My Experience - 9XX KH Racing Chip
- Thread starter barks944
- Start date
Well I have a later ('86) AFM and maybe loom so a later type DME and ECU would be all required..... possible.....
Thanks guys [
]
Mike
Thanks guys [
]Mike
AndrewS
Member
Yes, that's correct - the DME is the abbreviation used to describe the Bosch ECU (Electronic Control unit). DME is an abbreviation for Digital Motor Electronics (or Digitale Motor Elektronik in German).
By modern standards, the DME is very simple (if compared to a modern Porsche ECU for example). All the function of the DME is within the DME itself, it is not on the chip. The chip contains only reference data which is looked up by the DME to determine the Ignition timing and Injector pulsing. All the data is stored in binary format (i.e it is 8 bits per character). Each 8 bits can be represented as a HEX value (0-255 = 00-FF for those that may remember Hexadecimal from Computer Science at school?) The DME has many sensory inputs and also houses the ICM (Idle Control Module) that operates the Idle Stabilisation Valve (ISV).
When a performance chip is developed, the objective is to increase the performance, responsiveness and economy of the car by altering the reference data on the EPROM (chip). This reference data resembles a matrix where by values are looked up based on RPM, AFM voltage and other input data (such as Throttle Position Sensor, Speed & Reference Sensors, Coolant Sensor etc). Subtle changes to this reference data can alter ignition advance and/or fuel delivery to change the performance characteristics of the engine.
Hence, additional performance can be achieved especially when used in tandem with increased turbo boost, fuel pressure or fuel octane rating. However, there is no logic on the chip "" it contains only data used by the DME (may be referred to as 'operating code', but it is not 'procedural code' or software - it is simply binary reference data).
Regards,
Andrew.
By modern standards, the DME is very simple (if compared to a modern Porsche ECU for example). All the function of the DME is within the DME itself, it is not on the chip. The chip contains only reference data which is looked up by the DME to determine the Ignition timing and Injector pulsing. All the data is stored in binary format (i.e it is 8 bits per character). Each 8 bits can be represented as a HEX value (0-255 = 00-FF for those that may remember Hexadecimal from Computer Science at school?) The DME has many sensory inputs and also houses the ICM (Idle Control Module) that operates the Idle Stabilisation Valve (ISV).
When a performance chip is developed, the objective is to increase the performance, responsiveness and economy of the car by altering the reference data on the EPROM (chip). This reference data resembles a matrix where by values are looked up based on RPM, AFM voltage and other input data (such as Throttle Position Sensor, Speed & Reference Sensors, Coolant Sensor etc). Subtle changes to this reference data can alter ignition advance and/or fuel delivery to change the performance characteristics of the engine.
Hence, additional performance can be achieved especially when used in tandem with increased turbo boost, fuel pressure or fuel octane rating. However, there is no logic on the chip "" it contains only data used by the DME (may be referred to as 'operating code', but it is not 'procedural code' or software - it is simply binary reference data).
Regards,
Andrew.
I don't agree with this Andrew, afaik the chip is an intel 8051 micro processor which contains executable code as well as map data . The DME is an interface board with the 8051 at its centre. The 8051 reads in the available sensor (AFM/RPM etc) data and sets the outputs (injection/ignition etc). When re-programming the chip you can change both the executable code and the program variables (Maps). Most chips are simply re-mapped versions of the factory code, some like the KH/Vitese chips are re-programmed with new executable code as well.
Edit: To clarify, for example the DME needs to be re-programmed to work with MAF sensors. This is because the code uses a transfer function to convert the AFM voltage output into air flow. This transfer function cannot describe the curve of a MAF meter so the chip needs new executable code to handle MAF. This is what vitesse did with his MAF conversion. A chap on rennlist is developing his own code to do this, maybe he will release it at some point and we can all re-program our DME's and use MAF for cheap!
I swaped the engine for an '85 or '86 and it came with most of the bits, but as a couple of the plugs / sockets and vac lines didnt match, I changed all of the bits to match my old loom / ECU. I assumed that the ECU would like the signals from the same sources it was used to!
There are more wires in the newer loom, going through the bulkhead and my car does not have anywhere to put the extras.
Andrew - Computer Science! - I was 15 when the Sinclair Pocket Calculator came out! But I do understand what you are saying, thank you.
Mike
There are more wires in the newer loom, going through the bulkhead and my car does not have anywhere to put the extras.
Andrew - Computer Science! - I was 15 when the Sinclair Pocket Calculator came out! But I do understand what you are saying, thank you.
Mike
AndrewS
Member
An EPROM is not a processor. A processor is the portion of a computer system (DME in this case) that carries out the instructions of a computer program (the DME function which is stored as firmware), and is the primary element carrying out the computer's functions. An EPROM stores data - that's why it is called: Electronically Programmable Read Only Memory (EPROM). An EPROM is the data source for the processor. The DME has a processor, but when the EPROM is changed for a different variant (such as the 9XX KH chip), it is only the EPROM that is changed - the processor remains the same.
The car you were describing (your father's 944 8v - pre 88?) has a separate Intel 8051 Micro Controller which you can see in this picture (http://dorkiphus.net/Motronic/InsideDMELlabeled.JPG) - a 40 pin MC chip labelled as #1. This is an ML3.1 DME used in all 1985.5+ 944/911 models. #4 is where the EPROM is located.
The standard DME can not handle the voltage range of a MAF. That is nothing to do with the EPROM. That's why Vitesse intrdouce a daughtboard/piggy back to the DME to get around the DME's inherent design limitations (reading the much smaller analogue voltage range of an AFM).
An EPROM will store only data (which can be code instructions), but it is not a processor.Â
AndrewS
Member
For those that 'might' be interested in what the components parts of the Motronic are (see link to picture above):
1. S700 - The Intel 8051 Microcontroller that controls the DME. This is the brain the runs everything inside th DME
2. S703 - The ADC0809 8 bit A/D converter. This is how the 8051 reads analog inputs like AFM, Air temp, cylinder head temp It also read Wide Open throttle (WOT) and closed throttle switches, despite them being binary.
3. S100 - F12438 This is one part that's hard to get data on.
4. S701 - This empty socket is where the EPROM goes (for the reference or lookup data).
5. X700 - The Fuel Quality Switch. Changing this changes the fuel mixture/ignition timing (applies a change factor across the full map).
6. Indicates the "digital" part of the DME
7. Indicates the "analog" part of the DME.
8. T560 and T561 - There are two transistors labeled "8". These are ON588 NPN Darlington driver transistors that control the Idle Control Valve.
9. T504 - The Ignition driver transistor. Labeled 1012 in the schematic and on the part itself.
10. D1 - OF622 diode that's part of the injector driver circuit.
11. T404 - RBDT93 NPN Transistor that's part of the injector driver circuit.
12. T402 - RBDT65A NPN Darlington driver that's also part of the ignition circuit.
13. S300 - L4705 5V regulator. Supplies Vcc to the digital board.
14. S800 - CA139 Quad OpAmp. Two of these drive the O2 sensor circuit. The other two are part of the coil/tachometer circuit.
15. T480 - ON588 NPN Darlington driver that controls the fuel pump part of the DME relay.
16. S400 - Labeled simple 0127 - Injector driver circuit.
Regards,
Andrew
1. S700 - The Intel 8051 Microcontroller that controls the DME. This is the brain the runs everything inside th DME
2. S703 - The ADC0809 8 bit A/D converter. This is how the 8051 reads analog inputs like AFM, Air temp, cylinder head temp It also read Wide Open throttle (WOT) and closed throttle switches, despite them being binary.
3. S100 - F12438 This is one part that's hard to get data on.
4. S701 - This empty socket is where the EPROM goes (for the reference or lookup data).
5. X700 - The Fuel Quality Switch. Changing this changes the fuel mixture/ignition timing (applies a change factor across the full map).
6. Indicates the "digital" part of the DME
7. Indicates the "analog" part of the DME.
8. T560 and T561 - There are two transistors labeled "8". These are ON588 NPN Darlington driver transistors that control the Idle Control Valve.
9. T504 - The Ignition driver transistor. Labeled 1012 in the schematic and on the part itself.
10. D1 - OF622 diode that's part of the injector driver circuit.
11. T404 - RBDT93 NPN Transistor that's part of the injector driver circuit.
12. T402 - RBDT65A NPN Darlington driver that's also part of the ignition circuit.
13. S300 - L4705 5V regulator. Supplies Vcc to the digital board.
14. S800 - CA139 Quad OpAmp. Two of these drive the O2 sensor circuit. The other two are part of the coil/tachometer circuit.
15. T480 - ON588 NPN Darlington driver that controls the fuel pump part of the DME relay.
16. S400 - Labeled simple 0127 - Injector driver circuit.
Regards,
Andrew
AndrewS
Member
I would also add, that whilst the 8051 MC used in the DME dates back to 1980, programming it is not like using JavaScript or any other modern high level or 4th generation programming/scripting language. To read the programming used in the ML3.1 Motronic would require a disassembler - just looking at the binary data would be meaningless without a suitable disassembler. Changing any instruction code would result in the function of the DME being compromised (it would very likely not work at all). All EPROMs supplied contain only changed reference data (which is at very specific locations on the EPROM). There are very few programmers around these days that could program a native 8051.
Again, I can state with absolute confidence that custom EPROMS (for the Motronic DME) do not contain modified or changed instructional code.
Regards,
Andrew
Again, I can state with absolute confidence that custom EPROMS (for the Motronic DME) do not contain modified or changed instructional code.
Regards,
Andrew
Sorry, your right about the 8051 being separate from the EPROM. I'm not convinced about the not changed code part though. Sure if you don't know what the code does and you just overwrite part of it randomly its going to break.... But if you know what it does then you can modify it to do whatever you want. Theres a chap on the rennlist forums who is modifying the code to work the MAF sensors. Theres plenty of 0-5v MAF sensors which will work with the 944's DME. I built a MAF converter which I've had running, far from perfect and it was very hard to setup with just a wideband gauge but the engine would rev up and down and idle well. I don't see any reason, why with this modified code and a 0-5v MAF you couldn't run true MAF on the 944's DME without any hardware mods.
An interesting point though is that on my DME only part of the code is stored on external EPROM and part on internal EPROM, so there is certainly a limit to how much of the code you can modify by changing only the external EPROM on 86-87 DME's. I read that the 88- models disabled the 8051's internal EPROM completely and operate entirely from external EPROM. I'm not sure but perhaps only some generic libraries were stored internally.
What makes you so sure that none of the available chips contain modified code? I certainly had that impression for some reason. Although I cannot find the source now!
An interesting point though is that on my DME only part of the code is stored on external EPROM and part on internal EPROM, so there is certainly a limit to how much of the code you can modify by changing only the external EPROM on 86-87 DME's. I read that the 88- models disabled the 8051's internal EPROM completely and operate entirely from external EPROM. I'm not sure but perhaps only some generic libraries were stored internally.
What makes you so sure that none of the available chips contain modified code? I certainly had that impression for some reason. Although I cannot find the source now!
AndrewS
Member
Becuase I have studied many of the bin files from many different chips. To reverse engineer any instruction code would be a massive undertaking and require a lot of very specific knowledge about the 8051, the other components of the DME and their interoperability.
I guarantee you 100% that if I read your car's chip (with an EPROM reader) - the only changed data would be the reference data.
Regards,
Andrew
AndrewS
Member
Hi All,
The ProMAX chip was sent out nearly 5 weeks ago now; hence I too would be interested in how the product compares with the 9XX KH equivalent.
I know that Oli has received it, I would hope that the 9XX KH item has arrived by now (assuming that 9XX / JMG agreed to the trial)?
Regards,
Andrew
The ProMAX chip was sent out nearly 5 weeks ago now; hence I too would be interested in how the product compares with the 9XX KH equivalent.
I know that Oli has received it, I would hope that the 9XX KH item has arrived by now (assuming that 9XX / JMG agreed to the trial)?
Regards,
Andrew
Posts made and opinions expressed are those of the individual forum members
Use of the Forum is subject to the Terms and Conditions
