Show Posts
Pages: 1 2 [3] 4 5 ... 9
31  Forum 2005-2010 (read only) / Development / symmetric board edge connector? on: March 12, 2010, 10:11:45 pm
Hi guys a couple of questions about hw parts that I'll bet you DIY guys may have experience with.

I'm trying to make a small Arduino clone and am looking for the cheapest way to connect two boards edge-to-edge using a symmetric connector.

A simple example would be to use 2 female right angle connectors and then put (floating) extra long male pins between them.  But when I price that out its .50 for each female RA connector and then there are the male pins.

But perhaps you guys know of something even simpler.  Perhaps there is a card-edge solution?


And on another subject, I want to be able to use an ATX (PC) power supply to power a board, but I really don't need THAT much power.  So I settled on the hard-disk connector as the perfect solution.  But everyone just calls it the "molex 4 pin".  I can't figure out what the male side (to mount on the board) is really called so I can buy some!  Anyone every buy some of these?

Thanks!
32  Forum 2005-2010 (read only) / Development / Re: Available shields and shield ideas on: October 28, 2009, 07:35:47 pm
Hi guys, I've worked up some spreadsheets of components and prices.  I'm ok with the AT86 chips, esp since they also do Zigbee (802.15).  

The key question is "can the SW be done within a single AVR or is a separate dedicated 1281 required?"

Frankly, there seem to be a lot of hardware choices in the "not-quite Arduino compatible" department (dig into Inopia's link and you see a bunch), and there is the Arduino/XBee solution.  

So we have to fit in this ecosystem.  I think that that means 100% Arduino or Arduino Mega compatible and cheaper than the Arduino/XBee.  

 Looking at the prices -- this means no module (after all, the atmel module is not much cheaper than the xbee).  Just an AVR, the RF230, voltage level converter CPLD, (and maybe option to have USB -- or use the USB FTDI cable).

What is extremely interesting about this to me is that the 328 base cost of materials is $8 if we made 50-100 or so.  This is just those chips; the addition of the Arduino-compatible female headers, voltage regulator chip etc would push the hardware cost to about $11-$12.

At that price point (say $10-$15) you really could think about buying 5 to 20 of them on an garage-tinkerer's budget, for wireless sensor fun!

And of course we could also do a 1280 (its a $10 chip) version both for inopia's work and also perhaps as the "base station" for all of these smaller nodes.

This would be a lot cooler than the current XBee stuff... when I've looked at the projects (and don't get me wrong, they are cool!) they are mostly controlling [one thing] through an iphone; probably because it takes $90 to buy 2 of them so people only buy 2 and hook 1 up to the computer as the "router" and have one sensor.

I'd like to truly enable Arduino users to do sensor networks cheaply...



33  Forum 2005-2010 (read only) / Development / Re: Available shields and shield ideas on: October 15, 2009, 10:38:44 am
@Inopia: can you open source your work so our little device will have some pretty cool sensor network infrastructure?  Or are you working behind an iron curtain? :-)  I'm interested in cooperative robotic, formation flying RC airplanes, etc so the kind of advanced sensor network that you are talking about is an essential building block.

34  Forum 2005-2010 (read only) / Development / Re: Available shields and shield ideas on: October 13, 2009, 11:16:15 am
I have fear that they can only drive the chip at 4Mhz because the 230 is so sensitive to digital interference that it can't handle higher clock rates...
35  Forum 2005-2010 (read only) / Development / Re: Available shields and shield ideas on: October 12, 2009, 10:09:32 pm
Well, using the module won't be as fun :-(, but it will be a lot more likely to actually work :-)... but it does make the problem a LOT easier!

Anyone have an issue with the fact that its 3.3v?  It does not look like the pins are 5v tolerant.  So to expose the arduino pinouts at 5v we'd have to use the CPLD I've been talking about.  We could push all the Arduino pinouts into the CPLD, and then connect the CPLD to this module.  So you could actually reprogram the CPLD to remap the arduino pinout to the 1281 pinout. That will be fun! :-)

Given that there are a lot fewer pins on the 1281 then on the 1280, I'm wondering if there will be enough to push out to the arduino pins... we'll have to check to see if all CPU pins are exposed or if some are used within the module.

Finally, my major concern is that they seem to be operating the CPU at 4mhz.  Anyone else see issues with that?



36  Forum 2005-2010 (read only) / Development / Re: Available shields and shield ideas on: October 10, 2009, 09:01:21 am
@inopia:  That's too bad about the 128.  Yes I had heard that the 1280 isn't fully compatible either.  However, it is not so bad since the MEGA is out there so people are working on new libraries, etc.  
Actually, I can try to put the QFN-32 328 footprint inside the TQFP-100 footprint so that either chip can be chosen when the board is populated.

@novice:  Yes they sell bundles but they are not really any cheaper (and not stocked).  Also, all of the major RF chips have example implementations with PCB designs to make it easier to deal with the RF side of things.  By copying the exact design and being careful about RF on the digital side we shouldn't have to use an antenna specialist.  You also ask if we are aware of the modules.  Yes, we are.  This issue is that you can buy the RF chip for $2.36 but the module costs $17.  If you look at the XBee solution for Arduino the breakdown is like this: Arduino: $30, Xbee module $20, Xbee shield $20.  So the total cost is $70.  That is a LOT especially when you consider that the point is communications, so you need to buy at least 2 but 10 would be really cool.  So what does the difference between $2 and $17 pay for?  I'm guessing a lot is just "engineering/lab" markup... but I am sure that there will be some extra costs that we will discover as time goes on. :-)



37  Forum 2005-2010 (read only) / Development / Re: Available shields and shield ideas on: October 09, 2009, 03:53:30 pm
Inopia,  

Ah, I thought that price was for mega + wireless not just the shield!

I haven't examined the register map on the 128 so I don't know if it is 100% compat.  However, as I think I mentioned in my blog, the 640 and the 1280 even 2560 are pin-compatible if the TQFP-100 package is used.  So that would give us some price flexibility.  I think I have come full circle and realized that this original choice is the best.  However I am still a bit worried about software compatibility since I think that the arduino MEGA is not entirely compatible with the 328.  But I guess that since there is no 648 we really have no choice here.


38  Forum 2005-2010 (read only) / Development / Re: Available shields and shield ideas on: October 09, 2009, 08:17:38 am
Yes the CPLD will be useful first off simply as a voltage level converter.  But also for a lot of small logic like stepper motor control and PWM.  Yes, I know that the full 8051 is not NEEDED :-) but it would be cool if the price is low enough!

What were you thinking of in terms of board price?  Earlier in this thread you were talking $20.  But a ATMEGA1280 costs $15 itself so that price is impossible for that chip!  With the CPLD at $4-10 and the wireless chip around $5, the 1280 will cost as much as the other components combined.  On the other hand the 328 is only $2.70 USD ($2.56 lowest price, I'm sourcing some now for another project).  So the 1280 could double the cost of the board.  On the other hand, it gives you a LOT more functionality...

WRT MSP430, I want to stay within the Arduino family so we don't have to do a lot of portability software.  For this reason the 328 is also the better choice because it is directly register compatible with the Arduino, whereas the 1280 would be compatible with the Arduino mega...

Is your routing work in C or Java?

Best,
Andrew
39  Forum 2005-2010 (read only) / Development / Re: Available shields and shield ideas on: October 07, 2009, 10:48:12 am
WRT price, as cheap as possible since customers will need to buy a few!  But its always a tradeoff between price and performance.  Should I use the 328 at $3.00 a chip or the 640 or 1280 at 10 or 15 dollars?

I'm open to suggestions for the wireless chip.  I have experience doing SMT and QFN parts so that is not an issue.  It looks like all the chips have the same tricky antenna requirements.

I want the board to be usable in sensor networks within a home/office setting and for small cooperative robotics.  So I want a range > 10 feet! :-)  Otherwise I am pretty open to suggestions.  It would be nice to offload as much processing as possible from the AVR, so that it can be used for the job at hand.  Lets go over the options and discuss pros/cons.  Please feel free to add to this list:

1.  Nordic Semi: nRF24LU1+
PROS:
Has an onboard 8051 and RAM buffer so basically there would be 0 load on the AVR
Has a pin/algorithm compatible versions that use different frequencies.

CONS:
Can't find a source or a price

2. TI CC2500
PROS:
In common use
Arduino library exists

CONS:
older?

price (100): 2.38

3. Atmel AT86RF230
PROS:
Single manufacturer solution, provides software and demo boards
Zigbee

CONS:
Specs on the antenna seem pretty scary compared to other chips;  it seems to be a lot more sensitive to interference (or maybe the spec is just being honest :-))

4. Cypress CYRF7936
2.4 ghz Has been used in a RC airplane environment so has good range
Comes with software (but for the Cypress PSOC)

costs: $3.33

CONS:
not Zigbee

40  Forum 2005-2010 (read only) / Development / Re: Available shields and shield ideas on: October 06, 2009, 07:58:59 pm
I am planning on making a wireless + AVR arduino-compatible combination.  See
http://effluviaofascatteredmind.blogspot.com/2009/07/arduino-cpld-reconfigurable-hardware.html.

Actually, I did the complete eagle files for a shield but did not actually send it off because I wanted to integrate the wireless.  

41  Forum 2005-2010 (read only) / Development / Re: Big LED Ambient System on: October 18, 2009, 05:15:06 pm
Also you might be a bit disappointed by the color "mixing" with RGB leds.  They really are just 3 leds in one package and in most cases you can very much see distinct red, green and blue circles if you turn them all on and shine the result on the ceiling.  But don't worry, this is easy to fix with a diffuser -- trivially a piece of white paper will do the trick, but there are better choices...

42  Forum 2005-2010 (read only) / Development / Re: Big LED Ambient System on: October 16, 2009, 09:02:24 pm
Youran,

I've sent my boards to the UK, Canada, etc... I don't see why Belgium would be an issue.  Here's a link my blog about the new board http://effluviaofascatteredmind.blogspot.com/2009/10/lightuino-design-thoughts.html.  

These boards let you control 70 current sinks separately.... But you can always hook more than one LED up in series to the same current sink!  The chip takes 13volts maximum so you can do maybe 4 to 6 in series (depending on the voltage of the LED) without any other parts.  But if you put an NPN or N-channel MOSFET transistor between the current sink chip and the LEDs then the transistor will protect the chip from higher voltages so you can essentially put however many LEDs as you have voltage for in series (of course the transistor has to be able to handle the voltage).

But if you really just want all LEDs on or all off then the MOSFET + resistor is the simpler solution.

In terms of wattage, its best to try to use a combination of LEDs in series and parallel to consume as much of your voltage as possible (because as ajb said, what is left over is just dissipated in the resistor as heat).  For example, if you have an LED that uses 3 volts, then if you use a 5volt supply you would put the LEDs in parallel and the resistor would consume 2 volts; so 2/5 of your power is going into heat (at least -- the LED also generates heat).

But in that case, if you used 4 LEDs in series, that would sum to exactly 12volts, so in theory you would not need any resistor, and not lose any power in the resistor.  But of course, the LEDs aren't perfect so in practice you'd probably burn them out... instead in this case you'd need to drive them at slightly less than 3v and put some small value resistor in there so that the circuit has a linear (v=IR) component

This example was totally fabricated to show how using the 12 volt rail might be better than the 5 volt... in practice you'll need to see what voltage your LEDs dissipate (and it varies by color).

Also I think that computer power supplies generate -12 volts as well, so you could conceivably get a 24v out of it.

43  Forum 2005-2010 (read only) / Development / Re: Big LED Ambient System on: October 13, 2009, 01:59:44 pm
Yes, you need three resistors per RGB LED.  Or use a constant current chip.  This latter choice is preferable since it stops burnouts if one LED shorts, and also makes color intensity matching easier.

You'd hook the anode to the positive, and then each cathode line (R, G and B) to a separate resistor and then to the Arduino.  You can hook a couple of these up in parallel (the anodes together) and the all the R's together, G's together, etc.  Of course, you can't exceed the amount of current you can sink.  

You can't hook the RGB LEDs up in series... to do that you'd need separate R, G and B LEDs.

With a constant current chip like M5451 (allows 35 separate control lines), you leave out the resistor.  

WRT the power, use the 5v rail, because you can't hook the LEDs up in series and they take less then 5v to turn on.

Yes, the Arduino can be used when the computer is off.  It will run thru the sketch you last uploaded.  Every time you change the sketch you have to upload new code.  That takes time.  So don't change the sketch.  Instead use the Serial library to get data from the computer.

44  Forum 2005-2010 (read only) / Development / Re: Where to buy multicore - arduino boards? on: October 19, 2009, 09:53:13 am
@florinc:  Nice!

@Osgeld:  My point was simply that before you go to multi-core Arduino, go to a faster chipset.  I chose to highlight an architecture with the fastest chipset for under a typical hobbiest budget of $100.  But I agree, there's lots of good embedded choices in between.  Although it is interesting to note that they often cost more than a standard PC because of the lack of volume, or maybe because the market will bear it; the dev kits are targeted at embedded development companies which can spend $400-$1k for a single board without blinking an eye.
45  Forum 2005-2010 (read only) / Development / Re: Where to buy multicore - arduino boards? on: October 19, 2009, 09:45:49 am
@florinc:  Nice!

@Osgeld:  My point was simply that before you go to multi-core Arduino, go to a faster chipset.  I chose to highlight an architecture with the fastest chipset for under a typical hobbiest budget of $100.  But I agree, there's lots of good embedded choices in between.  Although it is interesting to note that they often cost more than a standard PC because of the lack of volume, or maybe because the market will bear it; the dev kits are targeted at embedded development companies which can spend $400-$1k for a single board without blinking an eye.
Pages: 1 2 [3] 4 5 ... 9