This is for the 1992-1996 trucks.
There is an output from the speedometer servo motor so that when the needle hits the peg the computer will not let the truck go any faster. This is the speed limiter for the truck. There must be a variable resistor inside the servo motor feeding information back to the PSOM (something like a TPS). I am not sure how the PSOM tells the computer not to go any faster but I guess it does somehow.


I got this from an Italian who work for the company that made the servo motor for Ford. He did the design (some of it) the electronic and wrote the software for it at that company while working in Italy.
He said if you remove the upper peg you will not have a speed limiter anymore. The bottom peg is to calibrate the speedometer to zero. I think the electronics and the software was for the PSOM. But they also made the housing, stator windings, rotor magnet and pinion for the servo motor.


I found this interesting so I though I would pass it along although I can not verify any of this but this is what he said and he is still living in Italy.

Here is one of his quotes.
“Well' it's a stepper motor. I know it well 'cause I realize the machines that wind the coils and mount the whole motor (installed in Reading, MA, if you care). Inside the motor there are two mechanical stops, for the minimum and maximum sweep angle of the motor axle. The "zero" post on the speedo is to let the motor execute it's homing procedure aligned to the 0 number printed on your tachometer.
Oh, yes, the other pin is to limit the speed, if you remove it your truck will run like a skulled dog!

I wrote "realize" to be short. In fact what I did was to design (some of) the electronic and write the software of the automated system, then I went to Reading, MA, to reassemble the system after the shipment from the company I used to work for, in the customer site, where they also produce the dial gauge.
You're correct in saying "assemble the entire motor" (housing, stator windings, rotor magnet and pinion).
About the attachment of the motor to the speedo plate, I don't know who perform this operation, probably Visteon somewhere around Detroit, MI.
Sorry if I wasn't clear enough, I'm Italian and my english is far from perfect.

Again, I didn't reassemnle anything, my machanical and electrical colleagues did it, I just restart the system, ran the run-off and teach the operators on how to operate it. “

Well maybe I am wrong and it is a stepper motor instead of a servo motor but is that not a servo motor.

 

Subford:

This is an interesting post. To me especially since my '95 F350 has recently met that 399,999.9 odometer ceiling and returned to 300,000.0. Could you inquire of your Italian acquaintance if he could offer any insight on this engineering faux pas?

Thanx,
dn.

 

Interesting, but not right, the ECU does the speed limiting buy using the PSOM signal. The pegs are there so the needle will not spin around and break the hair spring of the needle. I have heard of this before and know a few Mustang owners that munged up the speedo trying to get more speed.

I did build a circut up that fooled the ECU into thinking it was half actual speed, for that Mustang owner years ago, it took the PSOM/speedo signal and divided it by 2 then sent to the ECU. Took a few years for the race officals to figure out what we did, hehehe(bone stock car raceing, NO mods to car)......

 

I have ask the question for you but do not when or if I may get a reply.

 

I did not think that a speedometer had a hair spring of the needle to break in the 1992-1996 years.

 

The speedo of this year range is a stepper motor. It has four terminals, and is a three phase stepper with a ground, driven by a custom processor in the PSOM module that attaches to the four stud terminals.

The speed sensor (vss sender) is electronically identical to an ordinary ABS sensor, whereas it has a magnet in the center with a coil around it, and the changing proximity of steel or iron will introduce pulses into the wires feeding whatever. The "tone ring" in the rear axle has either tabs or slots depending on the style, and as these "wave" near the sensor's head they interrupt the magnetic field around the magnet which the coil picks up as a signal.

The VSS sending unit feeds the ABS unit, where the bi-polar signal is conditioned by an operational amplifer (op-amp). This op-amp in turn feeds digital circuitry which in turn feeds the processor of the ABS unit to control the rear brakes on older models, and all four wheels on newer models. The signal is also tapped right after the first gate of the digital circuitry and fed to the PSOM module as a digital signal with a 5V swing referencing electronic ground, and that signal is used by the PSOM module to increment the LCD display that's the odometer, as well as a serial flash chip in the processor that stores the mileage. This is why the odometer reads as it did when last powered off - it's stored in flash.

Depending on the year, that tapped digital speed signal may be fed to the PSOM exclusively, and the PSOM in turn feed the EEC, though I have seem cases where clearly the ABS output fed both the PSOM module and the EEC simultaniously, in parallel. In the 92-96 E-vans and F-trucks there is no fancy signal - speed is represented by a bunch of pulses. Go faster, get more pulses per second. Go slower, get less pulses.

When you power off the vehicle, the PSOM module uses power from the "batt" line to move the pointe to zero. If you are in an accident, the PSOM does not do this, leaving the speedometer displaying your speed at impact. Having endured this displeasure a few times, this "feature" doesn't work consistently. Sometimes it zero's out anyway, sometimes it goes part way down, sometimes it retains the speed at impact.

There is no hair spring either. If you were to take your cluster out, turn it upside down with the gauges facing you and shake it, the speedo needle will rotate around. It's counterbalanced by the design of the needle so it's difficult to do, but if you sling it around enough you can "set" the speed to anything you like. Stepper motors have an inherent "step" hence why they are called that and a certain amount of force is required to move the needle a single step. The coils, when energized, overcome this quite easily. Humans, if they shake the cluster hard enough, can also overcome this and set your cluster for "65" if you wish through deliberate shaking and rotating. Once powered, it will step back to zero by design. It's coded to step 256 times to the left when you turn the key to "zero" the speedo, so it's accurate at all times. It knows that 55 miles per hour is 166 pulses to the right after being zero'd out.

Wobbly needles in these clusters is either a weak stepper motor, likely with one of the three phases burned out, or the drivers on the board are either shorted or blown open. The stepper is expecting a three phase input, and if it does not receive it, it will act in pecular ways.

The mechanical stops prevent the needle and the low-power stepper from going past the display markings and that is all. Speed limiting is done completely by the EEC and earlier years did not have this feature at all - instead they only had RPM limiting. If you were to limit the RPMs to say, 5700 then through simple math based on tire ratio, gearing, engine power, you'll know what your maximum speed will be, assuming you can hit it with an incredibly heavy vehicle with a bone stock engine as you bought it.

Later years had speed limiters as well as rpm limiters, and different years produced different results.

The trucks always lagged in technology and code as compared to the cars, by about three to five years depending on the vintage we are talking about.

Around 2002 that's when code-wise trucks and cars were more similar, especially with the introduction of a full CAN network whereas the devices in the car (ABS module, body module, EEC module, and the cluster, all talk to each other serially. If you have an oscilloscope and a lot of time and patience, you can actually decode it to the point where you can see and understand that when you first turn the key to the run position, the EEC says "Hi, I'm an EEC, what are you?" and all the modules in turn announce their existance, their code revision, and say hello. These modules share everything - they are BFF's in shiny boxes, and have no secrets.

 

frederic,
That is more or less what I though before I read what the Italian said.
As I think I said I was just quoting him and did not say I believed what he said but was just passing it along for what it was worth.
There seems to be very little information out there on this and your response was what I was looking for.
Do you have any sources for this information or was it from your own hacking into the speedometer?

 

Very interesting discussion but over my head. However i thought the IAC was a stepper motor as opposed to a servo? (I know nothing to do with this, just an example.)

 

After doing some looking into this I find that the Stepping motor used is a two phase motor with no ground.
I did verify this on my test bench a few minutes ago that the two winding in the motor are not connected and there is not a ground.
Also stepping motors do not have feed back.

The stepping motor:


/

 

The wiring aspect (abs to psom to eec [or] abs to psom/eec in parallel) was discovered by dismantling wiring harnesses of various years, tracing wires from the firewall connector throughout the dashboard area. My doing this was mostly trying to find a more hackable EEC than the m2d2 unit that comes in there. See, I'm used to GM stuff that's well documented and it's a matter of firing up a 68HCxx decompiler and having at it. I struggle with the Fords.

The stepper motor analysis was based on measuring out a dead cluster that passed through my truck - three of the terminals had the same resistance to one terminal, and twice the resistance to each other, which is why I concluded it's a three phase stepper because that's how they measure out.

I would find it odd that it would have multiple shorts whereas the resistances would be so close in that it would appear this way, so I thought my conclusion was reasonable. This is based on one cluster, which I thought had a bad PSOM module. Maybe that's why the PSOM module was dead - the stepper was shorted in multiple places? Now I am baffled. If I still had the cluster I could de-shell the stepper and unwind it. That would confirm for sure what it is. If anyone has a spare/dead one, ship it to me and I'll carefully mill a slot down the side and peel the case apart and unwind it.

Regardless of whether it's unipolar or a three-phase, there is no feedback at least in the traditional sense whereas one would expect either am absolute or relative position sensor mounted on the back. With a little driver trickery, one can easily determine if the stepper actually payed attention to the instruction, by measuring "back EMF" as it steps. I seriously doubt the PSOM module has this capability. It also ensures a zero display by repeatedly stepping to the left with the expectation that the needle will rest on the stop and not fall off. I did an experiment a few years ago by trimming the needle so that it wouldn't hit the stop, and it rotated 270 degrees from it's current position to the left, then stopped. I counted 256 pulses in my pulse counter every time, regardless where the needle ended up. That's how I got that figure.

And it's cool passing information from any source, even if it's potentially inaccurate (as you thought your italian connection might be) because it sparks thought and discussion, and together we can reverse engineer it as necessary.

I reverse engineered various aspects because I wanted to convert my truck to EEC-V as well as "go digital". I've done the latter already as you know.



My only "gripe" about the conversion is the fuel flow sensor requires the transaxle module, which requires an automatic transaxle to "control" in order for that module to send any data to the digital cluster. I haven't been able to fool the transaxle module into thinking it's connected to something yet, but I also didn't give it much bench time.

My truck currently has two PSOM modules... the one that's integrated (and not seperate) from the digital electronics of the winstar cluster I grafted in, and a "spare" 92 F150 PSOM module that I removed all the unnecessary bits - the stepper, needle, housing, and so on. The reason for the cascaded PSOM module functionality is it gave me two things:

1. I can push the buttons to reprogram the dividing factor should I ever go with larger tires (the windstar cluster has this value fixed and it was easier to cascade PSOM modules than reverse engineer and rewrite code)

2. The spare F150 PSOM module drives both the windstar cluster AND the EEC (and why I was so interested in the series/parallel differences across years - I didn't want to blow the PSOM module every three months as it's mounted on the firewall now and a real PITA to get to... I have to remove the dash.

 

I think the IAC is a spring loaded solenoid. It has no feedback so I would not call it a servo.
It does not come close to how a stepping motor works either.

I think the IAC valve is pulse driven by the computer and the length of the pulse says how far it opens to meter the air bypassing the throttle plates.

 

The IAC is a variable position solenoid, with a spring return. It's actuated through a combination of pulse width and speed. There are times where it's simply fully powered and there is no pulsing, such as when you start the vehicle on a really cold day for the first time. Either that or my o-scope can't see the pulses. It's 100 years old so that's entirely possible.

 

donnor,

Here is the information that was sent back from the Italian acquaintance.
It seems I guessed wrong and he did not write the software for the PSOM.

He says:
"It seem I wasn't clear enough.
I wrote the software of the automated system that wound the coils and assemble the motor, not the sofware that control the motor itself.
But maybe I could be of some help.
As for the pinout, you have four pins because thare are two coils in the motor, each with two terminals. The control board send a pulse train to the coils (0-5Vdc), with a phase difference of 90° one from the other.
When a positive pulse energize the coil, this is creating a magnetic field that made the rotor (a little of ferrite magnetized alternatively +/- in 8 sectors) move by about 45°, then there are 3 plastic gears that operate a reduction on the output shaft, so that a movent of 45° on the magnet is a very little angle (about 0.5°) on the dial gauge.
So the 4 pins are ALL input to the motor, there is not a feedback on where the dial gauge actually is, it's not a closed loop circuit. The control board just execute a "setpoint" (an homing procedure) when you turn the key to the ON position, by executing a "negative" movement, to be sure that the starting point is with the dial gauge towards the "0" MpH post. Then the signal coming from the (on the wheel or the transfer case or the is converted in "steps" that are sent to the motor. Hope I was clear enough.

If you care to see some pictures of the process and the technical datasheet of the motors, have a look here: FRAEN Advanced Manufacturing Division - Series 6405-15xx Heavy Duty Stepper Motors you will see some images of the production department and "my" machines in the shopfloor in Wilmington, MA"

 

Most speedo's and tach's are now "air core meter movements" I found the patent 4356445 that shows how they operate.

Good info Frederic, you must be playing with them for some time. HAve you figured out how the eeprom stores the milage/ratios? On a side project, I building up a MPG gage and was going to use a "analog" meter to show instant MPG, a tach or speedo reprogrammed would be used, thats better than just bouncing numbers.

 

I have made a post with this Thread but it is be held up until a moderator reads it so I will post it again without the URL he put at the bottom.

Here is the information that was sent back from the Italian acquaintance.

He says:
"It seem I wasn't clear enough.
I wrote the software of the automated system that wound the coils and assemble the motor, not the sofware that control the motor itself.
But maybe I could be of some help.
As for the pinout, you have four pins because thare are two coils in the motor, each with two terminals. The control board send a pulse train to the coils (0-5Vdc), with a phase difference of 90° one from the other.
When a positive pulse energize the coil, this is creating a magnetic field that made the rotor (a little wheel of ferrite magnetized alternatively +/- in 8 sectors) move by about 45°, then there are 3 plastic gears that operate a reduction on the output shaft, so that a movent of 45° on the magnet is a very little angle (about 0.5°) on the dial gauge.
So the 4 pins are ALL input to the motor, there is not a feedback on where the dial gauge actually is, it's not a closed loop circuit. The control board just execute a "setpoint" (an homing procedure) when you turn the key to the ON position, by executing a "negative" movement, to be sure that the starting point is with the dial gauge towards the "0" MpH post. Then the signal coming from the speed sensor (on the wheel or the transfer case or the axle) is converted in "steps" that are sent to the motor. Hope I was clear enough."