Solenoid Hit n Miss

[cfellows]

I still have one more solenoid left from my Scotch Yoke project and I have a couple of 4.5" cast iron flywheels



So.....



I'm thinking another solenoid project, a Hit n Miss Engine.  Since the duty cycle on this solenoid will probably be 10% or less, I can hit it with 24 or more volts which would drive 9 or more amps through it.  That should give it enough oommph to spin the engine over a few times between hits.  I'll use an arduino with a hall sensor to calculate RPM's and, when the RPM's drop below a certain, adjustable threshold, I can fire the solenoid.

Gotta think this through a bit, and I want to finish up my scotch yoke engine, so it may be a while before I start making chips. 

Chuck

[black85vette]

I like your combination of technologies.   I may have to give it a try, but likely with a Raspberry Pi since I already have 4 of them.   But, they are not as good at real time control applications.   Maybe just "good enough"?

[cfellows]

Yeah, I have a couple of Rasberry Pi's.  I've played around with the WIFI and with Python, but haven't done anything with the GPIO.  The Arduino, as you say, is more straight forward to use with sensors and controlling things.  But the PI should work just fine.   

[cfellows]

It also occurred to me that an Arduino could be used to control the ignition on an internal combustion engine as well.  The governor assembly on a hit n miss engine could be replaced with just a hall sensor and the Arduino to keep a track of the speed.  When the RPM fell below a certain threshold, the ignition would be triggered.  To perhaps make it a little more interesting, the threshold could be  altered randomly between and upper and lower limit to slightly change the RPM at which the engine fires.  Just some idle musing.

Chuck

[cfellows]

Before I build a whole new solenoid powered hit n miss engine, I decided to test the idea on one of my existing engines.  This sad little vertical single is one of the first compressed air engines I built many years ago.  It has been used as a lash-up to test all manner of slave valve designs over the years.  So, it seemed fitting that I adapt my solenoid to it and see if it will work.



Before I can test it, I have to attach a hall sensor on the frame.  Sounds like another 3D printer project... And I need to make a magnet disk for the crankshaft.  Hopefully I'll have something running in the next couple of days.

Chuck

[cfellows]

I designed a Hall Sensor mount for this engine using Fusion 360.



Then printed it in PLA on my little 3D Printer.  Here it is mounted on the engine with the Hall Sensor in place...



Chuck

[cfellows]

Last night I soldered up the control board for my Solenoid Hit n Miss engine.   





It's a lot tidier, thanks to better design work and, undoubtedly, in part, at least to my new Hakko FX-888D soldering station...



Seriously, a temperature controlled soldering iron does make a world of difference.

I also completed the initial version of the Arduino software.  The algorithm will continuously count the number of milliseconds between revolutions.  When the elapsed milliseconds crosses a threshold, the solenoid will fire.  I'm also using a timer to control the duration of the solenoid on state.  I will undoubtedly have to tune the threshold and the solenoid-on variables and I will probably add controls so I can vary them as the engine is running.

Now all I have to do is connect up the hall sensor, solenoid, and power supply to the control board and give it a try.  I'm getting excited!

Chuck

[yogi]

Looking good Chuck! 
I really like what you are doing. Trying new, different things. I'll be following along with interest, to see how it turns out.

P.S. I envy your soldering iron.

[cfellows]

Thanks, Yogi.  Well, maybe this is the video you've been waiting for!  I got the engine working tonight.  I had a bad solder connection on the driver board and somehow managed to fry another Arduino Nano.  Luckily they are pretty cheap.  I decided to power the Arduino separately rather than from the 12 volt supply which is driving the solenoid.  I don't think the little voltage regulator on the Nano can handle the spikes.  Anyway, here's the video.

<a href="https://www.youtube.com/watch?v=N_iDxrgFgNo" target="_blank">http://www.youtube.com/watch?v=N_iDxrgFgNo</a>

I think now I need to add a couple of potentiometers to let me control the lower RPM firing threshold and the duration of the on-time for the solenoid.   Then, I'll get to work on a better model engine.

Chuck

[PStechPaul]

Seems to be running pretty well! It does have a sound similar to the hit-and-miss engines I've seen at Cabin Fever and Rough and Tumble, so you have definitely succeeded in that regard. 

I found a schematic of the Arduino Nano, and it looks like it uses an LM1117 regulator, with no input capacitor or other stabilization and protection of the input. It can handle 20V but there is no filtering to handle spikes, and that's probably why it was damaged.

http://download.arduino.org/products/NANO/Arduino%20Nano-Rev3.2-SCH.pdf

The regulator:

http://www.ti.com/lit/ds/symlink/lm1117.pdf

The design notes recommend a 10 uF tantalum capacitor on the input, and as much as 100 uF on the output. The Arduino schematic shows three 1 uF capacitors and one 4.7 uF on the 5V line. If you want to be able to use the circuit on a single 12V supply (such as a battery), and want to protect against spikes, it may be good to add a small inductor (perhaps 500 uH to 10 mH) as well as the recommended 10 uF capacitor. It's probably not necessary to use tantalum, but it should be a low ESR type, like ceramic multilayer.

[ShopShoe]

Chuck,

I like the way that's running now. Definite "hit" and very smooth and quiet series of "miss" cycles following: Sweet.

I understand your discussion about how the electronics work, but my knowledge is lacking when the forum discussion turns to adding caps and etc. and why. When the design becomes final will you post a schematic so my slow old brain can study it and figure out the additions or changes.

I have a solenoid or other type of coil-driven electric engine as a future wanna-build project, so your posts on these engines is instructional for me. (Not that I haven't followed and enjoyed all your work since I joined these forums.)

Regards, and thank you,

ShopShoe

[cfellows]

Seems to be running pretty well! It does have a sound similar to the hit-and-miss engines I've seen at Cabin Fever and Rough and Tumble, so you have definitely succeeded in that regard. 

I found a schematic of the Arduino Nano, and it looks like it uses an LM1117 regulator, with no input capacitor or other stabilization and protection of the input. It can handle 20V but there is no filtering to handle spikes, and that's probably why it was damaged.

http://download.arduino.org/products/NANO/Arduino%20Nano-Rev3.2-SCH.pdf

The regulator:

http://www.ti.com/lit/ds/symlink/lm1117.pdf

The design notes recommend a 10 uF tantalum capacitor on the input, and as much as 100 uF on the output. The Arduino schematic shows three 1 uF capacitors and one 4.7 uF on the 5V line. If you want to be able to use the circuit on a single 12V supply (such as a battery), and want to protect against spikes, it may be good to add a small inductor (perhaps 500 uH to 10 mH) as well as the recommended 10 uF capacitor. It's probably not necessary to use tantalum, but it should be a low ESR type, like ceramic multilayer.

I'm not entirely sure what fried the Nano.  Initially, the software was running correctly.  I had programmed the on-board LED on pin 13 to turn on and off simultaneously with triggering and turning off the MOSFET.  The onboard LED was operating as I would expect it to when the magnet triggered the hall sensor.  The only problem I had was that the solenoid wasn't activating. 

While troubleshooting, I had the circuit board upside down measuring voltages at various points and I noticed the main LED on the Nano, the one that's always lit, periodically flashing brighter then dimming again when I pressed on some of the connections with my probes.  I figured that was not good, and shifted my attention to pressing various connections to see if I could determine what was causing the LED to flash brighter.  It was at that point that I discovered the open solder connection between the MOSFET source and ground.  So, I turned everything off, fixed the open solder connection and tried firing it up again. The NANO LED light came on steadily at normal brightness, but the program wasn't working, at all.  I tried reloading the program from my PC, but the upload just hung.  So, I replaced the broken NANO with a new one, severed the connection between 12v and VIN pin and powered the NANO with the USB connection to my PC.  At that point everything, including the solenoid began working.  So, I still don't know what was causing the main LED on the NANO to flash bright, but I'm pretty sure that was what fried it.

Chuck

[Roger B]

Cool    That's a definite hit and miss action (4-5 misses?) I do enjoy your various solenoid and compressed air engines     

[PStechPaul]

Dang, I forgot to copy my post and the forum software gave an error and I lost it.

Post mortem analysis:

With the source disconnected from ground, the MOSFET could have turned on, or somehow the 12V supply connected through the solenoid and the gate to the output of the Arduino. If 12V was applied to a pin, the protection diodes would have routed it to the power supply and caused it to go to 12V, which would make the LED brighter and probably fry the Arduino.

Also, be careful about powering the Arduino from the USB port and the solenoid and MOSFET with an external supply. If something goes wrong, the USB port may be damaged, and possibly the motherboard as well. It might be good to use an opto-coupler to drive the gate of the MOSFET.

[cfellows]

First of all, I'm happy to report that my thyroid medication has started kicking in and my energy levels have skyrocketed.  I can only say, if you are feeling too tired to get off the couch, don't hesitate to have your thyroid levels checked.  Hypothyroidism is fairly common as we age and the synthetic hormone taken daily can make a world of difference.

And now to the solenoid powered hit n miss engine.  This little engine started out life as a compressed air powered, single cylinder, 4 stroke engine.  It was one of my early engines that used the slave valve.  I had always meant for it to be a hit n miss engine, but never got around to making the conversion.







I replaced the compressed air head with the solenoid and also replaced the 3" flywheels with these larger, 4.5" flywheels.  Using the Arduino circuit I had already put together, the engine runs quite nicely, if a little slower than I'd like. 

Today, I designed and 3D printed a semi enclosure for the circuit board and to hold 3 potentiometers. 



The potentiometers will control (1) the ignition timing, (2) the power pulse duration, and (3) the lower RPM threshold at which it will fire.   These three controls will simulate the timing, the throttle, and the governor speed.   As you can see, I don't have the potentiometers wired up yet and I have modify the Arduino software to incorporate the controls.  Not much too it and I should get it done in the next day or two.



There are three sets of wires from the controller.  The black and red wires go to the 12 volt solenoid power supply, the 3 wire harness goes to the hall sensor mounted next to the crankshaft, and the two white wires from the solenoid connect to the terminal block next to the MOSFET.  For now, the arduino is powered by a 5 volt USB supply.

Getting excited to see how this works out.

Chuck