I have a problem with one of the computers on a 1998 Ford Ranger. One of the computer’s functions is to reduce a variable frequency analog input to a lower frequency digital output. In my case the analog signal is being digitized, but the output frequency is not being reduced.
Specifically: the Vehicle Speed Sensor (VSS), a variable reluctance sensor, generates an AC signal and sends it to the Generic Electronic Module (GEM) via a twisted pair. The GEM is the computer that processes this signal by reducing its frequency and changing it to a digital 5V DC signal. This signal is sent from the GEM to the speedometer and the Powertrain Control Module (PCM), the vehicle’s main computer. The PCM controls ignition timing, fuel injection, transmission shifting and emissions control devices.

The problem is that the GEM is not reducing the frequency of the VSS signal. The un-modulated signal sent to the PCM indicates a vehicle speed much greater than its actual speed. The PCM is programmed to protect the engine by shutting off the fuel injectors at approximately 95mph, equal to approximately 212Hz. At a creep, perhaps 15mph, the VSS output is near 1KHz and it increases proportional to vehicle speed. This is normal, but because the GEM is sending its signal to the PCM at the same frequency, the PCM shuts the engine off. So something in the GEM has failed. But . . .

I have replaced the GEM and the VSS, with original Ford components. I have isolated the circuit by running dedicated wires from the VSS to the GEM, with the same results.

I’ve examined the reluctor for the VSS, which is inside the differential and is in perfect shape

With the speedometer/PCM wire disconnected the vehicle runs reasonably well, but definitely not right, and there is no speedometer reading.

The same symptoms occur on some Ford F-150 trucks when a faulty windshield seal allows water to leak into the GEM, which is near the left windshield pillar. This GEM is under the dash and dry.

On some Ford vehicles the GEM needs to be programmed to match the PCM. This is definitely not one of them.

The GEM also controls things such as the door locks, power windows, dash lighting, etc.

There are no more GEMs in the country for this vehicle. GEMS are very vehicle-specific, so nothing else will work.

I am the vehicle’s second owner and I know for a fact it has never been modified in any way.

So: 1) What’s killing the GEM?, and 2) How do I fix it? With no speed signal, the Check Engine Light comes on.

The truck will not pass a smog test with the light on. When it fails a smog test it can’t be registered for its annual new license plates.

In order to pass its previous smog test I built a Rube Goldberg device that convinced the PCM it was receiving a valid signal, which extinguished the light.

I can post photos and circuit diagrams should anyone desire.

Thanks,
Bob

 

Welcome to AAC
Well, I must say to post good pictures since you have done a lot to identify the issue.

 

Well, I knew I would leave at least one bit out, and it turns out I've missed a couple. First, Ford calls this safety strategy torque truncation. Truncation, because the PCM is supposed to shut the six injectors off progressively, one at a time. In mine, either the shutdown sequence is so fast it seems all are stopping at once, or all are shutting down at once. In fact, it may be the ignition shutting off, as both scope and scan tool show the two systems quitting simultaneously. Immaterial, though, because I know the ersatz speed signal is the cause.
Further updates as they surface, after nap time if at all. I haven't written about this in quite a while and my notes are almost as scattered as my . . . . .

 

Welcome to AAC
Well, I must say to post good pictures since you have done a lot to identify the issue.

Thank you for the welcome. I've browsed, and have been meaning to attend for some time now, but have yet to see a Remedial forum in the non-automotive fora.

Your comment, "I must say to post good pictures" is unclear. Are you asking me to post some pertinent images? If so, what would you like to see? I have photos of the GEM's circuit board, plus some scope and scan tool screen captures and their tabulated data if you think they might be helpful. How about if I upload high-res photos of the board so visitors could blow them up and find the damage I've missed? Anything, to help you to help me.

 

I have replaced the GEM and the VSS, with original Ford components.

You are certain the sensor or sensors affecting the output are producing the proper signals to the GEM?

 

At some point you might want to contact Ford and ask whether there are any Technical Service Bulletins (TSBs) or recall notices that cover these parts and this phenomenon.

 

You are certain the sensor or sensors affecting the output are producing the proper signals to the GEM?

Absolutely. Single sensor, original replaced with brand new OEM Ford, symptom and signal identical to original.

 

I asked you to post some pictures

 

I asked you to post some pictures

From my reply to your reply to my original post:
"Your comment, "I must say to post good pictures" is unclear. Are you asking me to post some pertinent images? If so, what would you like to see?"

Here's the top of the GEM circuit board. Is there an obvious (to the more educated/experienced/knowing than I) point I might begin looking for whatever performs the functions I've described? And any ideas on how to test whatever object might need testing?

 

I asked you to post some pictures

Addendum:
Here's a detail of the above shot.
Also, I spent quite a lot of time looking for datasheets for the processors (Is that the right term?) and found that most, if not all of them are proprietary Ford devices, even the TI one. It's been a year or more, though.

 

I left out a major item in my post and need to re-frame the request. Rather than ferret out the root cause of the failure and fix the original component, the easiest thing to do, had I the knowledge and skills, would be to build a device external to the GEM that would perform the same function.
So, given a variable AC input of 0 - ~2,000Hz, what might I construct that would modulate that signal to 0 - 212Hz? The 2KHz number is a rough guess. As mentioned in my post, 212Hz at the PCM equals approximately 95mph, which this truck will never see.
A close approximation of frequency/speed relationship the PCM needs to see is ~2.2Hz/mph.
I thought I'd seen and Edit Post function somewhere, but can't find it now. This forum uses a slightly different format from what I'm used to so please point out any gaffes and straighten me out.

Thanks.

 

At some point you might want to contact Ford and ask whether there are any Technical Service Bulletins (TSBs) or recall notices that cover these parts and this phenomenon.

Dick:
Thanks for the reply.
I’ve searched for TSBs and recalls and there are none that pertain.
I also have access to one of the best automotive repair databases extant, the iATN. The International Automotive Technicians Network has over 83,000 members world wide and is a marvelous source of both hands-on and theoretical information. This is where I found the information on the Ford F150 pickup truck GEM failures. I met a Ford engineer at one of the iATN conferences who is involved in the design of these systems. When I called him he explained to me that my system was so outdated he couldn’t tell me anything helpful. Figure he was working on a 1998 design in 1994 or 1995, and I contacted him in 2014, when he’s working on 2018-or so designs, that’s not improbable. To ask him to use his time searching for information for me wouldn’t have been proper. He did email OEM information containing an explanation of the system and diagnostic data, which I already had.
I have also spent many hours on the Internet chasing down leads, to no avail. I call this “research” and my wife calls it “wasting time.” I tell her it’s not my fault if a Google search for fuel injection information somehow, to my surprise eventually ends up at an article on a 1950's Aston Martin racing prototype, or maybe even the French Revolution.

Bob

 

I asked you to post pictures to check the caps. As I have repaired a couple of car management units with leaky caps

 

Rmpede, have you had any more luck on this?
How are you sure the GEM has the job of reducing the frequency of the VSS signal? You have a new GEM and VSS and have isolated the existing wiring loom, the output has still not been reduced is that correct?

Does the dashboard have any electronics in it apart from lights and simple gauges?

 

Rmpede, have you had any more luck on this?
How are you sure the GEM has the job of reducing the frequency of the VSS signal? You have a new GEM and VSS and have isolated the existing wiring loom, the output has still not been reduced is that correct?

Does the dashboard have any electronics in it apart from lights and simple gauges?

"How are you sure . . . " Factory specifications and empirical data. Using the Rube Goldberg device I mentioned in my original post, I can generate appropriate signals long enough to clear the diagnostic trouble codes and make the CEL go out. And no, there are no electronics in the instrument cluster that affect the GEM-speedo-PCM circuit.
I've attached a schematic of the VSS circuit input to the GEM and the GEM output to the PCM, speedometer, and cruise control. There is no difference in symptoms whether the ABS module is in or out of the circuit. Same with the speedometer and cruise control. As long as the PCM sees the appropriate signal the truck drives normally.
I've been trying to fix this problem for something like two years, and your questions have had me going over my notes . . . yet again. Trying to remember if I tested the signal ground between the GEM and the DLC (Diagnostic Link Connector), and I'm not sure. I'll do that today and document it. My memory has become less and less functional the last xxxxxxxxxxxx years. Well, it's not exactly less functional, it's just that its functions don't function as well as they used to. I think. You mentioned replacing caps as a fix, and I've done that a couple of times (Have you seen badcaps.net?), and I've looked at the few caps on this board and decided they were probably okay, for lack of bulging, and also because they were most likely okay on the new GEM.
I don't know if you saw my followup post, but what I really would like to do is build or buy a standalone A/D converter/modulator and bypass the GEM entirely. Any ideas on that? I was going to ask the appropriate person(s) if I could tidy up my original post and make it a request for a device that would fulfill my requirements, and stick it in one of the other fora, where more people would see it. The automotive area here seems to be, shall we say, looked down upon? I want my request seen by as many members as possible.

Thanks,
Bob

 

I guess what you need is a circuit that changes a Sinewave to a Squarewave & slow it down.

 

sounds like Schmidt trigger and decade counter should do it.

 

I guess what you need is a circuit that changes a Sinewave to a Squarewave & slow it down.

Exactly. Now, how do I find such a thing. or how do I build one?

 

sounds like Schmidt trigger and decade counter should do it.

Okay, I'll take your word for it. I have a very vague idea of what those devices are, and no idea at all about how to use them to build what I need. Could you be more specific? Specific as in, what parts do I use and how do I arrange them? I think I can figure out how to connect the inputs and output.

 

Did you specify the pulse frequency needed ?
Or did I miss that part.