter of writing the software to do it.

Outside of the DME the injectors are wired in pairs. Internally the injectors are actually fired from the same electronic driver. If I could install two electronic drivers I could drive each pair individually, I guess this would be semi sequential though? Otherwise you would need to wire each injector individually and run some additional loom.
 
ORIGINAL: barks944

Outside of the DME the injectors are wired in pairs. Internally the injectors are actually fired from the same electronic driver. If I could install two electronic drivers I could drive each pair individually, I guess this would be semi sequential though? Otherwise you would need to wire each injector individually and run some additional loom.
Click to expand...

I can't remember if we already discussed it, but there is no real benefit in converting to fully sequential, apart from reduce emissions at idle.
 
It allows for a more accurate delivery of fuel. For non-sequential injection you are squirting fuel onto the back of closed valves, which ultimately gets sucked into the engine anyway, but not in a controlled way - you will get a varied amount of it that makes its way into the engine with each induction stroke. When you are looking for very accurate tuning like in a high performance turbo engine or high compression ratio engine then you want to make sure that the fuel you're mapping for is what you get in the cylinder - you don't want one compression stroke to be lean then the next to be rich, then a sudden slug of fuel that has build up on the back of a valve in a droplet which doesn't atomise as it is sucked into the cylinder. It is all small % points of efficiency, which when considered in isolation are not significant, but all add to a system that overall is much better.
 
I'm not sure batch injection is innacurate. After all the 944 doesn't have much trouble maintaining a good AFR. Even fully sequential systems are basically batch firing at higher RPMs when the injector duty cycle is high. I could see the negatives of injecting into a closed port might have an effect on the dispersal of the fuel inside the cylinder more than anything. But apparently it can be good as the fuel evapourates in the hot port and this can actually be a good thing. One of the greatest benefits of fully sequential or rather having control over individual injectors I can see is that you can tune the AFR individually for each cylinder to correct flow imbalances.
 
Sequential fuel injection apparently allows for quicker response from the driver's input, which must be nice on our engines considering the rather sluggish response of the original ECU. This should represent a "real benefit", shouldn't it?
 
I think the big issue with the response of our engine is the lack of a *proper* TPS input. The original DME gets all its info from the slow to respond AFM uness you are going to WOT. If you give the system a step increase in throttle position such that the cylinders suddenly suck a load of air in, you get a situation where the mixture is lean and the ignition is too advanced. This is because the system hasn't detected the change in load thus hasn't changed the fuelling or ignition requirements. TPS lets your system see into the future and be predictive rather than reactive. If you have TPS then you can then squirt a bit of extra fuel in and retard the ignition ready for the increased cylinder fill. Even carb engines had a throttle squirter, there is nothing like this on the 944.

I suspect the first few cylinders must at least misfire when the throttle is jammed down hard. With a proper TPS you can see movement in throttle right as soon as it starts to move so you can start adding fuel immediately. Even with the original DME switching over to the WOT maps, where it ignores the AFM. you have still lost 200ms while throttle travels to WOT position. In this time you the engine has probably spun round 10 times and those cylinders have probably drawn more air than you have fuelled for and the ignition is probably too advanced too boot. You might even get knock from this.
 
Tom, do you mean the TPS unit itself is a poor/dated design or only the way the factory DME handles the signal it gets from it?
 
as I understand it the TPS is really only a WOT switch.
the ECU sees the results of other throttle position movement as a result of the airflow through the AFM and the resulting 0-5v input.
 
The TPS on an 8v non turbo is just that but on a turbo there is also a potentiometer as I used it on my ms setup, what the dme does with that I dont know.
 
I think the dme on the turbo only gets a wot signal lol. The klr gets the potentiometer signal and then makes a wot signal for the dme and only uses the throttle position itself for throttle based boost control.
 

ORIGINAL: barks944

I think the big issue with the response of our engine is the lack of a *proper* TPS input. The original DME gets all its info from the slow to respond AFM uness you are going to WOT. If you give the system a step increase in throttle position such that the cylinders suddenly suck a load of air in, you get a situation where the mixture is lean and the ignition is too advanced. This is because the system hasn't detected the change in load thus hasn't changed the fuelling or ignition requirements. TPS lets your system see into the future and be predictive rather than reactive.  If you have TPS then you can then squirt a bit of extra fuel in and retard the ignition ready for the increased cylinder fill. Even carb engines had a throttle squirter, there is nothing like this on the 944.

I suspect the first few cylinders must at least misfire when the throttle is jammed down hard. With a proper TPS you can see movement in throttle right as soon as it starts to move so you can start adding fuel immediately. Even with the original DME switching over to the WOT maps, where it ignores the AFM. you have still lost 200ms while throttle travels to WOT position. In this time you the engine has probably spun round 10 times and those cylinders have probably drawn more air than you have fuelled for and the ignition is probably too advanced too boot. You might even get knock from this.
Click to expand...

Just a detail, but the DME does not igore the AFM when it switches to it's WOT fuel map - it uses the AFM signal in conjunction with the WOT map. (I know there is not a load dimension to the WOT fuel table but it is still used in the fuel calculation.)

There is a difference in the way the DME uses the AFM at WOT in that the AFM signal has less significance in the fueling calculations than at PT, but it does still use it.

I have a MAP system instead of the original AFM but the DME doesn't know that - it just gets a 0-4.6v signal on the pin that the AFM used to connect to. At WOT I have correct fueling at 10psi or 20psi because the DME is taking this "AFM" signal and using it to derive the correct fueling to deliver the AFR that is implied by the WOT map.

I think that the DME does ignore the AFM signal for WOT timing though, so you need to set WOT timing for the highest level of boost you will run and accept that it will be suboptimal at lower boost levels - hopefully you will fix this in your project!
 
You must log in or register to reply here.

Posts made and opinions expressed are those of the individual forum members

Use of the Forum is subject to the Terms and Conditions

Disclaimer

The opinions expressed on this site are not necessarily those of the Club, who shall have no liability in respect of them or the accuracy of the content. The Club assumes no responsibility for any effects arising from errors or omissions.

Porsche Club Great Britain gives no warranties, guarantees or assurances and makes no representations or recommendations regarding any goods or services advertised on this site. It is the responsibility of visitors to satisfy themselves that goods and/or services supplied by any advertiser are bona fide and in no instance can the Porsche Club Great Britain be held responsible.

When responding to advertisements please ensure that you satisfy yourself of any applicable call charges on numbers not prefixed by usual "landline" STD Codes. Information can be obtained from the operator or the white pages. Before giving out ANY information regarding cars, or any other items for sale, please satisfy yourself that any potential purchaser is bona fide.

Directors of the Board of Porsche Club GB, Club Office Staff, Register Secretaries and Regional Organisers are often requested by Club members to provide information on matters connected with their cars and other matters referred to in the Club Rules. Such information, advice and assistance provided by such persons is given in good faith and is based on the personal experience and knowledge of the individual concerned.

Neither Porsche Club GB, nor any of the aforementioned, shall be under any liability in respect of any such information, advice or assistance given to members. Members are advised to consult qualified specialists for information, advice and assistance on matters connected with their cars at all times.

Porsche
Lockton
Pirelli
Officially approved
Oracle Finance
WEX Esso
Collecting Cars
Archant Dialogue
Back
Top