Hi
I am useing OEM ECU (MED9.1) and emu at the same time. That the MED is happy i share signals between both ECUs…means i connect signal wire e.g. from a pressure sensor to MED and EMU.
But sadly this results in a problem:
If MED supplys the sensor (5V and ground connected to MED) i see a noice on the signal reading in EMU. If i supply the sensor from EMU i see the noice in MED. The noice is stronger when throttlevalve is active. I tried different settings and can’t get rid of this.
For pullup sensors (e.g. coolant T with pullup integrated in MED) i can’t find a solution.
I assume those are balancing currents between the ECUs but i actually don’t understand why they occure. Both ECUs are connected to similar ground and powersupply.
Anyone has expiriences with this problem? Are there any tricks to reduce this problem?
Typical way to sharing analogue sensors is joint sensor ground between two…
Maybe this help
I think i already tried this without success. But the MED has several sensor grounds…i just connected one with the EMU. I try again connecting all sensorgrounds together. Thx for the hint.
One question for starters. Why do you want to use the OEM ECU in parallel?
I will try to explain the issue with a drawing, but first, we must assume some things. Let’s say that EMU PRO is the main ECU that does all the work, and OEM ECU wants to read some sensors. EMU PRO has significant currents flowing through it while the OEM is idling with very small, constant current consumption. We also have separate power connections for both ECUs.
Current issue:
The root of the issue is that the internal ground potential of both ECUs is different. Currents flowing through the devices generate different voltage drops around the system, causing the difference in ground potential. The sensor ground of the device has the same potential as the power ground. OEM ECU measures the signal in reference to its local ground potential, which is different from the potential of the EMU PRO. Moreover, the ground potential of the PRO is constantly changing, and all of that will be registered as signal changing in OEM ECU. That’s why you see more noise when DBW is working. You basically see the current flowing through the ECU as the sensor signal.
Solution:
Connecting the sensor grounds of both ECUs will not solve this issue. Connecting sensor grounds will equalize the ground potential to some extent, but it creates a path for the supply current to go through the sensor ground line, which is the same as connecting the sensor ground to the battery/chassis.
The only real solution is to have shared power supply wires for both ECUs or at least power ground wires. Power ground wires from both ECUs should be spliced together with as short wires as possible and then connected to the battery/chassis with a shared wire. That way, both ECUs should have almost identical internal ground potential, which means they have the same sensor ground potential and no ground offset issues.
I mainly decided to use the OEM ECU in parallel for stuff which is hard to reverse engineer or not readable with the EMU. In my current project main reasons are:
- Immobilzer communication to the vehicle (needed for starter activation, which is done by BCM)
- oillevel/temperature sensor, which is a SENT bus sensor (probably in future i can use GDI driver for that)
Since i have the OEM ECU in the vehicle cause of reasons above, i use it in addition for minor functions or simplify wiring or calibration activities:
- reading of coolant /IAT/clutch/brakepressure sensor, which is sent on OEM CAN stream (so i just take values for EMU from CAN)
- startermanagemente
- vacuum pump control for brake booster
- control of electrohydraulic engine mounts
- Basic CAN stream for the vehicle: i use a free programmable gateway between EMU/OEM ECU and vehicle, where i just modify single signals to the vehicle (sensor and status informations for vehicle from OEM ECU) that the vehicle is happy. Anything else is 1:1 gatewayed. With that i have less effort rebuilding the CAN stream, CAN timing is always correct and every response answer stuff is also covered
In addition i learned that EMUpro is a great logging device. I use it to verify or reverse engineer stuff by connecting it to OEM ECU while it is running…connect it to sensors and actuators and other stuff.
Okay thx for the explanation. Makes completly sense…i indeed discovered some differences in ground and power connection between both ECUs. With both EMU is “closer” to battery at the moment. I will update my wiring and see.
Update:
I did some wild wild west updates on my wiring:
- Connected both ECU grounds
- moved 12V for emu further away from battery
→ Now i have a stable reading in OEM ECU on my two problem sensore (brake vacuum sensor and clt)
Thx guys for help