Frequency divider for VSS

[Gasoline Fumes]

I have a question for the electronics experts. My Honda's alternator has two charging modes. A high-output mode of 14.5 volts and a low-output mode of 12.5 volts for better MPG. The ECU selects one of these two modes based on a few different inputs. One of these is vehicle speed. The alternator is always in high-output mode over 40 MPH. My speedometer and odometer are cable-driven, the VSS in the speedometer is only for the ECU. So what I want to do is cut the frequency of the VSS in half to get the low-output mode at speeds up to 80 MPH. Even 60 MPH would be fine. I know this can be done with a 555 timer in monostable mode as shown in this link. What I don't know are the values for the resistors and capacitors in the circuit. I guess the VSS signal is 8200 counts per mile, based on the MPGuino Wiki page. It's a reed switch that grounds itself four times per cable revolution. I just want to space those ground pulses apart a little more. Any advice on what component values to start with would be great.

[t vago]

I'm not sure you're going to want to use a monostable 555 as a frequency divider. It is true that you can build a sort of frequency divider out of a 555, but it's only good at actual frequency division for a narrow range of frequencies. You feed it a variable frequency signal, like out of a VSS, and you're going to get wacky results.

Suppose you make a 555 circuit to divide by 2 at some frequency F. Now, at input frequencies below F/2, the circuit will not divide at all, but will pass that frequency F. Between input frequencies of F/2 and F, the circuit will divide, but it will not divide reliably by 2 until the input frequency comes close to the design frequency F. At input frequencies above F, the circuit will tend to always pass an output frequency of F/2 regardless of the input frequency.

Now, if your instrument cluster does not depend on an accurate VSS signal, then the 555 circuit would be good enough. In that case, I'd recommend a starting point of R1 = 3.6K ohms, and C1 = 2.2 uF. That should get you a designed F such that you'll see division by 2 at about 55 MPH. That's assuming, of course, that your VSS does spit out 8200 pulses per mile.

[Gasoline Fumes]

Thanks, I'll give that a try! As long as the ECU can tell whether the car is moving or not, I think it will be OK.

Is there any particular 555 part number I should be looking for? Being inside a car in NY, it wlll be subjected to temps from about 0°F to 120°F.

My other idea was to manually engage the low-output mode of the alternator. I just have to figure out what signal the ECU sends to the alternator.

[Gasoline Fumes]

Originally Posted by t vago

At input frequencies above F, the circuit will tend to always pass an output frequency of F/2 regardless of the input frequency.

Now that I read that again, that's what I initially wanted. I just assumed dididing by 2 would be easier. I just really want to trick the ECU into thinking I never exceed 65 km/h or 40.3 MPH. Does that change your recommended values?

[jakobnev]

I wonder how it behaves if you simply disconnect the cable.

[t vago]

Originally Posted by Gasoline Fumes

Now that I read that again, that's what I initially wanted. I just assumed dididing by 2 would be easier. I just really want to trick the ECU into thinking I never exceed 65 km/h or 40.3 MPH. Does that change your recommended values?

Yes, it does. R1 become 3.3K ohm, and C1 becomes 3.3uf. Your car's engine computer would never see any speeds above about 40.31 MPH.

As far as simply disconnecting the VSS wire, I imagine it would throw a code.

[Gasoline Fumes]

Yep, check engine light will come on during deceleration.

[jakobnev]

Frequency Division using Flip-flops

If the ECU wants pulses of constant length, then a circuit with a 555 might be useful as a last step.

Yep, check engine light will come on during deceleration.

It's always a pity when the ultra simple solution doesn't work.