Author Topic: Switched power supply project  (Read 1090 times)

0 Members and 1 Guest are viewing this topic.

Offline okestTopic starter

  • Newbie
  • Posts: 3
  • Country: ee
Switched power supply project
« on: April 17, 2018, 05:25:16 pm »
Hello,

I'm making an SMPS as an university project. I've picked up a few good books but they are rather dated: "Power supply cookbook"(Also i've found version 2 online in PDF) and "Practical switching power supply design". Those are the only good ones i could find universities library but they are rather dated.
Basically the supply is going to be driving guitar audio amplifier heater elements(I wanted to make an SMPS to learn more about them and it just happened my tutor needed one for this purpose).
My nominals would be an input of 230VAC 50Hz. And the output would be 6.3VDC and max 8A(Probably going to be around 6A.)
Now the book lead me to believe my best bet would be flyback in discontinous mode due to my output power being in the 50-60W area(Less than 150W). Is that good for this purpose or should i look into some other topology? Keep in mind i'm not that familiar with SMPS and i'm using this project to learn. I understand the main principle of flyback.
The calculations i have made so far are(Assuming 80% efficency):
Pout=6.3V*8A=50.4W
Pin=50.4/0.8=63W
Vin(MIN)=207V*1,414=293VDC
Vin(MAX)=253V*1,414=358VDC
Iin(MAX)=63/292=0,22 ADC
Iin(MIN)=63/358=0.17 ADC
Ipeak=k*Pout/Vin(MIN)=5.5*50,4/292=0.95A       [where K=5.5 constant for flyback in the book for rough estimates]
Lpri=Vin(min)*dutycycle/Ipeak*f  = 293*0.5/0.95*50000=3084uH    (Seems rather on the high side?)
Now what i'm stuck at is picking the core. The book says to use EE core with core each side 47mm[Book is from 1994]. But searching around i feel like today there are better cores and a better way to pick a core? I have searched around the internet but there are different equations from site to site. Could someone elaborate a bit and/or point to a good source?

Again using another site there was Ipeak formula of Ip=2Pout/Vin(min)*f  which would give me 0.69A peak current. Also there was formula of Lprimary=Vin(min)*dutycycle/Ipk*f which would give me 4246mH. Which one is more correct to use?
« Last Edit: April 17, 2018, 05:42:02 pm by okest »
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21674
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Switched power supply project
« Reply #1 on: April 17, 2018, 10:22:47 pm »
This is handy:
http://schmidt-walter-schaltnetzteile.de/smps_e/smps_e.html

Note that it defaults to CCS (small k), just enter a different inductance and turns ratio manually to operate differently.

You may also find a forward converter is handy, just because of the low output voltage and high peak current.  This can also operate in, or near, DCM, where peak current mode control is still easy, and where peak output current is lessened thanks to the filter choke.

The concern with peak current comes from the quality of the capacitors.  Low-ESR capacitors are readily available today (you can choose aluminum polymer types if you want -- these have comparable quality to film capacitors, but operate at much lower voltages!), so it's not such a big deal.  Just be careful if you're doing it with salvaged parts, or cheap parts of unknown rating -- you may find far more output voltage ripple, and much more capacitor temperature rise, than you were expecting! :)

FYI, be careful with mains inputs.  Use an isolation transformer, of course, but also start small -- especially if you don't have any experience in this yet!  Make it for 12 or 24V input first, and make it work well.  You will learn every lesson you need, there.  Then you can increase the input voltage.

Set goals for yourself.  Create stages in your project, with clear expectations and outcomes.  Don't make it too rigid -- you don't know what circuit changes may be needed, or if you can satisfy certain criteria -- specifications may need to change as you realize what their consequences are.

And, don't neglect EMI checks.  Read up on how and why testing is performed, and obtain a LISN (you can build one from junk without too much trouble) to do basic testing.  It's not going to be regulatory grade, but that's okay, you don't have a test chamber and calibrated antennas anyway.  Just a precompliance test.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 
The following users thanked this post: okest

Offline okestTopic starter

  • Newbie
  • Posts: 3
  • Country: ee
Re: Switched power supply project
« Reply #2 on: April 19, 2018, 05:06:20 pm »
As not to get overwhelmed with this project, i'm putting flyback on the hold since i have been learning theres much to consider and i really don't have time right now plus the expertise.
So i'm going to put this project on hold and change the parameters.
Im planning on getting dedicated 12V 7A DC supply from the store/using PC ATX supply and switching it to 6.3V 8A with buck converter.
What kind of IC's would you recommend to drive mosfets? I was looking at all kinds of different regulators starting from converters to controllers.
I would have loved to use LM25085 but the package is so tiny(VSSOP) and also for my power ratings i think even that won't do and would require MSOP-Powerpad technology. Those packages are so small and seem unrealistic to test.
I don't mind losing a bit on efficency and the size of components if that would give me much better conditions to solder this thing together.
I feel like all the recent/listed IC's are meant for very small boards. Does someone have any good ideas what to use to control the buck converter?
« Last Edit: April 19, 2018, 05:08:05 pm by okest »
 

Offline strawberry

  • Super Contributor
  • ***
  • Posts: 1161
  • Country: lv
Re: Switched power supply project
« Reply #3 on: April 20, 2018, 09:05:45 am »
https://www.power.com/
There are many companies providing single chip solutions for all kind topologies
For 50W discrete SMPS is pointless to go below 90% eff

CV/CC synchronous buck converter IR2104S  TL494
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf