Lab power supply 0-20V 0-2A

[Harhus]

Hey all

this is the power supply i designed some time ago and built quite a few of them but this time i tought ill transfer it into eagle and build 2x for my self since power supplies are quite popular and i have a lot of vultures around me and im a nice guy and give away those things for free for a spare parts


i can post eagle files but before that ill try to make them 100% accurate if someone wants them now ask for them and ill post

[kam]

Well, i'm in the situation that i need a power supply now (my bench PC PSU was fried) but i'm not sure if i want a linear, a SMPS or a hybrid one... This one is linear right?

[Harhus]

yes it is a linear one but ultra fine and stable if u get multi turn pots and parts are ultra cheap exept for tantals they are like 0.35 cents each and offcourse transformer u can also use PWM to control op amps too instead of pots ( i know u love digital so there is hybrid )



well smps cant be used for audio electronics digitial electronics don't mind. all proffesional power supplyes are linear with digital control and digital error control

[kam]

Yes i know, and i will probably go linear. IO already have a 2x24V 8Amp transformer - that is a whole lot of 380 watts for power to play with. When i say hybrid i mean drop the volts down to a limit (for example 7 volts) with SMPS, and then go with linear for the rest 2 volts to get 5... With your experience what do you think? Would this have a negative result to the output? Should i go totally linear and add some more heat-sinking aluminum and fans?

[Harhus]

well if u wana use fixed output voltage of a smps power supply to get some lower voltage there are no problems.


but i wana name disadvantages and advantages of linear and smps power supplies

first linear:
advantages: lower ripple, almost no electromagnetic interference, very low harmonic frequencies, easy to design, very low chance that ur device will die coz of malfunction of power supply.
disadvantages: Heat generated into the transformer and output transistors, lower efficency since ohms loses, large peak current when turned on

Smps:
advantages: extremly higher efficency, no loses generetarded trou ohms heat, smaller size, lighter
disadvantages: ripple voltage at switching frequency, harmonic frequencies due to switching frequencies, extremly large peak current when turned on, extremly hard to desing a good spms and higher costs, electromagnetic interference which needs to be supressed(extra electronics) , if fails it can destroy stuff if not properly desinged (extra electronics).

bottom line yes SMPS are generaly better but when u want stable voltage and current for some high end supper sensitive stuff u need linear one or EXTREME enginering SMPS one then it becomes matter of cost vs efficency and cost wins linear one for now ...

[cheerio]

i use this one:
http://www.tuxgraphics.org/electronics/201005/bench-power-supply-v3.shtml

little bit customized old version of this. But i use a smps to bring the voltage to <30V. The ripple is pretty high +- 50mv. I need more than 2A so i guess i will redesign it some time this year.
But i do not like the size of the linear stuff. I want a small psu

[kam]

one more question, is it true that smps may not provide complete galvanic isolation with the mains?
Let me be more specific. I know that SMPS can be designed with series coil or flyback. A flyback smps with optocouple feedback is supposed to leave no galvanic connection with the mains. But i've heard that there is always some possibility to mess things up... is it true?

As for the "malfunction".. this is truly something that bothers me a lot with SMPS, especially with buck topologies. It is so easy to fry everything because of a malfunction, that really discourages me to use it.

PLUS (final) check this out:
http://www.pcbheaven.com/wikipages/Transistor_theory/?p=18

go to the bottom of the article with the oscilloscope screenshot. The yellow line measures 10mV p-p. Do you see the noise all around the signal? Do you think that the SMPS from a PC that i used as a power supply could cause this? The frequency of the signal is 1KHz


Cheerio: When you say "this one", you mean the first schematic?

[cheerio]

this one -> http://www.tuxgraphics.org/common/src2/article07071/
schematic should be the attached

it is pretty failsafe.

[Harhus]

well from noise on the picture, either ur power supply is close to the osciloscope and has bad emi suppression, long cabel's which messes up the the switching frequency filtering or just generally bad filtering try making a small LC filter and do the setup again. thats why u need good designed smps to work with analog and they dont come cheap

yes they are  galvnic isolated with flyback and opto but thats not the problem when smps fails it generates high voltage on the switching transistor causing it to short in 98% of the time this usually happens because of the input caps or u can get wild oscilations which i dont know which is worse imagine u are happen to be load when switching transistor that should handle 230AC fails


smps are easy to understand and make bare bone design but if u want good one it becomes complicated because of lots of side things u need to think about

but in a perfect world where price of making smps and linear are the same smps > linear only because of the efficiency

[kam]

i wish to start learning about SMPS, i have already download some pdfs and staff, but i did not have the time to read them yet. After i complete my transistor training, i will go on with SMPS. if you have something interesting to share with us, please attach to the forum. Any theory or design guide would be helpful.


BTW: I think i know where this EMI that you mention comes from. It comes from my PC. I had noticed this noise when i was recording audio for my videos. Some old videos have a slight high frequency in the background which i discovered that it comes from the PC when i approach myself near the cabinet.
Then i decided to put aluminum foil around the cabinet and ground it (the cabinet is made of wood only). Do you think this would help? Any other idea?

How stupid that i did not think of this PC EMI that could be the cause for this bad readings :///

[Harhus]

your idea is fine but that alu foil need to be thicker and with gaps something similar to faraday cage in mobile phones and yeah ground it

u probably know more then enough about transistors to design bare bone smps and refine it afterwards. i personaly if i had to start all over i would probably go into details of high frequency transformer .

then i would move to
http://www.youtube.com/watch?v=V17X-Xzbz-Y&feature=relmfu

then too
http://www.youtube.com/watch?list=PL80DBC78FA0498727&feature=player_detailpage&v=OdyqXQLj9Zk

thing is would like to write theory and stuff but i dont even have digital camera
my lab is hameg HM412-5 that i bought for 5 euros and reparied it for 5 cents and fix was ultra stupid 1 diode for voltage multiplier died that goes to crt and some mastech multimeter i got as a present and whatever i built someone comes and says oh plz i need it so much can u make one for me and im the type of person that if u ask me what time is it i would give u my watch
but there is one good thing, althou my sallery is kinda small coz i just started working at the power plant near my town and i live alone so i have jut enough to eat and pay bills there is 1 awesome thing, working at the power plant is like free electronic store they trow away lots of electronic parts  transformers and such i salwage electronic parts and rewind transformers and i recyckle i get the old burned copper wire give it to the maker of the wire and get 50% discount so i can make stuff which is all i wanted. so yeah it will be some time before i start going so i can spill some theory but NOTHING beats PRACTICE and EXPEREMENTING and what my father used to tell me when he was alive fuck the Engineer who cant build shit he designed. but ill gladly help answering questions and such

[kam]

I already watched the first video, and i think i will go through all the lectures from lecture #1. Excellent videos. Thank you very much.

Regarding your theories, you don't need a video cam to write good theories. I plan as well to write something regarding SMPS, but first i need to gain experience.

[George]

Kam have a look at Texas' design tool

http://webench.ti.com/webench5/power/webench5.cgi

Cheers

[cheerio]

Kam have a look at Texas' design tool

http://webench.ti.com/webench5/power/webench5.cgi

Cheers

so awesome! thx