Multipurpose Flyback Transformer Driver / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Multipurpose Flyback Transformer Driver

Move Thread

LAN_403

EEYORE

Sat Jan 23 2010, 08:10PM

Registered Member #99 Joined: Thu Feb 09 2006, 06:10PM
Location: florida, usa
Posts: 637

Hello,
I am planning to build a driver that can drive flyback transformers, but also hope it to be able for audio modulation. Right now, I am going to either go the TL494 route or the 555 route. I am not sure which way to go. Id like to drive a half bridge of IRFP450s for power.

Could I get some opinions? I believe both the 555 and the TL494 have been used for audio. The 555 would be the easiest route to go, but the TL494 seems to have more capabilities. If I went the 555 timer route, I would send the output to a pair of gate drive ICs (inverting and non-inverting) to drive a GDT. If I used the TL494, it looks like I would use non-inverting ICs.

I will ultimately build this onto a two sided pc board so that I can use the bottom as a ground plane. All ICs will be DIP so that I can easily change them out in the event they fail. The rest of the parts will likely be surface mount (resistors, capacitors, and diodes).

Could someone steer me in the right direction?
Matt

Dr. Dark Current

Sat Jan 23 2010, 08:52PM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

The problem with 555 is that it has only one output which is either on or off, and any higher deviation from 50% duty cycle is going to upset your GDT/mosfets. The audio modulation to pin 5 results in both frequency and duty cycle modulation, which I'm pretty sure would end up with flying MOSFET bits at any deeper modulation.

The TL494 is better in that it has two outputs which switch alternatively in a push-pull fashion, so the DC component on the GDT and flyback transformer's primary is always zero (OK, if you don't try to modulate with high frequency signals at big amplitudes, which is not the case with normal music).
The modulation is pure PWM/dead time modulation, so the frequency doesn't wander.

However, I wasn't able to get acceptable sound from a half bridge, the "half-controlled fullbridge" which has the other two switches replaced with diodes, sounded MUCH better, and after adding a series RC damping network on the primary, the quality is perfect.

Arcstarter

Sat Jan 23 2010, 09:46PM

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

If audio modulation is not completely necessary, i would suggest the IR2153 halfbridge driver... It does not need any GDT's or very many external components. You can drive the halfbridge off line, if you use like 40 turns on the LOPT, and you can push quite a bit of power. The duty cycle is always 50% (not sure of the tolerance), and you can change the freq with a pot, and timing capacitor. You can even supply the IC without a seperate AUX supply if you wanted. Obviously not with 120v input.

You could use a resistor to charge a reservoir capacitor up when you turn it on, which will start the circuit up, and a few turns on the LOPT core with a fast rectifier and curent limiting resistor to keep the reservoir cap charged... The IC has an internal 15.6v zener for clamping, but i would suggest you used some external 15v zener, maybe a watt or more. Because it is a halfbridge, you will not have crazy random HV spikes as much as a single device, so the AUX supply from the LOPT should be stable enough.

This would be a tiny but very powerful driver. If you really wanted to, you could audio modulate it with a bipolar transistor on the positive supply to the halfbridge, but i imagine that may be unstable, and it would dissipate some heat. As long as you kept the current low enough, it may not be too bad...

Dr. Dark Current

Sat Jan 23 2010, 10:02PM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

The IR2153 can be frequency modulated with a MOSFET connected as a varicap to the Ct pin, but this is far from ideal and works better for SSTC then a flyback. It can probably even be externally controlled into the Ct pin, if you make sure its voltage is between 1/6 and 1/3 of the supply voltage when low, but IR2111/2112 is better for this.

But I don't like the IR21xx chips and I don't use them anymore, because they like to explode for apparently no reason. I think this is because the floating supply ground can "fly" inductively below ground (from wiring inductances etc.) and this upsets the driver which turns on BOTH mosfets at once, you can imagine what happens.
This problem could probably be improved or almost eliminated by putting dc block PP caps as close to the MOSFET pins as possible, connecting a fast diode close to the IC from ground to floating supply gnd etc., but anyways I dont like those chips that much.

Mads Barnkob

Sun Jan 24 2010, 04:38PM

Registered Member #1403 Joined: Tue Mar 18 2008, 06:05PM
Location: Denmark, Odense C
Posts: 1968

Maybe you can get some inspiration from my yet to be completed TL494 driver

EEYORE

Mon Jan 25 2010, 07:14AM

Registered Member #99 Joined: Thu Feb 09 2006, 06:10PM
Location: florida, usa
Posts: 637

Hello,
I am not quite sure I understand how the half controlled H bridge works. It looks like you would need both FETs on at the same time, no? And if so, I dont understand why not just use a single FET.

Might I be better with just one FET and no half/full bridge work? Has anyone done audio modulation with just a single FET?
Also, thanks for that schematic Mads. Its about what I was thinking too. Where do I input the audio signal at? Pin 5?
Matt

Dr. Dark Current

Mon Jan 25 2010, 10:45AM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

Of course it does work with one FET, but for mains voltage operation you'd need an additional reset winding on the flyback to limit the off-state voltage, and you still need a MOSFET with ~3x higher voltage rating than your DC supply and if your mains voltage is 230V (325V DC), this starts getting a bit problematic. Also the coupling of the reset winding will not be ideal so the FET will be dissipating additional power from avalanching.

The advantage of the half controlled H bridge is that both transistors see just 1x supply voltage, and yes they both need to be ON at the same time, and the duty cycle must never reach 50% or more otherwise your flyback's core saturates.

If running off-line is not a requirement, you can still use a mains transformer and drive the flyback in standard single-transistor "flyback" configuration, if you don't mind the MOSFET getting hot.

EEYORE

Wed Jan 27 2010, 05:18AM

Registered Member #99 Joined: Thu Feb 09 2006, 06:10PM
Location: florida, usa
Posts: 637

Hello,
I will have to look into the half controlled H bridge since I HATE when things get hot!

Here is what I have so far for a pc board design. I am not sure how I would audio modulate this. Its pretty much the same circuit as referenced previously plus some extras.

Any thoughts? Suggestions?
Matt

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