Our Sunday hack-afternoon resulted in not only a working roach coach, but also gear for a second backpack — a wifi repeater. In the process we — hilariously, but less dramatically this time — fried another NextNet modem.

The team was me, Eli, Adam, Michael, Patrick, David and Jason. David brought a 12V motorcycle battery (off his motorcycle, which is hibernating this season), and Jason brought a little 12V rechargeable battery that he picked up at Active Surplus. Knowing that the motorcycle battery was already fully charged, we decided to hook the modem up to that one first. We tested the battery’s output voltage: 12.5V, safely in the 11-15V range that Tony recommended on our mailing list.

When we plugged it in the lights didn’t come on right away, as we would’ve expected. And there was a tiny little pop — we hoped it had come from a spark of the battery lead. We unhooked, and plugged the modem back into its wallwart… no luck, the modem was dead.

We tried diagnosing where we’d gone wrong… the only thing we came up with is that we’d reversed the polarity on the connector going into the modem.

That’s when we split up: one team would go fetch another NextNet modem, and the other team would start building backpack #2, a wifi-repeating backpack. The WRT54GL runs safely when plugged straight into our 18V cordless drill batteries.

Patrick and Michael did a beautiful job converting the Ryobi flashlight. Patrick took the gear home to test out how long the 18V drill batteries will power a WRT54G. (UPDATE: Eight full hours, with torrents running on one of the wifi clients!)

When we regained courage, we triple-checked the polarity on the connector, and plugged David’s 12V battery into the modem. (Though we used Jason’s modem this time — I was too chicken to risk blowing up mine for a third time in under a week.) It worked just fine! Then we tried plugging the modem and router into the battery at the same time — again, success!

The motorcycle battery is lead-acid, and probably not designed for deep cycle, making it a less than ideal choice: in terms of cost-efficiency, environmental-friendliness, and for the health and safety of the backpack’s wearer.

So we’ve yet to finalize a power source for the “backpack A”: the one containing both the NextNet modem and primary wifi router. We’ll either get some more Ryobi 18V drill batteries (so that we’re using the same power system for both backpacks) using a voltage regulator to knock it down to 12V, or we’ll use something like the Tekkeon myPower ALL MP3300. Anyone know of other (simple, inexpensive) options we should consider?

Once we’ve got the power stuff figured out, we’ll need to take it out onto the street, to see if we need to make any adjustments to the modem’s antenna, or add an external one…

Our first hacknight (last night) was… a blast. Due to confusion we had to change the venue at the last minute to — my home office. There have never been so many people (me, Edward, Patrick, Ana-Maria, Dave, Liam, Michael and Susan) in the room at once.

We’re building a wifi backpack, which we’ve affectionately called the WiFi Roach Coach (long story). It’ll be a battery-powered WRT54GL and Rogers/Bell/Inukshuk pre-WiMax modem providing connectivity. (It’s a NextNet Expedience RSU-2510-AV.) This’ll allow us to set up an instant Wireless Toronto hotspot anyplace where we can get a (pre-)WiMax signal. It’ll be especially useful to provide wifi coverage at events.

We started off well: we got a modem, router, two 18V Ryobi (cordless drill) batteries, a battery charger, and a 18V battery flashlight — from which we’d remove the light and plug into the two boxes. Patrick bought a nice blue backpack.

While the WRT54GL comes with a 12VDC, 1A power supply, I’d read that it can run on a wide range of voltages. Patrick unscrewed the flashlight, Dave hooked up the leads, Michael stripped the power supply cable, and we plugged it all in: success. The lights on the router came on and looked normal. w00t.

Next, we tried the modem. Same setup, different power cable. The modem has a 13VDC, 1.3A power supply — pretty close. In our giddiness and haste, we made the poor assumption that the modem would be similarly forgiving about the higher voltage.

We plugged it in, Dave hit the switch on the flashlight, and nothing happened. We sat. About 5 seconds later, there were a series of really loud pops — at least three, maybe as many at seven. A horrible smell filled the room, while a huge plume of smoke rose to the ceiling.

Oops.

So, we passed around another round of beers, and lamented. Ana-Maria — the only engineer among us — drew what we *should* have done… figured out the impedance of the router and modem, and added in the appropriate resistors to bring the voltage down to 13V. (Or this is what I think she said, anyway.)

We then (seven of us!) trekked back to the store where I bought the modem, hoping for a quick and painless exchange. Sadly, they were closed — we’ll have to try again another time. We bought potato chips (organic!) and walked back. We ate chips, made fun of ourselves, and cleaned up. Venceremos!

Thanks to Edward, Patrick and me for the beer, with special thanks to Patrick for carrying it all. (And sorry I forgot to tell you that the alleyway jogs at Markham.)

More photos from Dave are, hopefully, forthcoming.

Future hacknights will be held at InterAccess… sorry for the confusion this time. The next one will be sometime next month, though there may be a mini-hacknight soon, to finish the roach coach. Details on the discuss list.