How can make a chip for 14cux

[Belle427]

I cant answer that sorry.

[haaren]

davep said:

If by default you mean Map 0 (aka Limp Home Map], there is no defined data location for a RPM Limit in that map. The RPM Limit is tune specific so only appears in each of the five other maps.

Map 2 is indeed for open loop fuel management. The chip in haren's ECU must have a Map 2 tuned for TVR 400 Precat & 250 bhp, because the 6250 rpm value is unique to that tune. Given that the dates on the ECU case and PROM indicate this is a very early car, it could be that it started life as a non-CAT (TVR may have built a few 'precat' Chimaeras in 1992/3). Upon arriving in Holland the new owner changed the tune resistor to Map 5 for emissions reasons not realizing that the car would now run as a Land Rover 3.9 CAT.

Edited by davep on Thursday 21st August 12:54

I measured pin 5 and 27, this is 3880ohms. Resistor has withe wires so it's cat version. Car is registered as 1995, should always be with cats. I have a picture when it was just in nl, about 13 years ago, back then it already run on 2422. This guy did the rework and put it on a rolling road, it had 265hp. So the 250hp chip is tvr hp or it doesn't have it, camshaft and cat delete would be more as +15hp.
Don't understand how it can be map 5 and 2 together, rg always show map 5. Nevertheless, when I received the post with new chips, I will directly mount it. Hopefully a lot of riddles will be solved.

Edited by haaren on Saturday 23 August 08:43

[haaren]

Forgot to mention yesterday:
In map 5 you should see a target idle xxx rpm, in map 2 you should see co trim voltage.
Mine doesn't show any of them, so maybe it's a map 3,5 or something
Let's see what will happen next week...

O and I had 2 empty spaces in the ECU connector, but I expect this is normal.

Edited by haaren on Saturday 23 August 08:42

[davep]

haaren said:

Forgot to mention yesterday:
In map 5 you should see a target idle xxx rpm, in map 2 you should see co trim voltage.
Mine doesn't show any of them, so maybe it's a map 3,5 or something
Let's see what will happen next week...

O and I had 2 empty spaces in the ECU connector, but I expect this is normal.

Edited by haaren on Saturday 23 August 08:42

Looks like you are missing pin 31 on the 14CUX connector which is the ground reference for the diagnostic data link, hence the gobbledegook coming out of RoverGauge, probably.

You need a new connector, or reinsert a new pin socket.

[CGCobra]

Sorry for the late reply Dave, didn't see this post earlier which is strange as I have replied to other posts since

davep said:

If by default you mean Map 0 (aka Limp Home Map], there is no defined data location for a RPM Limit in that map. The RPM Limit is tune specific so only appears in each of the five other maps.

Well not quite the limp map, what I'm saying is that during initialisation, when running the 'reset' part of the routine some defaults are set-up and, IIRC, maximum RPM is one of these, I think the pointers are also set to map zero which is frequently called the limp home map although I must admit I don't recall any code which actually changes to this map during the fault sections. I could have missed it though, not something I was focussing on.
Connecting RG at this point will return the values read in 'reset'. My recollection is that the map is not established until first invokation of ICI which obviously means we need at least a spark (or possibly several) if not actual engine running to read the map related RPM limit. After connection I suspect RG does not request this parameter again hence will continue to display the value read on initial connection. IIRC the CO voltage may also fall into this category.
I have seen this many times on my own system since first spotting it a short time ago and can reproduce it consistently, just depends on what the engine is doing at time of RG connection.

davep said:

Map 2 is indeed for open loop fuel management. The chip in haren's ECU must have a Map 2 tuned for TVR 400 Precat & 250 bhp, because the 6250 rpm value is unique to that tune. Given that the dates on the ECU case and PROM indicate this is a very early car, it could be that it started life as a non-CAT (TVR may have built a few 'precat' Chimaeras in 1992/3). Upon arriving in Holland the new owner changed the tune resistor to Map 5 for emissions reasons not realizing that the car would now run as a Land Rover 3.9 CAT.

But I think Haaren has already established that he is running the map 5 resistor.
I don't think is is safe to infer what the map value is or that it is changing via reading into the RPM limit, if the map changed I'd expect the map value reported in RG to change also (which is doesn't) although this could also be a 'one time read' parameter value.

6250RPM may be the maximum available in any standard chip but the chip in use may not be standard. It would be quite easy for someone to have altered the max' RPM parameter in any (or all) map(s) on a chip, hopefully they will also have correctly created the map column settings up to this value also as Steve Sprint has in his chips


The above is why I think it is important to start with a known quantity which is what Haaren is trying to do. I have the 2 chips he's requested winging their way to him as we speak so hopefully we will see if that improves the situation. At least is will cross unknown PROM data of the list of things to check.

[CGCobra]

haaren said:

Forgot to mention yesterday:
In map 5 you should see a target idle xxx rpm, in map 2 you should see co trim voltage.

Humm, not true for Map 2. I'm running 2 and I see target idle light, RPM value (which changes due to coolant temp' eventually settling to 900) and also the CO voltage.
I suspect you won't see CO voltage for a closed loop map as the code for this calculates a short term trim value instead. Open loop maps use the short term trim setting to hold the CO voltage and hence add/remove a quantity to the fuelling value.

[CGCobra]

Just a few graphics to compare with RG results on my system. Particularly demonstrating the different max RPM values which dependent on RG connect sequence.


Image taken while connecting RG with engine off


At this point no information is shown for RPM limit although other parameters are being displayed

Leaving RG connected to the 14CUX engine is started.


Engine clearly running and RPM limit is now shown as 5403. This is the value captured in the tune routine prior to getting an ignition signal even though the ECU is now seeing ignition signals. ECU is now using a new values but RG has not queried it.

Also worth noting that I have MAF displaying the linear value and a value, 4%, is present unlike in Haaren's images which show zero for linear reading.
I have taken these readings in RG V11.3 as per Haaren so does not seem to be RG version affected. My value is around what I'd expect with only idle air mass reaching the engine.
Also my TPS is displayed at zero as per Haaren's images. Again what I'd expect for the corrected reading with no throttle pressure active.
Target idle light lit as conditions are allowing this and RPM is 900 (800 basic plus 100 idle adjust as manual 'box)


RG now disconnected then re-connected with engine still running.


Now the RPM limit is showing the 'map specifc' supporting my theory that RG only queries this value on connection.

In Haaren's images the one which shows an RPM limit of 6250 also shows map 5 as active. I believe this image was taken where RG was connected AFTER engine start and the fact that the RPM limit for this map is 6250 ($04B0)

@Haaren
If you could save a copy of your current chip contents via RG (File>Save ROM Image...) and send it to me I'd appreciate it. I'm intrigued to see how it's contents compares to other instances of the R2422 bins and I'd like to confirm that it has been tweaked.

[CGCobra]

By the way. While doing the above I was reminded of something I found some time back, just wondering if anyone could confirm whether this is an RG issue/feature or just my system.

I've found that the fuel pump controls do not work in later versions, I'm seeing this in both v11.2 & 11.3. If I use v10.7 they do work (both the 'once' and the 'continuous') so I don't think it's anything to do with my system but would appreciate confirmation.

[haaren]

davep said:

Looks like you are missing pin 31 on the 14CUX connector which is the ground reference for the diagnostic data link, hence the gobbledegook coming out of RoverGauge, probably.

You need a new connector, or reinsert a new pin socket.

Could be the previous owner has deleted this wire, he has made the USB cable direct in the loom and have direct contact from the dash. I will ask him if he has done this, the other is always empty?
Checked if there was a loose wire behind the connector, but this isn't so no loose wires on inside the connector

Cobra, I send you the ROM image. Please let us know if there are some differences

[CGCobra]

haaren said:

Cobra, I send you the ROM image. Please let us know if there are some differences

I have the file, I'd say the differences are not massive but I need to find some way to quantify. I'll look at this over the next day or two.

[CGCobra]

OK, here's my summary of the differences. I've compared Haaren's file with a file which I have titled "TVR_R2422_400_Precat.bin" which I'm fairly sure I downloaded from Steve S's site. I initially thought there was a great many differences in this file but it seems most are due to a few items squeezed into Haare's version making the remainder of the file uncomparable (or at least not easy to compare).

There are a few differences in what I will call the 'header' section, this code is read for all maps although some of the settings may be overwritten by map specific data.
Below I've listed the differences between the two files and also the comments from a later data.asm file where I can. Later data file has more entries so not always easy to get alignment.

Note - In all the images below Haaren's data is on the left and "TVR_R2422_400_Precat.bin" is on the right.

00CE Road speed sensor fault count, $0800 also used in later PROMS
0145-6 Data file notes indicate "; Used in ICI and main loop S/R" $2EE0 is same as later PROMs
015D Possibly used in idle control increment.
015E-F Data file simply states "Used in ICI"
0232 High speed MUX table, first entry.




There ar another couple of differences in the data for Map 2, obviously these won't make any difference if this map is not used.
045F Data file describes as "yet another fuel map value"
0460 Data file describes as "Startup timer&coolant threshold to inhibit closed loop"
These2 entries as per later data file.




There are an handful of differences in the fuel map area for map 5


I've only had a glance over these, in the early rows fuelling seems to be being decreased, in the higher row there is an increase.

After the fuel map 5 there are some data items.
0793-4 Data file describes as "RPM safety limit", value in line with 6250
079A MIL DELAY?
079F-A5 This section is additional in Haaren's file. Same data as later PROMs
The last one of these is 'interesting' as it is an addition in Haaren's file although the data is the same as later data files.

This is the end of the data which would be written by the 'data.asm' file.

The reason the the file initially looked very different is due to some more data being added in Haaren's file which makes all following data fail to compare even though it may be identical.

08AA Parameter ralated to idle control, setting as per later PROMs
0906-A Additional data, same as later PROMs although note indicates this is not used.




In summary I'd sat that Haaren's file was spawned from "TVR_R2422_400_Precat.bin" or both of these started life from the same file.
Where there are differences the data in Haaren's file is typically in line with what is found in later PROM's so it seems to be a bit of a hybrid (or Chimera).
What it does answer is why the RPM limit of 6250 is seen in some of the images as this is one item which has been altered in this file.

[haaren]

So it's a little mess my chip, something of all and nothing. Now I'm getting interested how it will react on the original chip and if I get a target idle. Waiting sucks

[CGCobra]

haaren said:

So it's a little mess my chip, something of all and nothing. Now I'm getting interested how it will react on the original chip and if I get a target idle. Waiting sucks

I wouldn't necessarily say that, in some ways I was expecting worse! The fact that the modified parameters are generally in line with what was used later means they are probably viable however without knowing exactly what they reflect they could be mixing apples and oranges. Also the fact that at least 2 strings of code have been added could create problems if not done properly without knowing the capability of the programmer there is no way to know. I get the feeling that someone has simply cut and pasted parts from different generations of PROM but as the underlying program code is different the individual parts may not sit together right.

I also wonder if different memory addresses were used to hold the various data items in earlier chips which would trip up Rovergauge. I'm thinking particularly of the fact you are not getting the MAF linear reading.
All things considered it must be better to simply forget that chip (or at least shelve it) and get the new chips in which are a known quantity.
I was hoping they would have been with you by now as they went out on Thursday last week, fingers crossed they'll trap in the next day or two.

[davep]

CGCobra said:

... I also wonder if different memory addresses were used to hold the various data items in earlier chips which would trip up Rovergauge. I'm thinking particularly of the fact you are not getting the MAF linear reading.
All things considered it must be better to simply forget that chip (or at least shelve it)

Map offsets changed (incremented) in 1992/93 during the transition from R2422 to R2967.

Good work CGCobra. For what it's worth I would have run the PROM dump through the dasmx disassembler, but that's being a bit nerdy.

Edited by davep on Tuesday 26th August 13:12

[blaze_away]

CGCobra said:

Belle427 said:

Some guideline healthy readings to expect in my limited experience are:

Maf direct 32%
Throttle position between 2% and 10%
Idle byass between 30% to 40%
Lambdas ideally cycling around the mid point plus or minus %10 ideal.

Good information. Do you know what sort of values you see for MAF Linear?

Here is my data for Direct v Linear MAF values

When MAF reads ......
DIRECT 0.3075
LINEAR 0.04018

[CGCobra]

davep said:

Map offsets changed (incremented) in 1992/93 during the transition from R2422 to R2967.

Good work CGCobra. For what it's worth I would have run the PROM dump through the dasmx disassembler, but that's being a bit nerdy.

Edited by davep on Tuesday 26th August 13:12

Thanks Dave
I just worked with what I've got, this kind of code and the toolsets which work with it are not really in my domain. I'm not really to interested to get too deep into this as, if I want to go further with the engine mangement, I'm probably better putting my efforts into a more modern system but is has been interesting getting involved with this relatively early system particularly given the great work others have put in.

[CGCobra]

blaze_away said:

Here is my data for Direct v Linear MAF values

When MAF reads ......
DIRECT 0.3075
LINEAR 0.04018

Thanks Blaze

So calculated into %ages that will be 30.75 direct, 4.02 linear. The linear is exactly what I got with the engine warm (graphic above), during warmup with RPM a little higher than idle I got 7% linear and 37% direct so I'd say my readings are exactly in line with yours which is comforting!