mouldy.org

I've been talking about setting up a software company for a long time. I'm finally getting organised about it and making progress.

I've set up a blog for it at sixproducts.com.

Nothing much has been happening over the last few days; other projects have been taking priority. Which isn't to say that no progress is being made on the business as a whole, but certainly nothing on the product development side. Right now, a lot of it is just waiting for phone calls and doing work that pays the bills but doesn't otherwise get me anywhere.

I've had a little bit of time to test the board. My usual trepidation applied. The board acted a little funny and got quite warm with the current limit set low - but that's not entirely surprisingly, because the input voltage was being dragged down to roughly the IC's minimum input voltage. With the current limit set to something sensible, it was definitely giving some output, but not the output that I was hoping for. My prototype board has a 1A current limit. With an 8V input, the output current was 1.4A, 5V in, 1A out. 12V in, 0.4A out. It's obviously doing something useful - if it wasn't operating, there'd be no output at all, given it's a buck-boost converter. I thought that my load (a 0.47R or 2R2 resistor) might be screwing things up, but the simulation doesn't confirm my results anyway.

I built the first prototype this morning. A few notes:

• Solder paste is annoying to apply by hand; it's best to think of it as tiny, sticky solder balls, rather than a viscous liquid. The best that I've come up with is to use a tiny screwdriver to chop off a little blob of paste, then try to drop it onto the board. It doesn't really smear, nor does it stick to the board or flux very well.
• It doesn't really matter how much paste you use, so long as there's a bit of a blob there. It'll spread out when you heat it.
• My SMT soldering tip is no good. It just doesn't get hot enough. This could be because it's not exactly the right tip for the soldering iron, but it seems to fit pretty well over the heater element. I had much better results with the full-sized tip.
• The soldering procedure goes something like: apply solder blobs, use tweezers to place part, use tweezers to hold down part, heat the pad/pin to cause the solder to reflow, then heat the other side.
• If you don't hold the part down with the tweezers you can sometimes get it to float perfectly into place over the pad, but just as often it'll float away. If you hold it down, every part will look bad, but you'll maintain your sanity.
• QFN parts are hard to solder by hand. You need to rely on the part floating into position, because the pins are too small to line up by hand. I also couldn't see the pads with solder paste all over the place. Pleasingly, the solder stays off the blank PCB, so the lack over solder mask isn't too much of a handicap. In hindsight, I should've gotten SOIC parts. They're slightly larger, but far easier to hand solder, and make it easier to lay out the PCB.
• If your soldering iron sucks (like mine did until I realised it was the tiny tip), it helps a lot to solder the side with less thermal mass (smaller tracks) first. This holds the part in position while you try to heat up the other side. Much of my design is high-current tracks, so the thermal mass of the tracks is significant (or at least, it was with the crap tip).
• I've had the board sitting in front of me on my desk for the last few hours, and I'm still marvelling at how tiny it is. The soldering was surprisingly easy - with the exception of the QFN part - so if you're umming and ahhing about it, I'd recommend just jumping in.

The board needs to be reflowed; the QFN IC is misaligned and only soldered on two sides. I also haven't tried to solder the underside pad. I bought a cheap hotplate and gave the infamous skillet reflow method a go. No problems there. It seemed to be a bit slow to start - there was plenty of flux boiling off, but no obvious solder reflowing. I got a bit impatient and started tapping some larger parts. The solder was already molten, and probably had been for some time. It took a while to cool off, too. The IC has probably had a lot more than its recommended ten seconds of soldering heat, but oh well. It appears to be pretty well aligned now, although there's still not quite enough solder on two of the pads. I also have no idea of what's going on underneath, so there'll be a lot of probing with a multimeter before I test it out.

It's been a week, and I don't have a tremendous amount to show for it - at least on the business side of things. The work situation is keeping me busy, and there'll be some major changes there soon.

I did my first laser toner transfer PCB today, and it went remarkably well. This is the first PCB that I've hand-made in probably eight years, and just about everything is different:

• Ammonium persulphate instead of ferric chloride
• Surface mount instead of through hole
• Laser toner instead of hand-drawn Dalo pen
• Extremely high levels of detail (0.1mm tracks) instead of, well, the crap that I can draw by hand with a Dalo pen

I used "Kodak Everyday Picture Paper for Inkjet Prints" in the 4x6 inch size. It was a clearance special at the supermarket. I figured that I probably won't be making boards larger than 4x6 inches anyway, so not using A4 paper won't be a major problem. I had to turn on the 'mirror' option when printing to keep everything properly oriented.

Who would've thought that so much random crap would come up on a daily basis?

• Spoke with the local supplier of my control chip. That should arrive in the next few days. 
• My would-be supplier that buys people's existing stock seems to have their website purely for generating sales leads and not for conducting any sort of commerce. Which is a shame, because I could do with a service that buys and sells people's electronics stock (eBay for components). Especially if they can get me obscure parts fast.
• Dan's Data has a review of a flashlight that uses Luxeon K2's. I panicked for a bit, and eventually realised that this doesn't hurt me at all. The light is kinda scary; three Luxeon K2's connected directly to a bunch of AAA cells. It relies on the series resistance of the cells to provide regulation, and does about 2.5A between three LEDs. I guess there are two things there for me: people aren't so likely to use a regulator in a small flashlight (there are 'good enough' regulation methods) and there'll be a bit more knowledge of the Luxeon K2 soon. Which is good for me. I guess I was a bit surprised at how quickly the thing came out; I can't even get the LEDs yet, let alone integrate them into a product and begin shipping.
• I've almost finished reading 'Bootstrapping Your Business' by Greg Gianforte and highly recommend it. Plenty of useful information, no fluff, easy to read. Good and information-dense, and tons of good ideas. Read it.
• I got thinking about heating etching tanks using ammonium persulphate. I've used ferric chloride in the past without heating, but have given in to agitating the board while it etches. Depending on who you read, ammonium persulphate doesn't need agitation, but it does need heating. Rather than using a heating element in the etchtant (which is difficult to build, messy and risky) I was thinking of putting a 'steam tray' underneath the etch tray. It'd just be another plastic container full of boiling water; it'd heat the etchtant without requiring any extra gear or cleanup.
• My passives have arrived. They're so, so small. I can barely read the markings on the side. I might have to get some sort of magnification to help with assembly and soldering. Alternatively, I could use larger parts; the board isn't that packed. Which probably means it could be packed tighter.

I can think of plenty of more fun ways to spend my weekend than working through food poisoning and laying out a PCB. Both are more or less done. I have a PCB layout that both looks like it'll work and is nice and small. I printed it out actual size and was somewhat stunned at the size of the components that I'd chosen. 4x4mm sounds all well and good in words, but you need to see it to really believe it. It's the sort of part you'd put on your fingernail and take a photo of for the press. As a result, the PCB has some very narrow features, and I'm not confident that I'll be able to toner-transfer it. I'll give it a shot anyway; it's not so tight that I can't draw the tracks in by hand if the laser toner doesn't stick.

It's official. I have now invested actual money on this business. I've ordered a batch of parts that I can use to assemble my prototypes.

I spent most of the day optimising and testing my design. Switchmode power supply design is hard, especially if you have space and cost restrictions. Moreso, because the 'one part in the world that can do what I want' is pretty marginal for that task. There's been some trimming of requirements, but I don't think they're unreasonable. The only loser is 3-4V operation (single lithium cell), where you might not quite get the full 1.5A output. The efficiency also blows. Running a 6W LED off a single lithium cell isn't exactly the sanest thing to do if you're expecting power efficiency.

I'm an obsessive record keeper. With the addition of contract invoices, parts acquisition, BAS, GST and sales to my repertoire of financial operations, I decided it was time to evaluate some accounting packages and see if any of them could do anything useful for me.

Excel

So far, my records are kept in Excel. It works well enough. I have some macros that will produce reports that I'm interested in. It's not pretty, it's not organised, but the information is there.

gnucash

This is a Linux-only app. And when they say Linux-only, they mean Linux-only. You can't run it on cygwin. I tried a few emulation environments and found them to be either slow, or more trouble than they're worth (for example, sucking up a quarter gig of physical RAM whether you're using it or not). I tried X forwarding to my server at home, but because it's using GTK, that's horribly slow. LBX helped somewhat, and I guess I could tolerate it if I had to. It runs OK over my home LAN (but notably, not great). I don't have a Linux box at home that has a keyboard and mouse attached, and I'm not getting one just for accounting.

I've been able to use LTSpice to verify that one of my two chip choices cannot do what I want. So good work there for Linear Technology, I suppose. I know not to buy one of their parts!

It's slowly becoming apparent why people don't just bang these designs together overnight. At the worst-case design parameters that I have in mind - three LEDs, 1.5A current, 3.5V li-ion cell - the current that the circuit tries to draw is absolutely massive. This isn't surprising; three LEDs is about 12V, giving 18W output power. From a 3.5V lithium cell, that's over 5 amps (even at 100% efficiency).

From the LTSpice documentation:

SwitcherCAD III is designed to be used by three different types of design engineers: those who know what they're doing, those who think they know, and those who are sure they know absolutely nothing about switching regulator design. The experienced designer needs a "what if" program that allows him to quickly alter aspects of a circuit to find an optimum design. The neophyte needs a cookbook approach that yields a reliable design based on the simplest of inputs. The "loose cannon" designer needs a program that will allow him to exercise his free will, but will be intelligent enough to alert him to fatal design flaws.