[*plonk*]

There is something missing on this forum (or at least I can't find it) that can be used on some other forums: the function 'ignore posts from this user'

Oh, for the days of usenet news; the crudest newsreader I ever used was more sophisticated than the best forum software.

*plonk*

A note/caveat to Jack's advice:  you can add coax, but not too much.

If you're using the 2.4GHz xbee (which is most of them), coax attenuates the signal quite badly.  I wouldn't use more than a few inches of coax, 10 or 12 max at 24.GHz.

The U.FL connector xbees can also use an antenna with coax (I think all the U.FL antennas have a feedline, but I could be wrong).

Don't forget that the xbee is a serial device, so it may be feasible to move the whole xbee to a different location.

More information would be helpful.  Like Jack says, what problem are we trying to solve?

-j

I find it interesting that new programmers often don't read/understand error messages.

No kidding.  I worked helpdesk in the computer center while an undergrad, and you wouldn't believe how many times I had these conversations:

#1:
User: My program keeps giving me a divide by zero error.  What does that mean?
me: you're dividing by zero.
User: Oh.

#2:

User: My program keeps giving me a divide by zero error.  Why?
me: you're dividing by zero.
User: no I'm not, I'm dividing by N.
me: Well, the value of N is zero.
User: Oh.

it appears to just be a single wire no complete circuit

You're half right.  It probably is a single piece of wire, but it has some specific characteristics - most importantly, its length is equal to 1/4 of the wavelength of the signal being radiated (as measured in the conductor, which is slightly different than the free space wavelength).  This geometry allows the signal to radiate.

If it was an inefficient design (incorrect length, etc), at least some of that power would be reflected back into the power amplifier (special transistor) output, which could damage the amplifier.  At best, the signal isn't getting out of the radio as intended.

-j

It's an antenna, not a random piece of wire.  The length of the antenna element is a function of the frequency.

Depending on the power of your transmitter and how bad the mismatch is, it can cause behavior anywhere from "it doesn't work" to "you just let the magic smoke out of the power amplifier".

There are xbees with coax connectors so you can use an antenna with a short feedline.

Another solution is to move the whole xbee, but there are limits to how far you can run the TTL level serial connection between the xbee and whatever it's plugged in to.  Those limits depend on noise, and a robot can be a noisy environment.

-j

I've seen a few videos on the milling process and they didn't look fine enough to allow for SMD mounting.  Maybe I'm seeing the wrong videos?

Or someone is using the wrong machine.  Our guys make PCBs for SMT components on a Haas TM pretty regularly.  Takes all freakin' day because the mill only has a 4000 RPM spindle (it's designed for machining, not PCB milling) but other than that it works great for small stuff.  I know they've done 100 pin LQPF, I think some no-lead stuff, and smaller passives (0603 for sure, and I'm fairly certain 0402).

-j

OSHpark will do shapes.  You pay for the rectangle that your shape fits inside.

The FG-100 read ...

300C

It was not even off by 1 deg at first, then went up to 302C after a bit.

So, which one is correct?

LEDs will fall silent

If your LEDs make noise, you may be doing it wrong...

Pretty difficult with just one MCU class one semester just my 2 cents. Only those motivated to do more will learn more.

Well, you've got to look at the whole program before getting the idea that it's only one semester.  Start with the freshman level intro to engineering where they'll do a project from a kit or an arduino project, add in a semester of programming and another of data structures, progress through the circuits and electronics labs where they play with resistors and power supplies and transistors, go through logic where they see gates and FPGAs, spend a semester in the embedded systems course, and finish off with a semester or two of senior design where they'll do something a bit more sophisticated, and they have a chance to come away with a reasonable education, especially given that they have to take literature and history and other non-technical stuff as well.

As for side projects for the motivated, we've got a student club building a satellite.  Delivery for payload integration on April 1, launch currently scheduled for mid-July.

-j

When you're in college, you're not there to learn about PIC or ATmega or any other specific product; you're there to learn how to use a microcontroller to solve a problem.  When you're in the real world, you'll do the math (e.g. like Osgeld says on cost of components vs. size of the run), look at the standard tools your company/group uses, or maybe just "use X because we've alway used X".  If you and your prof did your jobs, you'll know enough about solving problems with microcontrollers that you can adapt to the specific tools at hand.

-j

12 volts 50 amps running though 22 gauge wire connected by twisting the ends and a little bit of electrical tape covering it

Isn't that called a fuse?

-j

Even none polarized pins like these have same exact issue

Yep, that web page says:

They work with 22 to 28AWG wires, though 22AWG wires with thick insulation might be difficult to fit into the pins.

Which is probably from the datasheet, and more or less what I was getting at with the "thinner insulation".

There's not a "standard" other than the manufacturer's datasheet for that particular connector. I have some Molex connectors on hand that are for 0.1" headers with 0.025" square pins, and there are two different connectors - one for 22-24ga wire, and another for 24-30ga wire.  I think there's a third IDC variety as well.

Blah blah blah. The long, correct and sometimes painful answer is to read the datasheet, and any datasheets or specs that it references, and use what they say (including the $350 crimping tool designed for it).

The short answer is probably to find some 22ga with thinner insulation.

Thanks for the link, BTW; I noticed Polulu had a crimp tool a bit nicer that the one I'm using for these, that's still within my budget.

-j

Since I haven't seen these connectors and Sparkfun doesn't supply a datasheet, it's hard to tell for sure.  Sorta depends on what size your CAT5 wire is (most likely 24ga), and if it's stranded or solid. I would avoid solid conductor wire pretty much in any crimp connector, and any application where the wire will move.

It's possible you just need some different 22ga with thinner insulation.  If your CAT5 is smaller than 24ga, try 24ga wire.

-j

78H means 78 hex.  It's sometimes written as 78_H.

In C we write it as 0x78, not 78H (and definitely not 0x78H).

Did 0x78H actually get past the compiler?  It shouldn't  My gcc (solaris and linux) dies with "error: invalid suffix "H" on integer constant" when I tried 0x78H.

-j