SSTC building tips
coaster_chris, Sun Feb 18 2007, 01:24PMHi all,
I have built a spark gap TC last year. After some snooping around on the net and tips form some Dutch coilers I started on a SSTC based on Conner's schematics.
I've got all the components sorted now but I just can't find the courage to start building.
The reason for opening this thread is to ask you guys for some tips how I should start.
For example, how do I determine the VOC cap and resistor values on the HEF-4046? I know Steve explains this on his site but I don't fully understand, maybe someone could help me on that?
I have wound a secondary coil with the following specs:
Wire dia: 0.2mm (32 AWG)
No. windings: 2000 x
Former dia: 16cm
I hope you could give me some tips ‘n tricks.
Thanks!
Chris
Re: SSTC building tips
ragnar, Sun Feb 18 2007, 04:13PM
hi Chris,
if you've built and tuned a spark-gap TC, you'll have a reasonable feel for "twiddling" things and which way to turn stuff to coax more breakout, etc, that's a good start. Have you got 'all' the components, including the gate driver transformer core, etc?
I'll try and give you some tips:
First, if you try and build this circuit, I strongly recommend you connect pin14 of the PLL chip (which is the coil feedback, in the case of Steve's circuit) to ground with as short a link as you can. This will A) stop you getting confused as the PLL chip's frequency won't go all over the place, and B) stop whacko frequencies blowing up your bridge, as you work out what's right.
With a big secondary like that, at a stab, your Fres should be in the 100kHz ballpark. So, what you'd probably want to do is to have a look at the HEF4046 datasheet, go to the timing Rs and Cs graphs and look at the combinations that would give you 100kHz.
What's more fun, however, is to pop a ceramic capacitor (try, umm 2nF or 3nF) as the timing cap between pins 6 and 7, and then put two 50K potentiometers as your timing resistors. I really recommend those little 20-turn cermet encapsulated trimpots... they're fiddly, but really neat for this kind of stuff.
If you have a multimeter that can read frequency or an oscilloscope, connect it up to pins 3 and 4 (joined as output) at this stage, and twiddle those potentiometers until you're in that ballpark. If you don't have a multimeter... well, uhhh, last time I was playing around with a CD4046, if the timing cap is 2nF, and the timing resistors are each about 10kOhm, then that should be pretty close.
The trick is to wind a primary of maybe 20 (? at this frequency ?) turns around your secondary, connect it to your bridge.
With your gate-driver transformers, I suggest 20 turns is a good starting figure (? for this frequency ?)... use some Cat5E or something for easy five-filar windings. If you're driving two UCCs as illustrated in Steve's schematic, then a 1:1:1:1:1 winding ratio is fine; the UCCs will switch +/-12V, and you'll get +/-12V appearing on your MOSFET gates.
Don't worry about zeners on the gates yet.
Do worry about ten ohm gate resistors -- they should really neaten up the waveforms.
Don't worry about diodes across your MOSFETs, yet.
When you are ready to start twiddling those potentiometers on your PLL chip / driver, power up the bridge at 12V or 24V or something... that way voltage spikies shouldn't be able to destroy your FETs whilst you're getting a feel for the resonant frequency.
If you've got an oscilloscope, now's the time to dangle a probe near the top of the secondary (well out of breakout-range, of course) and watch the sinewave as you twiddle the pot. Because you shorted pin14 to ground earlier, there should be no surprises as the PLL pretends to 'lock' and the frequency jumps/changes.
If you don't have an oscilloscope... well, you can touch your breakout point with a knife or a screwdriver to see if it makes a little 1mm arc or so... and twiddle your potentiometers in the appropriate direction to get longer arcs. When you think you're close to the resonant frequency (another tip if you're scopeless; your current draw will be greatest at the Fres), maybe it's time to add a few more volts to your power supply to see if you can get breakout.
Once you've got breakout, and you're sure you're running at the Fres, and not a harmonic of it (shouldn't be too hard at all if you carefully measure your winding dimensions and calculate it), this is where you connect pin14 to an antenna (instead of ground) which picks up near the top turn of your secondary. Don't let the antenna get struck by lightning. Similarly, if your antenna is spraying corona, the signal diodes are probably getting blasted, too.
Now, see the 10K potentiometer in Steve's schematic, that goes between +ve, the top of a 100K resistor, and -ve? Twiddle that, wave your hand near the coil to see if it detunes properly, twiddle again, and so on and so forth until you think you're happy with the feedback.
You'll probably need to play with the position of the primary, the position of the antenna, and there are all kinds of other variables, but this should help you get on the right track.
Best of luck, and take photos for us =-)
Matt
ragnar, Sun Feb 18 2007, 04:13PM
hi Chris,
if you've built and tuned a spark-gap TC, you'll have a reasonable feel for "twiddling" things and which way to turn stuff to coax more breakout, etc, that's a good start. Have you got 'all' the components, including the gate driver transformer core, etc?
I'll try and give you some tips:
First, if you try and build this circuit, I strongly recommend you connect pin14 of the PLL chip (which is the coil feedback, in the case of Steve's circuit) to ground with as short a link as you can. This will A) stop you getting confused as the PLL chip's frequency won't go all over the place, and B) stop whacko frequencies blowing up your bridge, as you work out what's right.
With a big secondary like that, at a stab, your Fres should be in the 100kHz ballpark. So, what you'd probably want to do is to have a look at the HEF4046 datasheet, go to the timing Rs and Cs graphs and look at the combinations that would give you 100kHz.
What's more fun, however, is to pop a ceramic capacitor (try, umm 2nF or 3nF) as the timing cap between pins 6 and 7, and then put two 50K potentiometers as your timing resistors. I really recommend those little 20-turn cermet encapsulated trimpots... they're fiddly, but really neat for this kind of stuff.
If you have a multimeter that can read frequency or an oscilloscope, connect it up to pins 3 and 4 (joined as output) at this stage, and twiddle those potentiometers until you're in that ballpark. If you don't have a multimeter... well, uhhh, last time I was playing around with a CD4046, if the timing cap is 2nF, and the timing resistors are each about 10kOhm, then that should be pretty close.
The trick is to wind a primary of maybe 20 (? at this frequency ?) turns around your secondary, connect it to your bridge.
With your gate-driver transformers, I suggest 20 turns is a good starting figure (? for this frequency ?)... use some Cat5E or something for easy five-filar windings. If you're driving two UCCs as illustrated in Steve's schematic, then a 1:1:1:1:1 winding ratio is fine; the UCCs will switch +/-12V, and you'll get +/-12V appearing on your MOSFET gates.
Don't worry about zeners on the gates yet.
Do worry about ten ohm gate resistors -- they should really neaten up the waveforms.
Don't worry about diodes across your MOSFETs, yet.
When you are ready to start twiddling those potentiometers on your PLL chip / driver, power up the bridge at 12V or 24V or something... that way voltage spikies shouldn't be able to destroy your FETs whilst you're getting a feel for the resonant frequency.
If you've got an oscilloscope, now's the time to dangle a probe near the top of the secondary (well out of breakout-range, of course) and watch the sinewave as you twiddle the pot. Because you shorted pin14 to ground earlier, there should be no surprises as the PLL pretends to 'lock' and the frequency jumps/changes.
If you don't have an oscilloscope... well, you can touch your breakout point with a knife or a screwdriver to see if it makes a little 1mm arc or so... and twiddle your potentiometers in the appropriate direction to get longer arcs. When you think you're close to the resonant frequency (another tip if you're scopeless; your current draw will be greatest at the Fres), maybe it's time to add a few more volts to your power supply to see if you can get breakout.
Once you've got breakout, and you're sure you're running at the Fres, and not a harmonic of it (shouldn't be too hard at all if you carefully measure your winding dimensions and calculate it), this is where you connect pin14 to an antenna (instead of ground) which picks up near the top turn of your secondary. Don't let the antenna get struck by lightning. Similarly, if your antenna is spraying corona, the signal diodes are probably getting blasted, too.
Now, see the 10K potentiometer in Steve's schematic, that goes between +ve, the top of a 100K resistor, and -ve? Twiddle that, wave your hand near the coil to see if it detunes properly, twiddle again, and so on and so forth until you think you're happy with the feedback.
You'll probably need to play with the position of the primary, the position of the antenna, and there are all kinds of other variables, but this should help you get on the right track.
Best of luck, and take photos for us =-)
Matt
Re: SSTC building tips
Steve Conner, Sun Feb 18 2007, 04:23PM
Nice howto BP, thanks! I never managed to explain it that well (if at all!) myself.
Steve Conner, Sun Feb 18 2007, 04:23PM
Nice howto BP, thanks! I never managed to explain it that well (if at all!) myself.
Re: SSTC building tips
thedatastream, Sun Feb 18 2007, 09:31PM
I have modified both of my coils that use Steve's schematics to run the PLL from a regulated supply by using a 78L08 or a 78L05 voltage regulator. This is because the operting frequency is dependant on supply voltage, so variations can throw off the initial frequency set point. You can also use the internal Zener in the 4046 if you like.
James
thedatastream, Sun Feb 18 2007, 09:31PM
I have modified both of my coils that use Steve's schematics to run the PLL from a regulated supply by using a 78L08 or a 78L05 voltage regulator. This is because the operting frequency is dependant on supply voltage, so variations can throw off the initial frequency set point. You can also use the internal Zener in the 4046 if you like.
James
Re: SSTC building tips
ragnar, Sun Feb 18 2007, 10:50PM
Yeah, that internal zener, if you let it clamp at 20V, will allow you to run the CD4046B at up to 4.4MHz ^_^
ragnar, Sun Feb 18 2007, 10:50PM
Yeah, that internal zener, if you let it clamp at 20V, will allow you to run the CD4046B at up to 4.4MHz ^_^
Re: SSTC building tips
Steve Conner, Mon Feb 19 2007, 01:22AM
What? I thought it was about 7V, not 20.
Steve Conner, Mon Feb 19 2007, 01:22AM
What? I thought it was about 7V, not 20.
Re: SSTC building tips
Terry Fritz, Mon Feb 19 2007, 04:04AM
Hi,
Gary Johnson has an online "book" about solid state coils here:
Some might be a little out of date, but there is a ton of info there.
Terry
Terry Fritz, Mon Feb 19 2007, 04:04AM
Hi,
Gary Johnson has an online "book" about solid state coils here:
Some might be a little out of date, but there is a ton of info there.
Terry
Re: SSTC building tips
Marko, Mon Feb 19 2007, 10:14AM
It seems that he's talking about 'overclocking' the 4046 to his very limits, surely with external regulation. Internal zener is 7V.
Marko, Mon Feb 19 2007, 10:14AM
Steve Cønner wrote ...
What? I thought it was about 7V, not 20.
What? I thought it was about 7V, not 20.
It seems that he's talking about 'overclocking' the 4046 to his very limits, surely with external regulation. Internal zener is 7V.
Re: SSTC building tips
coaster_chris, Mon Feb 19 2007, 01:36PM
Woow guys... Thanks a lot!
That's some very helpful information!
Matt, you must have been sitting and typing for ever...
And Terry, that paper of Gary Johnson looks very helpful too at first glance.
For sure that I'll post some pics if I got something to show.
Now I need to go and find some time to start of, in between my school, job, girlfriend, beer and other fun stuff like that.
Chris
coaster_chris, Mon Feb 19 2007, 01:36PM
Woow guys... Thanks a lot!
That's some very helpful information!
Matt, you must have been sitting and typing for ever...
And Terry, that paper of Gary Johnson looks very helpful too at first glance.
For sure that I'll post some pics if I got something to show.
Now I need to go and find some time to start of, in between my school, job, girlfriend, beer and other fun stuff like that.
Chris
Re: SSTC building tips
ragnar, Mon Feb 19 2007, 10:45PM
Look, I'll take care of your job and your girlfriend whilst you build your TC. If you need any help you can just give me a call at (your) work. Deal? =P
ragnar, Mon Feb 19 2007, 10:45PM
Look, I'll take care of your job and your girlfriend whilst you build your TC. If you need any help you can just give me a call at (your) work. Deal? =P
Print this page