Author Topic: ATX Flyback Transformer  (Read 10086 times)

0 Members and 1 Guest are viewing this topic.

Offline engrguy42Topic starter

  • Frequent Contributor
  • **
  • Posts: 656
  • Country: us
ATX Flyback Transformer
« on: June 15, 2020, 10:53:20 pm »
I'm stumped. Pulled a flyback transformer from an ATX power supply, and for the life of me I can't figure out how it's configured. No markings on any of the pins, and only a "Delta" label on top, which is the OEM brand of the power supply. But I assume it's a special made for Delta cuz I can't find ANY spec sheets or references whatsoever. It looks to me like a Wurth with that yellow-ish tape around it. 

The ATX supply provides 12, 5.5, and 3VDC. As you can see from the image, the right side is the primary (with an unconnected center tap). (BTW, I soldered those wires for testing. The xfmr is normally mounted on the PCB). I think the secondaries on the left represent only two windings even though there are 6 pins, since if you look close you can see a couple sets of pins are soldered together. So I'm guessing the secondaries are like a 12VDC plus an aux. winding.

I applied a signal (sine wave) on the primary and get basically zilch on the secondaries. But I'm clueless about flybacks, so I'm not sure what I should get. It's just a transformer with the windings on a common core, right? Anyway, I especially want to figure out the polarities and ratios, but with nothing on the secondary I'm stumped how to figure that out.

Anyone?? 
« Last Edit: June 15, 2020, 11:08:49 pm by engrguy42 »
- The best engineers know enough to realize they don't know nuthin'...
- Those who agree with you can do no wrong. Those who disagree can do no right.
- I'm always amazed at how many people "already knew that" after you explain it to them in detail...
 

Offline engrguy42Topic starter

  • Frequent Contributor
  • **
  • Posts: 656
  • Country: us
Re: ATX Flyback Transformer
« Reply #1 on: June 15, 2020, 11:55:27 pm »
Nevermind...

I dunno, cockpit error or something. Got some nice clean sine waves on the secondaries now. The ratio for both windings is is 12.5:1, and I've got the polarities sorted by superimposing primary and secondary voltages on the scopely.  |O

« Last Edit: June 15, 2020, 11:57:07 pm by engrguy42 »
- The best engineers know enough to realize they don't know nuthin'...
- Those who agree with you can do no wrong. Those who disagree can do no right.
- I'm always amazed at how many people "already knew that" after you explain it to them in detail...
 

Offline engrguy42Topic starter

  • Frequent Contributor
  • **
  • Posts: 656
  • Country: us
Re: ATX Flyback Transformer
« Reply #2 on: June 16, 2020, 12:05:18 am »
Ahhh...okay. The cockpit error was that it doesn't like low frequencies. I had the waveform generator set for like 60Hz, but you need to get up past around 1kHz to get anything out of the secondary.

I imagine that will make sense if I think about it a lot...  :D
- The best engineers know enough to realize they don't know nuthin'...
- Those who agree with you can do no wrong. Those who disagree can do no right.
- I'm always amazed at how many people "already knew that" after you explain it to them in detail...
 

Offline wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
Re: ATX Flyback Transformer
« Reply #3 on: June 16, 2020, 12:18:46 am »
It's not even a flyback transformer. PSU most likely uses forward or some type of bridge topology.
Quote
Wurth with that yellow-ish tape around it.
Vast majority of SMPS transformers use yellow tape. And there is 99.9% chance it's a custom part.
 

Offline NiHaoMike

  • Super Contributor
  • ***
  • Posts: 9323
  • Country: us
  • "Don't turn it on - Take it apart!"
    • Facebook Page
Re: ATX Flyback Transformer
« Reply #4 on: June 16, 2020, 01:29:49 am »
Some small ATX PSUs (less than 150W) use the flyback topology.
Cryptocurrency has taught me to love math and at the same time be baffled by it.

Cryptocurrency lesson 0: Altcoins and Bitcoin are not the same thing.
 

Offline bin_liu

  • Contributor
  • Posts: 38
  • Country: cn
Re: ATX Flyback Transformer
« Reply #5 on: June 16, 2020, 01:38:22 am »
PC power supply is half-bridge or LLC topology.
transformer is customized.
 

Offline wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
Re: ATX Flyback Transformer
« Reply #6 on: June 16, 2020, 01:42:31 am »
Some small ATX PSUs (less than 150W) use the flyback topology.
Doubt that. Usually you won't find it even in non-knockoff laptop PSU. Flyback is no good for group voltage regulation. The only place you likely will find it is 5V standby PSU.
« Last Edit: June 16, 2020, 03:13:50 am by wraper »
 

Offline TimNJ

  • Super Contributor
  • ***
  • Posts: 1720
  • Country: us
Re: ATX Flyback Transformer
« Reply #7 on: June 16, 2020, 02:12:00 am »
FYI, I am fairly confident Wurth doesn't make much of anything. I believe they are mostly acting as a middle-man between Chinese magnetics suppliers and the Western markets.

Any SMPS manufacturer with reasonable volumes has either an in-house magnetics department and/or contracts with an outside vendor for custom designs. Magnetics are too important to try to make an off-the-shelf offering work.
 

Offline TimNJ

  • Super Contributor
  • ***
  • Posts: 1720
  • Country: us
Re: ATX Flyback Transformer
« Reply #8 on: June 16, 2020, 02:23:46 am »
PC power supply is half-bridge or LLC topology.
transformer is customized.

I think the majority of laptop PSUs are still flyback, but most laptop PSUs are <90W (hardly anything over 120W these days). Most laptops are shipping with a 65W adapter or less, as far as I cant tell.

90W is on the edge of where you might consider using resonant LLC with respect to price/size/EMI. Interestingly, high powered USB PD power supplies (90-100W) basically need to be flyback because of the wide-output voltage range requirement (3-20V). LLC struggles if operating too far away from its design point. So, there's now renewed interest in flyback at these power levels.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: ATX Flyback Transformer
« Reply #9 on: June 16, 2020, 03:37:09 am »
Doubtful it's a full wave forward converter: that would require a FWB on the secondary, and screw that at those voltages and currents.  It would be FWCT.

I was looking for the common ground lead in the second picture, but there it is, none to be found.  So it's not FWCT!

Seems it's a new power supply.  "ATX" covers a huge range of history by now.  They used to be BJT half bridge inverter, forward converters.  Then MOSFET, and now they seem to be largely 2-switch forward or flyback.  The above seems to be the latter.  Whether it's forward or flyback depends on the magnetizing inductance (high or low respectively), and whether the original application had a secondary series filter choke or not, obviously.

It's also consistent with an ATX supply in that there will be an aux supply for primary control power, optionally secondary control power, and motherboard standby.  No need for an aux winding on the main power transformer, hence the sparse primary side pinout.

Active clamp flyback is also pretty popular, though I'm not sure how common it is among offline supplies, and it seems to give awful efficiency at light loads?  At least that seems to be what I've observed from a few 1/nth-brick DC-DC modules I've seen.

The 2-switch configuration seems to be popular with active PFC, probably part of the motivation behind it; straight flyback or forward (half wave) would require very high voltages, making trouble with transistor selection, trace clearance and EMI.  Half bridge should still be attractive though, with LLC resonant being an option; perhaps the reduced EMI of 2-switch is worth the change after all.  (Or maybe it makes no difference because you just can't time the switches accurately enough to get balanced primary voltages.)

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline engrguy42Topic starter

  • Frequent Contributor
  • **
  • Posts: 656
  • Country: us
Re: ATX Flyback Transformer
« Reply #10 on: June 16, 2020, 10:29:14 am »
Don't ya just love how these threads always seem to wander off into "I'm smarter than you" discussions of irrelevant pseudo-factiods?  :D

Geez people. The ultimate manufacturer's name, wrapping tape color and perfectly accurate transformer type are irrelevant. At some point I'll probably reverse engineer this thing to find out how it's configured, but at this point it's irrelevant since I'm looking only for the transformer characteristics. 

So, back to the question of the thread...

I think the reason why my coupled inductor/flyback transformer/whatever the f*** you want to call it is giving no secondary voltage output at low frequencies is due to the fact that the waveform generator output impedance is 50 ohms, and the inductive reactance of the inductor drops below around 1kHz. And around 400Hz the measured volts drops to about 50%, which implies that it's like a 20mH. Cool, so now I can model it.

Though I'm sure everyone already knew that...  :D
- The best engineers know enough to realize they don't know nuthin'...
- Those who agree with you can do no wrong. Those who disagree can do no right.
- I'm always amazed at how many people "already knew that" after you explain it to them in detail...
 

Offline wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
Re: ATX Flyback Transformer
« Reply #11 on: June 16, 2020, 10:37:04 am »
wrapping tape color and perfectly accurate transformer type are irrelevant. At some point I'll probably reverse engineer this thing to find out how it's configured, but at this point it's irrelevant since I'm looking only for the transformer characteristics. 
It's very relevant. Because it directly affects how transformer was made. You cannot simply take it and use for different type of converter. Say flyback transformer is not suitable for bridge or forward topologies, even if you rewind it.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: ATX Flyback Transformer
« Reply #12 on: June 16, 2020, 10:55:21 am »
Close -- you want the 70.7% point, that's equal impedance magnitudes.  At 50%, it's 1/sqrt(3), or 50/sqrt(3) ohms, or 11mH.

That's pretty high so I'm thinking this board had filter chokes on the secondary side, i.e. it's a forward converter. :-+

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline engrguy42Topic starter

  • Frequent Contributor
  • **
  • Posts: 656
  • Country: us
Re: ATX Flyback Transformer
« Reply #13 on: June 16, 2020, 02:10:20 pm »
Well after a bit of research I think I found a schematic for a very similar 450watt ATX power supply, and it appears that the two secondaries of this transformer are for the 12VDC and 5VDC, and they tap off the 5VDC to get the 3VDC. And a coupled inductor between the 12, -12, and 5VDC.   :wtf: 

FWIW, there's also a separate transformer listed as a flyback converter for the bias and 5VDC standby. And a 3rd transformer for the gate drive.

Anyway, as you can see the secondaries of this transformer have a crap-ton of Schottky's on a big heat sink. Anyone know what those are for?

These things are so freakin' complicated... 
- The best engineers know enough to realize they don't know nuthin'...
- Those who agree with you can do no wrong. Those who disagree can do no right.
- I'm always amazed at how many people "already knew that" after you explain it to them in detail...
 

Offline Buriedcode

  • Super Contributor
  • ***
  • Posts: 1718
  • Country: gb
Re: ATX Flyback Transformer
« Reply #14 on: June 16, 2020, 02:19:11 pm »
Anyway, as you can see the secondaries of this transformer have a crap-ton of Schottky's on a big heat sink. Anyone know what those are for?

These things are so freakin' complicated...

Maybe start at the basics of the topology.  The fact you used 60Hz on a ferrite transformer shows you've made assumptions that aren't true (and worked that out!).
 

Offline engrguy42Topic starter

  • Frequent Contributor
  • **
  • Posts: 656
  • Country: us
Re: ATX Flyback Transformer
« Reply #15 on: June 16, 2020, 02:54:47 pm »
Anyway, as you can see the secondaries of this transformer have a crap-ton of Schottky's on a big heat sink. Anyone know what those are for?

These things are so freakin' complicated...

Maybe start at the basics of the topology.  The fact you used 60Hz on a ferrite transformer shows you've made assumptions that aren't true (and worked that out!).

 :-//

 :wtf:

Anyway, these Schottky's are made by STMicro, and I found a nice paper from them describing their use in this case.
« Last Edit: June 16, 2020, 02:59:13 pm by engrguy42 »
- The best engineers know enough to realize they don't know nuthin'...
- Those who agree with you can do no wrong. Those who disagree can do no right.
- I'm always amazed at how many people "already knew that" after you explain it to them in detail...
 

Offline engrguy42Topic starter

  • Frequent Contributor
  • **
  • Posts: 656
  • Country: us
Re: ATX Flyback Transformer
« Reply #16 on: June 16, 2020, 03:52:13 pm »
OOooooo....some yummy LTSpice SMPS transformer switching goodness.... :D
- The best engineers know enough to realize they don't know nuthin'...
- Those who agree with you can do no wrong. Those who disagree can do no right.
- I'm always amazed at how many people "already knew that" after you explain it to them in detail...
 

Offline mariush

  • Super Contributor
  • ***
  • Posts: 5171
  • Country: ro
  • .
Re: ATX Flyback Transformer
« Reply #17 on: June 16, 2020, 04:35:13 pm »
My advice would be to look a DiodeGoneWild's youtube channel, he has some videos explaining atx power supplies and how they work.












 
The following users thanked this post: orolo

Offline engrguy42Topic starter

  • Frequent Contributor
  • **
  • Posts: 656
  • Country: us
Re: ATX Flyback Transformer
« Reply #18 on: June 16, 2020, 06:14:11 pm »
Thanks, mariush. But for me personally, if I'm going to really understand something as complicated as a decent ATX power supply, a 10 minute video isn't going to do it.

My long term plan is to take an actual spare power supply and, component by component, test it and characterize it, then model it in LTSpice. I already have the front end modelled (input filtering, bridge, etc.), and next I'm trying to tackle the main transformer.

IMO, it's so much better to have an actual software simulation model of actual hardware that you can try things out on and do simulations and "what-if's" and test on the bench. And in this case, I already have a schematic that seems to be almost identical to the actual hardware, so it's a whole lot easier to reverse engineer.

BTW, if anyone really wants to understand down on the engineering level how a lot of this stuff works, you might consider the videos of Sam Ben-Yaakov. He's an EE professor (he also writes tons of IEEE papers) with some really in-depth explanations of this stuff.

Anyway, I guess I don't want to be a "10 minute video expert". There are more than enough of those out there.  :D
- The best engineers know enough to realize they don't know nuthin'...
- Those who agree with you can do no wrong. Those who disagree can do no right.
- I'm always amazed at how many people "already knew that" after you explain it to them in detail...
 

Offline Gyro

  • Super Contributor
  • ***
  • Posts: 10174
  • Country: gb
Re: ATX Flyback Transformer
« Reply #19 on: June 16, 2020, 06:23:53 pm »
Thanks, mariush. But for me personally, if I'm going to really understand something as complicated as a decent ATX power supply, a 10 minute video isn't going to do it.

My long term plan is to take an actual spare power supply and, component by component, test it and characterize it, then model it in LTSpice. I already have the front end modelled (input filtering, bridge, etc.), and next I'm trying to tackle the main transformer.

IMO, it's so much better to have an actual software simulation model of actual hardware that you can try things out on and do simulations and "what-if's" and test on the bench. And in this case, I already have a schematic that seems to be almost identical to the actual hardware, so it's a whole lot easier to reverse engineer.

BTW, if anyone really wants to understand down on the engineering level how a lot of this stuff works, you might consider the videos of Sam Ben-Yaakov. He's an EE professor (he also writes tons of IEEE papers) with some really in-depth explanations of this stuff.

Anyway, I guess I don't want to be a "10 minute video expert". There are more than enough of those out there.  :D

Sounds like you need to stop asking basic questions and go and get on with your modelling then.  :)
Best Regards, Chris
 

Offline engrguy42Topic starter

  • Frequent Contributor
  • **
  • Posts: 656
  • Country: us
Re: ATX Flyback Transformer
« Reply #20 on: June 16, 2020, 06:29:03 pm »
Thanks, mariush. But for me personally, if I'm going to really understand something as complicated as a decent ATX power supply, a 10 minute video isn't going to do it.

My long term plan is to take an actual spare power supply and, component by component, test it and characterize it, then model it in LTSpice. I already have the front end modelled (input filtering, bridge, etc.), and next I'm trying to tackle the main transformer.

IMO, it's so much better to have an actual software simulation model of actual hardware that you can try things out on and do simulations and "what-if's" and test on the bench. And in this case, I already have a schematic that seems to be almost identical to the actual hardware, so it's a whole lot easier to reverse engineer.

BTW, if anyone really wants to understand down on the engineering level how a lot of this stuff works, you might consider the videos of Sam Ben-Yaakov. He's an EE professor (he also writes tons of IEEE papers) with some really in-depth explanations of this stuff.

Anyway, I guess I don't want to be a "10 minute video expert". There are more than enough of those out there.  :D

Sounds like you need to stop asking basic questions and go and get on with your modelling then.  :)

I know, right? Dumb me, thought I could get some answers here to "basic" questions from the "experts".  |O
- The best engineers know enough to realize they don't know nuthin'...
- Those who agree with you can do no wrong. Those who disagree can do no right.
- I'm always amazed at how many people "already knew that" after you explain it to them in detail...
 

Offline Gyro

  • Super Contributor
  • ***
  • Posts: 10174
  • Country: gb
Re: ATX Flyback Transformer
« Reply #21 on: June 16, 2020, 06:33:24 pm »
Well what are the questions? Trawling back through the thread and filtering the arrogance, the only actual question I could find is 'why didn't my ferrite cored SMPS transformer produce any significant output at 60Hz?' and you answered that one yourself.  :-//

Apart from that, you've just snubbed the replies that anyone's submitted.
« Last Edit: June 16, 2020, 06:38:52 pm by Gyro »
Best Regards, Chris
 
The following users thanked this post: wraper

Offline engrguy42Topic starter

  • Frequent Contributor
  • **
  • Posts: 656
  • Country: us
Re: ATX Flyback Transformer
« Reply #22 on: June 16, 2020, 07:15:15 pm »
Well what are the questions? Trawling back through the thread and filtering the arrogance, the only actual question I could find is 'why didn't my ferrite cored SMPS transformer produce any significant output at 60Hz?' and you answered that one yourself.  :-//

Apart from that, you've just snubbed the replies that anyone's submitted.

So do you have any relevant technical input for the issue, or are you just here to complain? Because I'm still open to any relevant info on the characteristics and considerations regarding this transformer and how I can model and characterize it. I mean, I already gave a ton of info, including a schematic showing the actual circuit it's in. Why can't we discuss that, rather than go off on irrelevant tangents discussing who the actual manufacturer might be and what's the technically accurate identifier for this transformer?
- The best engineers know enough to realize they don't know nuthin'...
- Those who agree with you can do no wrong. Those who disagree can do no right.
- I'm always amazed at how many people "already knew that" after you explain it to them in detail...
 

Offline Gyro

  • Super Contributor
  • ***
  • Posts: 10174
  • Country: gb
Re: ATX Flyback Transformer
« Reply #23 on: June 16, 2020, 07:25:28 pm »
Sorry, I provide help to lots of people, your responses are too much of a turn-off to be worth the effort though.

You didn't clarify what your questions are by the way. A theoretical and practical text on magnetics and transformer design sounds like your best 'authoritative' option though.


P.S. Member treez might be able to help you, he's done a fair bit of SMPS.
« Last Edit: June 16, 2020, 07:28:37 pm by Gyro »
Best Regards, Chris
 
The following users thanked this post: HackedFridgeMagnet, wraper

Offline engrguy42Topic starter

  • Frequent Contributor
  • **
  • Posts: 656
  • Country: us
Re: ATX Flyback Transformer
« Reply #24 on: June 16, 2020, 09:48:26 pm »
Sorry, I provide help to lots of people, your responses are too much of a turn-off to be worth the effort though.

You didn't clarify what your questions are by the way. A theoretical and practical text on magnetics and transformer design sounds like your best 'authoritative' option though.


P.S. Member treez might be able to help you, he's done a fair bit of SMPS.

"Read a book" or "watch a video" is of zero help. That shit is beyond obvious. If you have nothing to add because you don't understand the subject then fine. stay out of the discussion.

Geez you guys give me a headache.
- The best engineers know enough to realize they don't know nuthin'...
- Those who agree with you can do no wrong. Those who disagree can do no right.
- I'm always amazed at how many people "already knew that" after you explain it to them in detail...
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf