The Mini SSTC

Page Created: 1/15/04

Updated 4/30/07

I built this small coil in about 1 day of work.  Its small and relatively safe.  It produces up to 7" of spark, though this is limited to my small heatsinks getting too warm.  If one was to copy the design but use larger heatsinks, the power could be much greater.  Please scroll down to the end of the page to see the finalized schematic!

Here are some pictures of the original setup.  It worked well but needed a really long antenna to work properly.  See the updates for the slightly revised version.

Schematic (not final, see end of page)

Update 1/18/04

I removed 2 primary turns for a total of just 4 turns.  This surprisingly helped with heating problems of the MOSFETs!  I also removed the toroid and now just have a wire pointing off the edge of the coil.  The coil runs much happier now.  The heatsinks have yet to get overheated.  Also, these changes brought about 7" sparks (compared with 6" previously).  Here are some snapshots of the new setup and the various sparks this little coil can create. 

Here the coil is operating with a moderate pulse width and rate.

 Here the coil is running at a low pulse rate and a low duty cycle.

 Now the pulse rate is maxed but still a low on time.

 Low pulse rate, but now the duty cycle is a bit higher, notice the thickness.

 The highest duty cycle at a low pulse rate produces flame like discharges that do burn things easily.

How the coil works

Here is an overview of the schematic and how the coil operates.  In the top of the schematic we see the low voltage power supply section composed of a small transformer, 4 diodes, some caps and voltage regulators.  The 7812 gives 12V at about 1A.  This regulator should be heatsinked as it supplies the power to the demanding gate driver section.  The 7805 supplies 5V for the logic ICs.  U1 is any typical 555 running at about 200khz or so (this is not critical as long as it is running relatively fast, or about 1/2 the frequency you intend the coil to run at).  Its sole purpose is to trigger the oscillation.  There is a 100K resistor on its output that makes the 555 look "weak" to the input of the 74hc14.  In fact, its *just* strong enough to trigger the 74hc14 (which is a hex inverter).  Now, also on the input of the 74hc14 we see an antenna with a series capacitor.  When the coil starts to oscillate, this antenna picks up a signal from the TC output and feeds it right back into the system. This signal looks "strong" so it overrides the 555 (U1).  Now the coil is running from its own noise and will be perfectly in tune at all times!  Also on the input are 2 1n60 germanium diodes.  Their function is to clamp the antenna's voltage to the 5V and gnd rails so that we don't fry our 5V logic chip!  The output of the 74hc14 is fed to the inputs of an inverting and non-inverting gate driver chip.  They work in opposition creating an effective 24V peak to peak voltage across the primary of the 16T transformer.  This transformer is just a small ferrite core wound by hand and is used to supply isolated gate driver signals to each of the MOSFETs.  This transformer is known as a Gate Driver Transformer (GDT).  Q1 and Q2 make up a half-bridge.  One fet is on while the other is off.  This creates a square wave across the primary of our TC of about 85V RMS (1/2 of the voltage supply, 170VDC).  The MUR860 diodes are fast diodes there to catch any reverse current that is back flowing from the tesla resonator.  D9 and C12 provide the DC for the half-bridge. The last piece of the puzzle is U5.  Its another 555, but its running at a low pulse rate.  This 555 controls the ON/OFF period of the tesla coil by enabling/disabling the gate drivers.  The adjustment of R5 and R6 allows for a wide variety of spark outputs.  You can also tailor the component values to your own liking.  This makes for a really cool and small SSTC.  With proper heatsinking on Q1/Q2 this thing should run for a long time.  Just make sure the MOSFETs are not overheating!

More Updates:

I'm adding an additional schematic that shows the latest revisions.  The largest change (and by far the coolest!) is the sub points added into the schematic around U5.  By allowing these 2 values to be changed an amazing variety of spark appearances can be had ranging from spark gap coils to a big flaming torch.  Observe the new changes

Revised Schematic (not final, see end of page)

Here are some substitutions I have tried so far

There are just so many awesome effects possible from the interrupter.  If one does not experiment, he is missing out!  I will try to get pics of each configuration later on. 

1/19/04

I made another change to the circuit that improves some things.  It seems the coil will work fine without U1 (the 555) even installed!  Basically, when the driver chips are enabled, the inverting chip will send a short pulse to the MOSFETs causing the oscillation to start without any problem.  The fact that there is only 1 signal going to the 74hc14 means there is no room for conflict here.  This ALSO means that the feedback is stronger... shorter antenna!  Now the coil will happily run WITH the toroid and produce up to 8" sparks now.  Here is the revised schematic:

Revision 3 (not final, see end of page)

Here we add some wetted salt to the breakout point:

4/30/07

After having this project out for a few years I've received quite a response from many other amateurs who have copied the design successfully.  But, I very often get the same questions over and over.  In particularly "why did you series up all 6 inverter gates?".  At the time I built this coil I was still pretty new to electronics, so I didn't see the harm in it, and it was easy to wire, but now I realize the extra gates add unwanted delays, so this last version of the schematic has fixed that part.  I also reduced the value of the gate resistors... I'm tempted to say leave them out all together, but 5 ohms should be safe.  Its also possible to completely eliminate the 74HC14 and feed the antenna directly to the UCC37321/2 gate driver inputs, but this might not work as reliably for some people, so experiment at your own risk.

Final Schematic (use this one to build it!)