I've been working on a project with Wifi lately and I could really use a breakout board for those pesky microchip MRF24..WB0Mx wireless boards (basically any ZG2100- or ZG2101- based board). This is the first place I looked, to grab a bare PCB rather than a populated one since I already have the modules and just want to prototype without soldering tons of small wires.
Which brings me to my question... why haven't you done WiFi yet? I know the TCP/IP stack is crap, but there are open source implementations for the Uno32 (using Microchip's stack and using a ported stack) and Maple amongst other things... The modules are super cheap and I think they'd be super useful for wireless data logging or messaging or whatever.
I'm trying to measure mains current and I'm having trouble picking an op-amp.
There are sooo many out there, and some are specifically for current shunts and whatnot.
Here's some details about my application: - mains is 120vac, 170 to be safe - max current is ~3A - the current plan is to use a 2W 0.1 ohm shunt resistor (0.9W at 3A)
I also have a question about isolation. The project is a smart outlet basically -- an I2C port expander triggers a relay, which has mains voltage on the high side. The circuit in the outlet (the expander and the power for the relays) is provided by a small 5vdc wall-wart. Because the whole thing is I2C, the only signal wires needed are SCK, SDA and GND. The remote controller board runs at 3.3v and the relays require 5v @ 100mA so it is necessary for each outlet to be self-powered. So, do I need isolation? On the one hand, the box is powered from that very same mains that's being monitored, but the wallwart uses a transformer so... And then since it's all over I2C, does it even really matter if that single op amp chip is isolated?
Anyway, Im looking for an op amp from somewhere with free samples for now: Analog, Microchip, TI, etc. Microchip doesn't seem to have a very good line, whereas TI has chips for current shunt monitoring that only seem to go from +80v to -16v (no 170vac). The kill-a-watt uses the same technique with a simple lm2902 so how hard can it be, right?
I just wanted to share a "project" i've been messing with lately. A while back I noticed some cheap ($7) 240 x 160 lcds on eBay with touchscreens. Obviously I had to buy a few of them. Since then, they've just been sitting on my desk shelf downstairs. Just recently I got around to hooking one up and making a little proto board for it.
The thing about these LCDs is that they don't have a controller -- you have to drive them yourself with a higher powered processor (PIC32). Luckily, Matt Bommicino from CafeLogic did all the dirty work and wrote up a sweet site (including firmware) about how to drive these screens.
The coolest part of this is that I left my PicKit2 programmer at home, so I had to run it all from a ChipKit Uno32 using the default bootloader. As you can see in the third picture (top right corner), the code uses about 25% of the processor time. This includes a stripped-down version of Microchip's graphics library (ie clock, buttons, text, etc).
Here's the kicker: If you already have a PIC32 lying around, this thing costs you under 15 bucks to set up! A graphic lcd with backlight and touchscreen for <$20?! I'll gladly share my board layout (once I fix it up for v2) if people are interested.
