VSS signal converter

[lionelhutz]

[sturgillbd]

Here is the frequency divider circuit for the optical VSS signal. Each output divides the signal frequency by two. If the VSS signal coming in is 8000 PPM, the first output will be 4000 PPM and the second output will be 2000 PPM. If the optical VSS signal output is higher than 5V, the 74LS74 could be replaced with a 74HC74 and the 5V regulator could be omitted. The MOSFET will allow a 5V signal to control a VSS input that is internally pulled up to battery voltage such as in the 1227727. The 7727 also worked when the VSS input was switched between 0 and 5V but I added the MOSFET to allow it to rise to it's pull-up voltage.

[1project2many]

Very nice. VSS input really is key to making a vehicle run correctly if using a 7747 or other C3. And it's also a big part of transitioning manual transmission equipped vehicles from coastdown to no-load conditions. Hopefully this saves some people from purchasing expensive aftermarket components.

Thank you.

[JeepsAndGuns]

Very interesting. I'm guessing this would be used for ecm swaps where the vehicle its being installed in has only 1 type of vss, or for ecm's that can only accept 1 type of input. Seems this would allow a lot more flexability when installing/setting up a vss.
Only thing that I think would make it cooler, would be being able to adjust the output pulses to adjust for different tire sizes/gear ratios.
For example, my 93 wrangler has the 8 pulse vss, but its driven by a gear in the tailhousing of the transfer case. Those gears are swappable for different tire sizes/gear ratios. But they only help to a point. Right now with current tire size and gear ratio, the closest speedo gear makes it read about 3 mph too fast. I know, not much, but it is annoying. Being able to adjust the output pulses could correct the speedo. But that would also mean it would have to have another output (adjusted for speed correction, and the same signal as the input) And not knowing anything about stuff like this, I'm afraid it would make it about 10 times more complex and expensive.

[RobertISaar]

i've been toying with the idea of using a small/cheap microcontroller for finer control of the VSS feed to the cluster due to the way the pulse divisor works in the P4 ECMs and not being able to get a divisor close enough for the tires/reluctor combo I use... as usual, my hangup is on the PC side of things, but I'm slowly working through that. from the looks of it, it could be done for ~$5 in part costs, so it shouldn't be too painful.

[sturgillbd]

Yeah, a little microcontroller could take the place of the divider circuit above also. Read an input, do sone math, send output on another pin based on the math just done. It could also multiply a signal frequency but would lose some accuracy. Any error on reading the input would be multiplied in the output. Ill try some tests with one at some point.

[Six_Shooter]

I figured you would. I remember reading that you were in the process of getting your AAS EET a while back. I received my AAS EET in 1994. I had all of the electronics courses completed by 1993 but took 3 credit hours my last two semesters. How many credit hours was/is your program. I had to have 81 credit hours for an Associates Degree. Don't be too critical of my schematic drawing skills, I just downloaded that software about three hours ago and am learning it as I go :)

I'm not sure what AAS is, but my program is a diploma program. I was in EET (Electronics Engineering Technology - (Telecom)), but as of January I switched in to EETcn, Electronics Engineering Technician, for various reasons. I will be done this semester, where as EET was going to be another 2 years, again due to various reasons. I do plan to continue on, but as part time, and probably switch into CET (Computer Engineering Technology), which is the 3rd sibling program in the trio, either that or something different like ESD (Embedded System Design), or maybe do all of it. lol

Here is the frequency divider circuit for the optical VSS signal. Each output divides the signal frequency by two. If the VSS signal coming in is 8000 PPM, the first output will be 4000 PPM and the second output will be 2000 PPM. If the optical VSS signal output is higher than 5V, the 74LS74 could be replaced with a 74HC74 and the 5V regulator could be omitted. The MOSFET will allow a 5V signal to control a VSS input that is internally pulled up to battery voltage such as in the 1227727. The 7727 also worked when the VSS input was switched between 0 and 5V but I added the MOSFET to allow it to rise to it's pull-up voltage.

I quickly looked at this but will take a more in depth look later since I need something like this very soon. The difference is I need to divide down from like 17000 PPM due to the VSS in the trans I am using (Nissan, not GM). Currently with my '7749, I am able to get it divided down to be pretty accurate, but will be switching to ECM(s) that do not have this capability. I'm thinking about starting with a 7492, 74290 or 74292, or just what ever I find will get me close to start with. I haven't put a pen to paper to start designing it yet, just thinking about it in passing.

[lionelhutz]

Very interesting. I'm guessing this would be used for ecm swaps where the vehicle its being installed in has only 1 type of vss, or for ecm's that can only accept 1 type of input. Seems this would allow a lot more flexability when installing/setting up a vss.

The idea of the MAX232 circuit is to feed a 0-5V or 0-12V squarewave VSS signal into the magnetic sensor input. This could allow PCM's with only magnetic inputs to accept the VSS. IT also allows the adjustable pulse parameter for the magnetic input to be used to match the VSS pulses.

On the PCM's with both input types, the DRAC VSS input only accepts 2000ppm signals. The signal from a different VSS might be electrically compatible with that input but have the wrong ppm count so the PCM will register a speed but the speed is wrong.

[sturgillbd]

The difference is I need to divide down from like 17000 PPM due to the VSS in the trans

With the divide by 4 output, you would be showing higher speed by 6.25%. If you can adjust your speed related items in the calibration by that amount, you should be fine. If you are depending on this signal for a speedometer, it will be off. With a microcontroller, the pulse train could be sampled for a time (lets say 250 milliseconds), and put into a counter. The program could then do some math based on a constant you put in and develop a separate output based on that constant. If you were off by 6.25% high on the input, it could make the period of the output 6.25% longer to lower the frequency by that amount. All of this would have to run in a loop in the program so it updates at a reasonable rate. It can be done but will take some code to be written. Luckily with the microcontrollers, there are C compilers which make development much faster. I have some code I need to find where I built a microcontroller project a few years ago. When the circuit was first powered up, it posed a prompt on the serial port that you could connect with a PC. You entered the data that you wanted it to run with and it took off. If after two seconds of powerup, nothing was entered, the program used the last entered data which was stored on the internal eeprom. If there would be interest in a project like this, maybe we could all develop it as a group.

Brian

[sturgillbd]

250 milliseconds would probably be too long to get reasonable updates on speed changes especially after the overhead of the processing. It was just an arbitrary value. The sample rates will have to be examined and lots of testing be done to make it pretty much a universal circuit.