UC3842 basics

[ben7]

The schematic.

Note: lol, the 100n cap marked 'C1' should be 'C8'

[ben7]

I tried using a different transformer, but I couldn't get it to work.
Its markings are:
22L-T07
TDK X 93®

Some measurements:
top to bottom of core : ~30mm
side to side of core : ~22.5mm
core thickness : ~6mm
primary side pins : 4
primary side windings : 2
secondary side pins : 5
secondary side windings : ?
insulation tape : black/white

Any ideas?

This was from a VCR psu (I think). Also, I soldered the wires on there, as the pins didn't fit in the breadboard.

[budm]

That I sense feedback needs just enough filter so it will not response to the noise and when the MOSFET first starting and causes the circuits to go into shutdown. Same thing for the opto feedback circuits.

[ben7]

Originally Posted by budm

That I sense feedback needs just enough filter so it will not response to the noise and when the MOSFET first starting and causes the circuits to go into shutdown. Same thing for the opto feedback circuits.

Tried different caps in feedback, no change. Tried different filter caps, no change. The only thing I haven't tried is to change the switching frequency.

[budm]

Oh, The feedback I am talking about is not related to this new transformer you are trying to get it to work. Since you have two primary windings, one of them will be for the drive winding, the other will be for supplying the DC to the SMPS IC once the circuits start running on its own.

[ben7]

Originally Posted by budm

Oh, The feedback I am talking about is not related to this new transformer you are trying to get it to work. Since you have two primary windings, one of them will be for the drive winding, the other will be for supplying the DC to the SMPS IC once the circuits start running on its own.

Yeah, I got the primary and AUX windings sorted out.

Oh, I think you have gotten confused, this circuit does work with a regular 5vsb transformer! I just tested it and it still works. It doesn't seem to like this transformer though :/

[ben7]

Hm, well, I have powered the 3844 with an external source, and it is working.

How could I get the AUX winding to supply a higher voltage without re-winding the transformer? could I use a voltage doubler?

EDIT: Um, yeah, my DMM says the AUX winding is giving out about 70mV! Does this mean the transformer is toast?!

[ben7]

Thats interesting, why can't I edit my posts in this thread anymore?

[ben7]

Could a transformer used in a quasi-resonant smps work in flyback mode? I think I found the problem, the TDK transformer was in an smps with a STRW6765 - a quasi-resonant smps ic.

[ben7]

I guess nobody else wants to contribute to this thread?

[rogfanther]

Well, I still could not find some uninterrupted time for building my first try in this, but expect to be able to in the next week(s). As you, I believe I will start with a 5VSB transformer, due to it having reasonable turn proportions already in place, later moving to rewound transformers for different voltages

[ben7]

Originally Posted by rogfanther

Well, I still could not find some uninterrupted time for building my first try in this, but expect to be able to in the next week(s). As you, I believe I will start with a 5VSB transformer, due to it having reasonable turn proportions already in place, later moving to rewound transformers for different voltages

Ah great!

I want to make my own transformers too, but to use OP's transformer design spreadsheet, I need windows office

I have at least two dozen transformers for experimenting with.

The Pirate Bay, here I come!

[tom66]

Originally Posted by ben7

Ah great!

I want to make my own transformers too, but to use OP's transformer design spreadsheet, I need windows office

I have at least two dozen transformers for experimenting with.

The Pirate Bay, here I come!

It worked fine for me with OpenOffice.org Math / LibreOffice Math.

[ben7]

I have a question.

Could the chip be used as a variable PWM source, by changing the voltage on the VFB pin, and yet still have a current limit?

[Th3_uN1Qu3]

It's not designed to be used that way. However, it is possible if you make the source resistor low enough. Make it so in normal operation the voltage feedback alone limits the pulse width, and the Isense threshold isn't reached.

I know i haven't been contributing to this lately, but there's been a whole lot of stuff coming in to be fixed, and i've hardly had time for my own projects.

I haven't been slacking off tho - i've been sorting all my parts nicely to have them handy, and i'm gonna have a bunch of stuff coming up. I'm kinda running on my spare tire tho due to the lack of computing hardware since my dv9000 finally bucked the kicket a couple months ago (or rather, i got fed up of fixing it all the time), the DV5 is used by my mom most of the time, and there's a lot of stuff that doesn't run on this little ibook i have now, most notably i don't have PCB design software and LTSpice...

It ain't all bad tho, as i'm more inclined to experiment with the physical parts instead of mucking around on the computer.

I recall i'd blown the 3842 in the little SMPS i built - i'll have to check, as i have like a dozen of them waiting in a ziploc bag hanging from a screw above my bench

Edit: That perfboard is a real messy piece of work... That's why i got lazy and never swapped out the blown 3842. That's it - from now on i'm making a PCB for every SMPS project i do... Oh wait, i'm back to square one - no PCB design software running on this thing. Hmm... i have to whip up a PC together, quickly.

[ben7]

Originally Posted by Th3_uN1Qu3

It's not designed to be used that way. However, it is possible if you make the source resistor low enough. Make it so in normal operation the voltage feedback alone limits the pulse width, and the Isense threshold isn't reached.

Source resistor? Which resistor is that?

Its not going to use a transformer, it is for a buck circuit (single inductor).
No, it isn't going to run off mains voltage either

[Th3_uN1Qu3]

The resistor in the source of the switching MOSFET. Anyway... UC3842 is designed for mains-powered applications as a primary side controller in a transformer-isolated power supply. The current sense resistor needs to be between the source of the power switch and ground, so it won't work for a DC-DC buck.

For a dc-dc buck converter there are other chips to use, like the MC34063, which is very easy to implement.

[kaboom]

Two schematics for you.

First, the 'normal' Panasonic VCR PS, as well as an example ckt with a STRW6765.

Maybe there's something in those that can help.


-Paul

[sam_sam_sam]

If someone get this project to work I would like to buy a board and make a 5 volt @ 5 or amp more power suppply

So I hope some will get one working I will keep my eye on this post

[ben7]

I would like to build one too, but just for the heck of it. I havent gotten bored of circuit design yet

Plus, I have tons more to learn

I fried my only UC384x yesterday, by a stupid mistake
Will have to put that on my parts list for next order...