Re: It's time for an SSTC
Zum Beispiel, Tue Sept 25 2007, 04:45PM

Whether the resonator will work or not depends on it's resonant frequency. If it's in the 100's of kHz it should work. IRFP460's seem to work well enough at these frequencies. My bigger SSTC is runs at ca. 260kHz.

For 230V mains I used about 10 turn primary and halfwave rectified input.

The antenna feedback seemed to work best when the antenna was some 40cm from the secondary. Though it's not very critical.

Hope this helps.

Re: It's time for an SSTC
CT2, Wed Sept 26 2007, 12:38AM

I built that circuit and it worked great, but about the diodes (on the antenna), I found that it didn't work when I substituted the 1N60 diodes for 1N4148 diodes it didn't want to resonate (that is without the 555 sending the start up signal). However when I added the 555 everything worked fine even with 1N4148's.

Re: It's time for an SSTC
vasil, Wed Sept 26 2007, 08:18AM

My mini SSTC used 1 N 4148 diodes without 555 and it worked ok.

Re: It's time for an SSTC
Dr. Dark Current, Wed Sept 26 2007, 02:37PM

Ok thanks for the replies, so will it work with the TC442x gate drivers?

Re: It's time for an SSTC
Zum Beispiel, Wed Sept 26 2007, 03:57PM

Should work fine. I believe the TC chips are lower power drivers than the UCC, though. They might run hot. If your operating frequency isn't too high and you use FETs that don't have a large gate C, they should work.

Re: It's time for an SSTC
Ken M., Wed Sept 26 2007, 09:35PM

I was only able to keep the antenna 8in from the breakout point, only problem was when I added a larger toroid the antenna was arcing towards the torroid, thats when I decided to have a current transformer instead of an antenna.

Re: It's time for an SSTC
Dr. Dark Current, Fri Sept 28 2007, 12:49PM

Ok since the TC442x's don't have the enable pin, how should I go around this?

Re: It's time for an SSTC
uzzors2k, Fri Sept 28 2007, 01:26PM

Use a transistor to pull the signal down, at the signal input.

Re: It's time for an SSTC
Dr. Dark Current, Fri Sept 28 2007, 03:21PM

I had the same idea but then one output will be low and the other high.. this might but also might not damage the fets and/or gate drivers, I don't know..

Re: It's time for an SSTC
Marko, Fri Sept 28 2007, 03:25PM

You want both of your drivers to be off when idsabled so you would have to use a lot of logic there. Both drivers would need to be non-inverting in order for you to be able to disable them with and gates. Then you would need to invert the input of one and gate again.

This is serious complication and lots of unwanted delays. If I had no UCC's I would probably just run CW until I get some. Drivers are pin compatibile apart from enable.

Re: It's time for an SSTC
Dr. Dark Current, Fri Sept 28 2007, 04:24PM

can you get samples of the ucc's?

Re: It's time for an SSTC
Munkey, Fri Sept 28 2007, 06:36PM

Yep, from they have fast delivery so you can blow em' up twice as fast
I did, the only thing which is annoying is that they only have the little SO-8 SMD packages. Have fun soldiering them

Re: It's time for an SSTC
Dr. Dark Current, Fri Sept 28 2007, 07:16PM

oh cr*p, there's no way I can solder this with a soldering gun
I'll probably have to run halfwave CW, until I find a way how to interrupt the TC's

Re: It's time for an SSTC
Zum Beispiel, Fri Sept 28 2007, 07:51PM

If you really need them, why not just buy some? They only cost like 2€ a pop.

Re: It's time for an SSTC
Dr. Dark Current, Fri Sept 28 2007, 07:54PM

well, the answer is simple.. because I can't buy them where I live and I don't want to pay 5x the cost of the ic's for shipping

Re: It's time for an SSTC
Munkey, Fri Sept 28 2007, 07:55PM

You don't have a soldiering iron? Even with a big tip I still did it. I would still get them, they're FREE!!

Re: It's time for an SSTC
ragnar, Fri Sept 28 2007, 11:38PM


Last time I checked, TI will deliver DIP8, SOIC8 and MSOP8 UCC MOSFET drivers, all of which are possible to solder by hand with a fine tip.

I tried a couple of techniques here:
http://4hv.org/e107_plugins/forum/forum_viewtopic.php?16319

P.S. Please don't over-exploit (exploit) the samples offered by TI. They are NOT 'free'. The samples you might receive are subsidized by the profit TI makes on the same chips that WE pay good money for.

If a group of regular abusers re-orders the chips without a followup purchase, it becomes uneconomical for TI to distribute samples of that chip, and they may stop offering them at any time.

At least try to make a quarter-effort not to blast them all by mistakes like "bumping the voltage knob".

Re: It's time for an SSTC
Munkey, Sat Sept 29 2007, 09:44AM

Another way that I found easy is to get some copper stripboard and stick it to that, heres a pic of mine:
Yes, I know it is wrong to exploit their free samples, the chances are that next time I am going to buy them along with a some other stuff.

Re: It's time for an SSTC
Experimentonomen, Sat Sept 29 2007, 11:23AM

Thats why i sample the dip version

Re: It's time for an SSTC
Dr. Dark Current, Sun Sept 30 2007, 08:28AM

the samples registration form on ti website is somewhat odd, it doesn't want to take my telephone number and I'm also requied to fill a field titled "Other:" in the mailing address, I have no idea what to write there...

Re: It's time for an SSTC
Munkey, Sun Sept 30 2007, 08:55AM

^ Yeah you are meant to be a registered business to get free samples, but I faked it all besides my address

Re: It's time for an SSTC
Dr. Dark Current, Sun Sept 30 2007, 10:02AM

yeah, but what's that "other" thing? I'm requied to fill it even if I've already written my complete address! :o

Re: It's time for an SSTC
Munkey, Sun Sept 30 2007, 11:17AM

Ummm, other? What category does it come under, address? Occupations?

Re: It's time for an SSTC
Dr. Dark Current, Sun Sept 30 2007, 12:14PM

"Please select one of the following:" ?! There's nothing to select




[mods] I apologise this thread is going off topic, I could create a new one titled "How to register at ti samples" if it's such an impossible task

Re: It's time for an SSTC
Steve Conner, Sun Sept 30 2007, 12:23PM

Sounds like your postal code hasn't been recognised by the server-side thingy that interprets the form.

Try entering it in some other format- I sometimes see European postcodes written with a national prefix

or just fill your postcode into the "Other" box again, and hope it gets bounced to a human being at TI who will figure out what happened.

Re: It's time for an SSTC
Marko, Sun Sept 30 2007, 12:36PM

I'm not sure if I remember right, but I think I just put a - into 'other' there and everything worked.

Regarding phone number, I also had trouble with it... just write it up full (country and city number) together without spaces adn let it add the +, then it should be ok.

Re: It's time for an SSTC
Dr. Dark Current, Sun Sept 30 2007, 01:16PM

uh, I've finally registered and succesfully placed a samples order (they had also the dip version btw), now I'll just have to hope that they interpret the address right 8-o

Re: It's time for an SSTC
ragnar, Sun Sept 30 2007, 04:40PM


Great, now boast about your deception on a public forum...

Re: It's time for an SSTC
Munkey, Sun Sept 30 2007, 05:42PM

Boasting? Deception? Try advice maybe. Ok it was the wrong advice. Boasting? WTF?

Re: It's time for an SSTC
Dr. Dark Current, Mon Oct 01 2007, 02:12PM

Ok, I think I'll go with secondary base feedback instead of antenna, because I feel it's a bit "safer" than the antenna especially when drawing arcs. I understand this is done with a current transformer, how should be the turns ratio and number of turns (pri/sec) on it? Can I connect it's output directly where the antenna was?

Re: It's time for an SSTC
CT2, Mon Oct 01 2007, 03:28PM

For the CT you can use the same core as you GDT, an it needs about 100:1 ratio. You can hook it directly where the antenna is (the other side goes to ground), but you need to add a resistor and/or blocking cap with it or else it may short your start up signal (there is a thread on this somewhere).

Re: It's time for an SSTC
Dr. Dark Current, Mon Oct 01 2007, 05:20PM

I think i found the thread
but what I'm still not sure about, the 100:1 ratio means I connect the 1 turn winding to the secondary base and the 100 turn winding to the input of the 74hc14 in series with a .01UF cap, or the other way round?

Re: It's time for an SSTC
Zum Beispiel, Mon Oct 01 2007, 06:10PM

Yes, exactly. You put the secondary ground wire through the CT once and connect the 100 turn winding to the 74hc14. Remember to use clamping diodes or you'll fry the sensitive 5V logic chip.

If you use secondary base feedback, you must make absolutely sure that you have the ground connected before you run the coil. With antenna feedback it won't explode immediately, but with CT no ground = no feedback = BOOM!.

I once managed to blow my bridge like this. The thin ground wire got cut inside the case, the bridge was left running on start-up osc (160kHz, secondary resonant at 260kHz). Crank the variac to around 50% and Poof!

Re: It's time for an SSTC
Dr. Dark Current, Tue Oct 02 2007, 03:30PM

I'm designing a mains-synchronised interrupter (I can never give up that characteristic mains buzz ) using one of the unused 74hc14 inverter gates. However I have no idea how big on/off ratio (min/max) usually the interrupter onthese coils runs, can someone advice? Also isn't 100hz break rate too low for a SSTC?

The samples should come in a few days (according to the ship date I received on order), so this is when the fun starts

Re: It's time for an SSTC
Dr. Dark Current, Wed Oct 03 2007, 05:07PM

Please excuse my stupid questions but what's the right way to wire up the CT? In the first diagram the cap could see high voltages across it and maybe detune the coil a bit (though I'm not sure about these) and in the other diagram I'm afraid the cap wouldn't do anything at all.




By "no ground" you also mean a loose connection (gap in the connection)? I have the secondary loosely connected to central heating radiator, there could be some minimal gap.

Edit- of course I meant CT not GDT

Re: It's time for an SSTC
uzzors2k, Wed Oct 03 2007, 05:58PM

The cap acts as a DC block, so the duty cycle stays even. In the second configuration it will never see more than 5V. I would burden the CT and add a resistor in series with it. Colin Heath had lots of trouble with buffer chips and strong CT feedback signals. 1k is about right I think, and 10 ohms for the burden. All of this is in the thread you found though.

As for grounding I think you'll be fine, as long as the CT is hooked up it should oscillate. I got a variation of the Skori coil working today with just a lousy ground connection to an ATX PSU. It's my first SSTC. Playing with flybacks can only go so far, the next step up would be a CCPS, which I don't have the ferrite for.

Re: It's time for an SSTC
Dr. Dark Current, Fri Oct 05 2007, 09:08AM

Samples came today, TI ships samples really fast! Let's start working on the coil



Update: well, I think it lives, but I have no way to confirm that I was getting some 1mm sparks to ground with a 25W bulb in series with the mains input, when my mom shouted from across the house that there's some interference on the TV. This was from like 15m away! I'm a little worried trying it with higher power, it might also fry some of the expensive electronic stuff stuff in my room

Update2- I'm a little confused now, there are 4 combinations how to connect the feedback path (you can reverse the CT and GDT) but the strange thing is that I'm getting some output from the coil on 3 of these 4 combinations, just one gives a little better output. Should I use that? Also the discharge makes some sputtering irregular noise, like it does not start every time.

Update3, lots of questions I now got it to make a softly buzzing ~1/2" discharge, but the problem is that my fets (for now just weak irf740's) are getting quite warm, but the low side one gets much hotter than the other one. Could this be because the are different batch or could there be a problem with my gate drive?

Last update for now: The hotter fet stopped working. Its gate is near short but d-s is open. I think it was defective

Re: It's time for an SSTC
Dr. Dark Current, Sat Oct 06 2007, 09:18AM

here's the design of my mains syncronised interrupter, if anyone is interested. It makes use of the remaining inverter gates in th 74hc14 so no additional ic requied. The on time can be smoothly varied from 0 to some 90%. And yes it's a bit diode rich but it works great

The 5v line should be loaded with a resistor (1k is ok) so it's voltage doesn't rise as a result of the additional voltages being clamped to it.

Re: It's time for an SSTC
Dr. Dark Current, Mon Oct 08 2007, 02:29PM

I'm getting some weird stability problems with the SSTC.
This is a very low power test. The problem is, the coil fails to start oscillating on some "bursts" especially when I put my hand close to the CT/control electronics. This is NOT a problem of the interrupter, I've looked at this on a scope. The problem seems to go away a bit when I un-ground the control electronics. Whem looking at the gate waveform it seems that in these "misfires" the frequency goes to 1/2 or 1/3 the resonant freq. but I'm not sure.

I don't know, maybe this will go away with higher powers?

Can anyone advise? I'm attaching a video of the problem.

]1191853682_152_FT31895_pict0186dsf.avi[/file]

Re: It's time for an SSTC
Zum Beispiel, Mon Oct 08 2007, 03:54PM

Maybe you are simply not getting enough feedback on lower powers. It might be very destructive for your bridge if the problem doesn't go away at higher powers.

Few things to try:

-Make sure your ground is good.
-Try adding some more turns on the CT
-Make sure your clamping diodes aren't damaged, or any other parts for that matter.

Re: It's time for an SSTC
Dr. Dark Current, Mon Oct 08 2007, 04:54PM

Woot, it lives! The trick was to add a resistor across the CT (I've used 4k7 resistor). Also it helped removing turns from the CT, not adding. I'm a little worried the CT could raise the supply voltage for the 74hc14, I think I'll zener-clamp the CT just to be sure. This is a discharge I get with 160V to the bridge and 8 primary turns:


I'll also be making a new resonator with more turns, this should help too.

Re: It's time for an SSTC
Ken M., Mon Oct 08 2007, 07:04PM

Whats the arc length, it looks like 1-3 inchs?

Re: It's time for an SSTC
uzzors2k, Mon Oct 08 2007, 07:18PM

Congrats! You beat med getting streamers!

Re: It's time for an SSTC
Dr. Dark Current, Mon Oct 08 2007, 07:24PM

That arc is not big, about 1", but it tells that the coil is working well. No noticable heating of the fets with the fres around 600khz. I'll be removing turns from the primary until the fets start getting warm, then as time allows I'll wind the new (lower fres) resonator, looks like a lot of fun is coming

Re: It's time for an SSTC
Dr. Dark Current, Tue Oct 09 2007, 08:01PM

well, with the new secondary which is 2" taller (total 10" wingind length) than the old one and has ~2x the turns, the spark length didnt increase much. Previously it was something like 2" and now it is 3" at most.




I'm using 4 primary turns with 160V to the bridge (with irfp250's), I think it's time to upgrade to 320V and irfp460's.

IIRC Steve Ward got 7" sparks with 4 turns and 160V, I don't know what the problem could be (if there is any).

Right now I have the primary just wound around the base of the secondary, should I make different primary for better results?

As for topload, should I use any? or is the nail I'm using now all right?

Re: It's time for an SSTC
Zum Beispiel, Wed Oct 10 2007, 09:15AM

A topload should help to get those bigger sparks. How much power are you drawing?
If it's low, playing with the coupling helps.

Re: It's time for an SSTC
Ken M., Wed Oct 10 2007, 12:42PM

Thats odd JM, Cuz I too have built this and I'm running 160v of fullwave rectified, with out a topload I get giant 9-13" arcs, no idea what I get with a top load since I was more afraid of my antenna arcing to the topload. =( Heres the video of the arcs

Re: It's time for an SSTC
Dr. Dark Current, Wed Oct 10 2007, 02:38PM

It says the file no longer exists

How big is your secondary and what type of primary are you using? and how moch power?

I'm drawing maybe 100W of power, with the interrupter at ~40-50% on time. The transistors are stone-cold on their little heatsinks, something must be limiting the power here but I have no idea what it is

Re: It's time for an SSTC
uzzors2k, Wed Oct 10 2007, 02:52PM


Coupling is a major part of determining power draw. I found Richie Burnett's site very informative.

Re: It's time for an SSTC
Dr. Dark Current, Wed Oct 10 2007, 03:03PM

One of the fets just went dead for no apparent reason, it wasn't even warm. I have no idea what is going on here


BTW here is a pic of the present primary arrangement

Re: It's time for an SSTC
Zum Beispiel, Wed Oct 10 2007, 05:12PM

Rising your primary up and spreading it over a larger area will allow it to draw more power.

The IRFP250 is a 200V device, right? If you are running at 160V (and you mean 160VDC, not AC, right?) you may get spikes large enough to kill them.

Re: It's time for an SSTC
Dr. Dark Current, Wed Oct 10 2007, 05:24PM

I think the spikes couldn't have killed it, I have 2x 1.5uF PP decoupling caps soldered right to the fets. Yes it is 160V DC, but I think its actually a bit less since it's rectified from a centertap of a 150W mains xfmr.

I'll replace the dead fet, remake the primary and see what it does..

Re: It's time for an SSTC
Ken M., Wed Oct 10 2007, 09:52PM

Secondary is 4x18, the primarys is 10 turns of 10AWG wrapped around a 1\2" thick rubber pipe couple which is wrapped around the secondary. I uploaded the vid to youtube, but it'll be a little while before its up.


Edit: The video finally got posted on youtube while I was at work, so heres the link. If your wondering why it goes through a variety of frequencys its cuz my interuptor has 6 settings all with a pot, and I went throught the whole spectrum.

Re: It's time for an SSTC
Dr. Dark Current, Thu Oct 11 2007, 06:25PM

I've remade the primary and now with 6 turns I'm getting more power draw than with 4 previously however I won't do a full power test until I find out why the coil starts mis-firing at increasing power level...


A strange thing is, I've tried zener-clamping the CT output and this reduced the output from the coil.. could this be due to the "slowness" of the zener diode?



EDIT- this is getting REALLY wierd! I connected a scope probe to the output of the interrupter to see if it wasn't the source of the mis-firing, and guess what! The mis-firing stopped this reminds me of the double slit experiment

Re: It's time for an SSTC
Tom540, Thu Oct 11 2007, 07:27PM

Maybe your CT needs more turns. I make mine by winding a twisted pair for 33 turns then series them for 66. Or when im not being lazy I wind 100 turns on them. Seems like more turns and then good clamping works better

Re: It's time for an SSTC
Dr. Dark Current, Thu Oct 11 2007, 07:35PM

Well I have the answer- it IS the interrupter. I've cranked it up a bit and it finally showed up also on the scope. I think there are some problems with the sync pulses from the xfmr possibly getting some noise from the output of the coil, it's time to rebuild it (the interrupter)I think...


update, I've fixed the interrupter and the problem still persisted.. so just out of curiosity I added a turn on the primary side of the CT, so the ratio is now 1:12.5 really low I know but it got rid of the problem

Re: It's time for an SSTC
Dr. Dark Current, Fri Oct 12 2007, 01:51PM

Yesterday the coil produced some 3-4" sparks with 6 turns on the primary and then it died again. Only one FET was dead (the low side one again) and the failure was silent. The fet was not shorted completely, in fact the coil could still make a little spark even with the dead FET (exactly the same failure as the last time).

Before I replace the fet and blow it again, does anybody have an idea wha could have killed it? It was perfectly cold when it failed, so the only two possibilites are too high peak current and voltage spikes, but I can't see how any of these could have been the reason for it's death...

Re: It's time for an SSTC
Tom540, Fri Oct 12 2007, 03:21PM

I'm thinking you should add more turns to that primary. I don't know what your values are but 6 turns sounds like not enough. It probably isn't as crucial since your not running it in DR mode, but it might help.

Try raising your primary way up and spacing out the turns to get more coverage. I have an SSTC that has a 6 or so turn primary and I have it about halfway up the secondary and the windings are very far apart.

Re: It's time for an SSTC
Dr. Dark Current, Fri Oct 12 2007, 04:10PM

Maybe I wasn't too exact, it's 6 turns for 160V supply voltage and I think that's enough. I remade the primary in a way that I've insulated the bottom 1/3 of the secondary and wound a primary on it, it's about on the bottom 1/10th of the coil.





What I don't get is why the fets die, this is beyond my understanding. However I need to fix this of course if I want to continue working on this coil.


Maybe I'll dig up my isolation transformer and see how the drain waveform looks like

Re: It's time for an SSTC
Dr. Drone, Fri Oct 12 2007, 04:33PM

Re: It's time for an SSTC
Dr. Dark Current, Fri Oct 12 2007, 06:39PM

Here's the gate waveform with ~400khz testsignal, the coil's resonant frequency is probably slightly below that:




I have no idea how it looks like with feedback signal since I haven't checked that, but I believe my scope would act up by picking up rf from the coil and display a distorted signal even if it would be clean.

The gates are clamped by 18V zeners so I'm sure I'm not overstressing them.

for sure




The streak on the secondary is just left over glue from a sellotape.

I guess I'm pushing the breadboard a bit beyond its component capacity limits, but it seems to work well.

The transformer on the back is just 230/115V for powering the irfp250 halfbridge. It's rectified and filtered by 220u cap, and two 1.5u PP decoupling caps are soldered right to the FETs.

The CT is a subject to change, since I'm getting stable feedback just with very low pri/sec ratios (now 2:25). A 4k7 resistor is connected across the CT output, then it goes through a 0.22u cap to the 74hc14, where it is clamped by 1n4148 diodes.


One thing I don't get is why always the low side FET fails and it is never shorted completely...

Re: It's time for an SSTC
Dr. Drone, Sat Oct 13 2007, 01:50AM

Re: It's time for an SSTC
Dr. Dark Current, Sat Oct 13 2007, 11:01AM

There's no FET on the breadboard, the GDT and half bridge are outside it.

Here's one observation: when the coil fails to oscillate (ie. no feedback signal to gate drivers), what happens is as the interrupter turns on, one gate driver chip is high while the other is low. This results in that the decoupling cap that is in series with the GDT primary starts to oscillate with it, which results in a nasty burst of low frequency sine wave on the output of the GDT. I'm wondering, couldn't this be what is killing my FET?


Update: I've completely rebuilt the circuit on the breadboard and added a 555 interrupter (to eliminate the old one was the source of the problems). I've also went for 320V and IRF740's since they are cheap to blow. So far it works (low power tests), but I've spotted something unusual. This is a scope shot taken by holding the probe near the secondary: (sorry for the poor quality, the wave on the scope seems to wander around)






notice the weird "dent" in the waveform, what could it be?

Re: It's time for an SSTC
Ken M., Sat Oct 13 2007, 06:40PM

Probably just a projection error, its probly nothing.

Re: It's time for an SSTC
Dr. Dark Current, Sun Oct 14 2007, 07:50AM

So far the coil works, sparks 4" or so (will be removing turns from primary, currently 14).

I tried running halfwave CW (connecting the gate drivers' enable input to +5V) but it did not work! It just made single spark as I applied power to the bridge and then nothing. What do I have to do to get it working halfwave CW?

Re: It's time for an SSTC
Zum Beispiel, Sun Oct 14 2007, 11:36AM

Do you have a start-up osc? Since with halfwave input there is no feedback for half the time it may fail to oscillate... A 555 running at ½ the resonant frequency should work.

Re: It's time for an SSTC
Dr. Dark Current, Sun Oct 14 2007, 12:08PM

ah, the start up osc, isn't it better if it runs above fres? This way I think there is less risk blowing the FETs with poor/no feedback

---
It works, the sparks are shorter but much more juicier, and also more VTTC-like, and I like that. Only if it didn't draw so much power


heh, do you think this thing can stand drawing arcs to ground? and do I draw them do the secondary base or to ground (before/after CT)?

Re: It's time for an SSTC
Dr. Dark Current, Mon Oct 15 2007, 03:14PM

Problems, weird ones!!!

The coil runs fine with a 500W bulb in series, but when I bypass the bulb, the spark starts emitting really bad sputtering sounds! Here's a video of the problem, first with the bulb, then I bypass it.

Can anyone help please?

]1192461278_152_FT31895_pict0185.avi[/file]

Re: It's time for an SSTC
Dr. Dark Current, Tue Oct 16 2007, 06:04PM

It works!

I found that all the stability problems lie in the CT and how is it connected up to the control electronics. Here's a connection that I currently use, any tips on improving it?
edit- I forgot a .1u cap in series with the CT




And of course some sparks The little IRF740's get just a bit warm so far.











How tightly can I couple the primary to the secondary without a risk of arc-over?

Re: It's time for an SSTC
Ken M., Tue Oct 16 2007, 07:51PM

IF by "tightly couple" you mean, how close can you make the primary and secondary? As I said before I have them ontop of each other.




How ever if you mean "tightly couple" as in magnetic coupling to increase mutual inductance I don't think too much considering the skori coil I have, has a gap of about 1/2" and another1-11/2" down the pvc pipe and I still get some primary to secondary arcing through the insulation.

Re: It's time for an SSTC
Dr. Dark Current, Sat Oct 20 2007, 11:44AM

OK, as I now got it working reliably, I wanted to draw some arcs. It didn't blow a FET as before, but as the streamers start hitting the ground electrode, they become very thin sparks. Why is this?

Re: It's time for an SSTC
Ken M., Sat Oct 20 2007, 12:15PM

Thin? Mine get thik as hell almost MOT style just really long, the only time when i got small sparks was when I had it set to think spark pulses flyback style.

Re: It's time for an SSTC
Dr. Dark Current, Sat Oct 20 2007, 01:49PM

Are you using CT or antenna? I'm now running halfwave CW with start-up osc and CT with a circuit I posted before.

Here's a video with ground sparks, they get almost invisible!
]1192888198_152_FT31895_pict0185sparks.avi[/file]

Re: It's time for an SSTC
Tom540, Sat Oct 20 2007, 06:28PM

It Looks like you're detuning the coil whe you bring the ground closer.

Re: It's time for an SSTC
Steve Ward, Sat Oct 20 2007, 07:12PM

This is in fact the main problem with using feedback from the secondary circuit. When you ground out the circuit, it tends to either 1) simply disturb the oscillations, and cause a minor "hiccup" in operation, or 2) completely destroy the self-sustained oscillation, and the SSTC stops once the spark connects with ground (instead of pumping more power in, so you get the nice flaming ground strikes).

Antenna feedback seemed slightly better for ground arcs, but it seemed to depend on a few factors, which i never quite figured out (since i left that route awhile back). Some of my SSTCs would work great for ground arcs, and others would make the "thin" weak sparks because oscillation would cease. It never really seemed too detrimental to the MOSFETS, but it was annoying. Any time i tried using CT feedback (without aid of PLL) this was the exact problem i had, so that is why i never posted any CT feedback SSTCs on my site.

Re: It's time for an SSTC
Ken M., Sat Oct 20 2007, 08:19PM

Hmm I guess that would make the difference, I was refering to antenna feedback since I have yet to rebuild it with CT feedback. Although From the skori coil the arc stays the same and its ct feedbak...

Re: It's time for an SSTC
Zum Beispiel, Sat Oct 20 2007, 08:44PM

When I play with the arcs I use an ungrounded chicken stick to draw arcs. I have a small toroid on the end of the stick and a sharp metal point pointing away from it.

Like this:

=========0----

== insulator
0 toroid
---- sharp metal thingy.

I get some very hot arcs this way

Re: It's time for an SSTC
Dr. Dark Current, Sun Oct 21 2007, 08:20PM

Well, I changed the resistor in series with CT back to 10k (I was using 4k7). I got a nice arc to ground and one dead FET as a bonus.

I'm going to rebuild it, I'll trash the CT and use an antenna probably.

Can anyone advise what SSTC self-tuning driver is the most reliable and allows ground arcs without flying silicon? PLL maybe?

Re: It's time for an SSTC
Dr. Dark Current, Tue Oct 23 2007, 01:56PM

I noticed one weird thing.. With no topload, sparks are small and power draw is low. With small topload, sparks got almost 2x big and the input power too.
Now I made a new bigger toroid, put it on- the sparks are still the same (maybe even little shorter) but the power draw is really large, probably 1.5x as much as with the small topload!
Where did the power go? I didn't notice any huge dissipation in the fets. Is there something I still don't understand, such as power factor? (I measured the input power "visually" by watching a large bulb in series).

On the same topic, I also noticed that (at very low power) drawing an arc from the topload results in very noticable input power decrease, while I think the heat generated by the ~1" hot arc is at least few times larger than the heat from the little ~1/2" purple discharge alone...

Re: It's time for an SSTC
uzzors2k, Wed Oct 24 2007, 07:33PM

This might be a bit far-fetched, but since no one else has answered I'll dare to speculate.

For some reason lower frequency coils give longer sparks than higher frequency coils for the same power input. I don't why, but it seems to be the norm. This would explain the better spark output since the topload would lower the TC frequency.

The reason a larger topload doesn't improve the spark size further is the increased capacitance. Energy stored in a capacitor is a function of voltage and capacitance. Higher capacitance means more energy per volt. With the same energy and more capacitance, the voltage will have to decrease.

I can't explain the increased power draw though. Maybe more power was being radiated into the environment around the TC with the larger topload, and there was therefor less to be "used" for sparks. This sort of explains the arcing too, as the spark is only a portion of the energy put in. Think of florescent tubes lighting from a distance for example.

Re: It's time for an SSTC
Dr. Dark Current, Wed Oct 24 2007, 08:06PM

I don't know, but this just came to my head- couldn't the larger toroid act as an one big shorted turn eating the power away?

Re: It's time for an SSTC
Zum Beispiel, Wed Oct 24 2007, 08:23PM

Yes, but so would the smaller one. Of cource a larger toroid would have higher resistance but I doubt it would make that much of a difference.

Re: It's time for an SSTC
Dr. Dark Current, Fri Oct 26 2007, 04:49PM

I've solved many electrical mysteries in my circuits before, but this one is really hard.

I'm now running the TC in a way that it's fed with halfwave mains input and an interruptor that turns it on as the mains reaches peak value (the idea was to get the same spark length with 1/2 input power).

The first mystery is that one of my fets was getting much hotter than the other. I thought I had defective gate driver, so I replaced them both. It was still the same.

To find out where the problem was, I swapped polarity of TC primary leads and gdt primary leads (to have the feedback polarity right).
The discharge made a really funny noise, but there was HF output on the gate drivers at the resonant freq.

I put one scope probe around 1m from the topload to see the RF envelope and unlike before, I saw a HUGE negative spike on each interruptor pulse. I think this even damaged my scope . (Update: my scope isn't damaged, one hard hit on the side repaired it. It's some 50 years old so I think it absolutely has the right to have a bad solder joint, especially considering what abuse it has survived so far!)

I would be grateful for any ideas. Why one of my FETs runs hotter? Why changing the polarity of primary and GDT did change the output so much? Why the HUGE spike? (it was showing up on my scope really hard even with all probes unplugged!)

It's also possible that the problem lies in my old breadboard, but I doubt that.

Thanks for any help!

Re: It's time for an SSTC
Firnagzen, Sat Oct 27 2007, 04:30AM

Just hazarding a guess- I'm by no means an expert- maybe one FET heats up more because it has a slightly higher resistance. They can't make every FET exactly the same, so maybe some defect in that one makes it heat up more?

Ps. Why shouldn't you mix up deuterium and tritium? The only difference is that tritium is radioactive, right?

Re: It's time for an SSTC
Dr. Dark Current, Sat Oct 27 2007, 08:30AM

Well, it is possible that one FET has higher resistance, but I doubt it. I'll try to swap them and see what happens.

P.S maybe you didn't get the point, it's a quote from an article on heavy water, just how many people are there who have deuterium and tritium in their closet?

Re: It's time for an SSTC
Firnagzen, Sat Oct 27 2007, 08:38AM

I wouldn't know how common either is. I've never tried obtaining 'em, have I?

By the way, if you suspect a loose solder joint in your o-scope, perhaps you should fix it? Just a suggestion.