ECU Part 3 (Ignition Timing)

[JMR_AW11]

I have just rewritten the club magazine article about how the ECU controls the ignition timing for the mk1a MR2.

Click
http://www.jmross.freeserve.co.uk/part3p1.html
For the first draft


Also Click
http://www.jmross.freeserve.co.uk/part2p1.html

for the previous stuff about the fuel injection.

Hope it proves to be of interest...

Jeremy

[jimi]

=D> Another excellent article Jeremy. My admiration for your work increases with every article of yours I read.
Thanks
Jimi

[dex]

Yes, another great article =D> .

I've never come across the bit about max combustion pressure occuring around 12degrees ATDC before, I'd always assumed it was related to rod angularity.

How many kb's of code have you had to wade through and is it all disassembled?

[JMR_AW11]

Yes, another great article =D> .

I've never come across the bit about max combustion pressure occuring around 12degrees ATDC before, I'd always assumed it was related to rod angularity.

How many kb's of code have you had to wade through and is it all disassembled?

Here is an extract from some Toyota Literature about the combustion pressure. I have seen figures of 10 to 20 degrees in other literature.

"For maximum engine output efficiency, the air/fuel mixture must be ignited such that maximum combustion pressure occurs approximately 10'-15' after TDC."

I'm not an engine guru BTW, I really just do electronics. All the theory stuff in the article is derived from reading books etc.

Therefore, any errors/omissions/suggestions about my articles are most welcome.

There is 4kb of ROM in the UK mk1 MR2 ECU.

I have reverse engineered pretty much all of it. Right down to how it generates and stores the diagnostic error codes in RAM! The only bit I haven't bothered with is the section on how it reads the various sensors, I am only interested in the result at the END of that section, i.e. the sensor value returned.

I can actually 'see' the main fuel and ignition program code running in my head as I drive the car (right down to the micro's internal registers and flags). Sad but true...

I plan to write a bit more on the ignition timing in the future. This will cover the various waveforms that occur on the control wires between the ECU and the ignitor.

[dex]

The 12 degrees ATDC comment was more about me learning something than knowing any better I'm no engineer.

4k After looking at quite a few Ford eec-iv codes I'm gobsmacked as to how small that is.

[JMR_AW11]

4k After looking at quite a few Ford eec-iv codes I'm gobsmacked as to how small that is.

I know what you mean. I learnt a lot about efficient assembly coding from studying the code in the early Denso ECUs! They used every trick in the book and I had to write a special disassembler program to cope with these tricks. Sections of the code are written at machine code level. A conventional disassembler would fall to its knees screaming 'illegal instruction, I can't cope!' in quite a few places for some Denso ECUs (especially the US mk1 MR2 ECUs, 17020 and 17021)

The 4kb limit must have been a real pain to them because compromises in the mapping had to be made in the later 17070 and 17140 UK mk1b ECUs.

When I rewrite the stuff on these ECUs you will see what I mean.

The mk1 SC ECU uses a different (bigger,faster) micro with a completely unique instruction set. The SC micro has 8kb of masked ROM (and lots of interesting secrets in its ROM too). ROM space was less of a problem as quite a few bytes are left unused. It didn't stop them using the same efficient coding techniques used in the early ECUs.

'Programmers' today with bloated high level languages like C (and hundreds of megabytes of memory to play with) have got it easy...

[dex]

When I rewrite the stuff on these ECUs you will see what I mean.

Looking forward to it

I definitely think this kind of information should go into the 'knowledge base' section of the site to save it dropping down the list (hint to admin).

[LimeyMk1]

Nice one Jeremy! Do you know the conversion between the 8 bit advance code and the actual advance/retard in degrees?

Would be very useful when I get round to setting up a base map on my MSII.

Ta

Chris

[JMR_AW11]

Nice one Jeremy! Do you know the conversion between the 8 bit advance code and the actual advance/retard in degrees?

Would be very useful when I get round to setting up a base map on my MSII.

Ta

Chris

I must admit I've been lazy on this. I haven't worked out the conversion.

The final ignition timing ends up as a 16 bit number. It is the sum of three 8 bit numbers.

One is the main 3d (or idle map if throttle closed) map, one is the coolant correction map and the other is the idle stability correction map (which is zero for when not idling).

I've just surfed thru the ROM again and dug out the fixed 16 bit ignition timing value for when the T to E1 wire link is fitted to the engine connector as per the instructions for setting the timing.
This should help me as it should be equivalent to 10deg BTDC as per the service info.

I'll try and work out the calibration from that. I could also do it with a scope on the IGT line with the engine running as I tweak the mapping but this would only give a rough conversion.

Edit: looking at the ROM code for the service mode, the ECU jumps over the fixed timing value and reverts to the normal mapping if you touch the throttle. The throttle IDL contact must therefore be in the throttle closed position during the service routine with the T to E1 wire link fitted.

Don't worry if you do touch the throttle, it will revert to the fixed timing once you take your foot back off...

[JMR_AW11]

I reckon I've worked out the ignition advance scaling for degrees BTDC etc.

I'm fairly certain I've got it right as there are some common offsets in some of the code and they are all in EXACT multiples of 1 degree.

The service ignition timing appears to be 10 deg BTDC with T to E1 link and it appears that the timing will jump to 16deg BTDC once the link is removed (at idle).

According to the ROM code the car does NOT have to be fully warm in the T to E1 service mode as the coolant correction is ignored in this mode.

However, the car has to be fully warm for the 16 deg BTDC to be correct (when the link is removed)

If it is not warm the timing will show MORE than 16 deg BTDC due to the coolant advance correction map.

I'll add a load of maps shown in degrees BTDC on my webpage as soon as poss...

[dex]

10 degrees BTDC is the fixed advance Ford eec-iv's use at underspeed rpm's so I think you're right on the scaling.

[ATSAaron]

I'm new to the board and VERY interested. Where can I find part 1?

Aaron