Getting a RPM Signal From the Alternator

Znago; some gen set manufacturers use that tap for high frequency ac for a reference to control the RPM. perry

Thanks Perry, yes I did forget to mention this is automotive....

Just looking to understand what type of signal it is so that I can then design a conditioning circuit to give me a square wave (as input to the microcontroller).

Ah, you're likely to get a whole variety of answers from people here about what they think you have in your hands. I recommend you get an oscilloscope and measure what this terminal will do across a variety of different alternator speeds and loads.

I would expect that you would get a sine wave AC signal with no DC offset. The amplitude will not be precisely constant but not varying too much over time. But I would not be surprised to find a rectified sine wave signal or even a sine wave with a DC offset so that the negative peak of the sine wave is at 0 volts. Without knowing how many poles your alternator has, you'll have to guess at the integer multiple the frequency of this W terminal produces in comparison to the alternator's rotational frequency.

Yes RedFred, you are absolutely correct and I plan to do that ASAP. Just that it is not very easy for me to get ahold of one (I know it sounds strange but true.....). So I am looking for info to get an understanding and a head start.

I would not be too much concerned with amplitude (I think) since anyway the output should be clamped to 5v probably by a zener or something (to go in to the micro).

If it is a true sine wave AC signal, is the circuit just a comperator that would give out say 5v when sine wave is positive and zero when sine wave is negative?

Yes, a comparator circuit will be your ideal interface for all three of the signal scenarios I proposed. A comparator circuit will permit you to easily control input impedance so that you do not load the W signal and effect any other circuitry. You can select a hysteresis range to reduce noise pick up. Naturally the actual resistor selection and topology will depend on what you find with the oscilloscope. If the signal is one of the three signals I proposed, then an AC coupled inverting comparator circuit will do your job. (Don't forget to clamp the inverting input to return with a diode to protect the negative voltages this will produce.) There is an outside chance though that W is the voltage/current being applied to the rotor. In this case it will be useless for an alternator tachometer.

Hi Redfred

Long time from your post. I find it very enlighting and maybe you can answer my question. I read that getting signal from the alternator for RPM measure has the drawback that the signal can be lost under full battery conditions. Don´t know if this is all true and if it applies to the W output of the alternator.

Thanks for you reply

FcoJavier

Yes, it has been a long time since I contributed here.

Naturally many things can attenuate a signal to the point of making it unreadable. Most operational amplifiers have a minimum gain of 10^6. So as long as the alternator signal remains above or just near the noise floor, a simple comparator circuit should be able to discern it from the noise. To extract a smaller signal from the noise will require a more refined analysis of signal and noise but you would be amazed what can be done.

Hi Redfred

I knew you were the one to answer my question!! Thanks a lot

Will have to test the engines under full battery conditions and see what happens then. Good to know I have to keep an eye on that.

So being that true, those Diesel RPM measure devices that connect directly to the battery should suffer a simmilar effect, or am I wrong? Would you reccomend one of those?

Thanks again

FcoJavier

I need to know much more information to render any opinion on a specific model instrument, let alone such a loosely defined group as "those Diesel RPM measure devices". I'm certain there are exquisite devices, cheap knock offs, and units that will work good enough if connected properly but mislead when connected improperly.

Redfred

You are right, need a lot more information. Thanks for taking the time to reply and thanks for sharing you knowledge.

Be good. FcoJavier

Hi Znago

I am looking to measure RPM from the W output.

How did your project go? Would be nice to hear about your results

Thanks

FcoJavier

If you don't have any dedicated output on gen-set, Frequency of generator itself is a signal of RPM ( RPM is proportional of frequency)

It is probably pretty dirty and at risk of surges & spikes. Once you filter it, surge suppress it and add a series resistor in front of your circuits you might be to the point where amplitude doesn't matter as long as it can be trusted to be above 5 volts. I don't have any experience with alternators but I suspect you can put a checkmark beside amplitude and work on the other issues.

Note: my "series resistor" is assuming you are only after frequency and/or phase information. If you are trying to draw power then discard my comment.

A little feedback after you get going would be nice. What you are doing sounds interesting.

http://www.amplepower.com/primer/tach/index.html#

http://www.google.co.uk/search?q=rpm+pickup+from+alternator&sourceid=ie7&rls=com.microsoft:en-gb:IE-SearchBox&ie=&oe=&redir_esc=&ei=wzvYS4uOOYGI0wT1n7yGCA

you could also use a frequency counter as used on some high end Digital Multi Meters to help getting you started

you can also get dongles to use on PCs much cheaper than a full blown o silly scope

http://www.picotech.com/10reasonswhyyouneedapcoscilloscope.html

http://www.picoauto.com/

http://www.zelscope.com/

http://www.virtins.com/page2.html

http://www.youtube.com/watch?v=1f9ss1hrCc8

http://www.sillanumsoft.org/prod01.htm

software featured in one of the utube videos

I can only speak for the American market automotive alternators. When a RPM signal is needed in a factory design, an alternator with 1, or 3 AC taps are provided. These connections are directly from the stator windings. Just before going to the diodes. I think it varies between alternator voltage above and below ground. ( 14v positive to 14 volts negative, relative to ground.)

OK so now I know that the signal generated is a simple sine wave of say -12v to +12v. Question is, which circuit to use to get a nice square wave out of it. My board will only have single 5v supply to work with, and I dont care about the negative cycle. Ideas? Thanks.

For that you will need help from one of the people better at electrical circuitry. Sorry.

I guess you are right. Thank you and all very much.

Heres a good circuit design for all of the features I mentioned for a zero crossing comparator circuit. If you use a LM339, you can use +5V instead of the diagram's +15V. There is one critical typo though in the diagram for you to use a LM339 family of chips. The common node of R4, R5 and the 10k resistor should be +5V.

Do you mean that R4, R5 and the 10K should all be pulled up to 5v? (and apart from that, the 15v supply changed to 5v of course).

By the way, what would be the equivelant of the LM339 in LTSpice? (i.e. an Linear part?)

Ahem, well you could try going to the Linear Technology site and look up which comparators they fabricate. That might be one method that would demonstrate a little initiative. You might post here the name of your computer and net access so that I or anyone else here could access your computer and show you precisely which buttons in LTSpice you should click to locate a suitable comparator.

Oh and by the way, your welcome.

_{(Walking away from this thread, again stunned by a refusal to think.)}

Sir!

I have been thankful many times in this thread. And please do not accuse me of a lack of initiative or inadequacy or inability to think on my own. I believe I asked a relevant question and apart from that, yes, I will go and see what replacement parts they offer. No I do not want you to teach me how to use LTSpice (I know where to find their forum thank you very much)

Lastly, even though you may think to the contrary, I am thankful for your help and appreciate your inputs. One day I hope to be able to contribute as you do. For now I am still learning, so pardon me.

I hope that one day you will be able to contribute here. And you have been grateful of the help granted you. I probably should have posted a separate recognition of that, because I do mean that you are very welcome. But we are all capable of making a thoughtless posting here. I believe that asking for a suitable part from another manufacturer was a thoughtless post. If you wish to learn how these block unit parts function you should investigate the data-sheets of these parts.

Oh a hopefully helpful hint, the LM339 originally designed by National Semiconductor is a long used part that has a small little quirk about it that hinges on the need for that 10k resistor.

(PREFACE: THANK YOU THANK YOU THANK YOU )

OK, I am posting two circuits. One which is comparator based, and another that just uses two transistors. It would seem, that they give the same results. And the one based on transistors would seem less costly in terms of footprint and cost. Anyway, appreciate views on if they are really on par or different in some way I do not see.

Tx

Yes! Nice work and insight using the high gain of two transistors to square up the sine wave. You are correct that two transistors will likely work here and will be slightly cheaper than the comparator circuit but let me point out what differences the two circuits will have.

If you add a second noise signal of 1V and a different frequency to your original signal (you do this by adding the second source in series with the original) you'll find both frequencies appearing out of your two transistor circuit. No amount of resistor changes will permit this lower amplitude signal to be ignored. With the comparator circuit though, the 10Meg resistor from the original diagram (sorry but your picture is too small for me to read the values you selected) can be reduced in value to make a hysteresis adjustment so that noise levels below a threshold can be ignored. This will restore the original tachometer signal to your next stage.

Another less noticeable difference will be that you'll find that the rise and fall times of the first circuit will not be as fast as the comparator circuit. This will become much more apparent if you add a grounded capacitor to each output that will simulate the wiring capacitance to your next circuit. The much lower output impedance of the comparator will charge and discharge this capacitor much faster.

"One day I hope to be able to contribute as you do. For now I am still learning, so pardon me."

We are all still learning. That is part of life. Currently you know all of what you know. I only know what I know. But together we both know how the A/C taps on an American alternator work. Stay tuned, enjoy the ride, and comment when you have an answer to another's questions.

feed alternator feed through a half wave rectifier that will give +12 with -12 missing if thats any good

http://www.allaboutcircuits.com/vol_3/chpt_3/4.html

http://www.ibiblio.org/kuphaldt/electricCircuits/AC/AC_12.html

http://www.discovercircuits.com/C/co-sinetosqr.htm

http://www.google.co.uk/search?hl=en&rls=com.microsoft:en-gb:IE-SearchBox&q=sine+to+square+wave&revid=81254961&ei=R7TZS4WcDZ-y0gTxzeVH&sa=X&oi=revisions_inline&resnum=0&ct=broad-revision&cd=1&ved=0CDwQ1QIoAA

this google link may be better.

http://www.google.co.uk/search?q=12volt+ac+converted+to+rpm&sourceid=ie7&rls=com.microsoft:en-gb:IE-SearchBox&ie=&oe=&redir_esc=&ei=57TZS5vSJ4OM0gTnjPBF

http://www.mcselec.com/index.php?option=com_content&task=view&id=106&Itemid=57

also this one

http://www.cruisersforum.com/forums/f14/rev-counter-pickup-3926.html

Its a sine wave (AC)