High Voltage Output Blocking Oscillators (JT like)

[Fraser]

OK, this thread is for anyone building blocking oscillators (aka Joule Thiefs) using hand wound coils or LOPTX (TV flyback trafos)

Here's my build of one with vids and schematics to help people get started.

I started with this info:
http://www.angelfire.com/80s/sixmhz/flyback.html

and made a few changes like adding input power filtering and using a different tranny. I use a 13005 which works well with little heating, after around 22VDC input they get zapped though so be careful.

the tranny base signal is biased using a R potential divider. But of course for more efficient operation one would want to do away with the divider as it constantly burns off supply power. However when doing away with the divider the circuits operating voltage window goes very narrow. I used the divider approach in order to give me a wide V window (2 ... 22V) and so I can use the device as a handy benchtop HV supply.

The LOPTX I used was labelled like this:

FCM-2015AL
2859-136-010
SAMSUNG GAb

[Fraser]

Here's pictures of the board and the finished unit.

Note the botomm pic shows a 0.5watt resistor I had to swap in to handle the power.

[Fraser]

A video of the device breaking down over a small jacobs ladder, hmmm ozone, lovely!:

the device powering out into a cap and gap breaking down via a 12V auto bulb:

another vid of the device in action (low room lighting):

[the_big_m_in_ok]

Here's pictures of the board and the finished unit.

Note the botomm pic shows a 0.5watt resistor I had to swap in to handle the power.

Fraser,
Is your transistor anything like an American 2N3055?  It's and industry standard NPN.

REEDIT:
I did see later the list that came with the first post.  I see a trend in NPN's that higher voltage and current tolerance often work netter.

[Fraser]

Fraser,
Is your transistor anything like an American 2N3055?  It's and industry standard NPN.

--Lee

Hi Lee,

No it has quite different specs, 13005 has better resilience to higher gate voltages and reverse bias and also faster, a 2N3055 would be zapped in this application at much lower supply voltage, but you would probably achieve oscillation with it (I haven't tried it though). If you try it then wind up slowly, may have to sacrifice one to find the max input V.

[the_big_m_in_ok]

Hi Lee,
No it has quite different specs, 13005 has better resilience to higher gate voltages and reverse bias and also faster, a 2N3055 would be zapped in this application at much lower supply voltage, but you would probably achieve oscillation with it (I haven't tried it though). If you try it then wind up slowly, may have to sacrifice one to find the max input V.

Good to hear from you Fraser,
A theoretical redesign idea occurs to me:  Wire several 2N3055's in series, collector-to-emitter, so that higher input supply voltages can be shared by individual transistors?  Base inputs can be paralleled.  What do you think?

--Lee

[Fraser]

I reckon that would work yes because fairly low operating freq. Also could handle fair power with minimal tran heating. But maybe a little more junction losses.

But if you wan´t a HV supply go for it!

I have no 2n3055 on hand to try, I´m betting you have a few of those trannys though.

Do you have an old TV flyback trafo? If so then I can talk you through finding pinouts. If you want an old TV just go to the local refuse site like I did, ask the fella if you can take a TV for spare parts, smile nicely or give him a couple of bucks, worked for me ( I smiled  ).

Also you might want to take the sting out of the base signal a little, feed the divider node with another R, maybe a 1k pot to tweak from 1k down. I think my failure mode (at 22V DC in) was base signal overvolt from the primary coil BEMF.

[the_big_m_in_ok]

"I have no 2n3055 on hand to try, I´m betting you have a few of those trannys though.Do you have an old TV flyback trafo?"

I have 3055's but no HV trannies.  I had the idea of making several of my own, but they would be low-current, and moderate step up/down ratio.  It's cheap and lacking in consistent precision, but "free" is hard to argue with.

"If so then I can talk you through finding pinouts. If you want an old TV just go to the local refuse site like I did, ask the fella if you can take a TV for spare parts, smile nicely or give him a couple of bucks, worked for me ( I smiled   )."

People in San Francisco, my hometown, leave obsolete CRT TV's on the street, These TVs have HV trafo's going to the CRT, but they're too big and I have no place to put them.

"Also you might want to take the sting out of the base signal a little, feed the divider node with another R, maybe a 1k pot to tweak from 1k down. I think my failure mode (at 22V DC in) was base signal overvolt from the primary coil BEMF."

I would take a diode to feed back to the heavier secondary of a 120 VAC tranny so as not to burn the primary when the whole trafo is turned around to step-up(I mean mains line voltage trafo's, that is. The same would hold true for HV types).  

[Kooler]

hello...
does this fit in here....
http://www.youtube.com/watch?v=ruJXtBby-pw&feature=channel

i just thought i would post it cause i get good results with a flyback coil from crt's also

[Peterae]

Great work Guys

Just for reference if you have a LOPTX or Line Output Transformer from a tv or monitor there is a code printed on it somewhere, they all should have one on.
If you then do a search for that number on this website it will give you the pinout of the transformer.

For instance Here's one i have used a lot it in the past
http://www.hrshop.es/index.php?command=viewSchema&filename=http://www.hrdiemen.es/products/img/esquemas/HR80033.gif

http://www.hrshop.es/index.php?command=search&by=tvmonitor&q=&x=12&y=8

Also the 13005 is used in some tv's and can be ordered with the ref MJE13005

[pdf]http://www.datasheetcatalog.org/datasheet/kec/MJE13005.pdf[/pdf]

Peter