History of MembersPage/PeepPaadam/PorscheTurbo
Older Newer
2006-09-14 05:35:22 . . . . PeepPaadam [Separate page for 944Turbo]


Changes by last author:

Added:
We are installing my V3.3 to my friends 944 Turbo 3.0 (yes, 3.0)

and we're planning to use cam mounted Hall trigger wheel. It's already used on 944T with Tatech EMS and has mounting for two Hall sensors (Honeywell 1GT101DC, datasheet here: http://catalog.sensing.honeywell.com/printfriendly.asp?FAM=solidstate&PN=1GT101DC) and aluminium trigger wheel that has two trigger points per sensor (180 degrees apart, because of the cam mount).

---------

23.07.06

Car is running very good now, we managed to tune it to 1.3 bar and it has absolutely no grip from 1st to 3rd gear!

Problem with the detonation was loose head nuts. The owner used stock 3.0 head-studs with 2.5 head but they were too long and he fabricated spacers for that. But those spacers had different thermal expansion characteristics and the nuts came loose.

Porsche should now produce something like 380-400 hp. We blew the doors off of a BMW 540 that was equipped with Garrett GT35R turbo, it means the car is quick:)

--------

Wiring progress

--------

The wiring is already done and we're going wasted spark with Bosch coil-pack (everything already waiting for install).

Fuel injectors are Siemens 780cc Low-Z.

----

Yesterday we installed VEMS on the 951 3.0 and the installation process was painless. Calibrated things (TPS, battery etc) and got spark with Bosch 2+2 coil-pack.

----

Little update.

Got the car running and driving pretty good, though we have some problems.

We can't get the fuel map to suit high and low loads: tuning the ve table doesn't give effect and if we get the AF ratio ok at let's say 0.4 Bar, it will become pig rich on the idle and low loads and floods plugs. To get the car running good at low loads, the reqfuel have to be 2!!

* We got the car running by setting the injectors in sequential mode (spark is still wasted. With one trigger of course)

* with batch mode it did run very poorly.

** double-check that all GND5 (EC36pin5, EC36pin21,EC36pin22, EC36pin32) and GND (EC36pin26) connected, near (0.2 .. 0.3m or max 0.4m) the EC36 connector. Not properly connected grounds can be one reason for running poor in batch mode (charge-stealing in the intake can be another).

Here's the configuration:

3.0 inline 4 Porsche 944 Turbo.

650cc Low-Z injectors

VEMS v3.3 (firmware 1.0.38), 1 Honeywell Hall sensor.

----

30.06.06

Car runs now pretty ok, but I have the temperature sensors little off: IAT shows 8 degrees higher than ambient and CLT is something like 5 (not verified). I know that I must use EasyTherm but haven't quite understood how to do it.

Sensors are all Bosch, CLT is 944 Turbo stock one, IAT is from Audi (only god knows from what model).

We can't tune it on the boost side before I haven't fixed the temperature thing.

Here are the excel sheet of the temp (both CLT and IAT) sensor impedance measurements: http://www.hot.ee/raceboy911/temp_andurid.xls

And here's the config file: http://www.hot.ee/raceboy911/kert130062006.msq

----

2.07.06

We're having a strange problem.

When going for a drive to tune, we can do it without problems for about 15-20 minutes, after that suddenly AFR spikes lean and no tuning helps that and engine dies.

We have the grounds all ok and checked number of times. Fuel pressure is rock steady also.

What could cause that kind of problem?

* 1.0.36 firmware

* publish mcd mct dumps

* measure again the impedance of injectors (DVM in 20 ..200 Ohm mode). Do the injectors get hot ?

* make MegaTune logs

** We'll check the pulsewidth and all the correction factors, eg. it is possible that coolant reaches max bin, which is (mis-)configured to take away all(or too much) fuel (instead of value = 100).

** with some experience you spot this on LCD, but with log you have more time

* also upgrade to latest firmware if no better idea (there is no known problem with 1.0.36 that can cause this)

Hook up DVM (voltmeter, 20V DC) across the fuel pump and fuel injectors.

* If the voltage drops, we'll check MegaTune logs and configuration.

* If voltage doesn't drop, some mechanical/fuel system problem is likely (maybe injectors clodged due to dirt ?)

----

05.07.06 Update

Here's the sayng: it's all in the details, isn't it?:)

The owner said the injectors to be High-Z, but after hassling with problems, and checking the impedance, they were 2,8 Ohms (Low-Z) after-all!!

Wonder how the VEMS unit held up this current and not giving up :D

Anyway, we got the car running under 1 Bar of boost yesterday, though with rather strange spark angles: @ WOT 6000 rpm it only wanted 3 degrees of advance, otherwise bit of knocking occured (we listened it with stetoscope of course, not ears, otherwise the engine would be destroyed I guess.

* is the base timing (trigger_tooth and ign_tdcdelay setting) verified ?

* what sparkplugs are used ? Sometimes experimentation is needed for turbo engines to find sparkplug type that is good for boost. In some cases a sparkplug will work during short runs and knock on the highway.

----

Knock problem was found. It appeared that the owner used original longer 3.0 block head studs but 2.5 head required them shorter and he used spacers under the nuts. As they had different thermal expansion characteristics, nuts loosened and caused headgasket to leak coolant, oil and god knows what else into the cylinder.

Problem solved by now.