Pulse multiplier for tachometer - opinions wanted on method

[eetech00]

Hi

I've been reading a little about outboard motors and tachs.
If I understand correctly, most tachs take a signal from the stator, correct?
if so, is this low voltage AC? Voltage level?

But you want to take the signal from a spark plug wire using a capacitive pickup sensor, right?

Also, with regard to the LM2917, changing the frequency vs pulses per revolution is trivial.

eT 

[Paul Price]

You can make your own sensitive gate SCR just using the transistors you have around, you need one pnp and one npn transistor, connect the pnp base to npn collector and the npn base to the pnp collector then add <100k resistors from base to emitter on both pnp and npn or it will be too sensitive. You can then trigger the resulting SCR via the npn or pnp base with your choice of voltage polarity. If you choose the pnp as the top transistor, then you ground the emitter of the npn transistor and your pnp emitter is the anode of the SCR.

[Paul Moir]

eetech00 - Yes what they do is usually take it from the battery charging coils wound into the stator.  The stator live under the flywheel which has a permanent magnet in it.  These coils are typically just rectified with a bridge rectifier and fed directly into the battery, so the maximum voltage across the coil is maybe 16v.  They're notorious for burning out their rectifier if the battery ever becomes disconnected while running, which also stops the tach (bridge fails shorted).  Some use regulators, and these have special provisions for the tach. 

The ignition system is completely independent, coming from another set of coils wound around the same stator and giving a few hundred volts to the CDI box.  Some outboards without charging coils will have provisions in the CDI box to give out a 12v pulse compatible with the charge coil tachometer.

The outboards I'm interested in have no tachometer provisions and no charging system, or if they do use a generator.  They often do not have stators at all.  To give you an idea, these would be just about any outboard made before the late 70s, and a few small ones after.

Paul Price - that's an exceedingly good idea; thank you.  I would much rather use some commonly available transistors than a very specific SCR, in addition to the flexibility the circuit gives.

[eetech00]

eetech00 - Yes what they do is usually take it from the battery charging coils wound into the stator.  The stator live under the flywheel which has a permanent magnet in it.  These coils are typically just rectified with a bridge rectifier and fed directly into the battery, so the maximum voltage across the coil is maybe 16v.  They're notorious for burning out their rectifier if the battery ever becomes disconnected while running, which also stops the tach (bridge fails shorted).  Some use regulators, and these have special provisions for the tach. 

The ignition system is completely independent, coming from another set of coils wound around the same stator and giving a few hundred volts to the CDI box.  Some outboards without charging coils will have provisions in the CDI box to give out a 12v pulse compatible with the charge coil tachometer.

The outboards I'm interested in have no tachometer provisions and no charging system, or if they do use a generator.  They often do not have stators at all.  To give you an idea, these would be just about any outboard made before the late 70s, and a few small ones after.

Thank you for that explanation.

It seems like the PLL approach is much more complicated than using an LM2917 chip. If the spark sensor can generate a low enough signal( 3-5v), we should be able to condition it with an op amp, and let the chip will do the rest.

What do you plan to use for the spark sensor?

What FSD does the tach gauge require?

eT

[Paul Moir]

eetech00 - I haven't really had time yet to properly investigate the F->V->V->F option.  Note I have to get the voltage output back to a frequency with a VCO because I want to use a normal everyday marine tachometer for the display.  So I guess the answer to your question on FSD is 400Hz or 600Hz or better yet 470Hz or 700Hz for a 7000RPM tach (though that is well above what these particular outboards ought to run at).  My general concerns are with accuracy:  I want to avoid having to calibrate the circuit.  With digital type options, I don't have to do this.  If the VCO can be implemented easily and accurately you're right, it may be the better way to go.

For the pickup, I'm hoping to do a few wraps of sold core wire around a spark plug lead.  This is similar to what the digital ones do, though in their case it's an antenna (assuming here; there's no ground return path on them).  I'm thinking I may have to put the detector circuit under the hood or use coax.  Still considering that one. The first step will be to get it working.

[Paul Price]

If you wind a few turns of insulated wire around a spark plug wire and connect each of the ends of this coil to a small diameter 75-ohm coax cable,  you have in fact created a precision pulse transformer, a shielded hi-frequency pulse transformer secondary output that can be fed through a small capacitor to a BJT amplifier or to directly to a sensitive gate SCR to produce a known-amplitude pulse (that can be further integrated or fed to a one-shot MVB) to then have a clean conditioned signal to feed into your pulse frequency (tach) measuring circuitry.

On the other hand, as you suggest, if you only connect the inner conductor of the coax to one end of the coil, you will have a antenna type connection, a capacitive coupling that can also be used to feed a JFET or BJT transistor front-end to your tach.

Automotive smog test-centers in L.A., I have noticed, use a clothe pin shaped (two half-cylinders of copper, the circumference which is slightly smaller than the diameter of a sparkplug wire) clamp device (it's about two-inches long)  to feed a coax connecting to the tach input of their equipment.

Keep in mind that a single spark plug discharge may yield several sparks (and so several pulses are fed to your tach) and so some RC integration or use of a one-shot or SCR  may be required to keep your tach error free. Smog test equipment requires only a temporary connection so a capacitive mechanical clothes pin clamp is the practical way to go.

[Joule Thief]

Automotive smog test-centers in L.A., I have noticed, use a clothe pin shaped (two half-cylinders of copper, the circumference which is slightly smaller than the diameter of a sparkplug wire) clamp device (it's about two-inches long)  to feed a coax connecting to the tach input of their equipment.

I have used these on CRO's without any difficulty.

http://www.ebay.com/itm/New-Hantek-HT25-8-High-Voltage-2nd-Ignition-Capacitive-Auto-Pickup-Probe-X10-4-/281362973737?pt=LH_DefaultDomain_0&hash=item41828a4429

[Joule Thief]

Hi

I've been reading a little about outboard motors and tachs.
If I understand correctly, most tachs take a signal from the stator, correct?
if so, is this low voltage AC? Voltage level?

Typical AC output from a stator falls in the 36 - 50 VAC range (dependent on RPM of the engine - higher RPM = higher AC output and higher frequency). Lawn and garden equipment with a typical charging system greater than 5 amps will have stator outputs around 8 AC cycles per crank revolution. I believe the same idea is used on diesel engine tachs.

[eetech00]

Typical AC output from a stator falls in the 36 - 50 VAC range (dependent on RPM of the engine - higher RPM = higher AC output and higher frequency). Lawn and garden equipment with a typical charging system greater than 5 amps will have stator outputs around 8 AC cycles per crank revolution. I believe the same idea is used on diesel engine tachs.

Is the output linear with respect to RPM?
Might be able to divide voltage down, condition with an opamp, then feed to a comparator.
Comparator output would be DC pulses reflecting frequency.

eT

[eetech00]

eetech00 - I haven't really had time yet to properly investigate the F->V->V->F option.  Note I have to get the voltage output back to a frequency with a VCO because I want to use a normal everyday marine tachometer for the display.  So I guess the answer to your question on FSD is 400Hz or 600Hz or better yet 470Hz or 700Hz for a 7000RPM tach (though that is well above what these particular outboards ought to run at).  My general concerns are with accuracy:  I want to avoid having to calibrate the circuit.  With digital type options, I don't have to do this.  If the VCO can be implemented easily and accurately you're right, it may be the better way to go.

Oh..ok.

One thing I didn't realize is that your Marine tach need a variable frequency input at some voltage level.
In view of that, then maybe the F2V converter might not be better.
I think all you really need is a good signal conditioning circuit that can interface a sensor,
attached to the spark plug wire, to your tach. Right?

For the pickup, I'm hoping to do a few wraps of sold core wire around a spark plug lead.  This is similar to what the digital ones do, though in their case it's an antenna (assuming here; there's no ground return path on them).  I'm thinking I may have to put the detector circuit under the hood or use coax.  Still considering that one. The first step will be to get it working.

What about a linear Hall Effect sensor attached to the spark plug wire?
Cost a dollar, I think..

eT

[TerminalJack505]

Here's another version of the circuit I posted earlier.  It would require less twiddling than the earlier circuit.

It makes use of the last comparator and uses a couple of BJTs to implement a SR latch.  A 555 timer is no longer necessary.  The top-most comparator resets the latch.

Technically it requires just two ICs (if you don't count the transistors as ICs) but it might be a little too much for your target audience to understand and build on their own.

[David Hess]

It would be worth checking out the service manual for the Tektronix DC504A frequency counter which uses a 4046 with external discrete VCO as a frequency multiplier operating from 1 Hz to 25 kHz.