Power in through Aux1

[Julien]

Hi guys,

the car is running !

I have to trouble shoot a few things but at least it is ideling

But I burned $250 solid state relay because the ECU!

On this 911 you have a châssis relay and the fuel pump relay in one box called DME relay. (This is known to be a source of issues so I replaced it with a solid state one made by an American company few months ago)

The chassis and the FP relays are controlled in low state. The FP coil is powered by the output of the chassis relay. And the coil of the chassis relay is on a permanent power supply and that’s where the problem seems to come from..

I addressed the chassis output to AUX1 (low state) 

And the power supply of the ECU is indeed to a switched 12v

problem is that when the car is switched off, the ECU stays on through the coil of the chassis relay!!!

the solid state one has gone and I had to fit an old genuine one that I saved..

For now I removed the AUX1 pin from the plug and I ground it manually after switching on but I’ll have either to install a relay for it or rewire the fuse box to take the coil power supply from a switched +12..

but I don’t understand that the ECU output are not better insulated to avoid that..

[Confused]

It's all in the documentation - relay coils, solenoids, etc that are connected to Aux inputs must be fed from a switched 12v feed, not permanent, otherwise the ECU won't shut down under some circumstances.

[Julien]

I sadly can confirm that… point is, this wiring was designed like this by Porsche

[Confused]

And that was designed like that for the standard Porsche ECU for a standard car, where you are tied to a certain component on a certain input/output.

The Link ECU has the flexibility to perform a huge number of varying functions on each pin, so the Link supports this, as explained in the manual (with, I might add, more technical reasoning about the inner workings of the ECU than they would be justified in sharing) as to why you might need to re-configure your vehicle to cater for this.

At the end of the day, this isn't the fault of the Link ECU, and all the necessary information is contained within the manual. It's a shame that it seems to have somehow damaged a reasonably expensive piece of kit in the process, and for that you do have my sympathies, despite my matter-of-fact replies thus far

[curly]

I had a similar issue on an AiM PDM32, where the power output I was using for the starter solenoid was staying on, despite my conditions to turn it off were all met. Found out they specifically designated High Power Output 1 for situations like this.

Here's the explanation they give in the PDM manual, see the attached picture:

"H1 output of PDM is a dedicated output protected by a series diode. This diode could be very useful to power devices that, when activated, can generate a current toward the PDM. Of course, H1 output could anyway be used as a standard High Side output.

A typical example of a device that needs such a protection is the starter motor. During the starter activation, a current flow from the PDM output to P1 of Starter. From P1 current goes to GND through the internal aux solenoid to P4 and then through the main solenoid to P3. During this phase, the internal aux solenoid closes the main switch of starter, connecting P2 to P4. The starter motor now is activated and its rotation switches ON the vehicle Engine. When the engine is ON, PDM out must be disabled. Internal diode block reverse current that would flow from P4 (shorted to P2) to P1 and finally to PDM out"

[Julien]

3 hours ago, Confused said:

And that was designed like that for the standard Porsche ECU for a standard car, where you are tied to a certain component on a certain input/output.

The Link ECU has the flexibility to perform a huge number of varying functions on each pin, so the Link supports this, as explained in the manual (with, I might add, more technical reasoning about the inner workings of the ECU than they would be justified in sharing) as to why you might need to re-configure your vehicle to cater for this.

At the end of the day, this isn't the fault of the Link ECU, and all the necessary information is contained within the manual. It's a shame that it seems to have somehow damaged a reasonably expensive piece of kit in the process, and for that you do have my sympathies, despite my matter-of-fact replies thus far

I fully agree, I just spent half an hour to take the main fuse/relay box (where this DME relay is) out of the car to see if I can rewire this coil supply.. I’ll work on it tomorrow.

I expected the Link outputs to be insulated from the rest on the ECU main power circuitry and I honestly didn’t expect this relay to be on permanent feed. I worked on the engine wiring schematic and for that part, I stopped investing at the coil ‘control’ because that was needed to make the new loom.

I’m sad to have blown my beloved static relay, that’s all..

1 hour ago, curly said:

I had a similar issue on an AiM PDM32, where the power output I was using for the starter solenoid was staying on, despite my conditions to turn it off were all met. Found out they specifically designated High Power Output 1 for situations like this.

Here's the explanation they give in the PDM manual, see the attached picture:

"H1 output of PDM is a dedicated output protected by a series diode. This diode could be very useful to power devices that, when activated, can generate a current toward the PDM. Of course, H1 output could anyway be used as a standard High Side output.

A typical example of a device that needs such a protection is the starter motor. During the starter activation, a current flow from the PDM output to P1 of Starter. From P1 current goes to GND through the internal aux solenoid to P4 and then through the main solenoid to P3. During this phase, the internal aux solenoid closes the main switch of starter, connecting P2 to P4. The starter motor now is activated and its rotation switches ON the vehicle Engine. When the engine is ON, PDM out must be disabled. Internal diode block reverse current that would flow from P4 (shorted to P2) to P1 and finally to PDM out"

In my case, it’s not an output staying on, it’s the ECU being powered through an output (and through a relay’s coil)

[Adamw]

This is explained in the help file article: Wiring Information > Output Wiring > Auxiliary Output Wiring > Wiring Auxiliary Outputs > Avoiding Aux Output Back-feeding.

This happens due to flywheeling diodes on the aux outputs, when the ECU is powered down there is a path from the aux pin through the diode back to the 14V rail.  It happens in all aftermarket ECU's with flexible outputs.  

 

 

If you move your relay control to an ignition drive or injector drive then these dont have flywheeling diodes so wont back feed.  

[Julien]

Thanks Adam, I spent a few hours reading through the help file on many subjects but not enough on the wiring side!

Your tip is very helpful

I’ve 1 spare ignition output, this will be the easiest way to solve my little problem.

You’re the best!

[Julien]

Since you solved my problem with the use of an ignition output I will let he fuse box and loom untouched so I worked on that relay.. an IC seems to have seen too much current and overheated at the point of melting the soldering…

I used a heat gun to carefully take it out, a soldering station to clear the board and apply new solder, some flux and heat gun to weld it back in place.. I which I had a proper heat station since the gun is blowing way too much air..

But after a couple of trials I managed to doit!

I tested on bench since I don’t have the car within reach and all seems to work again. I’ll order a new one anyway because I won’t take any risk with reliability but this may works until I get it!

 

 

[Julien]

So I made progress but still not solved.

I moved the chassis relay coil ground to Ign7

The +12v of the chassis relay coil is still permanent.

The chassis relay output powers all auxiliaries (coils, inj, boost valve, purge, etc..)

The condition for the chassis relay is “Batt voltage > 9V” I used that to trigger the relay as soon as the ignition key is on. 

Problem is indeed when I switch off, the relay remains energized and the engine don’t stop running, only way is to take the relay out..

I guess the only way will be to use a spare input connected to a switched +12 and use that as the condition for the chassis relay… 

[Adamw]

Can you draw us a quick schematic.  What is the ignition switch connected to?  

[Julien]

Looks like this, the top part is the DME relay (chassis + Fuel Pump)

Bottom is the power supply of the ECU

i added a relay on the power supply near the ECU because the genuine wiring has a bigger section on the permanent 12 than the switched 12… I guess I’ll add a link to the DI2 as drawn

 

 

[Vaughan]

chassis relay conditions should be on when Ignition sw active shouldn't they? If you force the chassis relay output off after you've turned the key off does that turn off the ECU? Also if you have DI 2 setup as Ignition Sw and it is going inactive when you turn the key off then it will at the very least stop your engine. If the Ignition Sw status is inactive a 100% fuel and ignition cut is applied.

[Adamw]

Yeah I cant really see an electrical path in your diagram that would keep the ecu powered up when that ign sw is off.  The only logic I can come up with is one of the aux devices must still have 12V when ign is off somehow - is it possible one of the aux devices isnt powered by that chassis relay?  There can sometimes be odd paths through other devices connected to the ign switch etc in some OEM wiring too. 

 

 

[Julien]

At the moment chassis relay is on the Aux2 and condition is Vbatt > 9V that’s why the ECU remains under power once the car has been started. The diagram a live is the solution I came up and that I’ll implement next weekend.hopefully after that I’ll be fine

thanks Adam

[Julien]

Problem sorted! With this diagram all works fine now.

my only remaining problem before we can start tuning is that the lambda doesn’t work

I have 2 link can lambda modules.

they are properly powered and they are showing when doing a search device.

I followed the instructions and set one as lambda 2, all fine up to this point but even when the engine is idling I only get 0 at the reading. The error display shows ‘OK’ and is green.. screenshots attached as well as my tune if it can help.’

the only thing ‘wrong’ I can think of is my canbus layout which is lambda 2 with 120ohm resistor - Lambada 1 - Link ECU with in eternal resistor - gaugeArt display with jumper

I didn’t tried to unplug the gauge because it works fine displaying lambda as 0 and all other parameters as per the ECU but maybe it messes with the lambda modules..

 

 

 

 

[Adamw]

Was that with the engine running?  Disabled/Diagnostics is usually because the engine isnt running.  

[Julien]

No, I did these after I stopped the engine… 

Ill check the status with the engine running and let you know.

I didn’t check in the runtime page but no fault were listed in the ECU diag during running.

Could you confirm their is nothing else to do appart from the can setup to get them reading?

[Adamw]

Setup looks ok to me.  Note to test you dont need the engine running, just go to >analog inputs>lambda 1 and enable "run when stalled".

[Julien]

A quick update, with the “Run When Stalled” option to ‘Yes’ both Lambdas are working. 

After the parameter was set back on ‘No’ they were still working with a fault 16 (heated too long) on both…

I’ll do another check from cold start Saturday and check for other faults with the parameter still on ‘No’

But at least I know we will be able to start mapping \0/

 

[Julien]

So I had error 16 33 and 34… according to the manual it was looking like a bad wiring. I had 12.5V but I read I could have been a current restriction so I wired he power supply to another source.

Now they heat up properly with the parameter « « run when stalled » back on NO. But I get under voltage and over voltage errors.. the status stays on Operating and I got readings. Can I disregard these messages ? I mean could they be only peaks while cranking and if they were constant the status would be on diagnostic?

[Adamw]

Keep an eye on it when running to make sure its happy.  Error 16 & 33 or 34 can often be caused by extreme inductive noise on the circuit - such as idle valve or VVT solenoid etc so those will be noisier when running.  It may need the capacitor added as per the alternative diagram if any of those errors cause the status to change to error when running.  

[Julien]

16, 33 and 34 are gone, it was definitely a power supply issue because the sensors wouldn’t reach the minimal temperature to switch on operating, there were always rebooting.

I didn’t save the new error numbers (was between 41 to 48) while cranking.

But the status remains stable on “Operating“ even if the error is still displayed in the error line. It was late so I did only 1 startup.. I’ll do more next weekend!

[Julien]

Hi guys,

I’m starting to check some logs to improve my settings. I noticed the lambda2 value is rounded and shows only 1 so I can’t check the discrepancy between the 2 banks under load that I’m trying to level. I didn’t find a parameter for that, is there one? Lambda 1 is 3 digits and perfectly fine.. see attachment

 

[Adamw]

4 hours ago, Julien said:

I noticed the lambda2 value is rounded and shows only 1 so I can’t check the discrepancy between the 2 banks under load that I’m trying to level.

That problem should have been fixed in the latest PC Link - so check on our website that you are up to date.  I believe it would only happen under an odd situation of freshly opened PC Link with no map open so it took anyone a long time to notice it. 

You can check your PC Link version in >help>about