Step 6: Test the prototype

Picture of Test the prototype
After you had built it, go ahead and test it!
If it does not work, don't get all frustrated or anything like that, go back and check your circuit closely, make sure there are no missing connections, and check the components are properly connected to the circuit...

Mine works quite well, unfortunately, the MOSFET is wasting large amounts of energy into heat, so it is blazing hot... I have to make the heat sink bigger to try keep it cool longer. But it still goes blazing hot after three minutes of continuous operation.
Oh well, that is my efficient design so far...

If yours working well with out very much problem, thats great!
lynchpman4 years ago
Why do you have to use a veroboard if the connections can just run through the breadboard?
jj.inc4 years ago
Why is that yellow cord hooked to your speaker, I am not saying this doesn't work because I'm sure it does, but I just wondered
Enoch234 years ago
why not use more mosfets. I have seen a scematic that uses 3 or 4 i believe, and they all have little heatsinks on them...
gobitz6 years ago
If you put a smaller heatsink on the mosfet and put it in mineral oil then the temperature should be quite lower, until it warms up the mineral oil.
You could make a crude water cooling loop with your idea. Simply gain a water cooling loop (a computer one), create a reservoir from something that iwll not melt, and have an input pipe and an output pipe. Then connect the fans and pump to a 12V DC supply and away you go. The pump may kill itself if you use mineral oil, but find something else. Anyway, the liquid is pumped past the MOSFET and then into a radiator, where fans dissipate the heat :D Should be really effective (note: SHOULD be, it may not work)(this is a disclaimer :])
Good thinking gobitz. I had forgotten all about this until you mentioned it.

Using a large container with a thin wall, filled with mineral oil, the entire surface of the container could be used as the heatsink. Perhaps a popcorn tin? It might not need to be that big, I don't know. I'd still attach a small heatsink directly to the MOSFET as gobitz suggested. I think it would help distribute the heat to the oil for better cooling.

As an electrician in the Navy I was given a basic knowledge of transformers and most transformers that I dealt with used the oil and case as their heatsink. Usually the case was around 2-3 times the size of the actuall transformer if memory serves me. I seem to remember that most of what is actually in a "pole pig" is just oil to help cool the transformer.

I like Gasburner's idea too but I think the oil is a much simpler and probably safer idea. No pumps, the oil works as an insulator, and less chance that a leak will cause any major problems. Not to mention it's probably cheaper unless you happen to have the spare water cooler already.

Theoretically the oil could eventually evaporate and as the MOSFET heats the oil you could end with a fire on your hands, but you'd have to run this nonstop for quite a while before it became an issue I think. Periodically checking your oil level and temp could easily prevent this though. You could easily have too little oil for this but never too much, so don't be afraid to use a bigger container.

Great instructable! I was actually reading this hoping I might be able to use it with a tesla coil. What do you think?
moshee6 years ago
hmmm. that heatsink you're using looks more like a giant piece of metal than a heatsink... could that be a problem?
 im sure it disperses heat much better than the mosfet alone
alexg506 years ago
would i be able to use a voltage comparator as a mosfet driver?(like the LM311N) my mosfets are also getting hot!
Hello everyone, I am currently using a computer power supply that i have adapted to a lab supply to power my speaker. I am having a problem with EM interference coming from the AC power grid, and tried running the speaker on some batteries. Like Plasmana, i used a 9V to power the TL494 side of the circuit, and am using a Yuasa NP7-12 (12V 7Ah) battery for the flyback. I can get the flyback and circuit running perfectly (with the exception of EM interference) when using the PSU, but absolutely nothing with the batteries. I think that the big battery needs to be charged, but i was wondering how you would check. It pushes less than one volt while running the flyback, but gives about 12 amps. If under no load at all and I test the battery leads with the Multimeter i get 12.3V and about 28A. Any suggestions? Thanks in Advance for any help you have to offer!
Maybe try using a small fan to cool the MOSFET.
whiplash6 years ago
Heya, i would just like to ask, could you be more specific in what you will need. could you tell me the different voltages for all the components, there are a lot of different capacitors and resistors out there. thanks
Arx6 years ago
Duty cycle shouldn't really be an issue. That fet has a nice low "Rds on" Either you're switching at too high a frequency for the fet (unlikely) or you're not switching it fast enough (taking too long between off and on) Perhaps the 494 doesn't put out enough current to drain the gate capacitance quickly enough? I'm just throwing out ideas, haven't really looked at the numbers. :)
rocketgeek Arx6 years ago
Gate charge (Qg) on the IRF540 is 71 nC, according to the International Rectifier datasheet. Current output of the TL494 is 200 mA... which implies a switching time of 355 ns... which means that the MOSFET is spending about 2.1% of the time partially on at a switching frequency of 30 kHz. Ouch! IR sells gate driver ICs capable of putting an amp or more into the gate for ~$1 on their site.
     BTW, that one does use a PAIR of IRF540s and they can get quite warm. It takes quite a lot of current because we're essentially shorting out the flyback transformer. The current has to be limited somehow. I've burned up lots flybacks, capacitors, and transistors due to crummy current limiting.
     If that is what's going on then this could benefit from using an SG3525. There's a flyback driver on my site that uses one http://kickme.to/lightningstalker It's the first circuit "How to Build a Transistorized Tesla Coil".
     You have to limit the current by using a shorter duty cycle or else get a bigger transistor. If it's that hot I'd try getting a temperature reading. It could be really hot but still within the temperature rating.
The fastest way to cut the heat is to switch the MOSFET harder, i.e. drive more current into and out of the gate.
endolith6 years ago
The MOSFET probably shouldn't get that hot. Maybe you need a transistor on the input to clamp it down and prevent it from being partially on? Can you look at it with an oscilloscope?
The key to turning the MOSFET on and off quickly is to supply lots of current to the gate. The TL494 can only drive about 200 mA... you really want an amp or more -- 2A-3A would not be out of line for the IRF540. See my response to step 2 for more detail.
Does anyone know how to make a bass plasma speaker because all the ones I've seen sound really tinny and don't have good bass responce? Thanks a lot Uberdum05
Why don't you use quite a few MOSFET's in parallel?
iBurn6 years ago
So plas.......how'd the board get that big ol black mark on it?