Speed sensor converter for sailboat (AC to square wave)

[dr_c]

Hello -- I have sailboat with an (existing) speed-sensing ‘paddlewheel’ sensor that I am trying to use with a new data display. I have gotten lucky because the sensor and the data display both have the same frequency per knot (~6Hz/knot) – however the existing sensor generates an AC signal whose frequency is proportional to the speed over water (similar to a variable reluctor pickup in a car, I suspect), while the data display is of the ‘hall effect’ style and expects pulses of approximately 11.2V DC in amplitude.

In other words, I need to convert a small-amplitude AC signal into a square wave of 11.2VDC.

Seems like a straightforward problem. . . Unfortunately, I’m a Chemical Engineer, not an Electrical, so phrases like ‘power supply current drain is independent of the magnitude of the power supply voltage’ or ‘input common-mode’ defy my understanding. Doing some reading, a ‘comparator’ seems like it would do the trick – when the sine wave is ‘positive’ relative to ground send the full saturation voltage to the output. Then stop when the sine wave goes to 0 or negative. Radio Shack has the LM339 quad comparator in-stock locally – using this chip, I devised the circuit in the attached sketch. Apparently there also needs to be a ‘pull up’ resistor? Not sure how I would size that, so I’ll ask for help!

Thank-you and please comment. I’d rather learn something about electronics than put a hole in the bottom of my boat for a new speed sensor.

Cheers,
Dr. C.

 

[ericgibbs]

hi,
A LM393 comparator can be used to convert a low magnitude sine wave to a square wave of 11.2V

Whats the magnitude of the sine wave signal.?

EDIT:
A simple circuit like this.

 

[dr_c]

Thanks for the reply - the speed sensor's output is speed-dependent. If I read the specs correctly, for the range of interest, the output voltage will be between 0-8 volts peak-to-peak (and will vary with frequency from 0-48Hz).

That is, the frequency is 6Hz/knot and the voltage output is 1V peak-to-peak/knot -- and I would be 'screaming' at 8 knots!

 

[ericgibbs]

Thanks for the reply - the speed sensor's output is speed-dependent. If I read the specs correctly, for the range of interest, the output voltage will be between 0-8 volts peak-to-peak (and will vary with frequency from 0-48Hz).

That is, the frequency is 6Hz/knot and the voltage output is 1V peak-to-peak/knot -- and I would be 'screaming' at 8 knots!

hi,
The circuit I posted will work over a frequency range of DC thru 1KHz.

You say about 8knots, is the boat around the 25ft overall length mark.??

Are you able to follow the circuit OK.?

 

[dr_c]

Thanks, Eric. Yes, the boat is a 1975 Sabre 28 -- length at waterline is just under 23 feet, so theoretical hull speed is, of course, actually under 8 knots.

The schematic makes sense -- two questions:
* What does the notation: "wiper = 0.25" mean? Some 'base' resistance in the variable resistor?
* Assuming I get the pinout staight, should I be able to use the quad LM339 in place of the LM393?

Thanks a bunch. Looking forward to building this!

-Tony

 

[ericgibbs]

Thanks, Eric. Yes, the boat is a 1975 Sabre 28 -- length at waterline is just under 23 feet, so theoretical hull speed is, of course, actually under 8 knots.

The schematic makes sense -- two questions:
* What does the notation: "wiper = 0.25" mean? Some 'base' resistance in the variable resistor?
* Assuming I get the pinout staight, should I be able to use the quad LM339 in place of the LM393?

Thanks a bunch. Looking forward to building this!

-Tony

hi,
As you can see the image was from a LTSpice simulation run.

The wiper 0.25 means that the pot has been set to 25% of it rotational value, thats all.

You can adjust the actual pot to suit your application.

An LM339 should be OK, I will double check.

OT: I used to sail a 25ft Westerly yacht.

EDIT:
Runs OK, Connect all unused INPUTS to 0V

 

[dr_c]

Thanks! FYI, this project is to connect the sender paddlewheel from a 'Signet Marine' SMARTpak SL80, to a Raymarine ST40 Bidata display.

OT: If I'm not mistaken, I believe the 25 foot westerly is similar to the Sabre 28. Kitchen is in a different spot, but both are nice cruising boats that feel bigger than they are. Good headroom down below and nice layouts.

-T

 

[Bill Hill]

Thanks for these posts.

I have a slightly different situation with an old speed trasducer that has a magnet on the paddlewheel that closes a reed switch contact each revolution. so 0V-12V pulses are generated in a similar way to a hall effect paddlewheel, BUT I think the pulses must be noisy. Modern pulse counting type speed instruments connected to this paddlewheel give erratic readings. I tried a simple RC filter to maybe subdue some of the noise, it seems a little better for certain speeds, but in the end is still unusable.
I hate to replace the transducer with a modern one as the current one is narrow, and has a special mount on the centerline so that it reads the same on both tacks. Can't do that with larger diameter new transducers.

Do you think the comaparator circuit might clean up the edges/spikes of a mechanical reed switch transducer so that olny the main pulse is seen?
And, will the comparator turn off if the signal drops to zero but does not go negative?

Thanks

 

[ericgibbs]

hi,
I would suggest an R/C filter at the input of the circuit shown in post #2.

If the pulses from the reed are 0v to 12V, you could adjust that trim potentiometer so that only the main pulse gets to the output of the LM339.

EDIT:
I would suggest these changes.

Added a 8.2V Zener diode at the R/c

 

[alec_t]

I think the pulses must be noisy. Modern pulse counting type speed instruments connected to this paddlewheel give erratic readings

That sounds like a classical switch-bounce problem. A de-bounce circuit, rather than an RC filter, might be the solution.

 

[KeepItSimpleStupid]

The LM1815 has been around for years. https://www.electro-tech-online.com/custompdfs/2012/04/snosbu8e.pdf