Feedback current transformer / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

Feedback current transformer

Move Thread

LAN_403

Frosty90

Mon Dec 01 2008, 05:59AM

Registered Member #1617 Joined: Fri Aug 01 2008, 07:31AM
Location: Adelaide, South Australia
Posts: 139

Hi all,

My new sstc is almost done (except for the new actual secondary!) but i still have a few questions.

I will be using current transformer feedback (but i could easily change to antenna, but i dont want to). Last time, i just hooked the secondary of the CT directly to the input of my driver circuit (a 74hc14) via a 1k resistor, and it worked well (it was a 1:60 self made CT, 60 just seemed like a good number). when running from halfwaverectifed mains though, i needed a startup oscilator (a 555) connected also to the input of the driver via a fairly low valued resistor (i cant remember what it was though). So my questions are 1) is 1:60 a good ratio for the CT? and 2) shoud it really be just connected like i hade before ? just via a 1K resistor?
Thank you everyone for patiently answering all my silly questions!!

cheers

Jesse

Arcstarter

Mon Dec 01 2008, 06:07PM

Registered Member #1225 Joined: Sat Jan 12 2008, 01:24AM
Location: Beaumont, Texas, USA
Posts: 2253

Well, small signal clamping diodes would be good if you don't already have them. I do not see why it would need a startup oscillator. Are you using the ucc pair? If so, they are supposed to be able to start the oscillations. At least on steve wards they did.

LithiumLord

Mon Dec 01 2008, 06:42PM

Registered Member #1739 Joined: Fri Oct 03 2008, 10:05AM
Location: Moscow, Russia
Posts: 261

Frosty90 wrote ...
shoud it really be just connected like i hade before ? just via a 1K resistor?

Actually the CT has to be loaded. As the CT is a transformer, an unloaded CT becomes just a step-up one, with primary and secondary voltages in phase. However you need the output signal to fit within the desired range and represent the current, so straight after your CT there has to be a clamping Zener.

Frosty90

Tue Dec 02 2008, 12:42AM

Registered Member #1617 Joined: Fri Aug 01 2008, 07:31AM
Location: Adelaide, South Australia
Posts: 139

Hi,
Arcstarter: I do have clamping diodes, and i am using tc4420's rather than ucc's. I need a start up oscilator as i was having problems with my last coil when running halfwave rectified, and also i am hopeing it will help alleviate problems with groundstrikes causing bad things to happen!!
Lithiumlord: with loading the CT, what sort of current should i be expecting out of the secondary base? and hence in the secondary of the CT? should i just then clamp the output with a pair of back-to-back zeners? I already have signal diodes clamping the input of the inverter gate.

Cheers
Jesse

LithiumLord

Tue Dec 02 2008, 01:49AM

Registered Member #1739 Joined: Fri Oct 03 2008, 10:05AM
Location: Moscow, Russia
Posts: 261

The idea is - your CT has to act like a transformer that has it's secondary shorted. Therefore it has to be clamped or shunted straight over it's output terminals, before the decoupling resistor - so you have to put your zeners straight on the CT's output pins. By the way you need the decoupling resistor between the clamp fork and the gate in any topology you use - as most diodes have a non-zero conductivity voltage threshold that usually exceeds the one of the gate's internal diodes and also induce some stray inductance, it's a good idea to limit the current there so there will be no chance of damaging the gate.
As for the startup circuit for a CT, it may get tougher then with the one for an antenna. The first, most simple but also the most inefficient way is an attempt to simulate an extra current by feeding the CT's output with the startup circuit for a while, just like in an antenna system. The better way is using a PLL of course. Somewhere between them (can be selected to get the fastest feedback response) are the two other ideas. The first is using a comparator to eliminate the histeresis. In that case the system will get very sensitive to any signal, so some initial noise your coil will pick up will be enough to start the system. If it is not, you can just add some minor dynamic displacement (like +-0.25v) into the comparator's threshold voltage, so if instead of the desired 0 - +5 or 0 - +12 your comparator sees a straight line (assuming you add some displacement for your CT to feed comparator with some midpoint voltage) this small displacement will give you a signal of your startup circuit on the output. When the CT starts giving off a full swing, that minor displacement signal will affect nothing, however the transition may make things a bit weird. The other way of a straight feedback is a counter. You'll have to make a counter that resets upon the startup and locks whenever it overflows. This can be done by means of a register for example - you just reset it after the system is disabled and upon the next startup you have to feed it with a logic 1 level, giving the startup oscillator's signal to it's sync output. Once it's upper bit output is set to 1, you just need a simple two-channel mux (can be made of two AND gates, an OR one and an invertor) to make the system switch from an open-loop mode to the feedback signal.

Actually you'd better take a deeper look at this. As you are going to use a mains modlation, the startup may get into a tricky thing.

Frosty90

Tue Dec 02 2008, 02:55AM

Registered Member #1617 Joined: Fri Aug 01 2008, 07:31AM
Location: Adelaide, South Australia
Posts: 139

Thanks for all your help.

Ok I will add the zeners to the CT output. You mentioned that the simple 'startup signal into the CT output' method is innefficient. Why is this so? I can understand how it is supposed to work with an antenna system., it simply injects some sartup 'noise'. But why should this be 'innefficient' when using a CT?

Cheers
Jesse

Frosty90

Tue Dec 02 2008, 02:59AM

Registered Member #1617 Joined: Fri Aug 01 2008, 07:31AM
Location: Adelaide, South Australia
Posts: 139

P.S.
Also, i was using a similar topology with my last coil and it seemed to work alright, but it did require a bit of experimentation to get the 'start-up level' right, but in the end i just did it as you would with an antenna, just connect the startup oscillator to the input of the first inverter gate.

Cheers

Jesse

LithiumLord

Tue Dec 02 2008, 12:06PM

Registered Member #1739 Joined: Fri Oct 03 2008, 10:05AM
Location: Moscow, Russia
Posts: 261

The CT needs way more startup current then the antenna, as all you need with an ant feedback is just to drive it with a small current from the oscillator that's only enough to charge/discharge it's capacity. The CT has both a higher internal capacity and also some coupled load, so it takes way more current to drive it, however it's output current is not much higher then the one from a long antenna, so the transition may become not that fast. However it's still yet another way so you can try it out, and, if it works well, leave like it is.

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob