Flyback driver

[electronics-whiz]

I have an old board i pulled from an old CRT TV. I got it a case from an old stereo now. It's big, and not real practical. I would like to make it so i can have some control of the output. From what i understand CRTs can do that, and that adjusts the brightness, i just don't know how to do that. I'd kind of like to use to run some old gas lasers, maybe some things like Jacobs ladders, lifters.

I've worked with electronics for years, know the safety stuff. I have primarily dealt with linear or analog electronics transistors, caps, etc. If this involves an IC you may need to explain a bit more.

I don't know how, but i'm thinking possibly a 555 IC and transistor(S) could do it along with some pots, etc to make the IC go.

I also have an air bladder activated kill switch i can use for safety. (It's meant for power tools, but the part you activate links via small air line.)

[electronics-whiz]

Cool info, gives me a couple of ideas, or at least realizing what parts i may want.

Idea one: if i want to adjust the output couldn't i just wire it up to a dimmer switch? Those pretty much work on a frequency type setup to control the output.

I have some old transistors and nice heat sinks from an old 1500VA UPS. I think their rated to almost microwave frequincies, and drove some fairly heavy duty transformers. A UPS driver board would be close to what i would probably need, but they seem to go from 50-75HZ i doubt that would be that effective with a flyback.

Generally what is the voltage for a fly-back primary? I believe it's relatively low voltage at high frequency like a SMPS. That's my primary guess why a dimmer may not work, and it would need some protection from kickback like a diode i would think.

[T3sl4co1l]

One of the few times I'll recommend a 555...

Nice thing is, if you have the PCB with the transformer, you can trace which pins are the primary winding, and what voltage.  Usually it's 160VDC (rectified mains).  One side connects to an electrolytic capacitor (or through jumpers, small chokes / ferrite beads, or low ohm resistors), the other to the HOT (horizontal output transistor).  The traces should be relatively fat and have good clearance, or even routed slots near connections.  So it should be pretty easy to follow.

Instead of MJE18008 or other switching transistors (anything of this general type can be found in CFLs, older AT/ATX supplies, etc.), you can use the original HOT, too.  It's going to want a lot of base current though.

Downside is, you need the +12V supply AND whatever the primary needs, so you probably want an isolation transformer* and rectifier, in addition to a 12V DC supply.  Don't forget to connect circuit ground to safety ground, or you'll build up static and cause breakdown of the isolation transformer.

*Uh, if we're making dangerously high voltage anyway, does it really matter that it be line isolated?... probably not. Just don't probe the line side without an isolation transformer.

I mentioned the HOT needs serious base drive.  They handle this in the real circuit by using a drive transformer, which matches quite nicely to a 12V supply, or something around there.  I should play with / draw up a circuit sometime which includes this change -- much better than trying to beat half an ampere out of a poor 22 ohm resistor and 2N4403.

Tim

[Paul Price]

The suggested driver above using 2N4403/2N4401 is over-complicated and uses too many 1K resistors.

Use the below circuit to drive the MJE18008, or better yet replace the MJE18008 and use a HV N-Chan MOSFET instead like millions of modern SMTP power supplies use. It is easy enough to find one free in a discarded DVD player SMPS.

The 100 ohm resistor in the MJE18008 circuit is likely to go up in smoke. A better protection circuit would be a fast HV diode connected to a .0033 cap/1600V with a 47K resistor to ground across it.  Anode of diode to collector/drain.

The MJE 18008 will certainly be smoked immediately if you set the duty cycle control to high.

 Try using a 18 to 20V PC 2A or better laptop wallwart to supply the output stage and you can get a few thousand volts safely isolated from the mains. The flyback transformer likely is connected to a voltage tripler/quadrupler diode circuit built into the transformer and you might have more HV than you expect.

Watch where you put your fingers.

[c4757p]

So, you expect a PNP transistor with its collector at ground to discharge the base of an NPN with its emitter at ground? You are familiar with V_BE, I'd assume. If the 555 manages to output exactly 0V instead of a bit above, it might just make it due to the lower V_BE at lower drive current... There's a reason he built the driver common-emitter...

[Paul Price]

The output transistor will be turned off when the voltage at the base drops below approx .7V. Should work good enough for fooling around with this HOT circuit.

In any case,  would work ok with a  .1 uF to 1uF connected across 22-ohm to remove stored charge in the B-E junction of the MJ18008.

Most TV circuits of this type parallel the 22-ohm resistor with a 1uF or so electrolytic and a fast diode might be needed (connected cathode to MJ18008 base, anode to ground) to limit the negative swing at the base of the MJ18008 to not reach base-emitter reverse breakdown voltage during turnoff.

[Paul Price]

Improved driver

[T3sl4co1l]

So you're improving a circuit with "too many components" by making it worse, then adding more components to fix it? 

The most important part of the circuit is having it not be the god damned 2N3055 POS that everyone else on the internet jumps to.   After that, yes it could still use improvement, but it's a better "getting started" circuit, at least.

If you check the range of duty cycle available with the resistors and pot shown, you will see it won't burn the transistor.  Assuming the FBT is doing its job, of course -- a circuit with a UC3842, say, is always safer, but takes more stuff, and probably not as likely to be on hand as a 555.

Likewise, the RC is there to absorb some ringing, and doesn't get hot.  Obviously, if it's burning, use a bigger resistor (or just forget it altogether, it's not important).  The 22 ohm 1W I would've thought would be a bigger offender, especially with the 2N4403 being practically in hFE limit.

The chain of 1k resistors is necessary, if you study the V_OH and V_OL of the 555 and the transistor Vbes.  Actually the top resistors should be smaller, for the aforementioned reason.

MOSFETs are good (and can... kinda sorta, be direct driven from the 555, but it's not as good as with a proper drive circuit), but the high voltage required (>800V) sucks for most purposes.  You only find those in newer flyback/forward PSUs, and even then, none over 1kV.  And those over 1kV suck on Rds(on), to the point where the BJT is actually starting to look better for performance reasons, despite its drive difficulty and slow speed.

It's noteworthy that my Trinitron has a MOSFET powering the high voltage (the FBT -- but not the HOT, which is a massive 40A BJT), but again, being that's high-falootin' equipment, they could afford to do that.

Tim

[Zero999]

What voltage does the flyback need on the primary? I have one working from 12V quite nicely and that's run from the existing primary, with no modifications.

If the flyback is run off the rectified mains then you could use a MOSFET + BJT cascode to drive it. The MOSFET can be a low voltage, low R-on type and the BJT, the usual high voltage type.

[c4757p]

Hmm, a couple extra 1k resistors, vs. a massive freaking electrolytic?

[BennVenn]

I saw a week or two ago (might have been youtube?) a man connect a flyback to the output of a modern smps downlight transformer. He wound a few turns around the ferrite core and had a globe in series for current limiting. Worked great and they're pretty cheap in hardware stores. Also dimmable for output control?

edit -

[T3sl4co1l]

Hmm, a couple extra 1k resistors, vs. a massive freaking electrolytic?

Well, it's only one or a few uF, hardly massive.  I'd be worried about it getting hot, honestly (in values like that, the ESR is going to be as much or more than the resistor in parallel with it), but the general idea is a good one -- boosting base drive, especially turn off drive, with a "speed-up" capacitor.  Sized appropriately, you could do this to drive it direct from the 555, assuming the 555 delivers enough current.  For best results, I'd recommend a large film capacitor (or ceramic) in series with 10-47 ohms.

Tim

[rickselectricalprojects]

if you want to run a flyback transformer one of the easiest and most powerful ways would be a zvs driver
http://www.instructables.com/id/The-ZVS-driver/
although i wouldn't recommend soldering it on a perf board because it draws at most 12 amps ( in my experience)