I've talked in the past about doing this and finally found some time to outline some half decent plans behind it. Let's start at the beginning though for completeness.
OBJECTIVE - To create a repeatable system on the bench that can simulate the inputs and outputs from a running engine while connected to an ECU... In order to test corner cases and various scenarios with help of a real-time emulator.
The complexity of this is moderate to high, but I have only dabbled on the edges of it it in the past as it would take some concerted effort to make such a beast. As such, I have decided to break it the project into two main parts and then carve it up from there - hopefully in a way where findings can be shared and that others can contribute their own expertise if they so wish. I know there aren't many 'ECU hackers' on this forum but hopefully there is enough that we can all get further down the path of knowledge more swiftly... I can only look at so much assembly code before my brain freezes up - I need a simulated environment where I can control load, rpm, air temp and such to do this properly.
I came back to this project as part of some work I was doing a 20V conversion into a UR which had me poring my head into German schematics for digital dashboards, patrizia lipp autocheck modules and such as I was determined to make as complete a conversion possible with all of the UR's original instruments working. Then I got to realise how this work was complimentary to the engine simulator.
Part 1 - Inputs & Outputs for the instrument cluster and autocheck
The scope of this part is to build a box that simulates all the switches and sensors that feed into the various connectors on the cluster.
These include -
- Oil pressure switches
- water temp / level alarms
- coolant temperature
- fuel level
- rpm
- speedometer
etc
This will serve two purposes -
a. Serve as a bench test platform for clusters and such... I should really try to make this as modular as possible so as to accommodate the UR clusters (and maybe B3/B4 mini-check as well) but that would be a detail
b. Serve as a means of having convenient instrumentation when elements of this are hooked up to the complete engine simulator in Part 2. This will require some connectivity between the two parts such as VSS & RPM for example.
Part 2 - Inputs & Outputs for the ECU
This has much greater scope, but in essence all that needs to be done is to provide the ECU with power, earth and valid inputs that make it think the engine is running. This boils down to G4, G28 and G40 but other inputs will have to be in range so as not to upset the ECU.
I will need to build emulators for those three primary inputs and have them synchronised in such a way that the ECU is fooled into trying to light sparks and open fuel injectors... Flashing LEDs will be suffice for the purposes here !
Ideally I would like to build one that is modular for RR/3B and AAN/ABY/ADU but that is a stretch goal. This part of the project will be broken down into a number of slices that focus in testing the key sub systems of interest... It is that area where some other folk may be able to help.
That's the kick off to the idea floated at least... Coming later is the 1st forays into Part 1 and that G22 simulator.
This thread is going to be really geeky so I apologise in advance !
OBJECTIVE - To create a repeatable system on the bench that can simulate the inputs and outputs from a running engine while connected to an ECU... In order to test corner cases and various scenarios with help of a real-time emulator.
The complexity of this is moderate to high, but I have only dabbled on the edges of it it in the past as it would take some concerted effort to make such a beast. As such, I have decided to break it the project into two main parts and then carve it up from there - hopefully in a way where findings can be shared and that others can contribute their own expertise if they so wish. I know there aren't many 'ECU hackers' on this forum but hopefully there is enough that we can all get further down the path of knowledge more swiftly... I can only look at so much assembly code before my brain freezes up - I need a simulated environment where I can control load, rpm, air temp and such to do this properly.
I came back to this project as part of some work I was doing a 20V conversion into a UR which had me poring my head into German schematics for digital dashboards, patrizia lipp autocheck modules and such as I was determined to make as complete a conversion possible with all of the UR's original instruments working. Then I got to realise how this work was complimentary to the engine simulator.
Part 1 - Inputs & Outputs for the instrument cluster and autocheck
The scope of this part is to build a box that simulates all the switches and sensors that feed into the various connectors on the cluster.
These include -
- Oil pressure switches
- water temp / level alarms
- coolant temperature
- fuel level
- rpm
- speedometer
etc
This will serve two purposes -
a. Serve as a bench test platform for clusters and such... I should really try to make this as modular as possible so as to accommodate the UR clusters (and maybe B3/B4 mini-check as well) but that would be a detail
b. Serve as a means of having convenient instrumentation when elements of this are hooked up to the complete engine simulator in Part 2. This will require some connectivity between the two parts such as VSS & RPM for example.
Part 2 - Inputs & Outputs for the ECU
This has much greater scope, but in essence all that needs to be done is to provide the ECU with power, earth and valid inputs that make it think the engine is running. This boils down to G4, G28 and G40 but other inputs will have to be in range so as not to upset the ECU.
I will need to build emulators for those three primary inputs and have them synchronised in such a way that the ECU is fooled into trying to light sparks and open fuel injectors... Flashing LEDs will be suffice for the purposes here !
Ideally I would like to build one that is modular for RR/3B and AAN/ABY/ADU but that is a stretch goal. This part of the project will be broken down into a number of slices that focus in testing the key sub systems of interest... It is that area where some other folk may be able to help.
That's the kick off to the idea floated at least... Coming later is the 1st forays into Part 1 and that G22 simulator.
This thread is going to be really geeky so I apologise in advance !
Comment