Small motor tachometer

Thread Starter

remout

Joined Apr 18, 2011
14
I'm exploring the differenet ways to make a tachometer for a small one cylinder 4 cycle motor. I have induced voltage from the spark plug wire and a hall sensor working. My next project is to trigger off of the primary side of the coil.

I built a full wave rectifier circuit (4 diodes). I hooked up a 5v led with resistor and got a blinking light. I expected the blink to coincide with the magnets on the flywheel moving past the coil. I next ran dc current from the rectifier circuit into a digital pin on an Arduino and immediately smoked the chip.

On a scope , with my limited ability to read it, I'm seeing a very high spike in the 200 volt range for 2-5 milliseconds. I need to bring this down to 5 volts so I can count the Hz.

Ideas?

Thanks!
 
Last edited:

praondevou

Joined Jul 9, 2011
2,942
What's the output voltage of the rectifier (not the spike). What diodes did you use?
Can you post a picture of the spikes and the voltage before the rectifier and a (handdrawn) schematic?
I would isolate this stuff from my MCU via an optocoupler.
 

Kermit2

Joined Feb 5, 2010
4,162
You might be successful with a large value voltage divider(around 50:1) feeding the input of a schmitt trigger logic gate to form 5 volt pulses
 

praondevou

Joined Jul 9, 2011
2,942
Well, if it's the 200V spikes you want to measure, then Kermit is right, just use a voltage divider. Your circuit diagram doesn't say anything. Where did you hooked it up to? And what's the waveform like? Apart from the 200V spikes.
 

THE_RB

Joined Feb 11, 2008
5,438
The primary side of the coil is a square wave that is either at 0v or 12v. It will also have some fast but high voltage spikes on the switching edges.

An RC filter into a 2 resistor voltage divider will give you a safer squarewave from 0v-5v, and a 4.7v zener on the output will limit any remaining spikes within -0.6v and +4.7v.
 

SgtWookie

Joined Jul 17, 2007
22,230
Hey Debe,
What was the time/div setting you were using? It would be good to know what the approximate frequency of the decaying signal was.

[eta]
Never mind, I think I have a reasonable approximation figured out.

Have a look at the attached simulation of an ignition system, and a rectify/filter/clamp/Schmitt trigger circuit to convert it to a tach signal.

The signal at Out1 really isn't terrible if the uC input is a Schmitt trigger, but that slow fall time could trip it up. Also, the impedance at out1 is quite high; one reason the TLC555 timer is recommended; the other reason is the rail-to-rail output of the TLC555.
As with all simulations, this is just a starting place; your mileage can/will vary.

[eta]
After reading the original post again, I see the duration of the HV spike is 2mS to 5mS - which is quite a bit longer duration than my simulation was set up for.

Can you narrow down the time of the spike a bit better than 2mS to 5mS? And does the ringing pulse look similar to what Debe posted, and what I have in the current state of the simulation?
[eta]
Whoops, see next post.
 

Attachments

Last edited:

SgtWookie

Joined Jul 17, 2007
22,230
This simulation is a bit more realistic; the last simulation would have been a single cylinder engine running at about 24,000 RPM :eek:

This one would be a single cylinder engine running at 3,600 RPM (full throttle for a Briggs & Stratton engine).

The 300v pulse on the points has a much longer duration now.

Not much changed in the rectify/filter/clamp section except R2 and C3.
 

Attachments

THE_RB

Joined Feb 11, 2008
5,438
Debe- is that scope trace from an actual motor? I've never seen ringing like that when the spark plug is connected.

The bulk of the energy is dumped through the sparkplug and is gone in one peak pulse.

The coil primary waveform is basically a 0v-12v squarewave with some short duration spike energy. Most of the old tacho circuits and auto tune up meters etc just use an RC filter into a voltage divider, and possibly a zener diode if still needed after the RC filter.

SgtWookie, I don;t think you can approximate a sparkplug with a 2M resistor. With the sparkplug, when the coils field collapses almost the entire energy of the back emf is dumped through the plug, there is very little energy left to make that ringing. There is also much less peak voltage at the primary than what your sims are showing.

Does your sim program hav a spark gap component? Or even something like a low impedance gas discharge tube (xenon etc)?
 

pistnbroke

Joined May 9, 2011
32
RB you know nothing about ignition systems there is no 12 0 12 0 pulse ....if its b and s its an electronically controlled magneto or if very old contacts . Any way it fires like a 2 stroke ..twice per revolution even though its a 4 stroke ..so expect the reading to be twice what you expected .....did you know you can have 4 stroke 2 stroke and 1 stroke firing patterns ...a potential divider based on 300v and a single diode will give you an output....sorry sarg see you did that ......
 

SgtWookie

Joined Jul 17, 2007
22,230
Debe- is that scope trace from an actual motor? I've never seen ringing like that when the spark plug is connected.

The bulk of the energy is dumped through the sparkplug and is gone in one peak pulse.

The coil primary waveform is basically a 0v-12v squarewave with some short duration spike energy. Most of the old tacho circuits and auto tune up meters etc just use an RC filter into a voltage divider, and possibly a zener diode if still needed after the RC filter.

SgtWookie, I don;t think you can approximate a sparkplug with a 2M resistor. With the sparkplug, when the coils field collapses almost the entire energy of the back emf is dumped through the plug, there is very little energy left to make that ringing. There is also much less peak voltage at the primary than what your sims are showing.

Does your sim program hav a spark gap component? Or even something like a low impedance gas discharge tube (xenon etc)?
Good points, RB.
Yes, LTSpice comes with a subcircuit for a neon bulb; it's modeled as a switch, cap and resistor with some parameters. You feed it the strike and hold voltages as parameters.

I'd actually modified a neonbulb symbol to be a spark plug a few years ago, and had forgotten about it.

Replacing the 2m resistor with a sparkplug that breaks down @ 2kv and holds @ 1kv does snub the ringing.

However, I really don't know for certain what the primary side waveform would look like nowadays. I can hardly remember seeing the display on a Sun engine diagnostic machine from 30 years ago; back then it seems to me that the secondary only had a single HV pulse - but I really don't remember what the primary side looked like.

I know that the later model GM ignitions fire the primary side multiple times; I don't know how many times.

I haven't taken apart a modern lawnmower engine; seems that they don't use points anymore, so you don't have a reason to pull the flywheel. And, since I haven't had a reason to troubleshoot a small engine's ignition system, I don't know for certain what it would look like.

I don't think that anyone could say for certain that one particular waveform would be valid for ALL small engines nowadays. If the primary side is electronic, all bets are off.

[eta]
My simulation model makes a number of assumptions, and I could be off by quite a bit on some of them - I was basically just trying to get a reasonably close match to the signal that Debe picked up on his 'scope. Except without knowing the time/div setting of his 'scope so I don't know the resonant frequency of the primary side of the coil, I think what my simulation shows is a reasonably close approximation.
 
Last edited:

debe

Joined Sep 21, 2010
1,389
Here you go Wookie Circuit & whats used in a CDI brush cutter Ryobi Ign. The coil is moulded on the side of the unit , the HV & trig windings are wound on the U core that the magnets on the fly wheel pass. Yes i just have to find out whats inside.
 

Attachments

debe

Joined Sep 21, 2010
1,389
Hi RB the scope pic is acros the primary windings of a magneto type ignition with points on a small 4stroke engine when running. Your splice wave form across the points is close to real life.
 
Last edited:

SgtWookie

Joined Jul 17, 2007
22,230
Debe,
Thanks for the schematic & image.

Any chance you could answer The_RB's question - did you have a spark plug connected to the secondary when you snapped the photo of your 'scope?

[eta]
Oh, I see you answered it (spark plug was connected, engine running) - thanks! We cross-posted.
 

pistnbroke

Joined May 9, 2011
32
The briggs and stratton is not a CDI set up ...its a magneto with a transistor instead of the contacts and condenser ...when the current in the mag is at max it switches off the transistor and thats the same as opening the contacts....Typical circuit below..not my construction I dont do birdsnesting !!!
 

Attachments

debe

Joined Sep 21, 2010
1,389
That circuit is from my post on Atom Ign modules. The birdnest is like that as a prototype to figure out what the component values were. This is a finished unit.
 

Attachments

THE_RB

Joined Feb 11, 2008
5,438
RB you know nothing about ignition systems there is no 12 0 12 0 pulse ....if its b and s its an electronically controlled magneto or if very old contacts .
...[/B]
Actually I do know a little. I have designed transistor ignition systems from scratch, also worked for many years in repair fixing tachos and auto tune up meters, and as a motorcycle enthusiast have spent many years diagnosing, fixing and installing ignitions on bikes from Harleys down.

The OP did not mention if it was Briggs and Stratton motor or not.

With a 12v ignition system the primary of the coil is at 12v whenever the points (or transistor) is open, and at roughly 0v whenever the points or transistor is closed. And there is some HV noise spikes on the transitions.

A CDI dumps a pulse from the capacitor into the coil so it's more a dynamic "pulse" than a squarewave, and a magneto can make more of a pulse than a squarewave too.

For direct (primary) connect tacho they will all work pretty well from a resistive divider, RC filter, and a zener diode never hurts and is recommended if it is a CDI or magneto.

So 2 resistors, a cap, and a zener should get it up and running. The older tune up meters used 2 resistors and a cap into a transistor base (which acts liek a 0.6v zener) and a reversed diode across the base/emitter to catch the negative spikes.

Debe, thanks for the info on the waveform. It's still much more ringing than I've ever seen form a 12v points system, it's probably a byproduct of the magneto, or possibly the sparkplug was disconnected.
 

Thread Starter

remout

Joined Apr 18, 2011
14
Guys,

Thank you very much! Interesting ensuing discussion.

It was indeed a B&S style ignition. This one happened to be from a Honda, 4 cycle, lawn mower.

I have it working, using an opto isolator.

Can I use this current, from the primary side of the coil, to power my Arduino? I was thinking of putting a diode and a capacitor after the - terminal on the opto isolator. As I'm sure you can tell, I'm new to this and that's about as far as I got.

Thanks!
 

SgtWookie

Joined Jul 17, 2007
22,230
Well, you might be able to steal a little bit of current - but not much. You'd need to use a large resistor to limit the current, or your engine would become hard to start or wouldn't run at all.
 
Top