Sorry missed this post... did not want to repost some first steps

Some updates:

I changed the fuel filter...the one that was on there was from 2000. Not a bad idea to replace it of course, but it wasn't the culprit.

I took a PSI reading from the fuel rails, and I'm getting 60 PSI.

I took off the Throttle Position Sensor (TPS) and checked it on the bench. Feeding it 5V, I seem to get about 1.6V when it is closed, and 5V when it is open. I thought it was supposed to be between .5V and 4.5V? In any case, the Motronic DME doesn't seem to be too confused. When the throttle is closed, it registers 0 degrees open, and floored, it registers 61 degrees. I think it may need some work to see it is open more than 61 degrees, but since we are concerned about idle / low range at the moment, I'm not going to worry about that yet.

We definitely are missing the Inlet Air Temperature Sensor (IATS). The DME is defaulting to -33 celsius for the air temperature. I don't see where one was ever placed in the past, and wonder if the fuel map was simply tuned to assume it wouldn't be there and ambient air temperature was constant. In any case, if I want to put one on, where would you suggest mounting it, and which sensor part should I buy? Something compatible with my Motronic DME of course.

Unfortunately, my EEPROM reader/burner has not arrived yet, so I can't read out the ROM so we can see what we are dealing with.

Mount the IAT in one of the air cleaner housings directly overhead of the ITBs on the RHS or LHS. The bigger issues is that you'll need to recreate the IAT Fuel Compensation map. Most likely this map already exists on the chip but was set to 0% fuel compensation across the temp range. This map adds fuel for cold IATs and removes fuel as IATs increase. The first task is finding the map and verifying it and inspecting its current settings.

I suggest not making any changes till you get the engine running as it did back in the day it raced.

Put it on one throttle body. If all the butterflies are in sync you will read exactly 1/6th the airflow. That will give you good data for the DME to work from.

What are the optimum zero and span settings for the TPS...........should it be 0 degrees at closed throttle and 90 degrees at wide open throttle? Is it possible that your system is only providing fuel for a butterfly opened at 60 degrees, while the throttle is truly opened to 90 degrees, resulting in less calculated fuel than the actual requirement? If so, I would concentrate on ensuring the TPS is set optimally.

regards,
Al

If I mount it inside the air cleaner "next to" one of the intake stacks rather than directly in the air stream will that be close enough?

I won't be making this change until I read the chip map and see what's going on in there. ROM reader should arrive today or tomorrow.

Al,

We aren't dealing with any wide open throttle issues yet. We are concentrating on idle / warmup and low partial throttle first.

Noted...my point being that if the required span is 0-90 degrees and your span is set at 0-60 degrees, the required fuel, anywhere within that range may be calculating approx. 30% lower than the actual throttle opening requires. If your throttle is cracked wide open, it is max air intake, but if the TPS is only reporting 60 degrees throttle opening, will it be requesting enough fuel for the correct mix? Sounds like it may be delivering insufficient fuel to provide the correct a/F mix, resulting in a lean condition throughout the range.

This is assuming that the true 0 to span should be set at 0 degrees to 90 degrees and related to the throttle butterfly position.
(I am familiar with Megasquirt only, which sets TPS zero and span by reading the Voltage change from butterfly closed to butterfly wide open.) I.E. 0-90 degrees

regards,
Al

Yes, that is good enough. Actually anywhere near the intake air stream is close enough. It could even be right outside the air cleaner on a bracket. These engine bays run temps that are about equal to IAT because we draw in engine combustion air directly from the engine bay. Also these engine bays tend to run about same temps as ambient air temp for the given day. The main reason the engine bay stays so cool is because the air within the bay is exchanged at a enormous rate because of the huge engine cooling fan within the motor.

The key is that as you go from closed throttle to WOT you MUST see the values increasing. For example: it would not be good if you tap out at 60% throttle. It's not good if 60-100% throttle are all the same value! Most EFI systems have a table that convert the 0-5vdc values into something else, like 0-100% throttle or AirFlow in Liters/minute per stroke.

Do you see the TPS values increasing as you go from closed throttle to WOT? they must keep increasing across the entire range of 0-100% throttle.

Yes, I do. It doesn't got to 62 degrees at like 3/4 throttle and then stay there. It keeps going up all the way to "pedal to the metal."

I've got my chip reader in the mail today, so I'm going to attempt to image the EPROM tonight. Stay tuned...

OK, I have a ROM image of the 3.9L "Offene Trichter" chip. Let me know if you want a link to download it to help me along with this experiment.

I have the bin and can tell this is not a stock 993 chip, the maps on this chip are relocated into different places than the stock 1995 993 chip. Do you have any details at all on map locations? If you do not have this it takes me several hours of time to locate and ID the maps, it's a very time consuming process.

How do you plan to edit the maps? what editor did you intend to use?

From what I've taught myself about the Motronic ROMs the Address Table begins at A458h and ends at A5A6h. I hope. I was going to start trying to decipher the maps this afternoon / evening but would love to divide up the work if anyone is interested/bored. :)

I also see this is different from the stock 993 image which I grabbed somewhere.

I have TunerPro RT, which may be slightly primitive for some things but good enough for lots.

I have a Moates Ostrich 2.0 for emulation / tracing. You think that will be enough to proceed?

I use TunerPro with Ostrich 2 on the 84-89 and the C2 cars without issue. The trick is to build the .XDF config file for tunerpro.

And the address you have for the main table of pointers is correct. But it takes a lot of effort to recognize and ID each map. And you have about 170 maps on this chip!

OK, Lots of progress in the past two days.

With the help of people who know far more about Motronic and Porsche's use of it than I ever will, we have made significant progress. We have located the idle maps for the car, and for some reason, the fuel was dialed considerably back from where it would normally be. Raising the values in this map has brought the AFRs far down and now the car is idling happily with the turn of a key. Why the chip was mapped this way back in 1997 by RS Tuning, we may never know.

There is lots more tuning to be done though as the part throttle map, at least on the low end, is also running lean. But we have progress and a good system to work on it now (TunerPro RT + Ostrich 2.0). We will then install the IAT and use that to fine tune the system for warmer and colder days.

This feels like major progress to me. I can now start worrying about some of the rest of the car. Like the alignment... I drove the car around the block tonight and it was wanting to dart left and right pretty badly. Don't know if that is just old slick tires or what. But it wasn't right. I'm pretty sure I checked wheel bearings and other slop in the suspension when I had it on the rack a month ago. Maybe the toe is just majorly off.

I was even confident enough to put the tail back on the car!

http://forums.pelicanparts.com/uploa...1342140465.jpg

That garage looks way to clean!

Good to hear progress.

Figure out 0-100% TPS? Didn't see it if so.

No, I haven't figured out of the TPS is ok or not yet. I ordered a used replacement and will compare on the bench against mine to see if they operate identically.

Sal, my wife's car is missing in this picture. And I swept the floor two weeks ago. :)