Fitting an ECU...

Cucullan

Is it hard to fit a new ECU to a 306 HDi? My brother has to get a new one after frying the one in his car trying to jump start it after his battery went flat. The garage are looking for almost €1000 for a new one incl fitting. I just saw one on Ebay for about €300 and might have a go at putting it in if its not too difficult. Here's the one on Ebay http://cgi.ebay.co.uk/ws/eBayISAPI.dll?ViewItem&ih=005&item=150050271058&rd=1&sspagename=STRK%3AMEWA%3AIT&rd=1

greglo23

there usually easy enough to fit. its more than likely fitted in the engine-bay or in the front passengers foot-well behind the dash. it only has one big multi-connector so it just slides in and out.

AMurphy

Can't see how he managed to fry it, can you explain.

However, I thought I read somewhere these ECU on the Diesels with the Electronic injectors were prone to failure.

Maybe it was a failed ECU flattened the battery, what age/year is it.

The other thing, make sure the part numbers of the item you are receiving and what you need/have are the same. Small changes and upgrades are made from time to time.

Maybe use PayPal, so if there is any need to reject the item, you can work through them.

You don't need to remove your ECU from the car to test the new one, assuming there is enough length and flex in the connecting bundle of wires.
Just plug it in hanging loose and it should work.

Mr.Diagnostic

Hi,

This ECU is not "plug and play". This is because the ECU must be matched to the immobiliser. The immobiliser is contained within a body module which is built into the fuse box inside the car.

You have 3 choices.
1. Pay Peugeot a grand.
2. Buy a complete kit second hand (ECU BSI and ign barrel/key)
3. Get someone to remove the immobiliser function from the ECU.


As an aside, I see an average of 3 cars each week with ECU faults that are likely caused by jump starting.

maidhc

The Injectors probably need to be coded to the ECU as well.

franksm

Can't ECUs be reprogrammed though to match the immobiliser, keys, car specifics, etc ?

I know that's the way it is with Mazda and Vauxhall - and with Vauxhall it's just a 60 quid job with a "Tech 2" programmer once the car is brought to the dealer.

prospect

Mr.Diagnostic wrote:

As an aside, I see an average of 3 cars each week with ECU faults that are likely caused by jump starting.

The handbook for my Megane states that you should NEVER jump start the car, or use it to jump start another car, for this very reason.

trotter_inc

I designed and made an ECU for an astra a couple of years ago, if the car is less than 5 years old Id strongly recommend taking it to a dealers to get it done...if anything goes wrong it will cost you more than a €1000 to fix your mess...

Mr.Diagnostic

Hi,

@ Maidhc, No, injectors on this one do not need to be coded.

@ Franksm, Some can be recoded and some cant, this one cant as it is locked to the immob when it is first coded.

@ trotter_inc, What can a dealer do that a good specialist can not? In fact, a good specialist can very often do a lot that the dealer can not. Where is the harn in re-writing a few bytes on an eeprom so as to avoid a large bill and circumvent the restrictive practices of car manufacturers. How could this cause a "mess".....

I would love to know more about the ECU you designed and built. Should Mr. Bosch be worried?

trotter_inc

Mr.Diagnostic wrote:

trotter_inc, What can a dealer do that a good specialist can not? In fact, a good specialist can very often do a lot that the dealer can not. Where is the harn in re-writing a few bytes on an eeprom so as to avoid a large bill and circumvent the restrictive practices of car manufacturers. How could this cause a "mess".....

I would love to know more about the ECU you designed and built. Should Mr. Bosch be worried?

I wasnt ruling out a "specialist"! Cucullan obviously doesnt know how to write anything to an eeprom, if he does start interfering with the ECU then he would cause a "mess"...

Apologies Mr. Diagnostic, we obviously dont know it all...

The ECU was for a 98 astra, built using a PIC16FXXX, coded in C, deployed using MPLAB... anymore questions pm me... sorry for hijacking your thread Cucullan.

Cucullan

I found this place in England www.ecu-testing.com that'll supply a re-furbed ECU with 2 year warrenty for €350 and he said it's an easy enough job just a matter of taking out the old one and replacing it. He said immobiliser etc will be fine because they're programmed the same as if a new one is been put in. What do you think Mr. Diagnostic?

Mr.Diagnostic

Hi,

They may be correct. A few ECU's come new in a virgin state and learn the immob code when first powered up. These types are then locked to that immob unit. It is possible to put some back to virgin state. I did not think yours was like that but French cars are not my forte. I can check that for you. The reason I mention this is because it is possible they are just removing the immob.

There are a lot of people in the UK who do (or claim to do) ECU repairs. Like all things, some are better than others. I have never used that crew but here are a few others that I know for you to try.

http://www.atpelectronics.co.uk/ Very good company who actually do ECU's for a lot of others.

Camco 88 Unit 129
Hoyle Street
Bewsey Industrial Estate
Warrington WA5 5LR
Tel: (01925) 445688
Fax: (01925) 444988 Another very good company.

Bluestreak, These do a lot of French ECU's 00441623753300

There are a few more whose details I dont have to hand so if you dont have any joy with these one's let me know.

Cucullan

Well maybe he didn't say exacly that the immobiliser will be working, I never asked if the immobiliser will be disabled and to tell you the truth I'm too worried even if it was disabled but he did guarantee me that all I have to do is install the new ECU and the car will start first time which is what I mainly want. So basically all I want to know is that if I was to get a refurbed ECU from a decent company in virgin state all I'll have to do is install it and away the car will go, Yes/No?
Thanks for the other number I'll try some of them tomorrow and see what they say.

AMurphy

prospect wrote:

The handbook for my Megane states that you should NEVER jump start the car, or use it to jump start another car, for this very reason.

As an aside, I see an average of 3 cars each week with ECU faults that are likely caused by jump starting.

Someone has to explain this one. "Dos'nt compute Laddie"

You connect one 12V battery to another 12V battery, you still only have 12V. Were it a case of connecting the batt in reverse, or connecting a 24V jump starter item I might agree.

And if you cannot jump start them, how about an external battery charger shoudl you run the battery flat?.

If you can connect a battery charger, I fail to see why not a jump starter (of matching voltage).
I don't hear of any EU cars sold here that come with "no jump start" warnings and I'm sure I would hear of the lawsuits, if nothing else, were what you describe occuring.

Integrated Immobilisers.... never had to deal with one of those yet, thank goodness.

prospect

AMurphy wrote:

If you can connect a battery charger, I fail to see why not a jump starter (of matching voltage).

Well,

I am no expert on the matter, I just remember reading this in the handbook of my car.

I can assume the difference is that when you connect your battery to a charger, it is no longer connected to the ECU. However, when you jump start the car, it is still connected to the ECU.

AMurphy

prospect wrote:

Well,

I am no expert on the matter, I just remember reading this in the handbook of my car.

I can assume the difference is that when you connect your battery to a charger, it is no longer connected to the ECU. However, when you jump start the car, it is still connected to the ECU.

I usually charge the battry in the car just place the charger clamps over the wiring clamps without disconnecting the clamps from the battery.

I'll ask someone in the trade here (used to be a service manager for BMW) if he knows of any restrictions about jump-starting bmw's at least.

I smoked (literally) an alternator regulator by inserting a battry wrong way round one day.... but not an ECU thus far... course there's always a first time.

DublinDilbert

All automotive ECUs have to pass a "jump-start" test and a "load dump" test...

The jump start test is incase the vehicle is jump started from a 24V vehicle ( Jeep / truck )... Typically the 24V is applied to the ECU for a number of minuets, then the voltage it dropped to 12Vs and the ECU must function as normal.

The load dump test is to test the EUCs against load dump. A load dump occurs where a large number of electrical loads in the car are switched off all at once ( say a few blew ), the alternator voltage would start to rise, and the altenator would have to reduce the field current in it's armature, but this takes ( typ 400mS ) as it has a very high inductance. So a load dump test would typically apply 40V for 400mS, after this period the voltage is reduced to 12V, and the ECU should function as normal.

The above tests are just 2 of a large number of tests that would be done on any ECU. These will ensure the hardware is ok under these situations, and may not even be carried out with the full production software in the ECU.


More than likely the hardware in the ECUs is fine, but what has happened is some sort of software malfunction/bug lock up, that's very hard to test for during software / hardware testing. It may also only happen on 1 in every 200 of these cars that will be jump started. More than likely some EEPROM or flash memory has been corrupted, and the ECU is not prepared to run, and goes into some sort of "safe" mode.


BTW when I'm jump starting a car, i never have the engine in the "good" car running, as that can really damage the altenator in the "good" car.

Mr.Diagnostic

Hi,

I think you are mistaken. When you say they must pass a test, who says they must? It would make no sense to subject 12v circuits to 24v. Also if the software was not loaded then the ECU could not be tested that way.
Corrupt eeproms are common, both from spikes and from removing supply. Damage from jump starting is very common, ask the owner of the BMW I am doing tomorrow. Merc's and all French cars suffer from this in particular. Vag dashes die when powered down as do Merc SAM modules. Years ago an apprentice who worked for me boost started a 750 BMW and allowed the smoke to escape from 5 ECU's.

DublinDilbert

Mr.Diagnostic wrote:

Hi,

I think you are mistaken. When you say they must pass a test, who says they must? It would make no sense to subject 12v circuits to 24v. Also if the software was not loaded then the ECU could not be tested that way.
Corrupt eeproms are common, both from spikes and from removing supply. Damage from jump starting is very common, ask the owner of the BMW I am doing tomorrow. Merc's and all French cars suffer from this in particular. Vag dashes die when powered down as do Merc SAM modules. Years ago an apprentice who worked for me boost started a 750 BMW and allowed the smoke to escape from 5 ECU's.

Mr. Diagnostic,

I'm not mistaken, i design automotive ECUs for a living. They do test standard 12V ECUs at 24V! All automotive ECUs must withstand 24Vs for a period of time, that's a basic requirement.

When i said all ECUs must pass a test, i meant that the OEM will ensure that every component that goes into a production vehicle passes very strict Conducted/Radated Immunity/Emissions testing.

You can read all about what sort of testing is carried out at: http://www.fordemc.com/docs/download/ES-XW7T-1A278-AC.pdf
If you go to page 62, CI270, you'll see the 24V jump start test!


I never said there would be no software loaded onto the ECU, of course there would be software, but this would typically not be the production software. There are various reasons you would not use the production software when preforming Conducted/Radated Immunity/Emissions testing.

I've no doubt that some issues do occur when jump starting, but i feel these days it's more than likely software causing the ECU's to lock up ( for various reasons corrupt EEPROM / FLASH ), rather than hardware failure.


why do you say it makes no sense to subject 12V ecus to 24V? ummm what voltage should ECUs be tested up to??

Do you think they just apply a nice 13.5V onto the ECU, and say that works, lets go home??

They will typically apply inductive switching spikes of up to 150Vs both positive and negative onto the power suppy lines to ECUs.