Oil Pressure gauge reads BACKWARDS???

[98 svt]

Wow Mark, that's some interesting stuff right there. Unfortunately, I don't have an ohm meter to perform the tests.
I definitely plan on sending you my gauge cluster over the winter though!

 

[NVENEM8U]

the 9 dollar sending unit I listed earlier will fix ur issues with the oil press gauge .....Mark good job with the gauge cluster stuff u've figured out.:rockon:

 

[mwolson]

I was able to duplicate your tach problem with the speedometer last night. I am now pretty sure that your tach signal is intermittant.

But first, a little background on VSS signals. A couple of weeks ago, I helped a guy set up a Speedcal on an OSS 02 Sable with a 96 Taurus VSS tranny. I put an O-scope on the VSS signal and we spun it with a variable speed drill. I noticed that the amplitude of the VSS sine wave went up with frequency to well over 1V p-p.

Here is what I set up on my bench.

I installed a freeware tone generator package from cnet.com called Expression Tone generator. I set it up to generate a fixed frequency sine wave by using the equation: sin([frequency]*t). I maxed out the volume and hooked one side of a stereo mini jack to the speedometer signal on a 120 MPH V6 cluster with a powered up speedometer. I was able to drive the speedometer up to about 100 MPH with this setup by entering the correct frequency into the formula. The max voltage the PC was able to put out was about 0.1V. My theory is that the voltage must go higher for the speedometer to work above 100 MPH. I will test that with a different signal generator sometime this weekend.

Here is what I learned. Below about 90 MPH, if you remove the signal, the speedometer needle will drop to zero. But if you suddenly remove the VSS signal above about 90 MPH, the needle will sweep all the way clockwise to under the needle stop pin. If I try to get the needle to move to a speed lower than about 90 MPH, it looks like the needle just pushes on the pin. It I feed it a signal that sets it to 100 MPH, it goes there, but then rotates clockwise back to under the pin when I remove the signal. But if I feed it a signal at about 90 MPH, the needle will rotate counter-clockwise to the 90 MPH mark. If I then remove the signal, it will drop to zero counter clockwise as it should.

The speedometer and tachometer must use current through two coils to create the magnetic fields needed to move the needle correctly. One coil creates a reference magnetic field that sets the reference for the needle rotation. Then the meter uses the frequency of the input signal to figure out what deflection field should be generated to move the needle to a certain angle. It seems that, with an absence of an input signal, the needle aligns to the reference field so it goes back to the pin. But, depending on which half circle the needle is on, it will rotate to the pin in a clockwise or counter-clock wise direction.

With normal operation, the signal field varies gradually up and down, so the needle will move up and down relative to the reference field, and we never see it go above the maximum speed that the needle goes to. The only time it can slam all the way to the back of the pin is if the signal is lost after the needle has rotated past 180 degrees from the pin.The only way to get the needle back is to run it up above 180 degrees from the pin and then back it down to less than 180 degrees from the pin before losing the signal.

The tach input needs a DC signal, so I was not yet able to try to reproduce this behavior with the tach, but I bet it will work the same way. I need to set up my fancy tach calibrator setup before I can try this experiment with the tach. I will try this with the tachometer this weekend.

But the data is pointing towards an intermittent tach signal somewhere between the PCM and the tach circuit in the cluster.

The oil pressure gauge issue appears to be completely different from the tach issue.

 

[cobraman1024]

Seriously, its been suggested twice that you change out the oil pressure switch for a new one. Its not such an outlandish idea that your pressure switch may have gone bad and its a whole lot simpler to change that out than it is to bench-troubleshoot your gauge cluster online in a forum.

 

[mwolson]

I pulled out the signal generator I use for calibrating tachometers tonight to see if I could get the tach to fail the same way I was able to get the speedometer to fail.

I learned that the tach input is a DC square wave that needs to be about 6V or more. An AC signal will not work to drive the tach.

I was able to drive the tach up to the 7 grand limit of the V6 cluster with my calibrator, no problem. I was even able to go all the way up to 9000 RPM which pointed the tach needle straight down. But when I removed the tach signal from the tach input, the tach immediately dropped to zero RPMs, regardless of where the needle was pointing, even straight down. So my guess was wrong about the failure of your tach.

FYI, I was able to AC couple the tach calibrator signal by using a 0.1uF cap in series with the signal to drive the speedometer. With this calibrator, I was able to make the speedometer needle wrap around the back to the needle, so this is a legitimate failure mode for a speedometer with an intermittent VSS signal. But it does not explain the failure of your tachometer.

Looks like I will need to get your specific cluster on my bench (when it is ready) to get a better chance of figuring out your tach issue.

I ordered a sweep function generator today in order to get more control over the signals I can generate in my lab. Should be here within two weeks.

 

[98 svt]

Damn Mark, thanks a lot.
To the other guy, the car is currently not in my possesion at the moment. The shop will be swapping the switch for me.
Mark, I wish I knew half of what you were explaining to me hahaha. I'm clueless when it comes to this type of thing.

Thanks again Mark

 

[mwolson]

I've been studying these clusters a lot. Ford put some custom integrated circuits inside of the tach and speedometer circuits that make it nearly impossible to figure out what is going on inside of them. That makes it hard to debug these strange problems when they crop up. But it is interesting to try to figure it all out.

 

[mwolson]

My new 3MHz function generator arrived today, so I was able to do some more testing of both the speedometer and tachometer.

The speedometer requires a sine wave of at least 7.5V peak-to-peak in order to drive it over the entire range of the speedometer. It can detect the VSS signal down to less than 0.5V p-p under 50 MPH, but the voltage must go up at higher speeds or the speedometer will malfunction. The speedometer expects the AC signal to start slowly and have continuously increasing, decreasing or steady frequencies. If you hit it suddenly with a high speed signal, it will not work. If you stop the signal suddenly, the needle will drop back to the proper side of the pin if it is in the first 180 half of the needle rotation, or it will go all the way to the back of the pin if you remove the signal with the needle in the second 180 degrees of needle travel. To get the needle back under control, you need to start out with a slow speed signal, then increase it to where the needle will be more than 180 degrees from the pin. At that point, the needle will go to the correct speed, and, if you sweep the signal back down, the needle will go back to the proper side of the gauge.

The bottom line, the speedometer expects a gradual increase and decrease of the VSS signal, with higher voltage at higher speeds. If it does not see such a smoothly changing signal, the speedometer may behave strangely.

The tachometer signal is a pure digital signal. I was able to drive the tach to any RPM with no problem. If I took the signal away at any point, it would drop to zero as it should. I was unable to duplicate your problem on my bench.

But at least I now have the ability to drive both the tachometer and speedometer with complete control. It should make debugging a lot easier.

 

[Mustang Matt]

That's an interesting design. I wonder why they wouldn't just control the speedo with a digital signal too? Maybe it would make the needle jump more?

 

[98 svt]

Very interesting mark. Thanks for posting your findings.