I'm considering rolling my own one-off buck converter. I need 14-20V in, 24V out, 3A. I am looking at using the LM2577-ADJ, which has a switching frequency of 52KHz. Is that going to cause an issue if I breadboard it or throw it together on a veroboard with questionable routing and sloppy soldering?

If so, can someone recommend a lower frequency circuit design that uses a 555 or something, that might be more suited to "rapid prototyping "?

Thanks

 

Sloppy layout can be problematic with any switching regulator of any design. (Sloppy soldering will likely just result in shorts or cold solder joints). You need to use a good ground, preferably a veroboard with a copper plane, and carefully lay it out so the high current paths are short and everything is well decoupled to ground.

 

Good layout practices are essential starting out at zero Hertz. Or DC.

I'd advise staying away from solderless breadboard, but do your sloppy best to solder something together. You should learn much form the adventure, assuming knowledge is your primary goal, and you have some troubleshooting equipment, at least a decent voltmeter but maybe also a scope.

If all you wish for is an inexpensive power supply then best buy one off EBay.

 

I'm considering rolling my own one-off buck converter. I need 14-20V in, 24V out, 3A. I am looking at using the LM2577-ADJ, which has a switching frequency of 52KHz.

That's the fundamental switching frequency. Switchers can have ringing and resonances in the range of many Mega Hertz. The simple switchers won't work on point wiring boards, they need a decent layout. We used to put a sample layout on the data sheet.

 

If all you wish for is an inexpensive power supply then best buy one off EBay.

+1 I see these for $5 all the time.

 

It's true. You're into a different world. Nice, slow, analog stuff is so simple that the only considerations are 1)fitting everything in the smallest space, 2) double checking ground paths, and 3) accounting for heat. You already know that even a 555 timer requires personal decoupling capacitors. Fast switching mosfets go through the roof for frequency considerations and the loop size of the current paths actually matters. If you aren't prepared to spend days learning the quirks, chicken out and go eBay or Amazon.

 

I don't know if anybody here is following my handheld power supply thread (see posts #26 & 34), but I've tried the amazon/ebay route and have been burned multiple times. One count of pure B.S. - a buck module advertized as a boost module, and two counts of pure crap design that blew up in my face.

I'm almost unwilling to give it another go. The stubborn half of me wants to do it myself to prove a point. Perhaps the logical side of me will win out after a pile of failed prototypes grows on my bench. But yes, you could say I'm in it for the experience at this point.

 

OK. There are a few people here that can help.
Some of the Simple Switchers have circuit boards available from T.I.
Go run their Simple Switcher solution and see if you get lucky.

http://www.ti.com/ww/en/simple_switcher/index.html?DCMP=simple_switcher&HQS=switcher

 

OK. There are a few people here that can help.
Some of the Simple Switchers have circuit boards available from T.I.
Go run their Simple Switcher solution and see if you get lucky.

http://www.ti.com/ww/en/simple_switcher/index.html?DCMP=simple_switcher&HQS=switcher

Ok, this webench thing is sucking me in. Much too much to absorb tonight; maybe even this week. I will get back on it in the morning. Thanks for the lead.

 

I'm considering rolling my own one-off buck converter. I need 14-20V in, 24V out, 3A. I am looking at using the LM2577-ADJ, which has a switching frequency of 52KHz. Is that going to cause an issue if I breadboard it or throw it together on a veroboard with questionable routing and sloppy soldering?

If so, can someone recommend a lower frequency circuit design that uses a 555 or something, that might be more suited to "rapid prototyping "?

Thanks

Do you mean boost converter? Or did you get the voltages wrong?

 

...
If so, can someone recommend a lower frequency circuit design that uses a 555 or something, that might be more suited to "rapid prototyping "?
...

I did a simple design here;
http://forum.allaboutcircuits.com/showthread.php?t=78805

using a cheap 34063 8pin IC, driving an external PFET. Frequency was low-ish, 26 kHz. Mostly that depends on your inductor choice (and the CT timing cap value).

You can easily adapt that design to give you a nice buck converter for 3A use.


Don't be too scared designing buck converters. The main layout issue is the circular buck path (diode, inductor, output cap) which needs to be short and close together. Eveything else is much less critical.

 

Do you mean boost converter? Or did you get the voltages wrong?

I'm officially an idiot. That's the second time I've made that mistake this week. BOOST is what I meant. Sorry for the confusion.

 

I did a simple design here;
http://forum.allaboutcircuits.com/showthread.php?t=78805

using a cheap 34063 8pin IC, driving an external PFET. Frequency was low-ish, 26 kHz. Mostly that depends on your inductor choice (and the CT timing cap value).

You can easily adapt that design to give you a nice buck converter for 3A use.


Don't be too scared designing buck converters. The main layout issue is the circular buck path (diode, inductor, output cap) which needs to be short and close together. Eveything else is much less critical.

Without looking at the IC datasheet (yet), I'm guessing I can make this work for BOOST, by moving the inductor up to the top of the schematic and moving the FET down to the bottom. Is that correct?

 

I'm in the middle of designing a boost right now... It can be done on a solderless breadboard (at least at the powers I'm designing at (2.5W), but you have to keep the high current paths small and tight. You also have to factor in inductance into your input and output caps and know how that is going to affect you when you go to the PWB. Your probing techniques also have to be suburb - for any switching power supply.

 

I don't know if anybody here is following my handheld power supply thread (see posts #26 & 34), but I've tried the amazon/ebay route and have been burned multiple times. One count of pure B.S. - a buck module advertized as a boost module, and two counts of pure crap design that blew up in my face.

I'm almost unwilling to give it another go. The stubborn half of me wants to do it myself to prove a point. Perhaps the logical side of me will win out after a pile of failed prototypes grows on my bench. But yes, you could say I'm in it for the experience at this point.

TI will probably give you a sample board for a simple switcher. We gave them away by the thousands. Make up a fake company name and ask them for a sample board.

 

Even low frequency analogue supplies need care when laying out the pcb.

I threw together a audio mixer on a pcb and it hummed really badly when I tested it.
There was 1 volt of hum on the output !!! which was completely useless.

I then laid out a second pcb taking care to star ground everything and it was silent.
In my case the main problem was the power supply on the same pcb was modulating the ground line and injecting hum into the audio.

 

I used something like that on stripboard years ago and it worked fine. Maybe efficiency suffered, I'm not sure. These days I'd probably hand draw it on copper clad board with an etch resist pen, not much more work than laying it out on stripboard and cutting the tracks.

 

Without looking at the IC datasheet (yet), I'm guessing I can make this work for BOOST, by moving the inductor up to the top of the schematic and moving the FET down to the bottom. Is that correct?

Going from memory I think the datasheet already has example circuits for boost converters.

It should be easy enough to use a FET in the boost circuit the same way I used one in the buck circuit. ie; using their circuit but replace a NPN with a NFET.

You will need to check the Vin voltage against the FET max allowable Gate voltage (Vgs).

There are a couple of different datasheets, different IC manufacturers that all make the 34063. Some sheets are better than others.

 

Going from memory I think the datasheet already has example circuits for boost converters.

It does and the simple switcher software will let you customize the design. The same demo board can be tweaked as needed to get different voltages and currents.