Wired Magic Trackpad hack

Ok, I've decided on my next project.

At work, I have a magic trackpad for my laptop, and it's sitting right next to a USB keyboard.

A keyboard that has a USB socket on the side.

2 inches away from the trackpad.

Which eats 2 AA batteries a month.

Seriously, Apple?

A 5 volt to 3 volt LDO regulator will certainly work. The question is, how best to add it to the trackpad?

Well, OSHPark can make any size circuit board you want, of course. So how about one the exact dimensions of a pair of AA batteries? On one end, there could be a little "nub" with something conductive, like tin foil or something, connected to the positive output terminal. The other end would have a spring to push against the screw-in plug. For the negative terminal, you simply need some sort of springy metal contact that reaches either up or sideways or whatever to touch the side of the chamber.

The last challenge is how to get the USB cable in.

I've been playing with 3D printing lately. I'm thinking that the metal twist-in plug could be completely replaced with a printed plastic plug designed with a hole in the middle. The pattern for the plug would need to have the same thread around the edge as the metal one - perhaps the metal plug could be simply scanned in as a starting point.

Of course, this would only be for power - the trackpad would still use Bluetooth to communicate. But that's fine. I just want to save the hassle of changing the batteries so frequently.

J1772 Hydra boards

Well, I think the v2.2 design for the Hydra is going to be the stable design for a while. I'm not sure what to do next, though.

If I had some hope that people would actually want them, I'd place a medium run order for 20 boards. That would have a pretty substantial impact on my costs, but I have yet to sell even one Hydra to anyone.

Step right up, folks, don't be shy...

High current wiring

The Hydra v2.2 boards came in today, but I've had trouble in the past with the high-current path of both the OpenEVSEs and Hydras that I've built so far. I've come to the basic conclusion that engineering things that handle 30A of electrical power is a very unforgiving pursuit.

But with help from the OpenEVSE community, I may have finally cracked it.

I had been using crimp on .25" QD terminals I got at Home Depot, but it appears that those just aren't good enough for that kind of high current application. Instead, I've gotten in a supply of Molex Avikrimp terminals (DigiKey WM18256-ND), and with my ratcheting crimping tool, those appear to crimp on much more firmly. It takes a lot more effort to close the tool, and the resulting connector just feels a lot more solid.

The downside is that they don't fit very well in the Cooper Bussmann power distribution block I've been using.

There is the Cooper Bussmann 16220-2 instead of the 11725-2. That has four screw terminals on the load side instead of 4 QD terminals. But it costs $55 instead of $17.

What to do?

Medium run orders in

My first two medium run orders from OSH Park are in.

We now have a lot of AD8495 breakout boards and Toast-R-Reflow power board kits for sale.

Come 'n get 'em!

Idea doldrums

I'm pretty happy with the stuff I've been able to design so far. But I feel stuck, a little. What's next?

I had thought of attempting to make a contemporary design of a vintage computer. Like a completely SMD CP/M machine. A Z-80 with 64K of RAM, say, 16K of flash and an SD card slot. But why? You can get a Raspberry Pi for $40 and run a Z-80 emulator on it. 

So I am in an idea vacuum. Anybody got any?

Woah! Toaster oven sale at Amazon!

If you've been procrastinating trying to build a reflow oven, this may help.

Amazon is selling the Hamilton Beach 31138 for only $23.99! With the Toast-R-Reflow backpack controller and power board at $50 for the pair (as kits), that's a reflow oven for less than $100!

Giddy-up!

More building fun

The second and final RS-232 board came in. The big change was an SSOP line driver in place of the SOIC one. I also decided to use the super-bright LEDs on it, just for fun. There's no missing them when they blink!

The first prototype on the bottom, the new one on top.

I still don't know what to do next. I can sell them for about $15 with low parts purchase quantities, but that doesn't include a case or a µUSB cable. I could lower that price if I were making thousands of them, but what market there is for USB-to-RS232 cables is saturated.

The other thing I built was the final version of the i2c RGB LCD backpack display. As a reminder, this is like the OpenEVSE display, but instead of just a SELECT button, I've included a 5 way tactile switch and a piezo element supported by the LiquidTWI2 "buzz" method. You can also change the i2c address of the GPIO chip with 2 way solder jumpers and you can change the common backlight pin from common cathode to common anode with another. In fact, this prototype is hooked up to a SparkFun RGB LCD display that has a common anode backlight. You just need to change an initialization constant in your LiquidTWI2 setup and it works exactly the same way.