Frequency Counter

I know this is a theory question I can't answer. Hopefully someone knows. My multimeter has a frequency testing function on it that is well calibrated -- in the wall it shows precisely 60 HZ. However, when I connect it to the negative side of the 1/4 coil (black wire that the tach shares), at idle it shows ~90 to 110 HZ, fluctuating. So, clearly it isn't measuring the tigger freqency, which at 1000 rpm would be about 17 HZ. Anyone have an answer on what I am measuring? Also, anyone know if I can use the frequency function on my multimeter as a electronic tach for loudhvx's lean drop idle mixture procedure? (like, should I use an inductive pickup into the multimeter or something like that)

Seems your meter is reading the falling and rising edges and it is unlikely to read them consistantly.

On most (but not all) meters you can "lock" the voltage or frequency range to the region you're looking for.

You're going to have to read your multi-meter manual, look under ranging, level lock or range lock or something like that. You can also try the manufacturer's help line or website.

Personally, I've found that a lot of DMMs have problems reading coil pulses.

Maybe you know if there is a small intermediate circuit I could throw together to filter the pulses into a cleaner wave, say a optoelectronic isolator or a transistor switch tigger circuit....

The primary signal of the coil is very dirty. It has many oscillations (called rings, or "coil ringing"). That will mess up the freq. counter.

It would require an isolator followed by a little filter to get the meter to read accurate rev/seconds.
Just a capacitor and a couple diodes and resistors should do it. If I have time I'll try to sketch something up.

I forgot, you could try this, it's a tach adapter I designed a few years ago:



1 diode and 1 resistor, but unfortunately it requires 12v (from the battery) to work.

Because you are using a meter instead of a tach, you might be able to get away with a 1K 1/2 watt. (It would use less power and stay cooler than the 300~600 ohm resistor.)

You could also probably use a 1 amp diode (instead of the PTC 205, which is a little expensive), but it needs to be a high voltage diode like a 1N4006.

Just leave off the diode for the 2-3 coil.

Post edited by: loudhvx, at: 2007/07/08 20:15

Cool idea, but I was able to find 2 circuits search online that claim to be able to do the signal conditioning without needing the +12 source.

The first is a straightforward voltage divider wiht capacitor filter. Opinions on if this would work well conditioning the signal to a square wave?

Second one coming in second post...

Second on is a similar RC circuit, but instead it has the RL time constant, but I have no idea how to select component values...can you help enlighten me?

If I didn't want to use 12v on the adapter, but I was willing to use ground, then I would try this circuit. The Zener voltage is not critical, as long as it is 12v or less, and the meter will work with such a low voltage.



The lower the resistor value, the more accurate the detection should be, but that also means more power dissipation, so you'll have to experiment a little.

Post edited by: loudhvx, at: 2007/07/09 20:18

So on the last schematic (resistor/zener), what amperage can be expected to go through the resistor. The coil is switching between 12V and 0V, with potientially higher short term spikes. So the amperage I'd expect to be 12/4700 = 2mA. This makes 25 mW. However, you seem concerned that 1/2W resistors may not be sufficient. Why? What am I calculating incorrectly?

The spikes are around 400 to 600 volts, and during the spark discharge, the voltage surge is around 50 volts. The duty cycle is small so 1/4 watt may work ok, and if that's what you have, try it. I just thought the 1/2 watt would be better at dissipating the spikes and surges.

BTW, these are just suggestions on what I would try first. They are by no means guaranteed to work. I don't really have enough info on the meter to make a better guess until you get back to me with the results.

I just hope nothing blows up.

you are using a Alternating Current freq. test on a DC circut.

that is dc has no sine wave freq because of the rectifing function in the reg/rct.

you may test the freq of the 3 phase alternator before the reg/rct on the 3 yellow ac wires coming from the engine case

the idle drop can be done by ear not as difficult as you are thinking and not needed to be absolutely precise it'll change with the barometer and temperature day by day.

a simple elec. tach or dwell meter is what you need..