Make a nonlinear tachometer into a linear (homemade)

[InterlinkKnight]

Hi, everybody. I am new on this forum, but I signup because I need help. What I need should be very simple to achieve but I know very little about electronics so here it goes:

I have a "big" motor attach to another small motor to use it as a generator. I use this generator as a tachometer, connected to a voltmeter gauge. This gauge is supposed to show how much RPM is turning the main motor, but the problem is that voltage increase is not linear to RPM. I make this graph so you understand better the problem:

Is not the end of the world as it is, but is not what I want. I want a real tachometer that is as linear as real tachometers.

I am sure there are generators that could give an even increase in voltage compare to RPM, but at this point there is no way I can replace the generator so the solution must be with electronics. And should be a simple-enough so a amateur with electronics can do it, and also should be a way to calibrate the effect so I can do adjustments (maybe with a potentiometer or by changing values of components). At this moment I have a potentiometer that reduce voltage of the generator (10V at max RPM) to something the gauge can handle (0.22V).

Why I need this? I am making a small car engine model, which is electric but reacts and looks like a real one. I have all mechanic and electric parts done, and only need work on the aesthetic. I will make a video when is finish and if you help me I will mention your name on the video (is all I can offer for your help). Here is a picture of how looks right now (it runs perfectly as it is with throttle pedal, starter, stall and even battery light):

Thanks.

 

[Harald Kapp]

Welcome to electronicspoint.

As you say, a good tacho generator would be the best solution. Lacking that, a linearization circuit is possible. For a better understanding the characteristic of the "generator", a diagram showing voltage vs. rpm in a chart (and as a set of measured point pairs) will be helpful. From such a chart we can derive the transfer function
V=f(rpm) and consequently find a correcting function g(x) such that Vlin=g(f(rpm))=k*rpm where k is a constant and the function overall now is linear. Can you provide this data (make the appropriate measurements)?

In the next step we need to decide whether the correction can be done with reasonable effort as an analog circuit or whether a microcontroller with a look-up table is the more suitable solution. Would you be comfortable witha microcontroller solution?

Also: what type of motor do you use as a generator? AC, DC, with permanent magnet or with wound stator?

And last not least: What are your requirements with respect to accuracy?

Saludos,
Harald

 

[Arouse1973]

Hello
Have you looked at frequency to voltage converters?
Adam

 

[Harald Kapp]

Have you looked at frequency to voltage converters?

Good point, Adam.

I think it will still be useful to know the transfer characteristic of teh "generator" though.

 

[Arouse1973]

Guten Abend Harald. Yes I agree its worth understanding a bit more.
Adam

 

[InterlinkKnight]

Thanks for the replies.

I thought before posting that a chart about voltage vs RPM would be nice, but I don't have a real tachometer to test. A laser tachometer would be great, but don't have it and are too expensive here (at least for this project).

The only values I can give are:
The gauge is very sensitive with range between 0.01V to 0.25V, from the moment it start moving to the top limit.

The generator output 1.5V when starting the engine; 5.25V when engine is idle, 10.10V at full RPM, but keep in mind I will later put a more powerful power supply on the engine so the actual full RPM is going to be a bit higher. My current adapter is 19V, and I will eventually put a 24V powersupply.

But anyway don't need to be super accurate. This project is only for fun. I put a tachometer on the engine just because is cool. No need to be accurate. Also, voltage is not very relevant because I can adjust it with a potentiometer in the moment. If any, I want to be able to calibrate it in the moment, like changing components capacities. Example: if the circuit do too much effect, I could change, lets say, a capacitor to less value and see if is enough. You know, try and error. If visually looks linear, that is enough.

I practically don't know anything about electronics, so a micro-controller I think would be too complicated.

The generator is a common DC motor that came from a CD players (the motor that spin the CD). Don't know the name of type of motor, but I think you guys know what i mean.

Hello
Have you looked at frequency to voltage converters?
Adam

No idea what that is. Again, I am practically 100% noob about electronics.

Thanks again for trying to help me.

 

[InterlinkKnight]

I search a lot about possible solutions, and cross with a component call varistor. So far what I understand is that the varistor is not linear, so don't work. Varistors seems to be made to output an even voltage, like a regulator.

But that make me think that maybe what I need is something similar, but linear. Some component or circuit that the more voltage is applied, the less resistance has. Like an automatic potentiometer. There is such thing?

 

[Harald Kapp]

Without a know characteristic of voltage vs. rpm it will be hard to design a useful linearization circuit. Lok into Adam's proposal and use the frequency to voltage converter method as the frequency of the generator's voltage is directly proportional to rpm.
Google gives hundreds of hits when searching for such a circuit.

 

[InterlinkKnight]

I search about frequency to voltage converter and look way too complicated. Some require even the computer, or programming. I was hoping making a resistor that change with voltage would be simple. Or that complicated too?

Remember, doesn't have to be accurate at all. I need something that simply do the opposite from what I have. A way to imagine would be: let's say you have a tachogenerator and want the gauge to be more sensitive at higher voltage, and less sensitive on low voltage.

There is such thing as a resistor that change it resistance depending on how much voltage there is?

 

[Arouse1973]

It's not that complicated they do a chip that does it all for you. Answer to the other question is yes, it's called a VDR(Voltage dependant resistor). But these normally don't have a very useful range which is why they are used for surge suppression, there other name is a Varistor.
http://www.ti.com/lit/ds/symlink/lm2907-n.pdf
Thanks
Adam

 

[InterlinkKnight]

you mean this?: http://www.circuitstoday.com/frequency-to-voltage-converter-using-lm331

I saw the varistor and as you say don't work because has too little range.

Was searching for other variable resistors and found this: http://www.instructables.com/id/Make-a-Voltage-Controlled-Resistor-and-Use-It/

Don't know if is what I need, though

 

[Arouse1973]

That's the sort of thing. But I guess we need to find out why your RPM gauge is not showing a linear sweep first. Can you post the information for the gauge you are using?
Adam

 

[InterlinkKnight]

The gauge is not the problem. The problem is the voltage of the generator is not linear to RPM.

The gauge is a simple voltmeter, that is linear from 0.01V to 0.25V.

I found something that might be the simplest way. Since I can't find a true resistor that is variable to voltage, looks like I can use light as middle-man:
http://www.instructables.com/id/Make-a-Voltage-Controlled-Resistor-and-Use-It/

Apparently I could put a photoresistor attach to a LED and the resistance will increase the more light there is. With less resistance, the gauge will go up, making it more sensitive at higher RPM (voltage). I will control all with potentiometers so all is in range.

No idea if will work, but looks like the simplest way to do it. What you think?

Don't have to be an accurate tachometer. And the simplest, the better.

 

[InterlinkKnight]

I spend the last hour experimenting with a photoresistor I found in an old trash TV (that have light sensor to compensate brightness) and put it with a LED. The idea is that the LED is connected with the generator and with a potentiometer to adjust how bright it is, and the LED lights the photoresistor in a dark enclosure. The more voltage the generator gives at higher RPM, the less resistance gives to the gauge, so the needle go higher at higher voltage.

It work! Is not perfectly linear, but is a lot closer to linear than before. The gauge now can work in all range. Now when the engine is idle the needle is only a few millimeters up from the 0, and at full RPM the needle can reach the top. There is another potentiometer in the gauge to calibrate the top value.

Is ok now as it is, but now I will experiment on how can i make it more sensitive still. Is still a bit too sensitive at lower RPM, and less sensitive at higher RPM. I can live with it, but maybe changing the LED I can have more effect.

I am thinking maybe putting another LED with a zener diode that turn on that led at specific voltage when reaching high RPM. Never play with zener diodes before so not sure if is too violent for my intention. If the transition from off to on is gradual should work.

Anyway, thanks for the help, guys. Any suggestion would be appreciated.

 

[Arouse1973]

Glad to see you have something working. It would be worth looking at a magnet and reed switch with a frequency to voltage converter, you may find this more accurate.
Adam

 

[InterlinkKnight]

I read a lot about that and looks very accurate, but is way too complicated for me. Maybe someday I will test that, but not in the near future.

Thanks for the tip, anyway.

 

[Arouse1973]

Ok good luck