Print PDF
Favorite
Email
This instructable will show you how to make an audio modulated plasma arc using a flyback transformer and a NE555 timer chip driving a MOSFET.
This driver can be used to both obtain high voltage arcs and to use the arcs to produce sound waves (plasma speaker).
I will update this instructable over time.
Note: Please turn up your sound volume, it sounds much better in real life but my camera does not pick it up too well.
You may not, except with our express written permission, distribute or commercially exploit the content. Nor may you transmit it or store it in any other website or other form of electronic retrieval system.
Remove these ads by
Signing UpStep 1What you will need
Parts list:
1x Flyback tranformer
A flyback transformer, sometimes called a line output transformer are used in older CRT TV's and computer monitors to produce the high voltage needed to power the CRT and electron gun. They also have other auxiliary windings built into them that the TV manufacturers use to power other parts of the TV, so they are usually customised by the manufacturers
You can get flyback transformers out of older CRT monitors and TV's. They are the ones that have a big chunky chassis. There are also other instructables on this website showing how to remove them from the chassis and circuit board.
1x Power MOSFET,
I used an IRF540 as that is all that I had lying around. I strongly recommended using a MOSFET with a higher drain-to-source voltage than the IRF540, which is only 100v.
Just for an example IRFP460 would be well suited for this and IRFP250N and IRFP260N would also work. Any MOSFET that is rated for high voltage, has a low on resistance and can take more than 15 amps would be fine.
1x Heatsink
The TV board you got the flyback from is a good source for heatsinks.
1x NE555 timer chip
I also recommend using an IC socket (8 pin) for the 555 so you can easily remove the chip without de-soldering it.
2x 47 ohm resistors
1x 22 ohm resistor
1x 470 ohm resistor
2x 50K potentiometers
3x 1nF capacitors
1x 220uF electrolytic capacitor (any capacitor around this value and a minimum of 35v is fine).
1x 10nF capacitor
1x 100nF capacitor
1x fast diode, such as UF4007
1x NPN and PNP complementary bi-polar junction transistor pair (if you are following the first schematic). BD139 and BD140 can be used here.
Stripboard
12V power source
13amp mains fuse (Optional to protect the power source/supply).
Audio source (This could be an MP3 player of some sort).
Solder and soldering iron + some spare wire.
Disclaimer
I am in no way responsible if you mess up with this circuit. If you mess up, receive an electrical shock or burn your house down whilst making this circuit you have no-one to blame but yourself. By following this guide in order to make this circuit you agree to accepting all liability if something were to go wrong.
1x Flyback tranformer
A flyback transformer, sometimes called a line output transformer are used in older CRT TV's and computer monitors to produce the high voltage needed to power the CRT and electron gun. They also have other auxiliary windings built into them that the TV manufacturers use to power other parts of the TV, so they are usually customised by the manufacturers
You can get flyback transformers out of older CRT monitors and TV's. They are the ones that have a big chunky chassis. There are also other instructables on this website showing how to remove them from the chassis and circuit board.
1x Power MOSFET,
I used an IRF540 as that is all that I had lying around. I strongly recommended using a MOSFET with a higher drain-to-source voltage than the IRF540, which is only 100v.
Just for an example IRFP460 would be well suited for this and IRFP250N and IRFP260N would also work. Any MOSFET that is rated for high voltage, has a low on resistance and can take more than 15 amps would be fine.
1x Heatsink
The TV board you got the flyback from is a good source for heatsinks.
1x NE555 timer chip
I also recommend using an IC socket (8 pin) for the 555 so you can easily remove the chip without de-soldering it.
2x 47 ohm resistors
1x 22 ohm resistor
1x 470 ohm resistor
2x 50K potentiometers
3x 1nF capacitors
1x 220uF electrolytic capacitor (any capacitor around this value and a minimum of 35v is fine).
1x 10nF capacitor
1x 100nF capacitor
1x fast diode, such as UF4007
1x NPN and PNP complementary bi-polar junction transistor pair (if you are following the first schematic). BD139 and BD140 can be used here.
Stripboard
12V power source
13amp mains fuse (Optional to protect the power source/supply).
Audio source (This could be an MP3 player of some sort).
Solder and soldering iron + some spare wire.
Disclaimer
I am in no way responsible if you mess up with this circuit. If you mess up, receive an electrical shock or burn your house down whilst making this circuit you have no-one to blame but yourself. By following this guide in order to make this circuit you agree to accepting all liability if something were to go wrong.
| « Previous Step | Download PDFView All Steps | Next Step » |
I am going to post an updated version of the schematic on here soon which will help with lowering the amount of heat the MOSFET generates.
I am using a 12v cpu power supply capable of several (around 12) amps. Is the capacitor across the mosfet important because I do not have it in my circuit. What does it do?
Long story short: 9v batteries were the invention of satan and should all be destroyed
An NPN and PNP totem pole BJT buffer stage? If so I think possibly the high peak current draw of the gate charging might have drawn down the 9v's so much that the voltage sagged. Also you might have shoot-through happening and both BJT's turning on at the same time causing the 9v's to short.
Also if you are using separate supply for the 555 and flyback (which I recommend) then remove R4 and make R3 22 ohms. This will help switch the MOSFET faster.
Alex.
The diode and the capacitor across the mosfet are to prevent surges? Or do they help keep the mosfet cool/increase arc length? I have a 500v mosfet so i dont think I need to worry about surges, but if adding those components will make the arc better then I am all for it,
The diode and capacitor is there to help reduce the back-emf voltage spikes that will be produced on the primary side when the MOSFET turns off (its also one of the reasons the MOSFET gets so hot because it is avalanching every cycle).
You *might* be ok with that 500v MOSFET but its always good not to let the MOSFET's breakdown voltage be the only thing protecting it. If your MOSFET does blow then you need a larger capacitor here.
I have also found putting a small MKP capacitor in parallel with the primary coil helps reduce the voltage spikes and MOSFET heating a little. It makes the arcs smaller but thicker though.
As for making the arcs smaller, putting anything there that reduces back emf spikes will make the arc slightly smaller but its a compromise between arc length and durability. The only reason you get such a high voltage on the secondary coil is becuase of this back emf spike, if you were to put a diode in reverse with the primary coil like you would with a motor or relay then you would remove all back emf spikes but arc length will be much smaller becuase there will be no "flyback kick" to further boost the output voltage.
I also recommend having a permanent arc gap set up for the arc to take place, rather than always drawing arcs and moving the arc around. I say this becuase the back emf spikes are different depending on how much you are loading the secondary coil of the flyback, and with no arc at all or if the arc is unstable the back emf will much higher than if you had a steady arc going.
Hope this helps.
i was thinking, what if you put the diode in between the primary and the transistor? like
+12V -> one side of primary -> other side of primary w/ small capacitor across it -> DIODE -> mosfet with backwards diode-> ground.
and also have zener diodes put together across the gate, so any overvoltage (<24) to further protecy the mosfet?
and havee you tried IGBT's? do they work any better?
too bad, cool inst. sent friend somewhere else...
You would also need to figure out which are the primary coil pins.
Alex.
Sunshiine
Alex.
Have you made sure all connections are soldered securely and that there are no short circuits anywhere? Also what MOSFET did you use? What is your power supply voltage and how much current can it deliver?
Did you try reversing the polarity of your primary coil connections then doing a sweep of the frequency range via the potentiometers.
Alex.
Do you have a frequency counter of some sort to test if anything is coming out of pin 3?
Also you will need to remove the MOSFET in order to get a reliable reading.
Really go over the circuit and solder traces as I am sure it is just a simple wrong connection somewhere, I always make at-least one mistake when I follow schematics!
You're actual circuit diagram has a mistake in it (the decoupling capacitors are placed incorrectly).
You're missing a FET driver. These power MOSFETs have a fairly large gate capacitance. Most logic ICs can't provide enough power or sink it fast enough for that matter to switch these. Also, you need to avoid keeping your FET open for longer than a few microsecond. Else it will heat considerably faster resulting in damage, up to the point where the junction is destroyed due to heat. Just pick any FET driver for this, a 48 kHz isn't all that fast anyway.
Add a large resistor (1M and up) between the gate of the FET and ground to remove the charge from the gate quicker and to remove the charge when the circuit is switched off. Else the FET might be open from the start resulting in a larger spike on the power supply.
While your coupling is logical. You should realise it's still a high voltage device, and in most cases music players are fragile. You will need to use an optocoupler like an IL3000. See the datasheet if you don't know how to use it. The circuit you need is on the first page most of the time. Obviously you should use different sources for the input and the output opamps if you do this, else it's a pointless endeavour.
What pair of matched transistors would you recommend for driving the MOSFET gate?
Should the decoupling caps go directly across the supply rails?
If you are going to use transistors to drive the gate, take a matched nmos-pmos pair. If you use BJTs you risk them being on at the same time. With FETs you have the threshold voltage to keep you somewhat safe. Though it's hard to recommend what to use. I use the highly scientific method of using whatever I find on my desk/in my drawers.
And as a rule of thumb you should decouple your power supply and all logic ICs that switch. The 555 pretty much acting as an oscillator here means it qualifies for that criteria.
The values are trickier. I generally tend to use multiples of 10 myself cause I can find a lot of those around the lab. But it really depends on your application. In this case, on the power supply rails don't be afraid to go high value. Just be sure the initial charging of the capacitor doesn't overload the power supply. But considering how much current the flyback needs that sort of irrelevant here. Hence, any big fat elco will do. On the other hand, you should add a small 1nF (or any arbitrary value, not too big though) ceramic to that. Though not strictly correct you can see the ceramic as a charge supply to provide for fast switching and the elco is somewhat "slower". The actual reasons are to be found in the non ideal behaviour (Troubleshooting Analog Circuits by R. Pease has a nice chapter on non ideal behaviour of capacitors explained in a simple way). In this case the 555 becomes trickier. Generally a 555 is fairly low noise so any small capacitor will do. But here you'd have to test it with a scope cause I'm not sure how a 555 behaves under heavy load.
What I meant with asking what type of transistor pair to use is would those small bi-polar transistors you get in disposable cameras be suitable, or are they too small for this purpose? I have found both npn and pnp's on them and after googleing they appear to be matched/complimentary pairs.
I think in future experiences I will use separate 12v battery's for both the flyback and 555/logic. Although I do not have a oscilloscope myself, I came across a video on youtube where the guy showed the waveform with and with-out the capacitors.
With-out the capacitors the waveform of the 555 output distorts when drawing arcs and with them it is much cleaner.
I have only managed to kill one flyback and that was with my ZVS driver @ 36v input, the ground pin arced to one of the low voltage windings.
As I said earlier, using a NPN-PNP push-pull configuration is problematic as driver. You risk short circuiting the battery by doing so. Anyway, you need to figure out how much current you need. Lets assume a 1 ohm resistance of the power supply and a 2nF capacitor. The capacitor voltage is 12V*(1-exp(-t/(R*C)). We assume a capacitance is charged after three times RC. So it takes 3 times R*C to charge the capacitance. 6 ns to charge the gate capacitance assuming an ideal step from the 555.
Sadly the 555 doesn't output an ideal step, in fact it takes about 200 ns for the signal to rise! This is why ideally you use a PWM chip designed for faster operation but lets stick with the 555. Modelling this isn't exactly fun. So lets assume it ramps up in a linear fashion. Now you can solve this with differential equations and nicer methods but lets use laplace for this. The system function is 1/(R*C*s + 1). The input can be modelled as 12V/(200ns) * t (u(t) - u(t - 200ns)) + 12 * u(t-200ns). The laplace transform of this is less fun sadly, since I don't wish to torture you, I suggest the usage of MATLAB or Maple. Multiplying the two gives an even more fun equation to return to the time domain. Now the response to that I'm not even going to type here due to its length. We know it's charged if the voltage over the capacitor reaches 12*(1-exp(-3)). This results in t being 192ns meaning the maximum frequency you could switch at is 5 Mhz (interesting result in case you wish to use a faster method to drive the FET in later more advanced circuits, though in that case you'll also have to consider the fall time). Now we need to know the current though. Taking the expression for the output voltage over the capacitor and the fact the supply provides 12V you'll see it starts with a peak of 12A and only drops to what we can consider 0A past 200ns. Meaning your BJTs will need to be capable of supplying a burst of 12A. Check the datasheet of the BJT to see if it's suitable for it as I'm not familiar with the discharge circuit of a camera. It's worth a try anyway if you have enough spare parts.
And the output of the 555 distorts without the capacitors for a few reasons. First of all, all logic ICs have switching noise. This is inherent to the way TTL and especially CMOS logic is built. Second of all the transformer will cause inductive spikes on the power supply rail. Now drawing arcs causes some additional effects. Not to mention the inductance in close proximity due to drawing a few ampere isn't to be ignored either.