I've noticed smps's sometimes have a small vertical mounted PCB.. Most likely it's the controller..but not every smps is like this.. (For reference, see pictures of open chassis smps's on ebay.)
My initial guesses are:
1) Better grounding scheme
2) Orientation to dodge electric and magnetic fields from power components
3) Enables the controller PCB to be changed either before or after production
4) Enables single sided PCB
5) Enables shortest connection to power MOS or other
6) Orientation to dodge hot components
Is a vertical mounted controller PCB the best way in a smps? And what is the main reason it's designed like that?
D
Didn't find your answer? Ask the community — no account required.
D
David L. Jones
For single sides boards - yeah, it can help a lot.
Don't think so.
Maybe, that's kinda handy.
Yep, can be a big benefit.
It doesn't hurt.
Don't think so.
As always, it depends.
If you are short on board area for some reason, it's a good choice. If you want future flexibility to change the controller type/package etc it's handy. Might be cheaper if your main PCB is single sided and your controller is some SMD package or something, saves double side loading. Get the main single sided board made cheaply, then the controller can be double sided or whatever and manufactured elsewhere, then simply inserted like a component.
Dave :)
E
Emanuele
D from BC ha scritto:
Another features is that You can make a 70micron solder (or much more) in a only Power board, with the vertical control board in smd. All points You write is good, only for the 2 i don't know but i see al lot of circuit with vertical board controller
Infineon 200W smps reference project for example
Emanuele
*** YOUR ELECTRONICS OPEN SOURCE ***
http://dev.emcelettronica.com;---------------------------------------------------------
Progetti Completi (Full Projects):
[IperKEY-ResKEY] Chiave elettronica a microcontrollore
(Microcontroller Electronics Key and key reader)
//TAGS:Pic, Eeprom, ADconverter, I2C
[IperCODE] Ricevitore RADIOCOMANDI e decodifiche
(Remote Control Receiver and decoder)
//TAGS:Pic, Radio control, routines, rolling code, UART, CCP1, LCD
W
whit3rd
Some parts of a power supply handle high currents, and the copper on the printed wiring board is thick to handle the current. It's cheaper to etch thin copper, and quicker to heat it to solder temperature.
So, a small board with surface mount components on both sides can be IR-reflow-oven soldered, while the large board, components crimped in place on one side only, with thick copper, is wave-soldered.
T
Terry Given
possibly, but probably not. many smps have terrible layouts.
again, possibly but probably not. If the layout is bad, then they clearly werent considering E & H fields.
No. why would you want to do this? The object of the exercise is to build things that work first time....
what it can give you is a standard SMT controller that can be used in various products, and the controller can also be tested before being soldered into the main smps, which can simplify the final tester.
likely. cheaper is better....
it can do.
possibly. usually heat isnt the problem for controllers, it is for the electrolytics, which usually sit right beside the heatsinks!
what it does do is make life much easier wrt voltage creepage and clearance. So if you mean electrically "hot" then Yes!
All the other posts make good points, and:
- often its far cheaper to do discrete SMT than use dedicated controller chips (eg UCC38xx), but this requires more space. My last smps used a Fairchild controller that cost US$1.50 in volume; the one I am now doing (using this vertical PCB approach) costs NZ$0.25 1-off, and every part is multiple-sourced, but it needs about 4x the PCB area.
- there is often a lot of wasted space inside a smps, and going 3-D helps fill it up - minimising waste space gives a smaller package
- heavy current PCBs have minimum track widths and separations that can be incompatible with fine lead pitch SMT. the last 4-Oz PCB I did (1-4-4-1 Oz 4-layer) had a min track width of 15mil, and a min spacing of 18mil per vendors instructions, but we still had problems with shorts (5% failure rate), so I upped the separation to 22mil, and 20k units later that seems to have worked pretty well (5 PCB failures). Some of the fine lead pitch devices I use have a 7.92mil gap between pads (so much for 8/8 design rules), and I dont use the really small stuff!
bear in mind most smps are designed for minimum cost.
Cheers Terry
Join the Discussion
Have something to add? Share your thoughts — no account required.
Didn't find your answer?
Ask the community — no account required
Report Content
You are reporting this content to the moderators. They will look at it
ASAP.