Maker Pro
Maker Pro

SMPS Build...

medictrode

Feb 21, 2013
16
Joined
Feb 21, 2013
Messages
16
Hello all...:)

I have a collection of power supplies on my bench, both simple commercial ones and homebrew Linear supplies based on 78xx or LM317 adjustable. They all work well, but for reasons of curiosity I have really been wanting to build a switching power supply...:cool:

So I have found many IC controllers and not sure where to start. I have a 24v DC Unregulated input and would like a 12-20v adjustable output, in the 3-5 amp range.

Any thoughts as to what components I Should try out? Any help would be super appreciated as there are so many ways to go.. Thank you all..;)
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
Linear Technology and National Semiconductor (now owned by Texas Instruments) both make a wide range of buck regulators. Many are self-contained in TO-220-style packages with five leads. In my experience LT devices are easier to use and operate more cleanly but are more expensive. Ratings up to 5A are no problem. There is no isolation.
 

medictrode

Feb 21, 2013
16
Joined
Feb 21, 2013
Messages
16
Thank you for the reply, I will check into some of those chips.

It's a lot tougher because there is so much info on linear regs out there, but still limited on DIY switching units. I'm sure I will figure it out eventually..:eek:
 

GonzoEngineer

Dec 2, 2011
321
Joined
Dec 2, 2011
Messages
321
Linear Technology and National Semiconductor (now owned by Texas Instruments) both make a wide range of buck regulators. Many are self-contained in TO-220-style packages with five leads. In my experience LT devices are easier to use and operate more cleanly but are more expensive. Ratings up to 5A are no problem. There is no isolation.

I have to agree with regards to Linear Technology. When they show a schematic in the Data Sheet, you can be assured it will work every time.

On the other hand, I find that the schematics from National Semiconductor are merely a suggestion......:rolleyes:
 

medictrode

Feb 21, 2013
16
Joined
Feb 21, 2013
Messages
16
Thank you all for the replys, much appreciated.

So I received one of the cheap Switching supplies. It was decent for the 3 bucks, and only had to re solder a few joints. It works well, and I was able to drive a TO-3 cased transistor to boost the current capacity. It regulated the voltage nicely.. :D

But then, as I am famous for, I made my life complicated. Someone gave me an IGBT module(CM1200HA-24J) and got the bright idea of trying to regulate it and seeing what kind of power it will handle. I handled it like a regular bi-polar transistor but it won't regulate so well. I have 24volts DC going into the collector, and a load tester on the emitter. The gate is driven by either a linear type of regulator or the switch mode I acquired. My thoughts were the fact that the IGBT has a high impedance and is so easy to drive, the regulator doesn't see any voltage drop on the emitter..:confused:

Maybe some sort of feedback circuit is needed? Or I am completely off on this one.. Hey it's just for fun. Can't learn anything unless you try..:) Thanks again all..:cool:
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
I don't think IGBTs are really intended to operate in the linear mode. They're always used for fast switching in any practical designs that I've seen. Check the continuous power dissipation figures on the one you have. You may be disappointed.

A linear regulator using an active device in the emitter follower (or equivalent) configuration can be done with or without feedback. If you need good regulation, use feedback. But generally the pass element is something big with a high power dissipation limit, such as a 2N3055. And a big heatsink.
 

medictrode

Feb 21, 2013
16
Joined
Feb 21, 2013
Messages
16
Thank you for the quick reply.

Yeah you are probably right. Looking at the applications of these modules it doesn't appear they do anything linear in nature. But hey it was free, fun to mess around with a neat looking device anyway. :). Plus I don't have anything that pulls that much current in the first place. Just all for fun.

I do have a few of the 2N5686 transistors that I want to experiment with.
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
Wow! 50A collector current, 300W dissipation, perfect for a linear power supply. Make sure you can drive the base hard. Guaranteed current gain is only 15 at 10A collector current!
Have a careful read of the data sheet - the VCEsat and leakage current and anything else unusual.
Watch out for burns in the shape of a heatsink!

Good luck!
 

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
What is the point of using a switching regulator, then increasing the current with a pass transitor? This just turns your switching regulator into a linear regulator with all the disadvantages of both.

Bob
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
Bob, I'm not really sure what he wants to do, and I don't think he really knows either.

medictrode, you know that feeding the output of a switching supply into the base of a transistor doesn't make that transistor switch? The output of a switching supply is DC. If you feed it into a transistor, that transistor will operate as a linear regulator. There is no need to use a switching supply to control the transistor in a linear regulator.

If you want to replace the switching element in the SMPS, those huge transistors will probably not be suitable, and in any case other circuit changes would be needed.

Perhaps you could explain what you want to achieve with this project.
 

medictrode

Feb 21, 2013
16
Joined
Feb 21, 2013
Messages
16
Absolutely, and I apologize for the confusion as this thread sort of made a wrong turn at albuquerque...:D This being my fault..


So my initial intention was to build a SMPS, which I am still working on just for fun. At this point I should have started a new thread, for that I apologize. So I had obtained that IGBT module, then things went in another direction. Not that I have anything besides a load tester that will draw this much current, even my HF rig, but I was curious to see what this IGBT would do in a homemade power supply. Taking the knowledge that you fine folks stated, it was determined that the module I had wasn't made for a linear supply...:rolleyes:

With that said, I again visited my friends shop and found this Module (mg300m1fk1). Looking at the datasheets it's more of a Huge Darlington setup, and wow did it work well. I drilled and tapped a piece of aluminium to mount it on, then wired in a LM338 Style Variable Linear reg to run it. I was able to regulate quite a bit of current with it, in the 150Amp range, which is all I was comfortable doing...:p

I would like to ask though, as well as that 338 regulator circuit was doing, I feel some sort of feedback circuit would make the regulation even better. Most circuits I find online don't have this, and was wondering if there are any good designs out there. To answer your other question, this is just another power supply for fun. Just seeing where I can go with it.

Thank you again for the patience, and I apologize for the thread derailment.. :)
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
Wow, that is an impressive component!

The general connection of a linear regulator with voltage feedback is shown here:
en.wikipedia.org/wiki/Voltage_regulator#Regulator_with_an_operational_amplifier

This is a voltage regulator only; it has no current limiting.

Current limiting can be added in various ways by monitoring the current in the circuit, using a shunt resistor or for very high currents, a Hall effect sensor. When the current gets too high, the circuit reduces the bias on the pass transistor to keep the current below the set limit.
 
Top