CT feedbacking Problem / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

CT feedbacking Problem

1 2 next

Move Thread

LAN_403

zzz_julian_zzz

Wed Nov 16 2011, 02:51AM

Registered Member #3964 Joined: Thu Jun 23 2011, 03:23AM
Location: Valenzuela City
Posts: 332

HEllo,

When I am using CT feedbacking, My oscillation stops when I reach 100volts input in my halfbridge. THe schem I am using is this

Why does the operation/oscillation stops whenever I put more voltage input from my Variac? What could be the problem? But when I use Antenna feedbacking the oscillation never stops even up to 220v input. =(

Marko

Wed Nov 16 2011, 05:14AM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Hi,

Well, all I could say is that your 'Schem' sucks (blame Steve Ward for this). Using a comparator in place of your 74HC14 gates might help a lot indeed - you can power the comparator from +12V and bias the CT burden to +5V instead of ground , then connect it across comparator + - inputs. It would probably also be a good idea to make the burden a few-ohm resistor and use diodes just as a safety measure.

This would be abasic fix for secondary base feedback, I never had good experience with it either using the infamous 74HC14. Though if you wish to build a DRSSTC you might want to switch to primary feedback and add OCD (HC14 has apparently better luck working with primary feedback).

Marko

zzz_julian_zzz

Fri Nov 18 2011, 03:41PM

Registered Member #3964 Joined: Thu Jun 23 2011, 03:23AM
Location: Valenzuela City
Posts: 332

Hey Marko,

Thanks for the informative input. You're right, S.Wards Circuits are far more difficult to make work. Uhmm, can i use antenna instead of a CT? also, I am using a discreet driver so do you think this would be problem? (discrete by mean, Antena=>INVERTER=>NAND gate(INTERUPTER)=>MOSFET halfbridge(N-P channel)=>GDT=>POWER HALF BRIDGE.

Marko

Fri Nov 18 2011, 09:19PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Hi sir,

well I'm not sure about your discrete gate driver, I would keep UCC's in any case if you already have them.

Also, replace the input inverter in your circuit with a fast comparator (LM311 should be OK for low frequencies). Look at the Steve ward's very first SSTC's - they used such an approach (+a starting oscillator). This should work with both antenna or a CT and I believe it's a far, far better approach than HC14 thing.

Marko

zzz_julian_zzz

Mon Nov 21 2011, 02:06AM

Registered Member #3964 Joined: Thu Jun 23 2011, 03:23AM
Location: Valenzuela City
Posts: 332

Sir Marko,

You said:
Also, replace the input inverter in your circuit with a fast comparator (LM311 should be OK for low frequencies)

My TC frequency oscillates around 200khz, is LM311 suitable for this range of frequency?

dude_500

Mon Nov 21 2011, 04:24AM

Registered Member #2288 Joined: Wed Aug 12 2009, 10:42PM
Location: Cambridge, MA
Posts: 179

In my experience, anything with an open-collector output such as an LM311 isn't nearly fast enough for tesla coil feedback applications, simply because of parasitic capacitances charging through the high-resistor of their output. I'd go for a slightly more expensive comparator with low propagation delays and specifically, a TTL logic output. I use an AD790 which is a bit costly, but works great.

Although there is really no reason to not just use the classic zener shunted current transformer fed right into a logic gate if you don't care about phase lead compensation. If you do, then you will need to use a comparator.

zzz_julian_zzz

Mon Nov 21 2011, 05:20AM

Registered Member #3964 Joined: Thu Jun 23 2011, 03:23AM
Location: Valenzuela City
Posts: 332

Hi dude 500,

thank you for sharing your experience with us, but what do you mean by this:

Although there is really no reason to not just use the classic zener shunted current transformer fed right into a logic gate if you don't care about phase lead compensation

I am very sorry for I have no any idea of phasing. What problems could it lead, and what are the benefits of having a correct phasing? I just followed the schematics above

, does this circuit have the correct phasing? how to check it and correct it? thank you very much..

dude_500

Mon Nov 21 2011, 05:58AM

Registered Member #2288 Joined: Wed Aug 12 2009, 10:42PM
Location: Cambridge, MA
Posts: 179

My comment on zener shunting is only applicable to current transformers.

The circuit you linked to is not phase lead compensated. Look at Finn Hammer's Prediktor controller for details on phase lead techniques. The idea is to read the current as a signal leading the actual current so that propagation delays in your controller and H-bridge can be corrected for. The benefit being that your reverse recovery diodes won't slip into forward bias which can cause a huge voltage spike on switch transitions as they reverse again.

It is usually not something done on small coils to medium powered coils (less than bricks), although there's no reason not to short saving on complexity.

Basically, my point being if you want to keep it simple just feed directly into gates, and if you want to then go all out go with phase leading into a high speed comparator, but there's really no reason to go half way and use a comparator without a leading compensator as there is almost no benefit to compared to feeding into a gate, and the cost is not warranted in my opinion.

zzz_julian_zzz

Mon Nov 21 2011, 08:08AM

Registered Member #3964 Joined: Thu Jun 23 2011, 03:23AM
Location: Valenzuela City
Posts: 332

Thank you very much again for your astude explanation. I am to go to "Easy mode" for I already have problems / headaches regarding feed backing. =l what do you think of my circuit being like this..(see below)

I was able to make it work for some couple of days, but when i tried to use c.t. instead of antenna, the system didn't work. then i again back to antenna system, the tc produces wimpy sparks.. =( from 14 inches(before) to 14 cm(after) only =(... please help me..

I need your comment on my circuit.. =( and what do you think make the circuit did not work again? i've checked the H-BRIDGE, and it works well with my VCO type driver. please help..

Marko

Mon Nov 21 2011, 11:16AM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Bipolar totem poles are voltage followers, so you'll only get like 4.3 V out of it if driven by 5V logic, and this will cause your output mosfets to sit in linear region and burn up.

You could use a level shifter made of a BS170 mosfet and a pullup resistor, or use a CMOS NAND gate powered from 15V but if you insist on 74HC14 input I'm afraid that loop gain will still not be closely good enough and you'll probably fry something.

Regarding comparator, I think LM311 should just borderline work at your frequency. UCC's have fairly easy to drive inputs, but if you find duty cycle distortion to be excessive you can use two comparators and two non-inverting UCC's (I usually design my systems that way). This way you'll have some deadtime built in which you could control by varying the pullup resistances!

BTW, comparators should work well with your discrete driver too provided you amplify the output properly.

A phase lead compensator would be nice, but considering to what I've seen so far your system should work without it fairly decently too.

The most common problem for noobies is obtaining a sufficiently fast comparator for this job, and hence they rather suffer endless trouble with HC14 thing.
Biasing of HC14 could likely be improved, but I'm not sure whether to even recommend that.

1 2 next

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