Moderators: IMOC Moderators, IMOC Committee Members
Luthor ChrisW Lewis Jay Munnsy Screech205 2bornot2b rookie-racer hnygt4 ben_mr2 PFrazier Scotster irish_dan Mrdb Spowy MrToo ast DMC Paul TZed9 << Beta testing offered mr2neil JasonB Gavaroo redred << Beta testing offered Murf jonp02 M0jo aka mike 2Crooky(Alex) Mrfil13 bubbles Dirk I need 3 pleasebobhatton wrote:Name added
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30 bobhatton
alan mc It doesn't have to be that big but I decided to use the existing fixings in the ECU. i.e. the fixing screws screw into existing tapped holes in the ECU metal chassis. Everything is soldered. It does mean that the board can't be easily removed again! As long as the sensor input is 0-5V the ECU can datalog it and mix it in with the usual ECU datalogging requests. Those are there for my use only for testing etc.(panic) switch. If there is a problem during a ROM update (eg laptop failure or cable failure halfway through an update 
) then the user would be stranded without this switch. It overrides into a safe sector of memory that holds a secure version of the factory code so the ECU will run as stock. The user can't alter this bit of factory code. All code and maps are locked to stock settings, but hopefully this switch will never be needed! I thought it was worth fitting for the cost of a switch. The switch contact is normally 'open' so it doesn't do anything in normal mode. (the switch is rated down to -30C temperature) Once it has a bootloader on board it can upload whole new ROM images or just new maps via the dataport. 
This needs redrawing and remaking after xmas.(the one Luthor1 tested for me) squeezed inside the ECU but it wasn't as accessible and it wasn't as securely fastened as this one.
You can see how much smaller the board could be. All the same chips are there, it's just a different layout on the new rev2 board. It would have to be blobbed in with gloopy adhesive etc.
alan mc wrote:bobhatton wrote:Name added
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30 bobhatton
31.
32.
I'm in.
, just needs to be modified to suite the original casing??? case then is it a working unit, and as far as the software to run and experimant with this what would you suggest?? 
..rossdymond3352 wrote:alan mc wrote:bobhatton wrote:Name added
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30 bobhatton
31.
32.
33.
I'm in.
Luthor ChrisW Lewis Jay Munnsy Screech205 2bornot2b rookie-racer hnygt4 ben_mr2 PFrazier Scotster irish_dan Mrdb Spowy MrToo ast DMC Paul TZed9 << Beta testing offered mr2neil JasonB Gavaroo redred << Beta testing offered Murf jonp02 M0jo aka mike 2Crooky(Alex) Mrfil13 bubbles Dirk alan mc rossdymond3352 algie
.. ..PW@Woodsport wrote:for what its worth jeremys done the same thing over on twobrutal,developed a v6 programmable ecu,tested it,promised us cheap
units and then disappeared off the planet.
to get his initial prototypes off the ground.
Now the only person that stood to gain financially was jeremy so we didnt need to do that,we even had members willing to donate that didnt even have
v6's!
Come on jeremy,tell us all whats happened? you have had the full support from two clubs on this and we feel let down.
and I 'gave up' asking you when you were ever going to plug it into an MR2 and tell me how it worked. I made the adaptor and someone tested the ECU...and then disappeared off the planet.
I know you got this because you posted me a V6 ECU. (even though it's not meant for a V6 ECU) (although it could only log one o2 sensor and one of the fuel readouts was wrong for the V6 ECU) we even set up a funding thread were some 25 members were willing to donate money into a pot for jeremy
(in case any readers think I did a runner with the dosh from 25 people)luthor1 wrote:Hi Jeremy, good to see you back!
Give me a call when you can, if you've still got my mobile, or PM me for it!!
Got lots to tell you.
(100x50x 23mm deep size) that I'd like to see running in a real car.'virtual' MR2 on the bench and show you changes to the rev limiter/mapping/boost and this can be seen on the datalogging. (or a bodged lid) then you could take it home to try. Otherwise there's going to be a wait for that new lid spacer to get drawn/made/ It should work great although it will look a bit slow compared to the internal datalogger you saw last time.jtjerry wrote:Hi
hello for everyone
that is my first post there.
how do you think, is that logger can work for my 1992 Toyota soarer 1jzgte? its ecu looks very similar to mr2.
best if mapper will work if not, perhaps logger will do.
I read that it work with camry v6 engine, so perhaps with I6 2.5 litre tt will work too?
most important for me is see what ignition and knock monitor.
(SDIP 64 package 0.07" x 0.75" pitch pins) then it can be datalogged but unfortunately I would need to spend time dumping out the ROM from the MCU chip and working out the unique datalog command set for your ECU type.(not much xmas/new year for me) and I just don't have the free time to look at other ECUs .. before toyota soarer I had Mitsubishi with 4g63 turbo engine. fortunetly it had data logger as MMCD and it work good. I think most important signal is knocking level as ecu correct ignition time as knocking rinse. "phantom" knocking. (say 1500 rpms) in Mitsubishi there was solution for it change stock program in ecu if it have separate eprom. that solve "phantom knocking"."think" you put poor gas so it back ignition for long time/distance (say 100-200 km). and again it can be cure by change in stock "program"."phantom knock" problem with toyotas and if yes can it be cured in program itself."phantom knock" in Mitsubishi, perhaps it help:- been following the thread for some time thanks to the heads up from Andy - great work and keep it up. I'd be happy to do some of the running around required too.- I only have my verniers at home and not my engineers studs so it's a rough measurement. (TPC 89661-1?380)- give me the clearance required and I'll get it done - lift some of the load 
- I grew it myself you knowGersen wrote:Hi Jeremy
As far as the lid spacer goes I can take or make an Autodesk type drawing and get our machinist to knock me one up.
It would be a relatively easy matter for me to design one myself if you let me have the extra clearance that's required
I have a 2nd Rev2 Turbo ECU ready and waiting so I can actually have this built with no other input from you other than the spacing measurement.
A quick measurement shows the hole centers are 144mm wide and 154mm high for M4 clearance
Let me know if this all seems right and I don't have a freaky unusual ECU
This is a hard offer
PM me when you get a moment.
cheers
Tom
I'm certainly open to ideas on the spacer as it has to be cheap to make but also it has to fit securely and not slide around etc.'guides' Also it's not the right finish.(or conductive coated) The screw hole tabs add strength as well. The nice thing is that the engineering company that can make it is only a short walk from my front door!
jtjerry wrote:Its pity you do not have time now.
want asking you about knocking in Toyota.
and in Mitsubishi community was problem described as
ECU senses in some circumstances knocking that was not knocking in low rpms range
second problem in mitsu was if ecu sense knocking it
I think you have most advanced knowlege there for anyone I know about how stock Toyota ecu work.
How do you think, can be
there is link to describing
http://users.wpi.edu/~ktarry/dsmtech/phantomknock.html
thanks
Jerry
the knock sensing in the Denso ECU is very protective of the engine. 4AGZE. I got this info from studying the ROM code for the ECU and also confirmed it from datalogging an ECU on a test bench with faked knock signals.(learned) ranges where it stores its memory of knock history as an 8bit number for each range. It gets stored in RAM. ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; Lets say 3000rpm for the middle of learned rpm knock range 2. (good cylinder filling eg with boost). (no learned knock correction for fuel/ignition and all mapping is at normal at 2000-4000rpm) really bad knock and 'learned' knock RAM value 2 pumps up rapidly to max and you get about 10degrees of ignition retard and some extra fuel as well to protect the engine. (this is the worst case scenario for knock detection level and most initial corrections will be a couple of degrees or less) The learned RAM value stays high/fixed but the ignition retard protection SLOWLY decays away as the ignition retard is slowly removed in tiny fractions of a degree (because there's no more knock) (that's a SLOW recovery!) Ignition timing/fuelling still track as per normal mapping values. (1 degree) and THE IGNITION IS RETARDED AGAIN by the new learned amount and the fuel is increased AGAIN accordingly. (even though there's no knock!) The ECU 'expects' knock to occur because it has data in its learned memory so retards the ignition to try and beat the knock beore it happens. Therefore it takes a while for the ECU to 'forget' the pain from the first big dollop of genuine knock.;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; It takes many cycles of the load threshold to slowly clear the initial knock learned value in RAM (in 1 deg steps) to totally remove the knock protection strategy if no more knock is detected. Think of this as a kind of short term memory about being cautious of being burned again. Eventually the ECU will get bolder again (until it sniffs more knock and the whole cycle repeats) Better to tune the engine so you only ever get small hints of knock worst case. It will probably affect the next runs even if you were able to remap the ignition on the fly between runs (eg with a Unichip). you would either have to reset the ECU or pump the throttle maybe 10 times to clear the RAM.(>5000rpm) the amount of knock correction available linearly decreases so by 6500rpm the knock protection only has a limited influence. .. To get the correct EFI duration in microseconds, just double it. (and fuel) correction kick back in every time the engine load is cycled up and down. Also the learned value (a kind of 'fear factor') gets slowly diminished each time until it fades away. .. 
'phantom' knock these ECUs also have this 'memory' knock to worry about! i.e. it could be modded to only respond to the initial knock signals from the sensor MCU but I would be wary of doing this!'phantom' knock, that is something much harder.(eg ign retard and fuel enrichment) It's extremely unlikely that I know more about preventing phantom knock than Denso/Toyota so this may prove a waste of time... Again, that would be dodgy. The stock sensor system is quite sophisticated for its age. Otherwise there is no detonation protection against poor fuel or reduced intercooling performance etc. i.e. it is clever enough to process the knock sensor data to decide which fuel and ignition maps to use if the fuel quality starts to deteriorate. This can save the engine. (assuming the stock knock sensor detects the knock efficiently)your knowlege about toyota ecu is simple amazing 
