Remanufacturing Electronic Control Units

Published:  24 October, 2014

By Thijs Jasink, COO of ACtronics Group

What is remanufacturing and how does it work? Remanufacturing always starts with research. At ACtronics we have 17 people in our R&D department who are doing all the work before the remanufacture of an ECU can start. They always start with buying broken core units from dealers. We aim to research at least 50 broken units from one ECU family to get a clear image of the actual problems on an ECU. Additionally, we buy a new OE unit and make hardware cables to connect it to our test machines. We investigate the anti-theft system, write test scripts to release this from the ECU to be able to test it, including CAN and other communication networks.

We do reverse engineering on some parts from the PCB, learn how to read and write to the ECU's memory and investigate hardware and software versions. We find reliable supply channels for the parts and/or machines needed to remanufacture an ECU and conduct environmental tests (shock aging) and, if applicable, vibrating chambers to test the solution. Then we can start with the actual remanufacture process for the customer. Remanufacturing involves replacing everything that could break in the ECUs lifetime. Repairing involves replacing only the parts that are broken, this could be just one resistor.

I would like to give a random example with a 'hybrid' ECU (hybrid in this sense means an ECU with two different technologies combined). This has a bare-die substrate (thick film on ceramic board technology), in a conventional housing, see Figure 1. There are around 100 aluminum wires which connect the bare die to the housing using an ultrasonic bonding technique. The aluminum wires can get loose from the path, which causes an error.

Repairing this unit may mean replacing only one wire. Usually when the first wire breaks, the ECU indicates an error. Sometimes such repairs are done with a manual bonding machine, sometimes by aluminum soldering, as illustrated in Figure 2 - a repair by a German competitor. In any case, both repair solutions have a very short or limited lifespan.

Testing an ECU and bypassing the anti-theft system integrated into the ECU, this needs to be removed before the ECU can be tested on a test bench. Manufacturers are beginning to integrate anti-theft systems in every control unit.

Testing control units

The most essential process for the remanufacture of ECUs is testing these units, whether for diagnosis before or after the remanufacture process. The testing has to be a proper simulation for a car, including a representative load from all actuators. Before an Engine ECU can be tested, the anti-theftsystem needs to be removed or cleared. There are several ways, varying from software simulation of the immobiliser circuit, placing additional software in the flash from the control unit to temporarily bypass or unlock the anti-theft system, or in the worst case, using the corresponding car key and immobiliser.

At ACtronics, we build our own testing machines. The latest machine is our Vision-5. This can easily, quickly and economically reproduce, expanding our rebuilding capacity. We use these machines to do the diagnosis (often by manual testing) and do the end of line test (mostly by automatic testing).

We often are asked to sell the machine, but this would give away our competitive advantage. To add a new ECU family to the testing machine takes between 3 and 6 weeks of work, writing specific scripts and writing very specific software for the anti-theft system and network, mostly CAN signals. If we would sell the machine, it would be useless without the specific software needed to test a specific product. For somebody to write such software, without the complete internal hardware design schedule, is almost impossible.

Where failure/warranty rates from 0.05% or smaller are acceptable for new OE parts during the assembly of the car, for remanufactured parts the practically acceptable range is between 1 to 2%, depending on the complexity of the control unit. This looks substantially higher but it's almost impossible to lower this failure rate in an economically attractive way. The reason why this failure rate during remanufacturing is so high is because there is a certain small percentage (say 0.5%-1%) of ECUs that pass the normal end test but would only fail on a long-term environmental/stress test. The costs of performing such long-term stress/environmental tests, in combination with the additional lead times needed to have every part extensively tested, are prohibitive. The other 0.5-1% are the result of tolerance misassumptions or human error. The missing knowledge about acceptable tolerances is something that can never be wiped out, unless cooperation with an OE/OEM manufacturer is possible.

Further, within ACtronics we see a clear relation between volume and warranty rates. As the company grows, better opportunities are available for better quality control and with higher volume, processes become more controllable and standardised, which result in a lower warranty rate. We believe that in the future, warranty can be below 1% for us.

Figure 2: A competitor's repair using soldering

Figure 3: ACtronics replaces all wires with a bonding machine

Related Articles


Sign Up

For the latest news and updates from Aftermarket Magazine.


Where should the next Automechanika show be held?


©DFA Media 1999-2020
Terms and Conditions