Go Down

Topic: Trouble with a boost converter (Read 2507 times) previous topic - next topic

aoeud

Argh, I'm just trying to increase my voltage a few times over...  Are there alternatives without having to source a new power supply?

Docedison

a solderless breadboarrd is your first mistake... try to "Dead Bug" the ic on a small piece of PCB material ESPECIALLY the components connected to the Catch diode, inductor and output filter cap... This is the Most Critical area next to the Vcc connection make your connections as short as possible AND BYPASS THE Vcc Right AT THE IC use the copper as the ground. You Shouldn't need more than a 2 - 3 cm bit of PCB material. If you can, another inductor of the size used for the switcher inductor can be placed in series with the output and another LOW ESR cap to ground on the output of the coil (About 100uF) should bring your ripple down to 20 - 40 mV or less depending on the actual total load current. IMO

Doc
--> WA7EMS <--
"The solution of every problem is another problem." -Johann Wolfgang von Goethe
I do answer technical questions PM'd to me with whatever is in my clipboard

aoeud

I don't have any PCB.  Could I just solder parts directly?  Nor copper for that matter, though I can scavenge aluminum.

So the idea is a slab of aluminum under the whole thing that serves as a universal ground, with plastic holding the non-grounded parts from touching it.  The bypass capacitor on the input voltage should have the cylinder holding the good stuff as close as possible to the lead and the grounding wire should be very close to the lead as well.  Then I add another inductor in line with the load, and connect the load input to ground via the 100 muF capacitor.

Right?

Docedison

Right all but something you can solder to, The Ground Plane. Seriously the single reason that I mentioned copper (PCB Material) is it's ease of soldering. Aluminum is Most difficult to solder to. A Flat piece of .2mm tin plated steel would be an acceptable substitute. Not because it is steel quite the contrary. It's just easier to solder to. Doesn't need to be big as I said a 25 to 35 mm square should be big enough for all the parts, certainly no more than 50 mm on a side (2 inches) the smaller the better as it (the size) forces you to make SHORT DIRECT connections which are the KEYS to correct operation.

Doc
--> WA7EMS <--
"The solution of every problem is another problem." -Johann Wolfgang von Goethe
I do answer technical questions PM'd to me with whatever is in my clipboard

aoeud

I ended up using a penny!

Haven't tested it yet, bad back has me bedbound for the next while...  Hopefully I didn't fry it or anything, I don't always remember the heatsink.

Docedison

With any luck in your construction, the Ground plane (the Penny) IS your Heat sink. At 10 mA current and ASSUMING your converter is 80% efficient then you will have 80% of .01 X 25V + .01 X 25V = .25W load power + [.25 X .2 (80% - 1.0) = .05 watt of heat... @ 80% Efficiency] Fifty milliWatts of power to dissipate, Looks Very do-able from here. Probably the single most important thing is that the lead length of the inductor, catch diode and output filter cap (Both ends) be AS SHORT AS POSSIBLE. The input filter from Vcc to ground must be there and the recommended value be used, again with SHORT LEADS. long wires to the points I mentioned can cause the converter to work in a different mode, one you don't want and one that will causse a great deal of heat to be made as well as the chance that you let out the Magic Smoke...

Doc
--> WA7EMS <--
"The solution of every problem is another problem." -Johann Wolfgang von Goethe
I do answer technical questions PM'd to me with whatever is in my clipboard

aoeud

#21
Aug 02, 2012, 10:40 pm Last Edit: Aug 03, 2012, 12:33 am by aoeud Reason: 1
Shortened the leads as much as I could, I think everything under a centimetre.  Very high wire density in there.  Took 3 hours, but that's okay, I need the skill.  Hadn't thought about the heat; I'm inclined to contain it in a small box so that I don't jar anything and cause a short, so I guess I'll make a small aluminum foil heatsink and attach it through the bottom.

Uh, hmm, if I'm going for 10 mA at 25 volts I'm going to need more than 50 mA in, which is more than the Arduino can supply for a single pin.  Do I need to get an external power supply at this point, or is there another way?  If it's relevant, my setup involves two Arduino units so that processing activities don't mess with timing; could I combine the 5V outputs at the part labeled "power" using a Y and 200 ohm resistors?

Edit: oh, there was talk of "oscillation" with my op amp.  Is that the same as this issue, so I need to attach it to a ground plane rather than a solderless breadboard?

mauried

Your circuit cant start.
If you look at the specs of the switcher chip you are using, the minimum peak switch current is 0.65 A , typically 0.9 A
which is well above what the Arduino can supply.
These types of switcher chips are designed to run from batteries which can provide the initial current to start the switcher.
Once its oscillating, the current drawn is then determined by the load.
You need to supply the switcher chip with a 5V source that has at least a 1 A current capacity to get the switcher chip started.

aoeud

I see the peaks listed now that you mention it.  Is it always the case that the peak current has to be provided for switching to begin for a boost converter?  To what else does that apply?  Anything switching?

So my next route is a wall wart and a jack, 5V, >1A, regulated, and just hook up +5V to V_in and the converter's ground to the wall wart's ground...  Then the 25V output goes into a circuit whose end is the wall wart's ground?  Or the arduino's ground?  Or it doesn't matter as long as it's not more than 40 mA?

Grr, going to take until Tuesday to get it.  I hate finding out I need something more on a Friday -_-

Grumpy_Mike

Quote
is the wall wart's ground?  Or the arduino's ground?


Both grounds MUST be joined together.

http://www.thebox.myzen.co.uk/Tutorial/Power_Supplies.html

winner10920

Perhaps a very large capacito can provide you the 1amp startup current, idk how large since it depends on how long it needs it

mauried

Only boost or inverting converters initially draw large amounts of current to start.
Buck converters (reducing voltage) arnt as bad as the switch is in series with the load.
But , any switching converter will have trouble starting when fed from a current limited
source like an Arduino output pin.

Docedison

You could add a PNP transistor and 2 resistors to make the pin switch the main 5V source off and on as required or for that matter and better a P Ch Mosfet would be an even better choice. There are several made by Supertex that are in TO-92 packages and have Rdson values around 2 to 3 ohms require no heat sinking and are easy to use there are a lot of varying types of packages... Siliconix makes a whole family that are SOT-23 size in the same current range and bigger ones in SOT-223 packages as well as a different line of devices in SO-8 packages for still higher > - 5 A Id current loads.

Doc
--> WA7EMS <--
"The solution of every problem is another problem." -Johann Wolfgang von Goethe
I do answer technical questions PM'd to me with whatever is in my clipboard

aoeud

#28
Aug 04, 2012, 04:29 pm Last Edit: Aug 04, 2012, 04:43 pm by aoeud Reason: 1
Oh you guys are so helpful! <3

Taking a look at all of this, really appreciate the power supply link (the Arduino wiki article was rather lacking.)  Actually, going to read all of your tutorials. :)

It happens that I was planning to eventually use a current closer to 1 amp so this stuff is really good to know.

Love this stuff.

mauried

If all you need to do is turn the converter on and off, the connect the Arduino output to pin 3 which is the shutdown control.
Converter runs when this pin is grounded, stops when this pin is high.
Connect the switcher chip Vcc to a decent 5v supply.

Go Up