[festy]

fair enough, but FYI it would seem that the more common '117 ECU will run the SZ firmware too, you don't need the rarer '141 ECU if that helps at all.

And on a different topic, I got hold of a S4 spider eprom dump yesterday (thanks to Gubi for sending me the file) and found something I wasn't expecting.
The S4 spider runs a '124 ECU (the same as the 75ie 1.6/1.8 cars) which is the same ECU as the TS, but with an internal ignition amplifier.
The real surprise was that the S4 spider runs the 1.6L version of the firmware, the exact same chip as the 1.6 unless my firmware database is wrong...

 

[Biba69]

festy, or?, what is the 75 QV and SZ? And which is, what I'll call, the common ECU?

Can I assume the SZ is the black part of the map?

 

[festy]

The SZ/RZ is the black map (the one that never gets above 14.7 AFR), the standard 12v 3.0 from the 75QV is the red bit.
So the SZ is a 'hot tune' compared to the 75, I don't know what the engine differences are betweem the two cars, but it gives you an idea of how the factory maps are changed when performance is the goal...

 

[AlfistaChris]

The SZ and 164 are different in shape and weight. The larger 164 has to be refined to meet emission laws while the lighter SZ can run more freely without filling the balloon with too much bad stuff. This is basically what tuning chips do: Find the points that are bad for performance and consumption just to keep emissions down. The top end flat out power does not count in the emission tests, this is why there is usually not much HP to find. But driveability can be much more fun with such a dirty chip. The ignition is retarded for this reason too, to reduce NOx, optimizing it gives better fuel consumption.

The 1.6 running the same software than the S4 is very unusual. I would check the data base again. But this Motronic does run with increased displacement very well. From this point it could well be if the emissions are low enough. Alfa technicians may just have copied it and called it a day...

 

[festy]

Here's another interesting comparison between a "standard" tune and the SZ. If you remember the Accellerator Enrichment (AE) map about 10 pages back (http://www.alfabb.com/bb/forums/engine-management/190352-what-stock-75-ts-motronic-can-cannot-do-21.html#post1116426) it was a very linear progression.
Here's the AE map from the QV and SZ (SZ is black again). The QV's AE map looks much like the twinspark, whereas the SZ map gives a significant amount more enrichment at lower AFM deltas.

 

[Fedezyl]

Probably for better throttle response at the expense of fuel consumption again

 

[festy]

Probably for better throttle response at the expense of fuel consumption again

Absolutely - i posted the comparison of those maps to show how the twinspark's AE map could be adjusted in a similar way, if improved throttle response was desired.

 

[BRUCE]

Fantastic work, a bit lost on a some things but getting the gist of things

A few years ago, I swapped over the QV ECU & AFM (141) with a standard 164 3.0 litre ECU & AFM (117). The difference was immense, the car was a rocket. The difference between the two were higher MPG (117) and lower rev limit, I think about 6000rpm? It seemed to come on cam earlier but appeared to run richer at lower speeds, petrol smell was evident.
Anyway, I changed it back to the QV ECU for MPG reasons.

Have you got a comparison between these 2 ECU's?

Further, the QV ECU now has a rev limit of 7200rpm, never taken it that far and supposed to run it on super unleaded but haven't notice any difference on 95 RON.

 

[festy]

The difference isn't with the ECUs themselves (they're almost identical from what I know) it's the program and maps on the chip that give the difference.
I know of two chip variants used with the 141 ECU - the '097/BF43 (used in 75 and SZ) and the '737/9AE3 which I have listed as being for SZ only - but I'm pretty sure that's wrong because it's a much less performance-orientated tune - unless it was crippled for a specific market, but it's more likely that it was for a 75 or 164....

For the 164's 117 ECU I know of 4 chip variants - '710/02EA, '310/49F3, '096/49FF and '482/5CB4.
Each one of those will behave slightly different - I would imagine the rev limiters are all pretty similar (except for the '097 which has the soft limiter set to only a few rpm less than the hard limiter - it's the for the SZ after all) but the maps for fuel, spark etc will all likely have small differences. I don't know what year/market each of the versions goes with, but would't be surprised if certain markets like NA, Japan, Switzerland etc got different maps to other parts of the world to meet specific emissions requirements, speed limiting etc.

It's a similar situation with the TS - there was only one type of ECU used (the '108) but I know of 7 different chips that were used, and each one behaves differently. Some differences are only minor, others would be quite noticable though.

If your rev limiter is set to 7200, then I assume you have an aftermarket chip fitted, i.e. Squadra or similar. In all the factory chips I've checked, nothing came stock with a rev limit anywhere near that high. Even the SZ has the limiter set to only 6600.

Without knowing the chip numbers in the two ECUs you used, I can't really do much in the way of a meaningful comparison - for all I know you could have had a '097 chip, which is the black 'SZ' part of the comparison maps I've posted recently. That one would be a very noticable difference over a stock chip/ECU

 

[festy]

Bit of an update on the realtime stuff...
My prototype board design just fits inside the TS ECU, but will need some changes to fit in the other ML4.1s - it will work fine in them, but you can't reassemble the ECU because the module hits the ignition driver. Not a major drama, I just need to change the design a bit.

The one module can be swapped between the TS, 33, S4 spider and 3.0 12v ECUs just by changing the jumper settings on the module.

The upload/download half of the tuner pro plugin is working pretty well, I've fixed most of the bugs and it's pretty stable. I'll start on the data acquisition part soon - that's what will handle the data logging, DTC reading/clearing, sensor interrogation etc.

I've patched various factory firmware versions for the TS, 33, spider, 164 12v, 75 12v and SZ so they'll be ready for realtime as soon as I write the XDF definition files for each of those.
I've written a couple of programs recently that will automate some of this, but there will still be a fair bit of manual work required...

Here's a pic or two of the realtime module - it's a little more compact than that infamous leaning tower

 

[LurntWubber]

Good work Festy - I've been lurking in this thread on and off and read it through a couple of times, althought technically im OK working in IT, I'm no programmer when it comes to this sort of thing!
I've been playing with a willem board and successfully rear my eeprom (After snapping pin 27 - DOH!)
Got 10x 27sh512's enroute as I intend to have some spare now!
Got a quick question in regard to the 2.0TS maps binaries.
Did you happen to stumble upon a respository anywhere for developers/mappers of the TS?
I have open access to a local rolling road and im very intrigued to try a few tunes in my TS.

 

[festy]

You're not the first person to snap an eprom leg off, InstiTS did the same thing with a flash chip I sent him (so like you he went and bought a bunch of spares).
If you're doing a lot of chip-swapping (which you have to without some sort of RT/emulator setup) it's a good idea to stick the chip in a socket, and move the socket between the programmer and the ECU.
When (not if) you break a pin off the socket, throw it away and use a new one - they're cheap enough to be sacrificial

No, I haven't found any sort of tuning repository for the TS (or any other ML4.1 for that matter) - if I had, I wouldn't have had to spend months working everything out for myself
Obviously there's some professional tuners that know enough about the TS ECU to be able to tune it (i.e. Squadra and the likes) but I honestly don't know whether they reverse-engineered it themselves, or bought an expensive tuning package that covers this ECU, or if they licensed the proprietary information from Alfa/Bosch, or if they just used an eprom emulator and guessed enough to do what they need to... but however they do it, that information isn't available to the general public that I know of.

What firmware number did you find on your chip? 604 seems to be the most common one, at least here in Australia.
If you're going to play around, start from the stock '191 bin. It's the only one I've created a decent xdf for so far.
I'd be very interested in seeing your results and bin changes, and if you need any help getting started with it all, just ask.

 

[LurntWubber]

Yeah I understand the intellectual property rights etc.
I just wondered if you had happened to read one of the squadra chips yet or attempted any attack on the actualy maps.

Basically the garage I have local to me are kit-car builders and intend to do a fair amount of rolling road tune ups.
These are mostly old pinto or zetec lumps on throttle bodies with megasquirt/jolt setups.
But I know a chap there has a couple of mi16 engines he wants to use in a caterham-esk custom kitcar.
And they to use the ML4.1.
He has a tuning house "generic" chip for his Citroen BX GTi (mi16), but after they had there rolling road installed a few months back, he found out this was causing a really lean mixture and really was no good.

I may have a fish around for some xdf for that bin - what are my chances?! lol

FYI I have a 191 chip in my ecu, appears standard and the ecu looked un-tampered with, but my rev limit appears quite high.
I have managed to read it even with the damaged pin.

Im currently only seeing 126hp at the wheels corrected, but ~140ft/lb torque. I will have to dig out the papers to check but thats it roughly.

 

[festy]

I have no intention of trying to read a commercial performance chip.
The companies that make these put a lot of time and effort into developing these tunes, and if I was to look at one then it would taint my work from that point on.
But with access to a dyno etc, with what I've put together so far there's no reason why you couldn't develop a tune that was just as good as an off-the-shelf commercial performance chip - and if you chose to share your result with us here then you'd make a lot of friends

What's the chip code for the mi16? I can probably put together a basic xdf to get you started pretty quickly if that would help.

 

[LurntWubber]

I will grab the chip from him some time this week and see what it is.
I will try and capture his stock map aswell as the "performance map".

I have no commercial intention for what I'm going to be playing with, so I will host the bin files on my web server, and anyone elses if they fancy contributing.

I can completely understand regd. the tuning house maps, hense my comment regarding intellectual rights.
I'm more interested in the values have been modified as opposed to the exact values anyway.

 

[LurntWubber]

Hopefully not diluting the cause here, but I wondered if you had any input on an error I can't seem to overcome with erasing my spare 27SF512 chips that arrived today.

I'm using a willem 5.0 board and all appears fine, mimiced the DIP settings for the chip and I think I have the jumper settings correct (very unsure).
But on each erase I get failures.

My biggest issue at the moment is that, after snapping the pin off my stock chip I don't have a fallback for my car!
Any help very much appreciated! (*pulling hair out*)

 

[festy]

Are you moving the "erase" jumper when the program is telling you to?

 

[LurntWubber]

Are you moving the "erase" jumper when the program is telling you to?

Hey man,

Yeah, I have a pcb5.0e.

Running v0.98d8
DIP set correctly for 27sf512
jumpers are set so the top set of 6pins make an "L" shape and the lower has the top 2 pins

Chip will read fine, blank check reports data is on the chip
but on erase it will just hang at 0% until it bombs out with an error of "device erase false"

I am absolutely loosing my mind on this one, been researching it for the last 3hrs all over the place.
I'm sure im doing something stupid - I've tested 3 tested good chips with no luck :/

 

[festy]

Have you got power connected to the Willem board? Are you using a USB cable for power, or via the power jack?

Does a read look much different if there's no chip in the ZIF socket?

As for the jumpers, I assume you're starting with the erase jumper off, connecting/removing it when the program tells you to?
Here's a few links that might help:
SST27SF512 Chip Erase
Standard PCB3B Willem Programmer User Guide
pgmfi.org • View topic - Willem Dual Powered Programmer and SST 27SF512
Ezoflash - Willem forum. Board: General. Topic: Cant burn 27sf512 with new pcb5.0lite

And you said you have the '191 chip in your ECU - I'm guessing it's a UK car? What year model is the car? I'd like to start putting together some information on where/when the different versions were used, and yours is the first '191 car I've come across.

 

[LurntWubber]

I have tried powering both from usb and from a 12v dc feed.
I am almost certain (although i will try again) that I have tried both with the jumper in place before erasing and at the first prompt.

I have read the forums you have linked me to already for the most part.
In the wiki's it looks relatively simple! But I'm just not getting the results.

I will look into sourcing an alternative 27sf512 for testing, from a local source hopefully proven working as my chips did just arrive air freight from china :/

FYI It's a early 1991 "H" registration UK model. (2.0TS w/CAT) - although the cat is removed now