Speedometer not working

[Silverado]

Lance/Dave,

Project is coming along nicely, with your help of course. I can not get my vehicle speedometer to work; I have a 92 chev pickup. If I understand this correctly, the VSS on the output shaft outputs a VR signal that goes to the Megashift and Megashift outputs a 12 or 5 v square wave signal. By changing the signal, this would not work to just put the Megashift between the VSS and the speedometer circuitry. Looking at a wiring diagram in the back of my Haynes manual, it appears that the signal goes to a "Vehicle Speed Buffer" that then feeds the dash speedometer, cruise, and original PCM. This was apparantly changed in 96 where the VSS was fed directly to the PCM. In order for mine to work, I should take the speedometer output from the Megashift and go directly to my speedometer after the Vehicle Speed Buffer. To maintain the cruise control, I should share the VSS output between Megashift and the Vehicle Speed Buffer. Does this sound correct?

Silverado

[Bernard Fife]

Silverado,

The vehicle speedometer typically takes an input of 2002 or 4004 pulses per mile, but the raw VSS signal is closer to 96,000 pulse per mile. So the OEM ECU (and MShift) divides the VSS signal down to get a signal appropriate for the speedometer (this dividing down is done in the 'buffer' that's mentioned in your manual). So as far as I know the raw VSS always went to the ECU in GM applications, the speedometer was then taken from the ECU. They do it this way so that they can use different tire sizes and axle ratios and just change the code, not the hardware.

The speedometer output circuit is on the PWM1 circuit, so you might check it against the assembly instructions to make sure it is right. Also check that all the soldering looks good. Be sure you have added the pull-up circuit components:

To use the speedometer output, you will need to connect a 'pull-up' circuit in the spot for D7 to force the output to be high or low (i.e., no to "float"). To do this, you will need to use a voltage source and a current limiting resistor. You can do this by soldering one end of a ~1.0K Ohm 1/8 watt resistor {brown-black-red, 1.0KEBK-ND} in the banded end of the location, while leaving the other 'free' end of the resistor standing up from the board. Then you run an 18 to 22 gauge jumper wire from the 'free' end of that resistor to a voltage source, typically either 5 Volts or 12 Volts. You can source this voltage from a jumper on one of the GPOx circuit jumpers (JP1, JP2, JP3, JP4). If you are not sure which voltage to use, 12 volts is usually safe and effective for most speedometers. You may have to experiment with different resistors to suit your speedometer's input needs.

I don't know if your speedometer wants a 12V or 5V pull-up, perhaps someone else will know (or has access to a factory service manual that may have this info).

I also don't know about cruise control. I suspect it also uses the 'divided down' output value (from PWM1/Amp 31) for the same reasons as the speedometer.

I do have some docs here (all for Corvettes though) and I'll have a look when I get a chance. If someone else can find out before I get a chance (using Google, a factory service manual, or direct experience), that's great too.

Lance.

[Silverado]

Lance,

Found some info on line about the speedometer used by GM. According to websites that I found, the VSSB (Vehicle Speed Sensor Buffer) converts the VR signal from the VSS to pulses per mile. These were also called DRAC modules in later years. This VSSB is a white box near the PCM behind the glove box. The only catch is that this sends differant pules to differant components: 2,000 pules to the ECM(and I would also assume the speedometer), 4,000 pules/mile is sent to the cruise and 128,000 pulses/mile to the ABS computer. To recalibrate this unit, you would adjust 7 jumpers based on a table for the correct ratio with your diff/tire size. So to keep all of these pieces working correctly, I should just share the VSS signal between the VSSB and the Megashift. Do you anticipate any problems sharing the signal?

Silverado

[Bernard Fife]

Silverado,

Okay, I see now. That's good info. I wouldn't expect sharing the VSS signal with the factory electronics to cause any issues at all. The MShift circuit is high-impedance, meaning it flows very little current regardless of the signal fed to it. So it shouldn't affect the VSSB at all.

In any case, even if it does, a 10K to 100K resistor inline with the VSS feed to MShift should fix it.

Lance.

[Silverado]

Lance,

I am sharing the VSS output for my speedometer input, dash speedo works fine, just can't get the speed to read in the Megashift. It was working at one time but very erraticaly, did not seem to matter what settings were in the filters. Now it does not work at all. While having someone else drive, I measured the AC voltage of the VSS out and found it to be steady changing with speed. Transmission works fine if shifted manually only. Where do I begin to track this down?

Silverado

[Bernard Fife]

Silverado,

First, I would return the VSS filtering settings to the defaults. Then I would go over the VSS circuit build instructions very carefully, and make sure your board is built exactly as described, and that there are no solder bridges, etc.. Be especially careful checking the transistors. Then I would wash the board in 99% isopropyl alcohol (and let it dry completely).

If those don't work, I would check to make sure the signal is getting to the board and subsequently the CPU. The best way to do this is with a scope. The next best way is a multimeter with a frequency reading ability. The AC voltage is not much use, but it only really tells you the amplitude of the signal, but it is the frequency that matters. However, I suppose you could check that you are seeing the same amplitude at the ampseal side of the VR circuit. A better way might be to check the resistance between all the components that are supposed to be connected (see the schematics here: http://www.msgpio.com/manuals/gpiohw.htm)

There is information on how the VR circuit works here: http://www.msgpio.com/manuals/vr.htm

Lance.

[Silverado]

Picked up a multimeter with frequency. I measured the VSS: at about 40mph freq was around 1K, when I stop it would reduce and then when full stop it would jump to around 15K. Not sure where the noise is coming from. The dash speedo remains at zero. When attempting to check the VR1 to the processor for frequency, I might have shorted to VB1. Now Megashift does not work at all. Fuses are still good, voltage at regulator. Will not connect up as stand alone or through MS2 CAN. Is it cooked or can this be fixed?

[Bernard Fife]

Silverado,

The crossing point detection on the VR circuit may be biased too close to ground, and may be triggering on noise in the signal. Better grounds might help this. But more likely, you will need to bias the circuit up a bit. The best way to do this is to lower the resistance of the resistor in R45 (VR1) or R53 (VR3). The specified value is 300K, and you can try from 47K to 220K to eliminate the false triggering on a noisy signal (the lower the resistance, the higher the trigger voltage).

If this was mine, the first thing I would try on a 'dead' board is loading new code. You will have to jumper the bootloader pins, then try burning new code with the downloader program. If that doesn't work, then next thing I would try is removing all the jumpers at the 25x2 header and try burning code again.

Lance.