Show Posts
Pages: 1 ... 330 331 [332] 333 334 ... 451
4966  Forum 2005-2010 (read only) / Development / Re: Shielded USB cable mfgr needed.... on: May 01, 2009, 12:12:22 am
ALL USB cables should be shielded, perhaps except for cables permanently attached to low-speed peripherals (like the highly-flexible cables on mice.)  It's part of the USB specification.

I vaguely recall hearing good things about, but it was probably in relation to serial cables; I don't know about USB.

Given economies of scale, it may be cheaper to buy off-the-shelf USB (A-B) cables and cut off the end you don't need.  Even in 1000 quantity.   If you do go the custom route, you might want to consider some sort of pluggable connector (0.1 inch centers socket of some kind) or some sort of metal stress-relief "thing"; bare wires attaching to PCBs and such are very subject to breakage...
4967  Forum 2005-2010 (read only) / Development / Re: Pinguino on: April 24, 2009, 07:26:30 pm
Theoretically, you can make an Arduino-like environment around any CPU with a C compiler.  It need not even be an open-source C compiler, or available on all operating systems (though those are big features for Arduino.)
I even think it's a good idea, more or less (depending on how close you manage to maintain compatibility with Arduino and its principles.  ARMuino, PICuino, Powerunino, Colduino, AVR32uino, Beagleunino, 8051unino, etc.  More power to them!  Yeah verily, we are all alike in more ways than we are different!)

That said, "Pinguino" is a lousy name.  My first thought was something with Ethernet built in, and my second was one of those ucLinux "picosystems."
4968  Forum 2005-2010 (read only) / Development / Re: Freeduino Suppliers? on: April 23, 2009, 05:21:05 pm
If you're ordering enough units on a continuous basis (say, you're a school), you might also look into what it takes to become a "distributor" instead of an end-customer, of the official boards.
4969  Forum 2005-2010 (read only) / Development / Re: Freeduino Suppliers? on: April 22, 2009, 09:28:40 pm
I think that most of the distributors of the Boards labeled "Freeduino" ARE the manufacturers of the boards as well.  This is certainly true of NKC,  MDC, and SeeedStudio for example (although those are three different "sorts" of Freeduino.  "Freeduino" as name is not controlled at all, so anyone can call anything a freeduino.)

This is part of the motivation for "Freeduino" - by knocking out a level of distribution, you (theoretically, anyway) increase profit margins (or decrease costs) by a significant amount.  I would expect that most of the Freeduino distributors might be open to discussions about buying bare board if you're talking about buying lots at one time.  (some already sell bare boards.)  Or you can take the design files and send them out to a PCB shop and get boards made on your own.  (so far, it looks like contract manufacturing to get complete board rather than bare boards or kits, has not been cost effective.)
4970  Forum 2005-2010 (read only) / Development / Re: Mega 1280 Daughterboard? on: April 28, 2009, 03:44:31 pm
- should be fairly cheap, the chip seems to be $10 or less
The problem is that by the time start with the $10 chip, design a PCB board, add pins to plug into the existing socket, and have the whole thing assembled, figure in the markups  at least one level of distribution, you're looking at a pretty expensive proposition...

The 328 is a painless upgrade.  I don't think the mega1280 would be (and ... how do you wire the conflicting pins like SPI?  Arduino compatible?  Or MEGA subset compatible?   Issues, issues...)
4971  Forum 2005-2010 (read only) / Development / Re: Whats wrong with my device? on: April 15, 2009, 03:56:05 pm
I checked the current, and with the dial at 200m it read 108.5.
That should be fine overall.  Let us know if the bad connection fix doesn't seem to fix it afterall...
4972  Forum 2005-2010 (read only) / Development / Re: Whats wrong with my device? on: April 14, 2009, 03:37:20 pm
The current and voltage are intimately related (V=I*R), and the current everywhere in a series circuit is the same, so it sounds like you have it wired right, even if your understanding is slightly off (the voltage across the LED will be approximately constant regardless of current, and if the voltage of the supply is greater than the voltage of the LED, in theory LOTS of current will flow, which is why you need the resistor.)  What value of resistor are you using?

While putting an LED directly across a 5V supply will probably burn out the LED, the AVR in the arduino has some internal resistance , and will USUALLY not do so.  But you can get yourself into trouble expecting that to scale to lots of LEDs, or in other circumstances.  You should have resistors, but just because your LEDs haven't burnt out doesn't mean you have the RIGHT resistors.
4973  Forum 2005-2010 (read only) / Development / Re: Whats wrong with my device? on: April 14, 2009, 01:03:21 am
My math says 20ma * 12 = 240ma per hour for the leds.. add a little for the arduino running at 16mhz.  On three AA batteries that should get a few weeks of use right?
Well, no.  A set of AA batteries is probably good for about 2500mAH of power, so they'd run out of juice in a 240mA circuit in about 10 hours.  On the other hand, it still shouldn't flake out after less than one hour.

OTOH, even 240mA is above the limit that is supposed to be allowed for the whole chip (200mA)

You DO have current limiting resistors of some sort in there, don't you?  An LED will happily consume more than it's rated current without a resistor, and doing THAT 12x over could be a problem in several ways.  I would be inclined to use ~100 ohm resistors for each TWO leds (in the side to the power rail), which would preserve your ability to do patterns and such and give you "full brightness" if you happened to only light one of each pair, and reduced overall current if you have both LEDs in a pair turned on.
4974  Forum 2005-2010 (read only) / Development / Re: Capacitor values for external crystal oscilator? on: April 16, 2009, 03:43:15 pm
I suspect that one of the reasons that it is so common to have a crystal circuit where someone has just "thrown on" the "typical value" 18 or 22pF caps, rather than carefully analyzed the PCB with respect to the particular crystal they are using is that initial designs tend to be built based on crystals from "hobbyist" sources where there IS no detailed datasheet for the crystal.

The two 22pF caps from the arduino basic schematic will almost certainly work with any crystal you are likely to find from suppliers similar to the one mentioned in this thread.  If you need "as much accuracy as you can get", you might want to buy a crystal from a supplier that does have a data sheet, and then carefully tune the cap value while watching a highly-calibrated scope and/or frequency counter.   If you have a particularly "weird" capacitor (ultra mini smt, for instance), you might worry about it more...
4975  Forum 2005-2010 (read only) / Development / Re: Capacitor values for external crystal oscilator? on: April 13, 2009, 12:12:05 am
Don's equations are correct, but the 30pF capacitance that you (the OP) quoted is pretty far outside the normal range of such crystals; digikey doesn't list any 16MHz crystals with a load capacitance above 20pF, for instance, and assorted Atmel data sheets and Microchip App notes say that 17pF is "typical."
Almost everyone throws a couple of 18-22pf caps in there and has it work right, despite the fact that you SHOULD take a bit more care...

4976  Forum 2005-2010 (read only) / Development / Re: Good Wire to PCB Option? on: March 25, 2009, 08:18:20 pm
See also: which includes some part numbers...

The big decision is how small your want the connectors.  getting them on 0.1 inch centers is possible, but seems to be pricier than the "normal" size of about twice that.
4977  Forum 2005-2010 (read only) / Development / Re: Shield-Duino on: March 25, 2009, 08:21:06 pm
One problem is that the Arduino form-factor is pretty limited on the amount of space you have at the edges for putting connectors, especially if you're trying not to interfere with the "tall" components on the main Arduino board...
4978  Forum 2005-2010 (read only) / Development / Re: Shield-Duino on: March 24, 2009, 03:59:40 pm
I see no value in all these shields that "bring up" the ICSP headers from the original board
The original (and followon) Zigbee shields use the ICSP header for power (grr.)  Several shields (mostly proto shields?) that pre-date diecimila  use it to extend "RESET" of the arduino to the shield (button, or for bootload-via-shield capability.)  I don't like it either :-(
4979  Forum 2005-2010 (read only) / Development / Re: Shield-Duino on: March 24, 2009, 10:29:24 am
this could be "multi" core, if designed correctly, right?
Yes, assuming SPI or some multi-drop serial protocol.  But mostly it probably equates to an easy way to make a stock arduino into a "dual core" arduino.  

What about having two set of headers?
So that one could stack an additional shield on top of shielduino.
Yes, that's the plan.  The original arduino shield connectors would stack, and the equivalent connections form the "coprocessor" would go somewhere else.  (next to the proto area in the pcb pictured.)  That lets you add additional shields to the original arduino, but doesn't let you add shields to coprocessor.
4980  Forum 2005-2010 (read only) / Development / Re: Shield-Duino on: March 24, 2009, 02:41:37 am
Works fine if your logged into picasaweb as me :-)
Fixed?  maybe?  It's just a rough draft picture anyway...
Pages: 1 ... 330 331 [332] 333 334 ... 451