Stepdown blocking oscillator possible? / General Science and Electronics / Forums

Forums

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

« Previous topic | Next topic »

Stepdown blocking oscillator possible?

1 2 next

Move Thread

LAN_403

Dr. Shark

Sat May 13 2006, 08:56AM

Registered Member #75 Joined: Thu Feb 09 2006, 09:30AM
Location: Montana, USA
Posts: 711

Dear 4HV community!
I am looking for a quick and dirty way to generate something like 12V from 220V for the logic part of mains-driven circuitry. Of course a small stepdown transformer from a wall-wart would do just fine, but I hate iron

and anyway, why should I do something simple if it can be done more complicated? So, the idea is to use a single transistor blocking oscillator to generate 12V. There are lots of schematics out there to step up voltage this way, but I am not so sure about stepping down. The trouble really is that 220V is not great for driving a transistor base or FET gate.
I am sure there are clever ways around this, so if you know of any, please enlighten me!

ragnar

Sat May 13 2006, 09:25AM

Registered Member #63 Joined: Thu Feb 09 2006, 06:18AM
Location:
Posts: 1425

How about some zener diodes and a resistor, Chinese-style? =)

Dr. Shark

Sat May 13 2006, 10:14AM

Registered Member #75 Joined: Thu Feb 09 2006, 09:30AM
Location: Montana, USA
Posts: 711

Umm, suppose I am drawing 500mA from the 12V line, then I would need to burn 100W somewhere! How do the Chinese do that? Series lightbulb maybe?

Nah, all of the mobile phone and digital camera wall-warts today are solid-state, so there must be some kind of easy way to do this. Of course they all use some kind of PWM controllers or other chips, but these must be powered some way, which is probably blocking-oscillator style. So I am sure there must be an easy way

ragnar

Sat May 13 2006, 10:20AM

Registered Member #63 Joined: Thu Feb 09 2006, 06:18AM
Location:
Posts: 1425

Fair enough =)

How about using zeners/resistor to provide the voltage for the base of your oscillating transistor / tiny SMPS, then?

Marko

Sat May 13 2006, 12:02PM

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

For 300V it is actually far harder than it seems.
I once tried to make a significant-power SMPS using a blocking oscillator, with a transformer from monitor SMPS, I interestingly made it work with one small HV transistor, I could pull tens of watts without even slightly heating it up.

When I removed the load on output (motor) it blew up.

Later I blew up tons of transistors (I ven tried to redesign circuit for mosfet's but unsucessfuly).

I tried to add zener+optocoupler regulation but this would just make circuit undergo slow relaxation oscillation and blow itself up. Feedback seems to work very poor in such conditions, it seems to just tear the transistor down into linear region rather than significantly affecting frequency or duty cycle.

This was all madness as I had no scope and I didn't want to blow another ton of BUT11A's.

Basically this was a copy of ATX supply auxiliyry SMPS. It is a blocking oscillator, usually with a mosfet or bjt (I preffered bjt's) with few more protection conponents and opto feedback (or sometimes 7805 regulator).

I think best idea would be to exactly copy this part (its generally simple, and any philosophy (like me puttong a big ferrite transformer there) will just ruin it.
Transformer must be exactly one from atx standby SMPS (ones that work with bjt's and mosfets are different, you can't swap them).

The thing is in the down-left corner, here are schematics using BJT's and mosfets:

Also found this while googling for another one (luckily I had it on a computer):

Stick around this and you'l have a very reliable small SMPS capable of giving amp or two at few volts (output voltage could be controlled by zener on feedback and/or 78** regulator, may even be made variable

I also have one schematic using only two transistors, based on simple astable multivibrator.
It was relatively efficient, capable of delivering maybe hundred watts (with BUT11A transistor) and main thing was that I could easily control duty cycle and make a good feedback.

Main problem was that I had a resistor powering the thing and it dissipated about 10W or more.
I guess I should have used additional winding for power supply, maybe do something for faster switching of main transistor, etc...

It is too unscoped, and I have no clue of waveforms and frequency. I think freq is somewhere between 30 -100kHz.
Feedback worked perfectly, the circuit always held eact output voltage equivalent to voltage drops of zener and optocoupler led.

Choke in series with primary is overcurrent protection. One I used was quite big, I could almost short the primary and leave the circuit working!

You could aslo use the circuit as mains-powered, simple regulated flyback driver, etc.

Pity I didn't continue to work on it.

Sulaiman

Sat May 13 2006, 12:53PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

IF YOU ARE NOT CONCERNED ABOUT THE DANGER,
then you can use ;

Capacitor from line to bridge rectifier feeding dc reservoir capacitor and zener.

ROUGHLY Idc = Vac/Xc = Vac.2.pi.50.C (for 50Hz here in UK)

e.g. for Vac=220 and Idc = 100 mA,
C=0.1/(2.pi.50.220) = 1.5 uF
(non-polarised, e.g. mylar)

ALL parts MUST be considered LIVE !!!!

Negligible dissipation in the capacitor.
Capacitive load for mains (good, counteracts normal inductive loads)

Put a fuse and a line filter in the ac line supply,
and a DANGER sign to remind yourself that you cannot
come into contact or 'scope this circuit.

Old TVs used a similar technique,
a mains/line Isolating transformer is required for 'scoping.

Dr. Shark

Sat May 13 2006, 12:53PM

Registered Member #75 Joined: Thu Feb 09 2006, 09:30AM
Location: Montana, USA
Posts: 711

Thanks for your input, Firkragg. Actually the ATX auxilliary PSU was what kind of inspired this thread. I think that you are right, I am best of exactly copying that (luckily I have a dead ATX with schematic around), and work from there. I just thought that maybe someone had already gone through the trouble of optimizing the design and might save me the work. MacSteve?

Marko

Sat May 13 2006, 01:01PM

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

I just thought that maybe someone had already gone through the trouble of optimizing the design and might save me the work.

How you mean 'optimizing'?

Any voltage 'mod' you do simply by adjusting feedback or changing regulator. Once you get main circuit working nicely you can do anything you want with it. :)

Steve Conner

Sat May 13 2006, 03:42PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

I've seen Chinese wall warts that do indeed use a step-down blocking oscillator. The only active component seems to be one tiny IGBT. But a design like that takes so much hassle to get working, it's just not worth the effort unless you're going to mass produce it. Why not join your local Switcherholics Anonymous

If you can't cure your addiction to overcomplex hobby projects, read this EDN article:

At the bottom of the page, you'll find the author reverse engineers a $1.50 Nokia phone charger that uses the stepdown blocking oscillator (or ringing choke converter as he calls it)

Marko

Sat May 13 2006, 04:14PM

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

The only active component seems to be one tiny IGBT.

Ooot!! I want some of such chargers!

Cell phone chargers use either blocking-circuit similar to those in ATX'es or more rarely they scitk in some controller IC's (usually on HV side).

Switches are usually small mosfets or bipolar transistors, and I truly never saw an IGBT there

Even the smallest ones would be hard to drive and probably too expensive for the chiense.

I did a lot of searching but I couldn't find any schematics f theese, so best picks are auxiliary ATX SMPS's.

If you just redo it with the same transformer and follow schematic exactly, you can have a very good, cool, few-ten watt power supply for free. Nice to place etc. on a DR/SSTC board to power electronics, if you can rewind the transformer you can get separate supplies maybe for bricks, etc etc...

1 2 next

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