raycm

I let the engine warm up then captured INPUTS and ACTUAL VALUES at idle-800, ~2000, then 3000 RPM, not driven. The WORD doc shows the results.

I know the RPM tracks the tach, to 5000 RPM. The engine and intake temps follow room temp then increase as engine heats. The battery also is correct.
The throttle output is in volts. I put in equation to make into degrees. I haven't measured angles to verify.

The question, how about all the other values?

I can modify the titles of the INPUTS. I can change the memory location where Actual Values are read and change the titles. Input and Actual Values can have equations to modify the outputs. These are the constraints.

FYI: I have found nothing for vehicle speed.

I'm asking to have you verify these values good or bad. If you can provide good values I may be able to modify to correct. Or at least I can remove bad items.

Thanks,
Ray

SMF

Ray,
i will answer with question.... my favorite one... how did you connect to obd1 . Is it hammer? Looks like computer screen.
thx
Marcin

DougB

Ray,

I wasn't aware that the 993 used a throttle position sensor yet you are showing throttle angle for input 0. I always thought input 0 was the MAF sensor. The tipronic cars have a throttle position sensor but I don't believe it's connected to the Motronic unit.

Also, I'm fairly certain there is no vehicle speed available from the 95 Motronic.

Really cool stuff by the way--my hat's off to you.

-doug

Lorenfb

"The throttle output is in volts. I put in equation to make into degrees. I haven't measured angles to verify."

That's correct. The '95 993 DME has a TPS input to pin 53. This pin is the WOT
input on the 964 DME.

"FYI: I have found nothing for vehicle speed."

That's correct, as there's no speed input to the DME,
just an ignition retard signal (shifting) on Tiptronic cars.

Lorenfb

The following DME data are states & not values:

1. Signal Heating (HS) - a switch closure indicating CCU temp **** not fully CCW
2. Ref Mark (TR) - an indication the reference signal is present
3. Fuel Switch (FS) - a switch closure indicating low fuel light

The above determined via a PST2. If not these, then the DIY
software is not reading correctly. So, without FULLY comparing
the data displayed on DIY software versus the PST2, then
"flip-a-coin" on the utility of it!

JasonAndreas

This should solve the MY95 LWF stalling issue at the expense of fuel mileage?

Lorenfb

"This should solve the MY95 LWF stalling issue at the expense of fuel mileage?"

- JasonAndreas -

Do What?????????????

JasonAndreas

If the heat is turned on, the MY95 ECU will not completely shutoff fuel during coasting so that the heat exchangers stay warm. With no fuel/ignition shutoff there is no voltage problem and then there should be no LWF induced stall?

Red rooster

Jason,
Why not just raise the "turn back on " rpm ?

Geoff

JasonAndreas

I've been wondering the same thing but my car is still in pieces so I haven't been able to try (sometime this month I hope to have everything here to put it all back together...). I assumed that all the aftermarket chips that advertise themselves as fixing the LWF stall but don't, modify that table. I know for fact that one chip does modify the table but they still require modifiying the AFM (modding the ISV would do the same). I don't have AC so the stall was never really an issue for me until I noticed that I could cause one by increasing the voltage load on the system. With the CCU heat option I was just curious if it might work. I've never tried and Loren's post was the first time I've seen anyone mention the input pin on Rennlist so I thought maybe he would know? There are probably better ways but this was just an idle curiousity.

Lorenfb

"I've never tried and Loren's post was the first time I've seen anyone mention the input pin on Rennlist so I thought maybe he would know?"

Sorry, I thought you were joking on the previous post.
I forget sometimes that I'm on the "good" forum & not the "other".

Yesterday, while doing "trips" for 993 readiness codes, I tried to identify the
unknowns from the software posted at the start of the thread using the PST2.
Here, I discovered the HS input (pin 41) from the CCU to the DME, which "closes"
when the temp **** is rotated off full cold.

"If the heat is turned on, the MY95 ECU will not completely shutoff fuel during coasting so that the heat exchangers stay warm."

That's interesting, hadn't been aware of that one. Maybe it's a "Euro only".
If that's the case, it would reduce the decel of the RPMs, thus reducing stalling
with a LWF. You could easily make a circuit (one-shot timer) to trigger the
pin 41 (few hundred milliseconds) everytime the idle switch closed.

You're correct about lots of chip claims for fixing stalling with a LWF, but NO real fixes.
The ideal way is to modify the idle code in the EPROM to reduce the RPM decel
when: 900 < RPM < = 1350 and idle switch closed.

raycm

Thanks to all for responses. I'll digest them and then repost. Anyone is wecome to fix stuff.
FYI: Throttle Angle and battery I know work. The other 4 values, I just put titles for what the Shop Manual lists as available in Inputs Menu. And 2 spots always are zero.

Thanks for doing this Doug. Your tops in my book.
The shop manual says vehicle speed is available. But it also says Throttle Angle is an Actual Value.(?)
It looks to me, the 964 sends a serial stream of numbers. So does the 993. But they are not all the same items and not in the same order. Making it a crap shot to figure out what function to assign each number.

Ray

DougB

Ray,

Try these:
Actual Value.MAF Sensor = 0x47;((n*500)/255);#.## Volts
Actual Value.Base Injection time=0x3f;;##.### ms;UINT16

The rpm, air temp and engine temp addresses are proper. I'm not sure about the others.

The basic injection time may be stored as microsecs/2--not sure. There's also an 8 bit version of injection time at 0x3e.

Unfortunatly the scantool software only reads 8 of the 16 input channels on the '95 ECU so you're missing some interesting stuff like the air temp sensor signal which is on a higher channel. I may address this if there's enough interest.

I'm looking at reverse engineered code from a '95 to determine this stuff and I'm a little puzzled because I see the MAF sensor on the first channel and no indication of a throttle position sensor on any of the channels. What's up with that?

-doug

raycm

Actual Value.MAF Sensor = 0x47;((n*500)/255);#.## Volts
0x47 had 46 at 800RPM. Looks like this will give reasonable output. I'll do.

Actual Value.Base Injection time=0x3f;;##.### ms;UINT16
0x3f had 3,2,2 at 800, 2000, 3000RPM. Looks bad. times 2?
0x3e had 41,35,28,25 maybe 800 1000, 2160, 3040RPM. Seems better number.

air temp sensor .
IATS? That looks fine. There isn't a compartment temp sensor like 996.

MAF sensor.
I listed values per RPM for all Inputs in the .doc file. I can't change where these values come from,right? 1st message.

throttle position sensor.
The shop manual in actual values lists throttle plate. But I found only in INPUTS at first location a number that varied from closed to open throttle. Engine not running or running. As voltage it follows the shop manuals voltage. So I made equation to make look like the angles 0 to 84. ~0.5V to 5V. No IDLE or WOT switches. It uses 66 degrees as WOT.

I, obviously, tried a lot of locations. Tell me what you want and I'll do it. I've done some driving and capturing.

I'd like to get ignition timing, and change which the manual says are available. I'm thinking I'll be able to tell if lower California octane is retardiing timing.?

Your ignition equation.
Actual Value.Ignition Advance=0x5d;(((n-0x68)*2075)/255)*-1;###.#°
the n-0x68 subtracts 104 from value. Is this correct? HEX 68
I thought it would subtract location 0x68.

Ray

pncarrerars

For those of you looking to cure stalling problems with lwf.

On my 964 I have been trying the new adaption feature. The adaption figure is nearly always 128 so it looks like nothing happened but by loading the engine during adaption (lights or slip clutch in gear) it comes out at 134. A usefully feature about this for people suffering stalling with light flywheels is that this causes a higher initial idle (about 1100) before settling down to 880 which would probably help cure any stalling problems.

Pete