Blogs Car Info Our Show Deals Mechanics Files Vehicle Donation

Does the drive belt move at a fixed ratio to the engine?

I bought a photo tachometer: it shines a light on a reflector (which one sticks to the target device).

I can’t get a good reading on a reflector stuck to the engine’s shaft (I’m too spooked to stick it that close) but I can get a good reading on the drive belt. I should calculate the ratio of the length of the belt to the diameter of the wheel that drives it to calculate how many times the belt rotates per rotation of the engine, but I wonder if it isn’t a fixed ratio, saving me the effort.

are you kidding ? off course it is fixed :slight_smile:
is it not easier to plug an OBD adapter for $5 and read RPM from your ECU ?

It is a fixed ratio, but it varies car to car. One idea, manually crank the engine by turning the crankshaft-pulley bolt with a ratchet/socket and measure how many times the crankshaft turns as the belt turns the complete route one time. from that ratio you can calculate the engine rpm from the belt rotation rpm.

So you’re not able to put that reflector right on the crank pulley then and still get the photo-tach gadget into position to see it safely?

But, you will not be able to count the passing of the belt markings to be one-for-one with the turning of the crank.
The diameter of the crank pulley is so much less than the length of the belt !
a 22" pulley will not pass the 91" length of the belt at predictable equal times.

Isn’t it sort of like trying to figure out a vehicles rear differential ratio? i.e. you drive very slowly (or put the rear wheels in the air) and measure how many times the prop shaft turns for each time the rear wheels go around once. There isn’t any need for it to be a one-for-one thing. In fact it won’t be one prop shaft rotation for one wheel rotation. More like 4 prop shaft rotations for 1 wheel rotation. You just need the ratio.

Great idea! Thanks! For my '87 Toyota pickup?

That’s a good idea.

Perhaps I’m merely a scaredy-cat. Unfortunately one has to look at an icon on the display to make sure one is getting a valid reading, which means taking one’s eyes off the target.

For any reason other than slipping?

I just measured. The belt is 45 3/8 inches. The pulley wheel is really hard to get to; it’s 5.5 or 6 inches. With the belt running at 434 rpm this is high but just under the state maximum.

If the AC compressor pulley, or any accessible pulley is measured and marked and the relative circumference of the crank pulley and other pulley calculated the photo tachometer should enable measuring a relatively accurate RPM reading.

If your fan is engine driven and not a clutch fan, put the sticker on the fan blade. If you have electric fan(s) unplug them.

For an '87, it is probably cheaper to just pick up a hand held tach/dwell meter on EBay. Under $20 and easy to hook up.

I couldn’t get a reading off the sticker on a fan blade.

Only if my time had any value. I had 1 of these, lost in a theft a few years ago. I used it once to confirm that it was running right. Since it’s sounded like it’s running right since I never used it again, didn’t think I needed it. Now that I look back, my idle has been slowly rising over the years, as recorded by the state’s emissions tests.

wild idea: make an oscilloscope from Android phone, measure frequency of spikes on one of high voltage wires with capacity pickup and substantial attenuation:

1 Like

Great idea! I have no phone but I have a computer and there are oscilloscope apps (which use the audio input). Thanks.

I measure the rpm of my Corolla like that. I just wrap one of the spark plug wires a half dozen times with some 30 awg wire, and hook the scope probe up. Some combo of inductive and capacities pickup. Works fine, give a good sized signal. I use a traditional lab 'scope, but a free phone-scope app should work fine for that too. It should have an accurate enough time base for engine rpm measurements. If anybody tries this, post some photos of the results. You’d think you could purchase a hand-held o’scope suitable for ignition system work for $20 or so. New. It’s not a demanding application for a hand-held gadget. But so far this product hasn’t reached the shelves.

Hmmm… I finally got my microphone input to work. What sort of voltage should I expect to get from a wire wrapped around a sparkplug wire? I don’t think I can count on my multimeter to respond quickly enough.

Voltage depends on may factors, such as number of turns, diameter of turns, and proximity to the HV wire. Could be a volt or two or hundreds.

Me, I’d put a pair of diodes back to back across the input to prevent voltages over 0.6 volts from entering the scope/audio input. And a series resistor.

Or just try one turn around the HV wire, keeping it at least an inch from the wire. My guess (and it’s only a guess) is that would be relatively low voltage… but I’d be reluctant to risk a burnout. A real scope has resistor attenuators at the input, but an audio input? dunno.

If wires insulation is not god, it can be very high, so I would plan on using Zener diode with a resistor to make sure you do not spike to something unsafe for your audio input:

Radioshack usesd to carry few of these, but I was not there for quite some time now, can not tell for sure if they still have it

add L-attenuator to get down from voltage you would be able to obtain from Zener-limited voltage to the one safe for your audio-input:

let’s assume you could get 3.3V Zener and audio-input would safely handle 316mV

in this case, I would go for 10:1 attenuator from your improvised pickup, something like 10K/1K resistors will do fine, as your audio input will have around 200K…500K input resistance I bet

essentially, it will be:

pickup wire wrapped around high-voltage wire -> zener to ground -> 10:1 attenuator -> audio input

andriy.fomenko: a pair of diodes are cheaper and easier to find. That limits the signal to 0.6 volts peak or 1.2 volts P-P.

that’s a great point!
way easier, and we would not care about how linear our pickup is :slight_smile:
Zener would probably give us better view if we wanted to see signal shape, but here we would care about peaks only

I just recorded the output of 1 loop of wire around the ignition coil. I count 55 spikes in 2 seconds, 1650/minute. It’s a 4-cylinder engine, so that’s 412 rpm? That’s way too low; 1650 seems too high. Do I misunderstand?

each cylinder would presumably show one half of the actual RPM as spark goes only every other rotation