Piezo crystal from spark ignitor to capacitor to spark gap?

Thread Starter

David Nolan

Joined Aug 4, 2021
30
download.jpg


Piezo crystal from spark ignitor to capacitor to spark gap?



So, I'm making a little combustion engine, it's a passion project and I've always wanted to make one, specifically using piezoelectric ignition, yup those little things in barbecue igniters. Now I've done the math and it can work and believe me many model engines use a piezo ignition system, they're just not common is all. You might be wondering why I'm not on an engine building forum and trust me I am but those guys aren't so great when your asking amateur level electronic questions. Now I've ran into one specific issue, the piezo crystals have to be struck by a spring, because striking them with something directly connected to the drive train would lead to say... 'rapid disassembly' of the crystals.

Now the problem with using springs is that although they are pretty quick, they just aren't quick enough, meaning that if your at high rpms, they're a little too slow. Now I admit it's only an issue at formula one level rpms but I am ambitious and I don't want unnecessary bottlenecks if I don't need them. So you can add in a predetermined delay but that leads to more issues when your considering that the rpm is dynamic. Bottom line relying on a spring is a bad choice. So I want to keep my piezo crystal ignition system, basically because I think it's cool, that's pretty much it. If I'm going to do that I need a middleman of sorts, that can instantly discharge a spark when called upon. So I was thinking of using the drivetrain to trigger a spring which will strike a crystal, which will charge a capacitor, then when the drive train triggers another switch the capacitor will discharge and create a spark exactly when I need it.

So, guys going to be honest, I know pretty much nothing about electronics, well I know the basics but this stuff is pretty much still magic to me and I am in no way capable of designing a circuit like this myself, fine with the mechanical stuff but really need someone to tell me what components to buy and how to connect them up in a circuit. If it helps I think a piezo crystal releases about 12,000 volts when struck.

So more concisely...

I need someone to design a circuit for me that consists of a capacitor that can be charged by a piezo crystal ( the same one's in those cheap barbecue lighters) that when triggered by a switch (a normal trigger switch will do but if you folks think I need a mosfet for that super fast on and off stuff, that's fine by me I honestly have no idea.) will create a spark over a spark gap ( No need to think of a actual spark plug just two wires separated by a few 'mm' will do. ) that doesn't destroy the circuit in the process. Now I'm pretty sure Piezo's create AC current so maybe I need a diode for DC shenanigans . That's your guys call. Please try to explain what your doing and just tell me what to buy. That would be fantastic! Else I'll just be overcharged by some pcb designer on Fiver.

Thank you!

- David
 

scorbin1

Joined Dec 24, 2019
103
I honestly don't have much experience with this sort of system used as an ignition source, but I imagine the main issue is going to be how hard do you have to hit the crystal to get enough power out. I just watched a documentary on spark plug design and they mentioned a company that used piezoelectric ignition systems but I can't remember who it was, a patent to reference would be great right now. I believe they used a combination of springs to achieve proper timing without storing it, and I think that might be where you will run into issues. Storage has inherent losses and you would have to turn the stored charge back into a high voltage spark, which will have losses as well. Essentially it would be a magneto system with a piezo crystal in place of the magneto. The big problem I see is that magneto's constantly produce power where piezo electric crystals only produce power when struck. Two problems stem from that

1. can you get enough energy out of that crystal in the 10's of ms it's producing power to make up for the losses inherent to storing and converting the stored charge back into a high voltage spark and still produce the spark?
2. A magneto connected to a capacitor would always be charging the capacitor so current only flows into the capacitor from the magneto, however a piezo-electric crystal not producing power would leach voltage from the capacitor and produce vibrations instead. You would either need some diodes to prevent current backflow(which would also have some losses) or a second points system to time the connection between crystal and capacitor to only when the crystal is producing power.

I did not do the math, and perhaps these are not as big of an issue as I suspect, but these are my thoughts so far.

edit: Looking back I don't think I read your post thoroughly enough. I assumed that in wanting to use piezo electric that you were trying to avoid modern electronics and keep it simple. Now I that I look back I assume you don't care too much about the type of components used, as long as it's sourced from piezoelectric, correct? If that's the case, perhaps another approach to piezo electric power generation could be the answer. Are you dead set on using that specific type of crystal? I'm thinking of perhaps some sort of mechanically driven crystal slapping mechanism, if you will. Like a paddle wheel of sorts that would continously strike the crystal as the engine turned. I think if that would be acceptable it could be very nice and simple and should only require that mechanism, a capacitor and some sort of points to open the connection to the primary of the coil when the spark is desired.
 
Last edited:

Thread Starter

David Nolan

Joined Aug 4, 2021
30
I honestly don't have much experience with this sort of system used as an ignition source, but I imagine the main issue is going to be how hard do you have to hit the crystal to get enough power out. I just watched a documentary on spark plug design and they mentioned a company that used piezoelectric ignition systems but I can't remember who it was, a patent to reference would be great right now. I believe they used a combination of springs to achieve proper timing without storing it, and I think that might be where you will run into issues. Storage has inherent losses and you would have to turn the stored charge back into a high voltage spark, which will have losses as well. Essentially it would be a magneto system with a piezo crystal in place of the magneto. The big problem I see is that magneto's constantly produce power where piezo electric crystals only produce power when struck. Two problems stem from that

1. can you get enough energy out of that crystal in the 10's of ms it's producing power to make up for the losses inherent to storing and converting the stored charge back into a high voltage spark and still produce the spark?
2. A magneto connected to a capacitor would always be charging the capacitor so current only flows into the capacitor from the magneto, however a piezo-electric crystal not producing power would leach voltage from the capacitor and produce vibrations instead. You would either need some diodes to prevent current backflow(which would also have some losses) or a second points system to time the connection between crystal and capacitor to only when the crystal is producing power.

I did not do the math, and perhaps these are not as big of an issue as I suspect, but these are my thoughts so far.

edit: Looking back I don't think I read your post thoroughly enough. I assumed that in wanting to use piezo electric that you were trying to avoid modern electronics and keep it simple. Now I that I look back I assume you don't care too much about the type of components used, as long as it's sourced from piezoelectric, correct? If that's the case, perhaps another approach to piezo electric power generation could be the answer. Are you dead set on using that specific type of crystal? I'm thinking of perhaps some sort of mechanically driven crystal slapping mechanism, if you will. Like a paddle wheel of sorts that would continously strike the crystal as the engine turned. I think if that would be acceptable it could be very nice and simple and should only require that mechanism, a capacitor and some sort of points to open the connection to the primary of the coil when the spark is desired.

Well yes, that's what I had in mind, I may use multiple Piezo's for a single ignition if need be, see I can't use a slapping mechanism like the paddle wheel you described because as the engine rpm gets higher and higher the crystal is more likely to explode when struck by the mechanism (Presuming it's connected directly to the drivetrain, without a spring). so you need to use springs and have the engine cock or load the springs, that way the amount of force the crystal receives is uniform. I'm going to need maybe 50 piezo crystals at the minimum, per cylinder. But again, what I really need help is with the circuit. The actual mechanics is my area of 'expertise' The issue with the springs is that they are too slow, by that I mean the delay between me telling the spring to fire and the spark actually occurring is delayed as the striker is sprung towards the crystal
 
Last edited:

AlbertHall

Joined Jun 4, 2014
12,345
A common system in I.C. engines is to advance the spark at higher r.p.m. by rotating the plate holding the points. May be you could implement a similar system.
 

Thread Starter

David Nolan

Joined Aug 4, 2021
30
A common system in I.C. engines is to advance the spark at higher r.p.m. by rotating the plate holding the points. May be you could implement a similar system.

I considered doing that, triggering the springs earlier and earlier as the rpms become higher and higher, this makes things pretty complicated. I think using a capacitor would be the easier option.
 

Sensacell

Joined Jun 19, 2012
3,432
The idea of storing the high voltage output to delay or control the timing is not practical.
The main problem is that the spark voltage can be 10 to 20 kV, which is up in the range where electronic components are very large, expensive, and inefficient.

Switching this level of voltage with electronics is really impractical, for the most part, vacuum tubes are the only viable switching elements that can handle it.

Piezo crystals are high voltage and high impedance devices, totally not the right choice if you want to harvest and process the energy with semiconductors.

Consider an alternator-driven CDI ignition system, the energy is harvested from the crankshaft with a low impedance, low voltage coil (alternator) - at a voltage where semiconductors shine.
The spark energy is derived from a relatively low voltage applied to a transformer (ignition coil) that acts to transform the voltage up to ~20kV where a crude mechanical switching device (distributor) routes the energy to the appropriate spark plug.

The entire concept is identical to what you propose, but the architecture is designed to take advantage of what the components do well, no extreme performance is required from any of the components, so the system is cheap, compact, and reliable.
 

Thread Starter

David Nolan

Joined Aug 4, 2021
30
The idea of storing the high voltage output to delay or control the timing is not practical.
The main problem is that the spark voltage can be 10 to 20 kV, which is up in the range where electronic components are very large, expensive, and inefficient.

Switching this level of voltage with electronics is really impractical, for the most part, vacuum tubes are the only viable switching elements that can handle it.

Piezo crystals are high voltage and high impedance devices, totally not the right choice if you want to harvest and process the energy with semiconductors.

Consider an alternator-driven CDI ignition system, the energy is harvested from the crankshaft with a low impedance, low voltage coil (alternator) - at a voltage where semiconductors shine.
The spark energy is derived from a relatively low voltage applied to a transformer (ignition coil) that acts to transform the voltage up to ~20kV where a crude mechanical switching device (distributor) routes the energy to the appropriate spark plug.

The entire concept is identical to what you propose, but the architecture is designed to take advantage of what the components do well, no extreme performance is required from any of the components, so the system is cheap, compact, and reliable.

I want to stick to Piezo ignition. Your sure that a capacitor dealing with 10-20 KV has to be expensive or large. I've seen small capacitors give off sparks before, so this confuses me?
 

Ya’akov

Joined Jan 27, 2019
9,070
I want to stick to Piezo ignition. Your sure that a capacitor dealing with 10-20 KV has to be expensive or large. I've seen small capacitors give off sparks before, so this confuses me?
Capacitors aren't ideal. Every capacitor is also an inductor and a resistor. The very high voltage, very low current output of a piezo crystal will not charge a capacitor to anywhere near the voltage of the spark it generates. The capacitors you have seen produce sparks are going to be rated on the order of hundreds, not thousands, of volts and there will be a lot more current.

You could possibly use an energy harvesting scheme with several piezo crystals generating power at a relatively high frequency to charge a capacitor that would act as an accumulator, but it is exceedingly unlikely you could find an arrangement where a capacitor is employed as a delay for an advanced spark from the piezo.
 

Audioguru again

Joined Oct 21, 2019
6,673
High voltage produces a spark or high current also produces a spark.
Your piezo being hit is mechanical. Ignition has used electronics instead for years.
The piezo ignition on my BBQ wears out soon. The electronic ignition on my furnace lasts for many years.
 

Sensacell

Joined Jun 19, 2012
3,432
I want to stick to Piezo ignition. Your sure that a capacitor dealing with 10-20 KV has to be expensive or large. I've seen small capacitors give off sparks before, so this confuses me?

How does a sparkplug fire?
A rapidly increasing voltage applied causes the air gap to ionize and break down- a spark.
How would you create this rapidly changing voltage?
If you were successful in charging a capacitor to 20 kV, it would be a DC voltage, you would need a switch capable of handling this voltage to initiate the spark- at exactly the correct time.

Maybe you have seen sparks when a capacitor is shorted to discharge it?
A small capacitor can contain enough energy to make a spark, but a method of starting the arc is required, a fixed air gap with a voltage under the breakdown voltage will not create a spark.
The gap needs to be forced to break down, and that requires a rapid voltage spike significantly over the gaps breakdown voltage.
 

scorbin1

Joined Dec 24, 2019
103
Assuming you can generate enough continous power from a piezoelectric array you could potentially store the charge in the capacitor, and use a points type system to interrupt the primary side of the coil to induce a high voltage spark. As others have said, it's definitely not going to be efficient or practical but you're not building an engine for practical use. Even with an array, a single synchronized pulse from an array is not going to charge the capacitor near enough in such short duration due to the internal resistance of the capacitor. You would have to produce a continuous pulse train, with a duty cycle as close to 100% duty cycle as possible. With the very fast transient voltage produced by a crystal, even repeatedly striking the crystal at a rate on the order of hundereds of strikes per second, the duty cycle of that pulse train is probably going to be on the order of a fraction of a percent up into the single digits maybe. I can't find any documentation with specific numbers and don't have a crystal to take measurements, so that is all in theory and surely prone to some degree of error but I think it's going to be a significant problem. I'm honestly a little intrigued with this idea and I'm curious what it would take to acheive this, however impractical it may be, so I think a thought experiment is in order. I'll definitely share what I find. If anyone has links to any resources regarding crystal waveforms and/or the relation of power out to size of crystal and striking force, that would be helpful. That being said, I don't think the paddle wheel concept, or at least something very similar, is going to be optional. If you can't produce an, at least mostly, continous stream of pulses, then you will have to somehow disconnect the connection between the capacitor and the crystal when it isn't producing power requiring a SECOND set of points, otherwise the crystal will consume power from the capacitor and create mechanical motion(vibration) when it isn't producing power. I am still not sure that even a continous stream of pulses would produce the required duty cycle though. The transient pulse from the crystal is pretty sharp meaning that the duty cycle is very short, and essentially you would have to produce the next pulse the instant the first pulse is finished, perhaps even faster. That's going to be a whole lot of pulses. Perhaps diodes could be used to prevent backflow into the crystal, but I have a feeling the losses in the diode would be pretty significant.
 

timm27

Joined Dec 11, 2020
11
Modern automobile (petrol) engines typically use a coil-pack ignition coil to fire the spark plug. This works pretty much in the same way as the single ignition coil. The coil is connected to a 12 volt supply which then behaves both as an electromagnet an inductor and a transformer. Then at the appropriate time the 12v supply is disconnected - classically the distributor points open, nowadays this will be a transistor or FET in the ECU. The current collapses rapidly to zero and the voltage across the points (essentially across the coil) rises very rapidly to a few hundred volts. This is the primary winding of the ignition coil. The secondary winding probably has 100x as many turns and the hundreds of volts across the primary become tens of thousands of volts across secondary coil and the spark plug and the energy stored in the coil is suddenly dumped into the sparkplug, giving a nice fat juicy spark to ignite the mixture. It can do this a few hundred times per second. A piezo ignitor carefully designed might give you a few big sparks per second, but magnetics are surely the way to go.

Your engine is really impressive and mounting an OTS coil-pack to each cylinder might mar its good looks, but a single coil and a distributor definitely would not do it justice. If there is a magnetics guru on the forum maybe they could dive to a nice design, perhaps based n a wide gapped ferrite core?
 

MisterBill2

Joined Jan 23, 2018
18,174
If there are already piezo-type spark igniters for model engines then you need to see how they do it. For that 4-cyl engine in the picture, 6000 RPM means 3000 sparks per minute, =3000/4= 750 sparks per minute for each plug. That will take a cam to lift the spring and let it snap to hit the piezo. Study mechanical diesel injector pumps for ideas about that. Certainly mechanical advance will be required as well. The problem is that it takes as much spark energy to fire in a small engine as is needed in a big engine.
 
Top