1kW smps project (based on MicrosiM design)

MicrosiM

MicrosiM

Administrator
Staff member
SMPS LAYOUT for best results

Hi luchni


If you really would like to make a new PCB layout, I advice you to inspire from the below layout

In order to get a robust results.

And in order to put the driver in front of the MOSFETS + making the SHORTEST signal path from IR2110 to the G of MOSFET. witch is necessary

Spend some time now, and you will make an excellent LAYOUT

Or you can proceed with that one.

Thank you
 

Attachments

  • DSC00925.jpg
    DSC00925.jpg
    204 KB · Views: 1,581
luchni

luchni

New member
Yes, i thought of something like that, the largest components are layed diagonally across the board, but that would be a completely new design. As i'm using ludo's template, i know it's not perfect, due to the signal lines being long and passing by the power lines etc (little imperfections), but the main thing is that it's working :). Is it possible to add output voltage regulation to this circuit? That would be really nice for most users, so if i do another pcb design (i say maybe because i will be very busy in the next two months) i could squeeze that in there.
 
MicrosiM

MicrosiM

Administrator
Staff member
This SMPS can be made regulated, but that needs a new desing

you can modify ludo layout.
 
V

vikramjoshi

New member
Hello MicrosiM,
For an stereo amp with output of 2 x 275 watts rms at 4 ohms, I suppose this smps will be enough ?
For applications needing only 200 to 300 watts ,will this smps be fine.
 
K

kubeek

New member
I am working on a similar power supply, which should be ready to be changed to a regulated one.
Will post the first schematic and pcb later this week, still waiting for the transformer layout to be done.
 
MicrosiM

MicrosiM

Administrator
Staff member
vikramjoshi said:
Hello MicrosiM,
For an stereo amp with output of 2 x 275 watts rms at 4 ohms, I suppose this smps will be enough ?
For applications needing only 200 to 300 watts ,will this smps be fine.
Click to expand...

This SMPS will power at least 2x 375W Amplifiers. This is guaranteed result

@Kubeek

Thanks for sharing!
 
V

vikramjoshi

New member
Hello Ludo3232,
Will be etching pcb today, do you have bom for this smps.
Smps board size I am getting is Height-5.512 Inches and width-7.698, Is it correct. I need to inverse the print for Dry film process.
 
Last edited:
luchni

luchni

New member
Hi everybody, just to let you know, i came to some setbacks with the pcb, but i will make it right. This will be more time consuming than i thought. The rectifier brigde has been replaced to a square 35A type, not the rectangular one anymore. The components are squeezed a little tightly which got me another 2cm of space on the pcb to play with. The protection circuit is now entirely on the main pcb. I will have a few questions, as i am not very skilled in understanding how this circuit works, but i suspect there may be some mistakes on the current pcb.
 
K

kanon

Member
ee or etd ..?

sorry: For the bad ingles..

well got this in an ATX I have no idea which etd29 .. or etd34.
One more I can calculate with exel file to make it to 160V input and output 13.5V. how much amperius can get from this transformer.
I have 1 horizontal and 1 vertical

tanks.
 

Attachments

  • 108_0558.jpg
    108_0558.jpg
    215.1 KB · Views: 624
MicrosiM

MicrosiM

Administrator
Staff member
kanon said:
sorry: For the bad ingles..

well got this in an ATX I have no idea which etd29 .. or etd34.
One more I can calculate with exel file to make it to 160V input and output 13.5V. how much amperius can get from this transformer.
I have 1 horizontal and 1 vertical

tanks.
Click to expand...

I think thats an EE Core type, you can measure the dimensions and compare to ETD39, ETD29.

Excel calculator cannot calculate unless you have the right informations to be entered into the formula.

have you counted the number of turns when you dissemble that core?

Thats about 400W Max I expect
 
luchni

luchni

New member
while replacing the old rectifier with the larger one, the pcb got a little bigger, but that will be corrected. also the current sense lines are a piece of cable (like ludo's prototype), thinking about moving the ne555 from the protection circuit a little higher (then you have to use cable). have to add the ultra-fast diodes and caps, and move the ETD core more to the right. if you spot some mistakes, please say right away.


 
K

Khron

New member
Just a few ideas:

1) How about trying to reverse the two AC inputs a bit, so the "snakey" one wouldn't have to be so "snakey"? Specifically, the line from (after) the fuse goes to the small trafo, and the other line to the R /relay.
2) Maybe move the small rectifier bridge a couple "clicks" closer to the small trafo? Like, line up the ~/+ pins with the voltage regulator's pins. That way, you could move the capacitors a bit more to the "north" as well.
3) You could move the NE555's timing cap to its "south", so you can push the timer more to the "north".
4) I just noticed, there's less creepage distance between V+ and the upper FET's gate drive network, than around the big cap after the small rectifier bridge. Eagle lets you easily set an "isolation" distance for polygons :)
5) I see a loooooong wire through two jumpers from pin 10 of the SG3525 to the protection circuit. Move the SG and the IR groups more "west", and you can take that long trace more directly, on the "east" of the 3525

I'm working on a PCB for this myself, but it's with a different (larger-surface) stand-by transformer, and i'm putting the pwm section on a daughterboard (and likely, the protection part as well).
 
Last edited:
luchni

luchni

New member
Thank you for those observations, your ideas are welcome. Things from 1) through 5) will all be done. You have a good eye for these things, and i am very glad there is still enthusiasm around this project.
 
K

Khron

New member
Let's just say i've had my fair bit of practice at these things :D And i've developed a sort of "allergy" to (too) long snaking tracks, and 90' corners.

I'm most worried about 4) though. You COULD try to change the packages for those R's / D's with the next step up (i think they're 10mm now, try the 12mm or 15mm pitch versions; you've got those long traces anyway), and it looks like you have room to fatten up the V+ trace (and if you lengthen the R's/D's, you can fatten the GND as well).

Also, you could try extending that "current shunt" wire jumper a bit, and take the current sensor tracks through underneith the ETD. Shorter length again, and gives you space between the SG and the ETD.
 
Hyper2283

Hyper2283

New member
Some quiestions

First of all I want to congratulate you for your hard work. Build this source and I have some questions. I circled parts of the picture attached, which I do not know value resistors and capacitors near the IRFP460. In connection with the voltage capacitors near the IR2110, which is the voltage, and are those who are at high voltage? Protective plate, which is the value of R1 and potentiometer? I appreciate your help. Excuse my English.
 

Attachments

  • Clipboard01.jpg
    Clipboard01.jpg
    317.8 KB · Views: 970
K

Khron

New member
Top left = RC snubbers; those values will have to be determined experimentally, to minimise the "ringing" across the FET's.

And regarding the caps in the mid-section, there's a 7815 regulator in front of those. 16v caps is "pushing the limits", but 25v (and over) caps should be just fine.

This is how far i've got with my board design. Had some business to take care of last week, so haven't done a lot, but i should have some more time now.
 

Attachments

  • Untitled.jpg
    Untitled.jpg
    68.6 KB · Views: 513
Last edited:
You must log in or register to reply here.
Top