Another rpm / tacho calibration query

[julianf]

The diesel tachos that work off the W terminal -

They have a single potentiometer for calibration.


I'm trying to understand this - surely that would only be any good if the crank and alternator had the same diameter pulleys?

What I mean is, if calibrating to a single speed point, how could the gauge know the multiplier from any pulley diameter? It couldn't.

Is it just accepted that they're a rough idea, rather than any pretence at accuracy?

 

[500e]

Surely the ratio stays the same, thinking back to bug eye Sprites and other cars even Chronometeric ran from a gear box that screwed onto the back of the dynamo.

 

[julianf]

Say i have a 1:1 crank pulley / alternator pulley setup (ie equal diameters)

I run at a 100 rpm where the W terminal gives 10 pulses per second (made up numbers)
I double the speed to 200 rpm, so 20 pulses.


But, say i have a 1:2 ratio.

I run at 100rpm and get 10 pulses
I run at 200rpm and get 30 pulses.

1:10 ratio -

100 rpm 10
200 rpm 110 (or somthing)

whatever - my maths might not be right, but, if the gauge works out the rpm by the pulses, the gearing of the pulleys is going to alter this. It could be a linear relationship, or various exponential curves.


So far as i can tell!

 

[Misterg]

my maths might not be right, but, if the gauge works out the rpm by the pulses, the gearing of the pulleys is going to alter this. It could be a linear relationship, or various exponential curves.

It's linear - that's what the calibration pot is for

example:

2:1 ratio between engine and alternator (alternator faster)

100 Engine RPM = 200 Alt RPM = 20 pules/sec using your numbers
200 Engine RPM = 400 Alt RPM = 40 pluses/sec (which is twice as many as 100 engine RPM)

 

[julianf]

Even if my maths is incorrect, and my mention of exponential wrong, the gradient of the linear graph would alter, right?

So when the gauge gets double the number of pulses, how does it know if the rpm has doubled, tripped, or what?

 

[My Old Landy]

When I fitted one to my Land Rover it required an second tacho to set the pot. You ran the engine at 1,000 rpm verified with a tacho off the crank pulley. Then adjust the pot until the fitted tacho agrees. You then up it to 1500 rpm and tweak it again. They are just an indication and I found not that accurate but good enough for the LR 90.

 

[Misterg]

when the gauge gets double the number of pulses, how does it know if the rpm has doubled, tripped, or what?

Double the pulses = double the RPM, so the gauge will display double what it did before. If you set it right at one RPM, it’ll be right at the others.

 

[m_c]

Even if my maths is incorrect, and my mention of exponential wrong, the gradient of the linear graph would alter, right?

So when the gauge gets double the number of pulses, how does it know if the rpm has doubled, tripped, or what?

The gradient of the graph does not alter.

The ratio between alternator and crank remains constant. It doesn't miraculously alter as speed changes.
There is no difference between an alternator outputting 10 pulses per alternator shaft rotation, while driven at a 1:2 ratio, giving 20 pulses per crank revolution, than there is a sensor directly mounted to the crank giving 20 pulses per crank revolution.

If the crank speed doubles, then the alternator speed also doubles.