Sorry to anyone here looking for dessert, but I’m talking about the tiny and bizarrely inexpensive computer called Raspberry Pi. The kit I got included a T-shaped board called a cobbler, or breakout board. It allows you to connect pins on the Pi to a breadboard for doing wiring and other super nerdy stuff. Now, the Pi has 40 of these pins and the stroke of the cobbler’s T also has 40 connectors, so that makes sense. However, the cross of the T has four more connectors, a positive and negative for each of 3.3V and 5V power. I don’t get how these are mapped to pins. Do they duplicate some of the other connectors? In particular, only one pin on the Pi is labeled 3.3V. How can there be both positive and negative on the other end? Maybe one goes to ground?
I feel like understanding this is key to my not frying up my Pi, and possibly perishing in the resulting fire. So, I guess need answers sorta fast?
They just copy the 3.3 V and ground. There’s no negative voltages on the Pi. This is commonly done and not a big deal.
The main thing is that whatever you connect to those pins cannot draw more than the 3.3v regulator on the Pi mainboard. A few hundred mA, maybe. Enough to run a small microcontroller or external chip, I suspect.
You’re right, DPRK, there are two 3.3V. I missed the second one because it’s separated from the first. Stupid me. I’m using the cobbler from CanaKit, like the the one here, just a bit different from the one that DPRK is showing.
SamuelA basically confirmed what I thought, i.e., the extra four connections are just copies. I guess it could be 1 or 17, and 9, 25, or 39, and 2 or 4, and one of the many grounds on the 5V side. It doesn’t matter to me which, right? Also, just to confirm, the +/- columns on either side of the breadboard don’t add anything but convenience, right? I could wire up the same circuits without them. Well, maybe not me, but someone who knew what they were doing could.
Yeah, it’s common breadboarding practice to use the two rails on either side as power and ground. I believe my breadboards are even marked that way, but I’m too lazy to look.
The advantage with that setup is that you always have a power and a ground location close to where it’s needed in your circuit. And ignoring current limitations for a moment, it means you can have as many power and ground connections as you need.
It just makes your breadboarded circuits much more tidy, really. Easier to setup, diagnose, add add.
The thing the OP doesn’t realize is that in reality, the RP has a plane of copper. One entire layer of the PCB (traditionally it’s near the top but it could be lots of places) is mostly just interconnected copper, broken only by vias, that is connected to the 3.3 volt power source.
This pin on the connector is tied to that plane. You can get lots of current through that type of pin - on the order of amps.
So all the power comes from the same place. It’s totally ok to subdivide that power. The actual limiting factor is the voltage regulator component on the Pi is limited in how many milliamps it can supply.
A bit of googling says the limit is 400 milliamps. That’s a bunch. An arduino draws only 15 mA.
Thanks for the link. I completely concur with everything SamuelA said, in particular that they are just duplicates of the other power pins, that “-” = 0V = ground, and that you should always make sure your circuit does not overload the power pins or any of the I/O pins.
Since you probably have a digital multimeter handy for working with electronics, one thing you can do is put it in a mode that beeps whenever there is an electrical connection; then you can touch one lead to, e.g., 3.3V “-” and the other to a ground pin, to confirm that they are wired together.
As for those +/- rails on the breadboard/protoboard, that is completely standard and just for your convenience. Note that each one of those power rows is all connected together, unlike the columns labelled 1 through 30 which give you groups of 5.
A Raspberry Pi is closer to being a desktop computer in terms of computing ability, whereas the various Arduino systems are closer to being the kinds of embedded systems you’d find running a microwave oven. Power draw scales with computational ability.
The great thing about the Pi is that you can put it on the LAN, ssh into it, and run the compilers and all of the other development tools on the same piece of hardware all your peripherals are hooked to. It’s completely self-hosting. For the Arduino, you have to run specialized software on a host computer and transfer binaries over. That said, I’m sure doing that with the Arduino is much less of a pain in the ass than doing it with Real Serious Single-Board Computers from Real Serious Companies like TI, if my experience with the LaunchPad is anything to go by.
And, of course, you can use the Pi for a lot of other things, too, like being a media player or a compute node or a web server or a firewall or a proxy…
Well. Kinda. The problem with the Pi is that it’s a desktop computer. All the complexity, maintenance requirements, non-realtime operation. It comes down to what you are trying to accomplish.
If you are trying to do something simple, the Arduino is the easiest way. If you’re looking for Real Serious solutions, you can use PICs and this cool compiler called PICC that costs $450 for a license and is worth every penny. It makes building a basic embedded system very rapid and only slightly harder than using an arduino. You can slap a pic and the 8 passive parts (a few 0.1 microfarad caps, a couple resistors) it needs in order to run onto a circuit board for very little money and effort and program in your firmware just by plugging a cheap emulator pod into a connector on the board.
These kind of bare metal solutions tend to be very reliable, respond and startup instantly, and so if you are trying to run a microwave oven, that’s what you gotta use.
TIs stuff…well…the thing is, TIs stuff is enormously harder to use, yes. But their chips have blistering fast performance in comparison to PIC, and you can use them in extreme environments. Perfect for the military market. I’ve done this as well, and while it took me a lot longer to setup, the actual chip is extremely reliable and fast.
There’s a new kid on the block, though it’s expensive enough that it’s not really for hobbyists. Intel has a Pi-like product that has a GPU meant for machine learning. Called the Nvidia Jetson. There’s some incredible stuff you could do with that. You could make a quadcopter that can fly on autopilot, see around with a camera, and do acrobatics using straightforward machine learning techniques.