In order to gain access to the IDM signals we built a special signal evaluation wiring harness which allows easy access to all the wires. The IDM plugs directly into th e harness in order to tap into some of the board level signals inside the IDM. The evaluation IDM is mounted on a piece of LEXAN plastic that has been treated with a static protective spray. A few signals of interest were tapped and brought out onto a break-out-board so that a Digital Logic Analyzer could easily be attached to whole mess. The signals are CID, Fuel Delivery Signal, TTL-CID, TTL-Fuel Delivery, TTL-IDM Feedback, MCU-FF, ??, ???.

Figure 1-A shows what may be an enable signal. This signal can be somewhat tricky to find as it takes place somewhere after IDM power up but prior to the IDM Feedback. A trigger will be added to the IDM power supply to get a better idea how this signal relates to other conditions. Figure 1-B may be a binding signal. It seems that this signal is present some time after IDM feedback. Note that this signal exists even though the engine is off. The TTL side of the of the IDM has no activity when this signal is present. Both Figure 1-A and Figure 1-B were obtained using 20us/d. These are two very different signals with a unified purpose.

Its hard to see in Figure 1-C, but cursors A and B are in two different time samples. You can see from the screen shot that the interval between the two cursors is 53.4 us. Take a moment and look at Figure 1-A, notice the pulse width of 46 us. The scope has a difficult time extracting the exact time domain of the signal due to the impedance characteristics of the CID and Fuel Delivery line this results in the conflicting numbers 53.4 us and 46 us. These are in fact the same signal. Note the signal names on the left of Figure 1-C.

The start up signal is very clear with this divide time. Note that the signal in Figure 1-B is not carried on by the TTL counter parts of the circuit. This leads me to thinking that the signal in Figure 1-B is some sort of binding signal. It probably only exists in the CML domain between two special interfacing chips. The exact purpose of these chips is not fully understood yet(other than to deter modification?). One thing that is strange is the TTL counter parts of the startup signal are inverted but the normal signals are not. This makes the chip in question even more interesting. about this please pass them on.

The results of this test show that the CML chip has a few tricks to play. Note how the TTL counter parts follow the line signal. This is not the case in Figure 1-D. Consider that in Figure 1-D both CID and Fuel Delivery have a low voltage potential but the CML signal makes it onto the TTL counterparts. In this test run the high voltage potential was used and a signal of roughly equal time domain was generated. The CML chip responded very differently to this signal. The more testing that is done on this CML chip the more it looks like a signal HASP.

Figure 1-F shows that IDM Feedback aligns to the Fuel Delivery Signal. Though not shown here, The IDM stretches the pulse width of IDM Feedback in the event of a fault. We do not fully understand the additional pulse width encoding and code number correlation. Further study is needed here but is not likely unless we find good reason to do so. We may provide a screen shot of the encoding in the future.

Once again we see that IDM feed back aligns with fuel delivery. CID and Fuel Delivery signals are not manipulated in any way by the CML chip. MCU-FF logically aligns with the TTL fuel delivery signal, allowing the signal to pass to the Fuel Injector transistor matrix. Note the transients on the CID signal. This is cross talk noise from the high voltage injectors. This is verified by studying the relationship between the transients and the Fuel Delivery Signal. The reason MCU-FF is tapped is because it is possible to simply cut this trace from the host MCU, jump a wire to +5 VDC which enables complete pulse width modulation of the injector driver transistors(sink). This does not enable injector pulse width modulation. In order to gain complete pulse width modulation control over the actual injectors we need to figure out the limiting circuit which controls the Left-Right feeders(source). Currently it is known that the Left Right feeders and the injector driver transistors use a balanced current network via current sense resistors. If the Left-Right Feeders current sense resistor supplies a bias greater than the injector driver sense resistor an injector fault is detected. Early IDM models used a programmable digital potentiometer to calibrate the current references. Late IDM models seem to have omitted this functionality.

Figure 1-F is a screen shot of the IDM while the engine is running.. A few things to note here are that the TTL counterparts of CID and Fuel Delivery Signal are not transformed or manipulated in any way other than standard propagation delay. Note again how TTL-Fuel Delivery Signal aligns with Fuel Delivery. This may be the easiest way to verify that the IDM is functioning normally. This was NOT taken from the test setup. This was taken from the actual truck with the evaluation wiring harness and IDM platform. This was the first test run of both the harness and platform. I know now that i should have had each fuel injector's driver and left-right feeder control logic brought out to the break-out-board. After studying the signals i also have a better idea on how to setup the Digital Logical Analyzer for this configuration.

A little explanation of these signals and their names follows. Note that i use the term TTL throughout this document. Though the IDM technically uses High Speed CMOS, i generalize and use the term TTL both for historic reasons and problem solving.

• CID and Fuel Delivery are the same signal found between the IDM and PCM.

• TTL-CID, TTL-Fuel Delivery, and TTL-IDM Feedback are simply the board level TTL signals.

• MCU-FF, ??, and ??? are immediate signals of interest

  MCU-FF appears to be some sort of MCU validation output control which connects directly with a 2-input logical AND gate. The second input to this gate is the TTL counterpart of the Fuel Delivery Signal. If a certain code is not presented to the IDM under proper conditions the IDM will enter a diagnostic like mode. Under this mode the logical relationship between MCU-FF and TTL-Fuel Delivery Signal overlap exactly 90uS. The output of the 2-input AND gate is directly wired as a master enable in a matrix of 8 2-input AND gates. This matrix effectively drives the fuel injector transistors controlled by the MCU using paralleled injector driver control bits and a master enable bit. Under this Diagnostic like mode of operation RPM codes are permitted to pass to the injectors but Pulse Width Modulation Control codes are rejected and instead the fixed Pulse Width of 90uS is used. The other two signals named ?? and ??? are pins that belong to the very interesting CML chip used in the IDM. This CML chips seems to be a DIFFPECL-TTL / TTL-DIFFPECL transceiver that binds with a 300mv, 6ma signal but in the same token utilizes rail-to-rail swing logic. Very interesting to say the least. There is a possibility that the chip used in the PCM may have binding status output/ I have not found evidence of a signal on ?? and ??? yet. This leads me to believe that the chip used in the IDM does not have parallel binding status output. But i do believe it to have a serial binding out. This means that the validation code is probably only issued once per power up and possible power down.

Here we have a signal sniff from Key-Off through Key-On Engine-Off, Key-On,Engine-On. You may download this file and study it yourself. You will need to download the software first, which is free to use even if you do not own the logic analyzer. Get the software here http://www.pctestinstruments.com/ simply use it in demo mode. Here is the project file. Disregard the cylinder numbers, they were not created correctly and need to be properly aligned.