Hello everyone! I'm new here, although I've checked this site's maths/physics threads out several times. I'm also kind of new to circuit building, so feel free to laugh at me if needed. I'm having problems with a dc boost converter using MAX1708 (datasheet). I like long posts, so if you're one of those TL;DR guys just skip to the end :v

I'm trying to build a battery pack for cellphones or anything running at 5V and I'm using some 18650 batteries I got from a broken laptop. As I don't want to connect them in series and as I don't just want to use one of those 2$ converters from eBay, I've decided to build the circuit myself. Connecting couples of series-connected batteries in parallel would lead to some problems:
- it's a waste of energy, as the LM78S05's I'd use to convert 7.4V to 5V isn't as efficient as the circuit I'm trying to make;
- if one battery gets damaged, the other one won't work;
- ...I couldn't run it with an odd number of batteries.

So I've got a couple of those maxim ICs thanks to a friend of mine. Its dimensions make it really hard to solder it normally, so I designed the simplest pcb ever (really, look at it) to test it out.

(it can be aligned with a perfboard, the distance between holes is the same)
As apparently I don't like nice-looking things, I've soldered 16 cables to it the worst way I could. Contacts are ok though, I triple-checked all of them.

The 100% professional duct tape helps preventing it from scratches and/or shorts.
After this magnificent work, I've built the circuit in the datasheet (page 9) on a breadboard, but, as the title says, it doesn't work.

TL;DR part starts here :v
I've built the circuit at page 9 of the datasheet, using one of the diodes (this one) listed at page 11, 220uF capacitors instead of those 150uF ones in the datasheet and a 2.2uH inductor I already had (basically collected from scraps, but pristine). As R1 (the resistance connected to the SS/LIM pin), I'm using 60kΩ, which should limit the current to around 1A. C3 = 100nF. I just connected ONB and ONA to ground. The input and output voltage are totally equal (there's just a voltage drop caused by the diode), even trying with different loads. Any ideas about what the problem could be? Where did i fail? Any help would be appreciated :v

 

I would request your attention to Page 13 - Layout considerations!

Check FB is connected to ground.
PGnd and Gnd should also be connected (Star).
Out (Output Sense) and -3.3/5 should also be connected as per the circuit.

 

When you can PCB's, why didn't you make a whole boost converter PCB ?

 

You are unlikely to have success when using those long cables. The circuit operates at 600kHz, at which frequency the inductance of the cables presents considerable impedance. As R!f says, see P13 of the datasheet. Did you also note the design recommendations for the inductor and capacitors?

 

Thank you for your fast replies!

I would request your attention to Page 13 - Layout considerations!

Check FB is connected to ground.
PGnd and Gnd should also be connected (Star).
Out (Output Sense) and -3.3/5 should also be connected as per the circuit.

Just checked every connection, they all look fine. Gnd/Pgnd pins were already connected that way!

When you can PCB's, why didn't you make a whole boost converter PCB ?

That's a good question! Unfortunately, when I did that (this is actually an unfinished project I started about a year ago) it was my very first time using Eagle and a whole converter pcb design looked too complex for me (I basically just hoped the circuit would work anyway). Moreover, shipping times are a pain.

You are unlikely to have success when using those long cables. The circuit operates at 600kHz, at which frequency the inductance of the cables presents considerable impedance. As R!f says, see P13 of the datasheet. Did you also note the design recommendations for the inductor and capacitors?

I'm gonna replace the inductor and every capacitor as soon as I can get those listed in the datasheet (or equivalent ones). Cables are what I was starting to think the problem could be! I guess I'll probably need to design a full pcb as R!f said. Just one question (I honestly lack the knowledge to answer myself): if the problem was the impedance/inductance of the cables, wouldn't the circuit work badly instead of not working at all, like, outputting unstable voltage or something?

 

if the problem was the impedance/inductance of the cables, wouldn't the circuit work badly instead of not working at all, like, outputting unstable voltage or something?

Who knows? The chip manufacturer might be able to say. But if you adopt all the recommendations in the datasheet regarding layout and components (apart from the cabling) then you should be able to tell us yourself .