Small flyback SMPS's / General Science and Electronics / Forums

Forums

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

« Previous topic | Next topic »

Small flyback SMPS's

previous 1 2 3 next

Move Thread

LAN_403

Dr. Dark Current

Tue Apr 24 2007, 11:38AM

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

Steve Ward wrote ...

I have one question for you, with standard flyback drivers the usual flyback spike goes ~10-20 times the supply voltage, that would equate ~3-6kV with off-line operation.

Standard? You design the converter to do whatever voltage you want.

I meant in the simple TV flyback drivers such as 555+mosfet, which actually have no regulation, there the flyback spike goes really high with no load on the flyback

Another question, shouldn't it work without the snubber? If the fet is avalanche rated, shouldn't it avalanche and survive the spikes?
Sorry if my questions seem a little stupid

Marko

Tue Apr 24 2007, 11:48AM

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

Another question, shouldn't it work without the snubber? If the fet is avalanche rated, shouldn't it avalanche and survive the spikes?

Probably, but its just considered *not right* to avalanche the mosfet under any circumstances.

Snubber can also be built to eat up much higher energies that would blow the mosfet.

Still, snubber doesn't see much use in this circuit in normal operation.
It mostly absorbs peak transients wich occur under short-term heavy loads like initial charging of output filter cap.

uzzors2k

Thu Apr 26 2007, 05:45PM

Registered Member #95 Joined: Thu Feb 09 2006, 04:57PM
Location: Norway
Posts: 1308

Exams are done so I can finally have some fun. I have read through both the application notes and the UC3842 datasheet, and I'm ready to build a UC3842 flyback converter. Last part of the design is the switching mosfet. I NEED (would prefer) to use an IRFP450, as I have well over 250 of them. I plan on building a RCD clamp to clamp the voltage to maximun 410V. Can it be done? I only plan on an output power of 25W.
Also I can't find any good formulas for calculating the capacitor and resistor size. Without any formal electronics training, all the technical talk is too much. Does someone have a dumbed-down formula?

I can measure the inductance of the primary and I know the peak current, but stuff like "Reflected Voltage" keeps coming up. What is it?

Btw, I found this in my junk-box.

Fully integrated power MOSFET and PWM. Monitors and ATX' have all sorts of goodies!

Marko

Thu Apr 26 2007, 06:11PM

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

Really, IRFP450 is one insane overkill except maybe for 100W+...

I used a tiny mosfet, IRFU410 has 7ohm Rds, 200pF gate wich could be driven much more easily and faster than IRFP450. And it was still enough and didn't get too hot at 20W.

RCD snubber is generlly hard to design properly, and it will just 'damp' sthe voltage spike rather than clamping it to a particular voltage.

That's why I prefer a TVS or zener, TVS is especially robust if you can afford it.

Be careful about winding orientations; follow the board wich the transformer was mounted on before or test it with a signal generator.
And have good luck!

BTW, I hate those fairchild power switches.
When they die they are impossible to replace.

That's why I choose only the most popular and readily available controllers like UC384*,
TL949, SG/UC352* and etc...

Steve Conner

Thu Apr 26 2007, 09:01PM

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

Reflected voltage is just the output voltage of the supply multiplied by the turns ratio. It is the peak voltage that the switch would see if you neglected leakage inductance in the transformer.

If you consider the transformer as ideal, then clamping the voltage on one winding clamps all of the others too, in the same proportion.

Firkragg, you're not the only person who hates those TopSwitch things :P

uzzors2k

Thu Apr 26 2007, 09:20PM

Registered Member #95 Joined: Thu Feb 09 2006, 04:57PM
Location: Norway
Posts: 1308

Well, I found it easier to just make my own formulas and I made a nice Excel spreadsheet out of it to ease up design. It works by calculating leakage inductance energy, and then with the clamping voltage it calculates the size of capacitor required to store the energy. Based on the duty cycle it further calculates the off time, and then with the capacitance finds the Rclamp value. I set it to 25 RC time constants, what is usual? Watts dissipated in the resistor = power in inductor / discharge-time. Does it look right?

Here is the schematic too, untested but it should work. The primary inductance and therefore the RCD values are based on an estimate. I'll measure the real inductance of the primary once I make the flyback inductor. Does everything look right, no mistakes? I only have two UC3842s in my junkbox, so I can only fail once.

Mod Edit: fixed multi-post issue.

EDIT: I forgot a minimum load! I was thinking something like 290 ohms or less. Also I decided to drop the RC time constants down to 6, which should be enough.

Marko

Fri Apr 27 2007, 08:44PM

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

EDIT: I forgot a minimum load! I was thinking something like 290 ohms or less. Also I decided to drop the RC time constants down to 6, which should be enough.

There is no reason why would you need any kind of 'minimum load'.

If you are building a single output cnverter you should really use an optocoupler to stabilize the voltage.

You can regulate the auxiliary winding voltage and rely on cross-regulation, but I found it to be not closely as good as direct use of an optocoupler. (with multiple outputs, you wouldn't have anything to lose).

With about 5 watts of load I could get the voltage to drop for 0.5..1V, and it seems that resistance of my secondary plays biggest part in it.

Larger transformer with fatter multifilar windings would probably cross-regulate better.

Dr. Shark

Mon Apr 30 2007, 11:48AM

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

I looked into the UC3842 again, did you know that it can be sampled from TI?

I already considered using it about a year ago when I discovered it in almost all of the old CRT monitors I had taken apart. There it is used to control the main (200W +) supply, so it sure is a good chip to use.

The method of deriving the power for starting up the chip by dropping line voltage over a resistor is not nearly as 1337 as the ATX style blocking oscillator, but who cares as long as it works, and I might add, as long as you have an isolation transformer if you need to scope things.

uzzors2k

Mon Apr 30 2007, 02:40PM

Registered Member #95 Joined: Thu Feb 09 2006, 04:57PM
Location: Norway
Posts: 1308

I actually got around to building it today, and woe and behold it worked! For a little while. Snubbing is a major problem and the snubber resistor was smoking in seconds. I ran it without the snubber just for a test run, and within a few seconds the IRFP450 became dangerously hot, so it definitely needs a snubber. Relying on avalanching isn't smart for long term use...
This is the schematic of the current setup.

Am I calculating the RCD clamp values right, or could someone look at my excel sheet? I've put a screen shot in the zip so if you don't have Excel you can still see how it works. I think the calculators are made right, but I can't fill them out properly.

The duty cycle is essential for everything, but how do I know what the max duty will be? My guess is 41.5% (time it takes at 100kHz for current to reach 0.6A through a 2mH load, at which the current sense resistor will trigger current limiting.) But I'm not sure.

Dr. Shark

Mon Apr 30 2007, 03:21PM

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

Do you have a layout for that circuit or did you build it on a breadboard?

I can't really comment on your snubber problem, but have you read the "usual suspects", e.g.

and the UC3842 App notes? It seems they all have this second snubber/clamp thing going from the drain to ground in addition to the drain to Vcc snubber. It might be important to have that!?!

previous 1 2 3 next

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