Considerations for a bench powersupply for microcontroller development

[Einar]

Hi forum

Im doing some microcontroller development, mainly for education and as a hobbyist.

So far I have powered it all from usb-power or 5V wall warts. At his point I want to be able to supply variable voltage levels both for testing and to try different voltage levels as power source. I also would like to be able to draw more than the odd 1A you can get from the usb port.

My thought is to get a variable bench power supply.

Are there any certain things I need to consider except the obvious power/current ratings?

Would a base version of an el cheapo brand for 50€ be sufficient or do I need better filtering/protection?

Regards 

[dannyf]

You probably want to ask yourself if you really need more than 1amp for mcu development - those things consume rarely over 20ma by themselves and if you go over 200ma, something is seriously wrong.

High amperage ratings may not be a good thing here.

If I were you, I would use an adjustment 3-terminal regulator (LM317 or the likes) in to220 package. I would put a 10 - 20ohm or so power resistor in serial with the output, before the feedback divider. The resistor means that if there is a mistake, the circuit never draw too much current - a beefy power supply here can be a bad thing.

Hope it helps.

[David Hess]

I would look for a supply or supplies which provide variable +2.5 to +8 volts for digital logic and floating variable +/- 0 to 18 volts for analog circuits.  Then I would duplicate it.

Something like an HP 6236, HP 6237, Tektronix PS-503, or Power Designs TP325, TP330, or TP343.

[SirNick]

I'm with dannyf on this one.  A simple linear regulator is all you need.  You can build yourself a small PCB and clip, adhere, or screw it on to a breadboard.

I've got this on my "one of these days" project list:  1.8v, 3.3v, and 5v (switching), 12v (higher current, but unregulated -- for LEDs, servos, etc.), +/-12v (linear), adjustable 1.2-9v (linear).  A lot of simple switching regulators provide configurable over-current trip points, which can be limited to something sane like 500mA.  If you draw more than that, something's probably not right.

[diyaudio]

Hi forum

Im doing some microcontroller development, mainly for education and as a hobbyist.

So far I have powered it all from usb-power or 5V wall warts. At his point I want to be able to supply variable voltage levels both for testing and to try different voltage levels as power source. I also would like to be able to draw more than the odd 1A you can get from the usb port.

My thought is to get a variable bench power supply.

Are there any certain things I need to consider except the obvious power/current ratings?

Would a base version of an el cheapo brand for 50€ be sufficient or do I need better filtering/protection?

Regards

For microcontroller development, your first aim is to obtain a power supply with low noise floor, its okay to use a few LM317`s with fixed 1.2v, 3.3v, 5v and a separate variable 1.2v to +/- 15v 1A outputs diy (if you cannot afford a bench power supply),  but seriously when you start interfacing with acquisitions systems with op-amps and 14-bit > ADC`s you really need to "step up" your game, so what im saying if you can afford get a decent power suuply like a rigol dp832 or something.

     

[paulie]

If this is for educational purpose then all that about linear & switcher is good advice. If you just want to get something working so you can continue with the micro projects it's foolish to play with that stuff. Better to just pick up a $1 buck converter if 1.2-12v is enough otherwise one of the $2 buck/boost if you need up to 30v. These can supply up to several amps when used with a  generic 12v supply or a PC USB port. A couple cheap caps tacked on the output will solve any noise issues although I can't imagine that being a problem with MCU work.

http://www.ebay.com/itm/1-DC-DC-Buck-Converter-Step-Down-Module-LM2596-Power-Supply-Output-1-23V-30V-/181409861491?pt=LH_DefaultDomain_2&hash=item2a3cdecf73

http://www.ebay.com/itm/DC-DC-Boost-Buck-Converter-Step-Up-Step-Down-Supply-Module-3-35V-to-2-2-30V-TE-/111290455503?pt=LH_DefaultDomain_2&hash=item19e96d95cf

[diyaudio]

If this is for educational purpose then all that about linear & switcher is good advice. If you just want to get something working so you can continue with the micro projects it's foolish to play with that stuff. Better to just pick up a $1 buck converter if 1.2-12v is enough otherwise one of the $2 buck/boost if you need up to 30v. These can supply up to several amps when used with a  generic 12v supply or a PC USB port. A couple cheap caps tacked on the output will solve any noise issues although I can't imagine that being a problem with MCU work.

http://www.ebay.com/itm/1-DC-DC-Buck-Converter-Step-Down-Module-LM2596-Power-Supply-Output-1-23V-30V-/181409861491?pt=LH_DefaultDomain_2&hash=item2a3cdecf73

http://www.ebay.com/itm/DC-DC-Boost-Buck-Converter-Step-Up-Step-Down-Supply-Module-3-35V-to-2-2-30V-TE-/111290455503?pt=LH_DefaultDomain_2&hash=item19e96d95cf

irrationally throwing capacitors at a noisy circuit (especially a SMPS) doesn't solve anything.

[paulie]

It certainly doesn't if there is no problem to begin with. I suspect diy"audio" lives in a different world than mcu guys. Even when working with ADC applications noise is rarely a problem and if it were then caps, irrationally thrown or not, actually do work to reduce noise. I've measured up to 50-60db broad spektrum reduction (be careful where you put those probes) with just a small monolithic and a 1000uf aluminum.

Of course if this is medical or other mission critical stuff then blowing a few hundred on a lab supply might be wise.

[mathsquid]

High amperage ratings may not be a good thing here.

I've learned that the hard way.  I've done some simple projects that have LCD displays.  I'm not always careful, and I've connected them incorrectly a few times.  That was never a problem when I was powering them with USB or AA batteries, but when I made a CNC controller and powered it with an ATX power supply, I connected something backward and roasted the LCD panel.

I'd recommend either going with a 317-based solution or using the current-limiting feature of a bench power supply.

[SirNick]

For your typical MCU, switching regs are fine.  USB isn't usually the cleanest source of power, after all.  If you need really low noise, well... you either know how to build an LDO circuit with sufficient isolation, or your are going to be in trouble when it's time to move your circuit to its own PCB and average wall-wart PSU or whatever anyway.

Really really quiet, high-performance power supplies have their place, but if you need one, you know you need one, and why.  Until then, while it might give you warm fuzzies to know your lab PSU has killer specs, it probably isn't really doing anything you couldn't do with an LM7805, LM317, MCP33063, TPS54226, or any of a few dozen other run-of-the-mill parts.

Just my 2c.

[salvix]

I got this one and I'm very happy with it (I had similar requirements as the OP): http://www.banggood.com/DC-LED-Digital-Controlled-Step-Down-Driver-Power-Module-p-910096.html

In practice, it thermal shuts down at ~2.8A (at 12V), which is plenty for me.

Great review from Julian Ilett:

[legacy]

You probably want to ask yourself if you really need more than 1amp for mcu development - those things consume rarely over 20ma by themselves and if you go over 200ma, something is seriously wrong.]

It depends, my 68332 board usually eats 290mA! It's a 1990 technology 

[andyturk]

I've got one of these and it works great for small circuits (and even bigger ones if you only need 5V):
 
B&K Precision 1651

You should be able to find one in good condition for around $100.

[FreddyVictor]

I got this one and I'm very happy with it (I had similar requirements as the OP): http://www.banggood.com/DC-LED-Digital-Controlled-Step-Down-Driver-Power-Module-p-910096.html

In practice, it thermal shuts down at ~2.8A (at 12V), which is plenty for me.

Great review from Julian Ilett:

this seems like a great option !

I went this route

[diyaudio]

I suspect diy"audio" lives in a different world than mcu guys.

I find it very assuming that some members on the forum intelligently comes to the conclusion that as a result of my alias "diyaudio" im an audiophile OR a narrow minded audio only personality (audio fool), it says alot about you.

I think I should update my avatar to one of those animated pictures with bouncing tities and see what happens next.. for your information my interest is not limited to 20khz.

 

[paulie]

Actually my comment was more a result of your excessive concern over power rail noise with microcomputers that the avatar. Wasn't the only one in this thread with that viewpoint either. Anyway I would be very grateful if you would follow through with that animated gif suggestion.

[diyaudio]

Actually my comment was more a result of your excessive concern over power rail noise with microcomputers that the avatar.

yes, and its remains relevant.

Anyway I would be very grateful if you would follow through with that animated gif suggestion.

I'm sure you would be. 

[SirNick]

I find it very assuming that some members on the forum intelligently comes to the conclusion that as a result of my alias "diyaudio" im an audiophile OR a narrow minded audio only personality (audio fool), it says alot about you.

I think I should update my avatar to one of those animated pictures with bouncing tities and see what happens next.. for your information my interest is not limited to 20khz.

Seems like a reasonable conclusion that your interests might be skewed to analog audio, based on your chosen name.  Kinda like how it would be reasonable to assume I'm male.  If you absolutely don't want the stigma, don't use a name that implies a connection.  (Preferably, keep the personality, and just understand when such a natural conclusion is drawn.)

At any rate, your suggestions weren't wrong, and certainly something to keep in mind, but (here we go again with the assumptions...) the OP seems to be at a level where your concerns aren't likely to apply.  14-bit+ ADCs?  Yeah, for that you'll need quiet power rails.  However, that's not exactly standard kit on a typical ATmega or 8-bit PIC.   I suspect our guest of honor here will get by just fine with reasonable amounts of ripple and noise.  No need for a super snazzy PSU.

[Rasz]

I suspect diy"audio" lives in a different world than mcu guys.

I find it very assuming that some members on the forum intelligently comes to the conclusion that as a result of my alias "diyaudio" im an audiophile OR a narrow minded audio only personality (audio fool), it says alot about you.

and

14-bit > ADC

hilarious




back on topic:

get yourself one of those cheap Chinese buck converters, but with two pots (voltage and current limit). You can connect them to PC, old printer or laptop power supply. It will be absolutely fine for what you do.

[mazurov]

In my opinion, for working on low-power circuits, MCU or not, fast current limiter is more important than anything else. I prefer Agilent E3640A for MCU work, even though I have several much nicer cheapies standing right next to it on the bench.