DRSSTC building feed (+pictures!) update 3.2.2013 / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

DRSSTC building feed (+pictures!) update 3.2.2013

Move Thread

LAN_403

dingo27

Tue Jan 15 2013, 09:10PM

Registered Member #890 Joined: Tue Jul 10 2007, 10:06PM
Location: Slovakia
Posts: 180

Hello all, i want to start thread with my TC, where i will post updates for my work. Current status is "made as it is, working, but have no idea if it is working ok". I am only electronic enthusiast, not any science teacher or so, so i am making mistakes (and i like to learn from them). Thank you in advance for telling me where i am wrong.

I have been building and tinkering with that for years, trying many things from basic gate drive to current situation. Many pictures will be added (mostly to picasa webalbums). Coil was named as "testing coil", because it is not wound as carefully as my other coils.

Currently TC is 22cm on 40mm dia pipe, wind with 0.2 mm copper wire, appx. 1100 turns.. Resonating frequency was a huge mystery for me for a long time (See measurement at the bottom). Primary is inverted cone type, made with 8 turns, bottom diameter 60mm, upper diameter 105mm.

Driven by fullbridge of IRG4PC50WPBF IGBT - warp speed, 600V 55A /220A pulsed, VGE 20V, Gate capacitance 3.7nF (datasheet www.irf.com/product-info/datasheets/data/irg4pc50wpbf.pdf). Sturdy, can drive it even in fullwave, but that is not my point.

Driving circuit is 74HC14 fed with current transformer (base feed currently, works well with primary feed too), interrupted through 74HD132 gate with simple 555 from Steve Ward page (really great one, but i must build improved one).

Power circuit for GDT is currently made with 2x BS170, and BC139-16 & 140-16, this is a main problem in chain. Needs a lot of improvement, current goal is to use discrete components, because good DIP drivers are not so greatly available here.

Plans to do -
1. low power requirement, i hate working with huge currents and voltages. So - a boost converter currently developed currently in working state (needs to find its own limits with good transformer). Currently could made 130V from 18V, after that my gatedrive circuit died, Aim is for at least 200V
2. 30cm+ sparks, i dont care if that will be 1x in second, better longer with few strikes, than shorter with high bps.
3. Protective circuits ! OK state LED that will check voltages from all rails.
4. Nice build. I never builded one in any sort of functional and nice state. Nice driver box with power circuit under TC.
5. Rebuild my old fullbridge for use with 2x2200uF 450V Caps, bridge was ok but i could not mount new caps. i used 4x330uF 400V before.
6. Input voltage - 15-18V, 2A max.

---------------------------------------------
So, First step would be to gather resonating frequency. That was made this evening with a scope and my older SSTC driver, with one brick of IGBT.

test setup- power supply, voltmeter, driver and TC, also spare toroid , smaller TC and boost driver can be seen

self resonating driver with IGBT.

Driv1

Simple schematics - on photo is simple interrupter, which does not work, so i omitted it. Basically - feedback from antenna is cut by 1N4148 diodes and fed into 74HC14 gate, which drives BS170 mosfet. That with 100 - 150 ohm resistors works like level shifter from 5V to 15V logic. Two NPN and PNP transistors then drives a gate of mosfet (or IGBT, be careful to use diode if your IGBT does not have one). On antenna can (and usually will) be corona, be careful to not touch it, or you can get RF burn. If driver does not work, swap inputs of primary, or add another schmitt gate.

So, Resonating frequency was gathered from 74HC14 and gate of IGBT, both showed 490kHz. One mystery solved. I will post schematic to that, it can generate needle-like 5mm spark from sharp breakpoint. Driving transistor can be any mosfer or fast IGBT, but needs great cooling. I was using that mostly for testing performace of transistors.

SDS00004

This is what gatedrive from that circuit looks - could be improved with larger foil and aluminum filtering caps for power part, also ringing could be (maybe) improved with series resistor to IGBT gate input.

Next step - building better GDT power circuit with discretes. Probably P/N mosfet. Output voltage at least 18V.
-------------------------------------------------- -------------------------------------------------- -------
19.1

Boost converter ready, pushing up to 250V.

schematics by Ronald Dekker, but i used different setup, read down.
Boost2

I made a little improvement, added voltage comparator that compares output to known voltage (output divider 1Mohm : 20kohm, input divider 50kohm pot on 3.4V) - if output voltage is greater than one pre-set, output will drive a gate of BS170, and resets 555, so output will be zero. 555 need a pullup resistor for reset pin. pin 5 is unconnected. Reset pin than works for info LED, that shows if pre-set voltage is reached (if it is lit, reset is on) Works great, but there are a little things i will do other way next time.

Some things i learned -
- even smaller choke can produce high voltage, toroidal or I form works well. Smaller value can be used, currently i am using 27uH from PC powersupply (one on picture was getting hot, saturation-overcurrent? )
- use larger heatsink! Not that this does not work, but it gets hot, i need other than IRF740 mosfet - possibly with 450Vds, low Rdson. I had IRFB4127, but works only up to 200V...
- Second smaller pot was mentioned to work as a max voltage limiter (without load, i can get over 275V limited by X2 cap at output), but its value is too low! should be 20k instead of 10k.
- I always use bleeding resistor on output, but in screw plug, not on PCB. Reason? Sometimes i put bleeding resistor to supplied thing as well, so i can simply take it off and use only one.

Of course i tested it with current TC, i don't have my IGBT fullbridge assembled, so i used only mosfet one with IRF830 (bad, bad ones, use other transistors!!!) Sparks length up to 16cm, could be more if i used other mosfets. Double resonant mode was not possible to start - i guess due to high Rdson (1.5ohm) of mosfet. That would make sparks even longer. But i had strange phenomenom, tc was tunning happy for 5-6 seconds, than sudden stop for second, than running again. This was happening constantly, Any advice why?

dingo27

Sun Feb 03 2013, 10:02PM

Registered Member #890 Joined: Tue Jul 10 2007, 10:06PM
Location: Slovakia
Posts: 180

(first things first - should this be in projects thread? If so please be kind and move my thread... THank you in advance. )
Since last update i Was tinkering a lot with my TC (well with 3 of them), and now three bridges.

Last time i ended up with my boost controller - good thing is that if I use proper filtering caps (of high XXXXuF range) on output, the heatsink doesn't get hot, it has room temperature. I used it with my current IGBT bridge and was able to produce longest sparks i could achieve - sparks that were getting out of toroid, not a breakout point! I played a little with that but next day i was not able to produce same results until now, and i don't know why :(

So currently i am working on my low power driving circuitry, which has terrible habit to produce random signals out of nowhere (well they are from somewhere but i don't know from where :) ). So i bypassed all my TTL gates with 560uF caps, and added antiparralell schottky diodes (BAT43, maybe should be for higher current) to my output BD139/140, which destroyed ringing nicely

SDS00016

Ringing destroyed with use of schottky diodes
(i have no idea of what a blue line is, but if should be gate voltage of IGBT)

I also thought of redesigning my IGBT fullbridge, so i disassembled it, redone GDT wiring (before that it was a birds nest of wires, but worked greatly - now twisted pairs seems to work not-so-great as before that, but that might not be it.) I added 4.4 decoupling WIMA FKP1 caps for both halves (i did not have any better caps and more space for bigger one) and also smaller (1/4w?) resistors with antiparallel schottky diodes to each gate. Also changed Faston type connector for driving primary coil for screws, but some from those changes had worsen the performance of bridge. Currently i can only think it is the screw connection i did, because i desoldered one leg of schottky and performace was same. Hard to tell, maybe i need chrome screws which always worked better for connections in my case, but i did not had long enough one.

Redesigned bridge, but performance was worsened. The brass leverers are not used in final design, brass is not a good conductor with HF circuits.

I also wanted to build another bridge to see how smaller gate capacitance can impact performance (or i was just bored). Link for photo of halfbridge.

Worked marvelously, i could drive with it even my worst nightmare 2.2MHz TC (which is hard to get feedback from, needs to be driven from generator). If consists of (now) old IRF510 mosfet which has i think 180pF gate input, 2.2nF WIMA FKP1 caps (2 for decoupling, 2 for voltage divider), and 3 electrolytics, which the biggest one is 1500uF / 450V low impedance one (i wonder if that low impedance matters with such a low power mosfet). The big elyte is important, without it the voltage was getting crazy. Also, smaller elytes should be low impedance too ( before that i had some 470uF/200V from PC powersupply and that worked better than current ones, but they had shorter leads which did not allow me to mount bigger elyte. I need to order pair or better ones, but possibly smaller. 47-100uF could be enough.

Shame on me i don't have any spark photos (but it will be), currently i am working on transferring my driver into metal box, first part of driving board is done, but i still need power and interrupter board (modules are easily manufactured for me, and driving board was a test of making board with laser printer)

That board has everything needed to run GDT, so it would be enough for SSTC, but output transistors does not have heatsinks so it would be only short run. I have headers for 4 enable inputs , so 2 will go for interrupter (for each channel) and two for safety voltage comparator (if voltage for GDT driving circuit gets below 12V, it stops oscillating). Two power resistors needs to be more powerful ones! At least 5W for 15V gate drive, they get hot (i mean 5W ones get hot, this ones get hot even at 7V). This was only for testing purpose to see if bridge is working.

I expected that i would get pretty clean waveform , but no - there is ringing again! Eagle eyed person could notice i forgot to add schottky diodes to transistors ( well i did not had any place for it anyways), so i need to solder them from bottom, and also put decoupling caps across ICs.

SDS00026

Yellow line - new, PCB version, Blue line - old testing board version of my driver. At least rising edge seems a bit sharper on new design, so it was not a total waste of time. Board also seems foolproof, for a moment i put 15V on 5V rail, and only thing that went away was indicating diode. For a moment i was questioning my decision of soldering IC directly without slot, thankfully both of them are fine.

Next mission will be to destroy ringing ( but on already etched board? i will see what i can do), Also, i need to implement zero crossing circuit, so a mosfet will turn on only if there is no current going through it, but how to implement it, any ideas? I believe that is the most peroformace weakening factor in my design, retuning coil again and again, which cannot be healthy. Maybe a divided frequency that would be taken from current CT of promary ground, so that interrupting signal would be always synchronized with waveform....

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