Mazda 3 ECU swap question

**newguy90** · 21st July, 2019, 04:54 PM

Hello everyone!
I have to replace the ECU of my Mazda 3, year 2003, 1.6 L gasoline. The one from my car is DENSO 279721-2240 and the ECU that I want to put in place is DENSO 279271-3990. Both ECU's are from the same type of engine (1.6 L , gasoline). Do you know if these two ECU's are compatible with each other? Can I safely make the swap?
The old ECU ( 279721-2240) has 93c56 EEPROM chipset (contains IMMO software) and the new ECU has S93A56 EEPROM chipset.
So, do anyone know if the IMMO code can be cloned from my old ECU to the new one?
Thank you!

**rafal** · 21st July, 2019, 04:56 PM

why you can not to buy same ecu as ori?

**newguy90** · 23rd July, 2019, 09:31 PM

Originally Posted by rafal

why you can not to buy same ecu as ori?

Actually, in the beginning, I did just that... I bought a used ECU with the same PCB code (279721-2240), and it turned out that the car has exactly the same symptoms like before the ECU swapping. I don't know if the second ECU has the same problem like my original ECU (could I be so unlucky? ), or if the problem is not the ECU. The Mazda dealer said that the culprit is definitely the ECU, because they couldn't find any other problems. So I manged to get a 3rd ECU ( the one with PCB Code 279721-3990), hoping that this will solve my car's problem. But, after I did some research over the internet about IMMO /ECU's & EEPROMS, I wonder if the IMMO software of these two ECU versions can be copied from one to another.
Thank anyone who can enlighten me on this matter!

**newguy90** · 28th July, 2019, 10:48 AM

No one can help me? Really?

**kiszka** · 29th July, 2019, 12:23 AM

Originally Posted by newguy90

No one can help me? Really?

you are aware that by making copies of the eeprom
you did a copy of the fault

**narkeleptk** · 29th July, 2019, 12:44 AM

Originally Posted by kiszka

you are aware that by making copies of the eeprom

Originally Posted by kiszka

you did a copy of the fault

Copy of the stored fault codes yes, copy of the faults themselves not likely.
What issue's do you have and what is your DTC's?

I do not know the cross reference for the ecu's programming but as far as the eeprom goes, read both and compare structure. If they look the same its probably ok. I would try only swapping immo related data tho.

**newguy90** · 4th August, 2019, 03:00 PM

Originally Posted by kiszka

you are aware that by making copies of the eeprom
you did a copy of the fault

Ummm, I don't think so... As far as I know the 93c56 EEPROM holds just the IMMO and keys data. No DTC appeared after the ECU has been swapped, although the car was acting the same as with the old ECU!

**newguy90** · 4th August, 2019, 03:43 PM

Originally Posted by narkeleptk

[FONT=Verdana]
Copy of the stored fault codes yes, copy of the faults themselves not likely.
What issue's do you have and what is your DTC's?

I do not know the cross reference for the ecu's programming but as far as the eeprom goes, read both and compare structure. If they look the same its probably ok. I would try only swapping immo related data tho.

I'm planning to do just that... I'm awaiting on my EEPROM programmer to arrive ( I ordered a cheap clip on EEPROM Programmer from china, so it takes a couple of weeks until arrival).
The issue is the following: The car died while coasting downhill and refused to start ever since. I pulled over and I plugged in the cars OBD a code reader and discovered 5 DTCs:
P0031
P0037
P0102
P0403
P0443
I've cleared them all, and then I tried to start the car. The engine cranks, but doesn't start. I took it to a Mazda dealer and after testing it for almost a day, dealer said that ECU is the problem... A new ECU is 2000 US dolars, so I took it home and I start investigating the problem myself. With the workshop the workshop manual and Autodata by my side, I checked fuses, wires, sensors resistance, voltages, checked for shorts to ground and batt+... almost everything I could think of!
What I found was: The ECU doesn't provide 12 volts to the following : Lambda sondes (1 and 2),EVAP, CMP, CKP, EGR, MAF and all 4 fuel injectors. All these sensors are wired to the ECU (ports 2T and 2BG) It only provides 12 Volts to the Ignition coils. As I said I checked all the inputs of the ECU (it receives 12V from Battery and Ignition) and all grounds are good also. If I take a live feed from battery and hook it to the MAF power wire the car starts and runs just fine, no DTCs at all! In fact I'm using the car hacked like this for almost a year now!

But I'm not comfortable knowing that something is wrong so that's why I want to try a new ECU to see if the car will be fixed, so that I will not need to drive a 'hacked' car ...
Any help will be appreciated! Thanks!

**narkeleptk** · 4th August, 2019, 11:09 PM

I have not looked up any diagrams or into your problems so I may be wrong but you may want to double check yourself. Usually those type of control signals from ECU are ground signals and not 12V supplys.
Usually the 12V supply comes via a fused connection from fuse box.

also you do not need a programmer. just swap the chips around.

**newguy90** · 5th August, 2019, 07:47 AM

Originally Posted by narkeleptk

I have not looked up any diagrams or into your problems so I may be wrong but you may want to double check yourself. Usually those type of control signals from ECU are ground signals and not 12V supplys.
Usually the 12V supply comes via a fused connection from fuse box.

also you do not need a programmer. just swap the chips around.

I don't want to mess something up in the process of unsoldering /soldering the EEPROM, so that's why I want to copy EEPROM contents. I think it's safer this way, don't you agree?
I attached the electrical diagram if you want to check it out.
Thanks!

**narkeleptk** · 5th August, 2019, 12:40 PM

No fear of risk here either way sorry. I always desolder eeprom's to read them. I have never cared much for reading in-circuit with clips.

..
So you have the 12V supply to pcm but it does not activate the internal B+ circuits.
You ran a 12V jumper which powers the MAF and as a side effect it ended up activating the ecu's B+ circuits (i'm assuming since you said codes went away and runs normal)

Seems like an easy repair. Probably just a diode inside ecu bad where ever the B+ comes in.
OR
Since you said you tried other ecu with same results, it could possible be a wire break inside ecu harness where the wire joins the connector. These are hard to find sometimes because when your probing from back of connector you will test before the break and when you test from the front you can push the terminal just enough to make the connection for false read.

If you where using a DMM then sometimes hairline breaks in the wires will still read as 12V but fail under load. Make sure to try testing the b+ supply on connectors terminal side using a test light to simulate load.

**newguy90** · 6th August, 2019, 12:31 PM

I didn't probe from the wires side of the ECU connector, but from the ECU side by carefully inserting a thin needle in the connector pin holes using an Digital Ohmeter, of course. I found no breaks, shorts to mass or B+ as I previously said. I wil try to check all the ECU harness again by using a test lamp this time, as you suggested.
I also tested for continuity between 2T , 2BG and injector's pins (2E, 2F, 2G and 2L) of the new ECU which I want to put in the car after I'll copy the IMMO from the old one, and I found that all of them are connected (0 ohms between all of them).
So I wonder what could be the problem? Is it a bad diode insde the ECU, like you said, or maybe a false signal / missing signal which makes the ECU to stop from feeding 12 v to all of the sensors/parts mentioned in my previous message?
The only other thing which I could find is that the voltage at the CLUTCH SWITCH (1P) and NEUTRAL SWITCH (1AB) is only 5.91 volts, instead of B+ as it should be according to the workshop manual.