Author Topic: ATX Flyback Transformer  (Read 10089 times)

0 Members and 1 Guest are viewing this topic.

Offline engrguy42Topic starter

  • Frequent Contributor
  • **
  • Posts: 656
  • Country: us
Re: ATX Flyback Transformer
« Reply #50 on: June 22, 2020, 11:03:45 am »
So Tim, it sounds like you're saying that the gate voltage will, due to the gate capacitance, decay as the gate voltage charges the gate capacitance? And that seems to match with the trace below where I plotted gate voltage vs. Vout, and the gate voltage peaks decay until they get around 4V, which I assume is the Vgs threshold?

But I don't follow what you're saying after that, and how to resolve this.

Anyone? 
- 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 T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: ATX Flyback Transformer
« Reply #51 on: June 22, 2020, 11:05:10 am »
Ohhh you're so close!  Gate charge is drawn from the supply.  Repeated charging means current draw.  Gate drive voltage is essentially supply voltage.  What's supply voltage doing then?  And what is the shutdown and startup threshold for supply? :D

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 #52 on: June 22, 2020, 11:17:58 am »
Ahh, so we're gonna play games...  :D

Anyway, if I can ever get the server to respond, here's a plot of Vcc vs Vout. After startup it decays to 4V, which is the point of shutdown. So is this a threshold setting issue or a "don't let the Vcc get so low" issue?
- 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 #53 on: June 22, 2020, 12:44:14 pm »
Okay, well, (fingers crossed) I think I've resolved the minor shutdown/cycling annoyance. Instead of powering Vcc from the incoming bridge rectifier, I added a fixed 9VDC supply to power Vcc. At this point I'm not really interested in the idiosyncracies of a particular controller, but rather getting the basic design principles down. Since I'm trying to model what's actually a CM6800 I'm not gonna spend a lot of time on a science project with a different controller.

In any case, it seems to be working fine. Slammed the 2.4A, 30 watt load and it's fine. I'll let it run for the full 0.5 seconds and see if anything else pops up.
- 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 #54 on: June 22, 2020, 03:14:16 pm »
Dumb question...

On the DC side of a full wave bridge rectifier, if you ground the negative lead (which apparently LTSpice must have or its brain explodes...) that no longer gives you a full wave rectified output, right? Only half wave?

Because after I apply my 2.4 amp load, it maintains perfect 12VDC output for about 25mSec, then decides to start decaying until it drops to around 10.8 volts, and from then on it has ripple between 10.5 and 11 VDC. And the period of the ripple is about 16.667 mSec.

I go to all that trouble of making a full wave bridge, and it ignores 50% of my input waveform. That sucks.
« Last Edit: June 22, 2020, 03:19:33 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 T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: ATX Flyback Transformer
« Reply #55 on: June 22, 2020, 03:31:40 pm »
Well great, the schematic disappeared again.

Well, it only doesn't-work if you ground the AC source.  Which I don't recall being the case, and it was also transformer isolated.

You phrased it like it's 12V DC out of the FWB, which, are you now talking about a completely different circuit?  I don't follow.

Yes, you need a global ground, otherwise inverting the node matrix is equivalent to dividing by zero.  Easiest way to avoid dividing by zero: put a ground somewhere useful, and don't connect capacitors in series with no resistance in parallel with them, and don't connect inductors in parallel with no resistance in series with them.

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 #56 on: June 22, 2020, 03:36:21 pm »
Well great, the schematic disappeared again.

Well, it only doesn't-work if you ground the AC source.  Which I don't recall being the case, and it was also transformer isolated.

You phrased it like it's 12V DC out of the FWB, which, are you now talking about a completely different circuit?  I don't follow.

Yes, you need a global ground, otherwise inverting the node matrix is equivalent to dividing by zero.  Easiest way to avoid dividing by zero: put a ground somewhere useful, and don't connect capacitors in series with no resistance in parallel with them, and don't connect inductors in parallel with no resistance in series with them.

Tim

No, I'm talking about the 12VDC output voltage decaying and showing the ripple. I still have a 160VDC input to this coming out of the FWB, and a separate 9VDC source feeding only the Vcc input to the controller.

For some reason the server is giving me problems again, not allowing me to upload cuz it says the file is too big or took too long to upload.
- 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 gnif

  • Administrator
  • *****
  • Posts: 1716
  • Country: au
  • Views and opinions are my own
    • AMD
Re: ATX Flyback Transformer
« Reply #57 on: June 22, 2020, 04:34:24 pm »
Anyone else notice the image attachments are all screwed up? Sometimes it shows the correct thumbnail, but when you expand it's a different image.  :wtf:  I think the server was having issues yesterday with uploading too.

Everything is loading fine for me and there are no errors reported on the server's end. Likely cache, browser or ISP issue. I actively monitor the server for slowdowns, network congestion, and load issues. Everything is running smoothly with good response times.
« Last Edit: June 22, 2020, 04:36:06 pm by gnif »
 

Offline engrguy42Topic starter

  • Frequent Contributor
  • **
  • Posts: 656
  • Country: us
Re: ATX Flyback Transformer
« Reply #58 on: June 22, 2020, 04:53:32 pm »
Anyone else notice the image attachments are all screwed up? Sometimes it shows the correct thumbnail, but when you expand it's a different image.  :wtf:  I think the server was having issues yesterday with uploading too.

Everything is loading fine for me and there are no errors reported on the server's end. Likely cache, browser or ISP issue. I actively monitor the server for slowdowns, network congestion, and load issues. Everything is running smoothly with good response times.

Cool, thanks.
- 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 T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: ATX Flyback Transformer
« Reply #59 on: June 22, 2020, 05:34:36 pm »
I cleared my cache and it still persists.  Cloudflare cache setting..??

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 #60 on: June 22, 2020, 06:25:41 pm »
Okay, well here's what I have so far.

Dedicated 9VDC connected to the Vcc of the controller. Full 160VDC feeding the switching circuit. No ground on the 120VAC, only on the secondary of the associated isolating 1:1 transformer.

And the 12VDC output is flat at 12VDC for about 20mSec after I switch on the 30watt, 2.4 amp load to the 12VDC. Then it decays, and settles at just under 11 volts with +/- 0.5 volt ripple, with a ripple frequency of 60Hz (not 120Hz as I'd expect for a full wave). BTW, the 160VDC input voltage at Vin looks identical in waveform to the Vout shown in the trace. 
- 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 T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: ATX Flyback Transformer
« Reply #61 on: June 22, 2020, 07:12:10 pm »
Oh, you did have a ground on the AC side. Yeah, that.  Poor bridge, it's shorted out...

Other than that, I'd guess it's hitting current limit, i.e., for the values shown, it's running at the design maximum current, whatever that is.  This is best illustrated by Ith saturating to its maximum 2V, i.e. it's commanding all the current it can.

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 #62 on: June 22, 2020, 07:15:39 pm »
Oh, you did have a ground on the AC side. Yeah, that.  Poor bridge, it's shorted out...

Other than that, I'd guess it's hitting current limit, i.e., for the values shown, it's running at the design maximum current, whatever that is.  This is best illustrated by Ith saturating to its maximum 2V, i.e. it's commanding all the current it can.

Tim

So if I removed the 120VAC ground, why is it still showing 60Hz ripple on the DC??

And what exactly is limiting this? Is it running out of duty cycle? Is it a limitation of the controller?
- 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 #63 on: June 22, 2020, 07:21:59 pm »
Oh wait, I see. If there's a ground on BOTH the AC and DC a diode gets shorted. So no grounds at all on the AC side? I thought I tried that before and LTSpice freaked.

EDIT: Ok, thanks...no AC grounds at all is working great. Maybe that will give me more "oomph" to "oomph" my 12VDC output so it doesn't languish at 11V.  :D
« Last Edit: June 22, 2020, 07:35:30 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 #64 on: June 22, 2020, 10:37:55 pm »
Wow, well I guess I answered my own question again...Looks like this circuit isn't being limited, at least for the load I'm applying. In fact all I did was tweak the primary of the switching transformer and now the 12VDC output under load is flat as a pancake. And the Ith volts are hanging between 1.3 and 1.4 volts, not the 2 volt ceiling.

Previously I had the primary set to 400mH, and secondary at 10mH, giving a turns ratio of a little over 6:1. I then brought that down to 300mH, for a ratio of about 5.5, and as you can see in the attached plot it's doing fine.

And that's one area of this DC switching and the associated transformers that I haven't grasped yet. With AC it's pretty easy and straightforward. But with this DC crap it's about inductance magnitudes and relative magnitudes and stuff. And it's a whole freakin' science project to figure out what transformer parameters to use.

Anyway, that's for another day. At least at this point I have a working forward converter, so now I can go back and tweak things per the actual power supply design.   
« Last Edit: June 22, 2020, 10:39:40 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 nctnico

  • Super Contributor
  • ***
  • Posts: 28429
  • Country: nl
    • NCT Developments
Re: ATX Flyback Transformer
« Reply #65 on: June 22, 2020, 11:13:16 pm »
My advice would be to look a DiodeGoneWild's youtube channel, he has some videos explaining atx power supplies and how they work.
But these videos look like they are from flyback power supplies and not forward converters. With some Googling you can find good tutorials on how forward converters (and thus many of the ATX power supplies) work.
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline engrguy42Topic starter

  • Frequent Contributor
  • **
  • Posts: 656
  • Country: us
Re: ATX Flyback Transformer
« Reply #66 on: June 23, 2020, 12:32:59 am »
DAYUM !!!!

So I thought this power supply I put together was, well, probably kinda wimpy. Yeah, I hit it with a couple amps and it just had a tiny blip then back to flat 12VDC. But I figured if I hit it with more power it would probably explode in a flaming mass of LTSpice metal.

So I figured, "okay, just for grins let's hit it with some real power". So I did a simulation with applying 6 amps/72 watts to the 12VDC output. It probably wouldn't survive the initial dip.

And how did the power supply respond??

It replied "HA !!! I sneer and spit on your measly, wimpy 72 watts !!!"  :D

DAMN, this thing is a MONSTER  :D

A brief hiccup for 10mSec or so, then a nice smooth recovery back to a totally flat 12VDC. And Ith is still hanging below 2 volts after about 0.1 second.

I'll let it run for a while, but it appears like we have a winner. It looks like tuning the switching transformer is a key component in all of this, so now I have to figure out why it worked so well.
- 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 #67 on: June 23, 2020, 10:14:43 am »
And after letting this run overnight, here's the result of a 0.5 second simulation. Some ripple starts at 100mSec after applying load, and it's 120Hz and about 0.25V pk-to-pk.
- 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 #68 on: June 23, 2020, 03:17:13 pm »
Okay, well this is INSANE !!!

This little 12VDC power supply I kludged together is an absolute BEAST !!!

This time I slammed a 12A, 144watt load on the 12VDC output, and it did a nice, smooth recovery, and settled at 11VDC with some ripple.

Damn. This is crazy.
- 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