|
Metajo Cum Dumpster posted:Oh crap, so it has to be inline with the circuit? If you have an instrumentation amplifier or feel like building one with your favourite op amp you don't need to purchase anything else.
|
#
?
Jan 8, 2010 18:17
|
|
Told you I was a beginner. Thanks.
|
|
| # ? Jun 3, 2024 17:22 |
|
|
Metajo Cum Dumpster posted:Oh crap, so it has to be inline with the circuit? Haha, that thermocouple is a steal at the price of a thermistor (SFE's ability to correctly fill an order strikes again!), but isn't as easy to interface with. What are you reading the output with? What are you doing with it? What temperature ranges are you trying to measure? The thermocouple should output a small voltage depending on temperature, and usually you need some sort of high-quality amplifier to get the output high enough to be useful (hence the instrumentation amplifiers). You might expect a thermocouple to typically change ~40uV for every deg C; this is a tiny change and can be hard to deal with. Not to mention the possibility of parasitic thermocouples when you try and connect it together. From my (small amount) of experience, there are a lot of considerations with thermocouples! If you are using a microcontroller, something like the Thermocouple Amplifier Digital MAX6675 might make reading the thermocouple a bit more straightforward.
|
|
#
?
Jan 8, 2010 19:18
|
|
|
Wow, if you order that amplifier from Digikey it's almost the same price after shipping than the unit price before shipping from Sparkfun. Even if you're getting it shipped to Canada and paying in Canadian dollars. gently caress Sparkfun. Anyways loads of instrumentation amplifiers have application notes in their datasheets for how to use them with thermocouples, but I didn't know they made amplifiers designed for them.
|
|
#
?
Jan 8, 2010 19:34
|
|
|
BattleMaster posted:Wow, if you order that amplifier from Digikey it's almost the same price after shipping than the unit price before shipping from Sparkfun. Even if you're getting it shipped to Canada and paying in Canadian dollars. gently caress Sparkfun. It's not JUST an amplifier. You read the temperature over SPI.
|
|
#
?
Jan 8, 2010 19:55
|
|
|
Delta-Wye posted:It's not JUST an amplifier. You read the temperature over SPI. Oh that's pretty awesome.
|
|
#
?
Jan 8, 2010 20:02
|
|
|
Delta-Wye posted:Haha, that thermocouple is a steal at the price of a thermistor (SFE's ability to correctly fill an order strikes again!), but isn't as easy to interface with. What are you reading the output with? What are you doing with it? What temperature ranges are you trying to measure? Thanks. Honestly I had no idea what I was going to do with the thermistor, just ordered it with a bunch of other cheap sensors and resistors so I could go through and tinker with them later to learn. Googling that IC lead to this schematic, it look like it'd work on an arduino instead of a pic (w/ diff code)? http://www.ivcity.com/jesse/thermocouple/info.html
|
|
#
?
Jan 8, 2010 20:27
|
|
|
Metajo Cum Dumpster posted:Thanks. Honestly I had no idea what I was going to do with the thermistor, just ordered it with a bunch of other cheap sensors and resistors so I could go through and tinker with them later to learn. The schematic is pretty much overkill - all you should need to reproduce is the three lines connecting the digital amp to the PIC (chip select [CS], serial clock [SCK] and slave out [SO]). The datasheet as all the info you SHOULD require, although deciphering a datasheet is a skill in-and-of itself. With such a simple SPI device (you're not even sending data to it) you can probably bit-bang it manually using digital ports pretty easy, although I'm not sure how much direct access an arduino would give you. You pull CS low, and toggle the CLK line while reading the SO line as you go - as far as interfacing goes, it's super straightforward. Hell, with the CLK line (the clock is the major differences between synchronous and asynchronous communcation, aka SPI vs RS232) you should find that the timing can be fairly loose and still will work well. Battlemaster posted:Wow, if you order that amplifier from Digikey it's almost the same price after shipping than the unit price before shipping from Sparkfun. Even if you're getting it shipped to Canada and paying in Canadian dollars. gently caress Sparkfun.  Sparkfun: $11.95 price $10.76 10-99 (10% off) $9.56 100+ (20% off) Mouser: 1: $13.92 25: $10.59 50: $7.05 Digikey: MAX6675ISA+TDKR-ND Price Break Unit Price Extended Price 1 14.05000 14.05 25 10.60000 265.00 50 9.54000 477.00 100 9.27500 927.50 MAX6675ISA+-ND Price Break Unit Price Extended Price 1 13.86000 13.86 25 10.46000 261.50 50 6.60000 330.00 100 6.11910 611.91 MAX6675ISA+TCT-ND Price Break Unit Price Extended Price 1 14.05000 14.05 25 10.60000 265.00 50 9.54000 477.00 100 9.27500 927.50 (as an aside, I sometimes hate digikeys selection. They have lots of varieties of chips, sometimes hundreds with common chips like the 555, and very little information to differentiate them. Often times even the datasheet is mum on the differences between the small variations.) Delta-Wye fucked around with this message at 20:56 on Jan 8, 2010 |
|
#
?
Jan 8, 2010 20:45
|
|
|
Oops I was rushing it because Sparkfun usually does blow for pricing especially on kits and prototyping stuff. It turns out the price I was looking at was for one version of the part that you had to order 5000 minimum of. Well Sparkfun wins this round  quote:(as an aside, I sometimes hate digikeys selection. They have lots of varieties of chips, sometimes hundreds with common chips like the 555, and very little information to differentiate them. Often times even the datasheet is mum on the differences between the small variations.) Common parts like the 555 are often manufactured by loads of different people with little difference. The 555 in particular usually only varies in maximum frequency and operating voltage range from part to part. You see that for common op amps, too.
|
|
#
?
Jan 8, 2010 21:08
|
|
|
Delta-Wye posted:The schematic is pretty much overkill - all you should need to reproduce is the three lines connecting the digital amp to the PIC (chip select [CS], serial clock [SCK] and slave out [SO]). The datasheet as all the info you SHOULD require, although deciphering a datasheet is a skill in-and-of itself. With such a simple SPI device (you're not even sending data to it) you can probably bit-bang it manually using digital ports pretty easy, although I'm not sure how much direct access an arduino would give you. You *can* bit-bang on the Arduino by accessing various registers directly, but there's an SPI library so you might was well use that, your code will be 10x shorter. http://www.arduino.cc/playground/Code/Spi (As an aside to your aside, I once had that happen when I bought an i2c EEPROM that has an 8-byte page size, but was referencing a datasheet for a nearly-identical part that had a 16 byte page size. That caused some errors in my program  )
Mill Town fucked around with this message at 22:11 on Jan 8, 2010 |
|
#
?
Jan 8, 2010 22:08
|
|
|
Delta-Wye posted:Words Thanks, I'll have to grab that the next time I order something. Sparkfun seems really hit or miss with pricing. The Arduino boards from what I've seen are averagely priced, dunno about the others. Stuff like bulk wire costs 20% more than same amount at Jameco, leadfree tinner/cleaner on digikey is half less than sfe, single resistors/caps look about 50% more than jameco. Digikey's endless filters confuse the hell out of me.
|
|
#
?
Jan 8, 2010 23:29
|
|
|
When searching Digikey be sure to check the box that says "In Stock" to make things easier. When you search for a part, you only see the filters if there's more than one page of results. If you don't know what paramters to choose, just click the link to see the first page of the results you have selected and start browsing. They recently put pictures inline with the search results which makes things a bit easier.
|
|
#
?
Jan 9, 2010 00:05
|
|
|
That thermocouple and amp could be fun with the arduino. You could learn a bit about pid controllers! I think it would be a better idea to go with spi option as pointed out already vs analog as will be way cheaper.. and you get to learn spi stuff too. If you want to go the analog route however, the AD595-AQ is popular and this kit could help you out: http://store.makerbot.com/electronics/electronics-kits/thermocouple-sensor-v1-0-kit.html Or since it's out of stock, grab just the pcb... or just steal the schematic!  Regardless of the amp you get for this, make sure it says something about being "cold junction compensated." Otherwise the temperature reading won't be accurate.
|
|
#
?
Jan 9, 2010 00:29
|
|
|
macpod posted:That thermocouple and amp could be fun with the arduino. You could learn a bit about pid controllers! What does a PID controller have to do with reading a thermocouple?
|
|
#
?
Jan 9, 2010 00:46
|
|
|
Delta-Wye posted:What does a PID controller have to do with reading a thermocouple? With a thermocouple and arduino acting as a PID unit, you can do neat stuff like.. make your own controlled soldering hotplate: http://mightyohm.com/blog/2009/01/diy-pid-controlled-soldering-hotplate/ Or expresso machine stuff: http://www.arduino.cc/playground/Main/BarebonesPIDForEspresso
|
|
#
?
Jan 9, 2010 01:51
|
|
|
I've got way too much reading to do. I don't even know what PID and SPI are. What book should I advance to after I've finished the really intro ones? Right now I'm trying to finish off "There are no electrons" and Forest Mimm's hand-drawn book.
|
|
#
?
Jan 9, 2010 02:21
|
|
|
If you're going to be working with microcontrollers a lot you'll want to read about I2C and SPI at least, because a lot of parts interface using one of those two protocols.
|
|
#
?
Jan 9, 2010 02:37
|
|
|
How would I go about actually programming/compiling to my arduino in C? I know they provide their Sketchbook development environment, but I want to strip away as many layers (short of assembly, at least) as possible. Also, I'd like to be able to compile executables, which I can't do within the Sketchbook environment.
Cyril Sneer fucked around with this message at 06:00 on Jan 9, 2010 |
|
#
?
Jan 9, 2010 05:57
|
|
|
avr-gcc and avrdude; arduino.cc provides a makefile if you still want to use arduino libraries but it's way out of date and doesn't work with the current distribution of said libs so you're better off just making your own for whatever libraries you may want to use
Blotto Skorzany fucked around with this message at 06:02 on Jan 9, 2010 |
|
#
?
Jan 9, 2010 05:59
|
|
|
So, I'm quite new at microcontrollers and picked up an Arduino a week ago (and I still get really excited when I make buttons and LED's do things that still seem complicated to me). The ultimate goal is robotics, which I've wanted to do since I was a little kid. What's the general limit of useful update rate for Ardupilot-like navigation with GPS? I see kickass tiny receivers with 20hz rates and it gets me all excited with the possibilities.
|
|
#
?
Jan 10, 2010 09:09
|
|
|
GPS isn't pinpoint accurate so the importance of the update rate is going to depend on how fast you are going. Under realistic real-world conditions there can be up to around 15m of horizontal inaccuracy with gps alone and up to around 5m horizontal inaccuracy with WAAS (Wide Area Augmentation System) support. http://users.erols.com/dlwilson/gps.htm http://users.erols.com/dlwilson/gpsacc.htm http://users.erols.com/dlwilson/gpswaas.htm You can certainly get a receiver with a high refresh rate, but if you aren't moving fast enough to escape the error thresholds, it is a little overkill. With that said a fast refresh rate can have some benefits to you, for example you can do some data smoothing to round out errors and potentially get a better bearing, then again some receivers do this for you already! These relatively large in robot term error thresholds are a reason you shouldn't try using GPS for tracking indoors. Another reason is because gps signals are pretty weak so you may not get a signal. Are you thinking of using the Ardupilot down the road? If so perhaps you want to pick up one of the gps receivers they sell for compatibility reasons.
|
|
#
?
Jan 10, 2010 17:05
|
|
|
I've been planning on picking up an ardupilot for a while, as I've been into RC stuff for a few years and have a whole lot of planes that'll make incredible platforms for it. I've got FPV gear as well. The only reason I haven't sunk the ~$200 into the board/compatible gps/ accelerometers is I figured it'd be a good idea to learn as much as I can with a basic arduino just so I know what the hell is going on. I know it's relatively plug and play, but I'm trying to avoid jumping in headfirst. I've got some decent RC truck platforms that will happily run outside, which is why I thought GPS, and ultrasonic for object avoidance. The plan is to start out like most people seem to with a little 3 wheeled avoidance platform and move to something more interesting like the aforementioned truck platform.
|
|
#
?
Jan 10, 2010 17:38
|
|
|
I was going to just edit this in, but it started getting long, so doublepost. I'm making a arduino powered water pump/temp and humidity sensor. The object is to pump a second or three of water with a RC pump onto a sheet that's used to humidify a bunch of tarantulas. I want to have a 20x4 LCD readout with the humidity and temp in 2 different places, then 2 pumps that turn on when a humidity threshold is reached. If I can make it do minor logging so I can hit a button and flip through different readouts on the LCD like average humidity/temp, max/min and current stats. A little piezo alarm when various things happen like low water reserve, low temp, etc. The shopping list thus far is; * A second Arduino (so I can keep dicking around with my first one) * Adafruit protoshield for the final design * 20x4 LCD (most seem to use up 6 pins, which isn't too bad) * Temp sensor (2) TMP36 The first problem is the water pump. I don't know if I should use a RC fuel pump, aquarium pump and a relay or a continuous rotation servo running a tiny little belt drive pump. The final option seems easiest, because I've got servos laying around, I just don't know where the hell I'd find a little tiny unpowered pump. The other problem is humidity sensors are amusingly expensive. This one is the best I've found thus far. Are there any problems with these 3 leg style sensors and extending them? They'll probably need to sit about 2-3 feet away from the arduino.
|
|
#
?
Jan 10, 2010 20:41
|
|
|
This one is a bit cheaper. http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=480-3166-1-ND For three dollars more you can get it on a breakout board at Sparkfun if you don't feel comfortable soldering smd.
|
|
#
?
Jan 10, 2010 20:48
|
|
|
Hypnolobster posted:
You won't need two Arduinos. Just buy another ATmega for tenbux or so. You can find the pinout on the Arduino website somewhere, or the proper datasheet on Atmel's website.
|
|
#
?
Jan 10, 2010 21:13
|
|
|
ante posted:You won't need two Arduinos. Just buy another ATmega for tenbux or so. You can find the pinout on the Arduino website somewhere, or the proper datasheet on Atmel's website. He may have trouble programming a fresh Atmega without a preflashed bootloader and interface hardware. Though if he actually has a project in mind that he wants to keep after constructing it he may as well make his own board for it and just keep the Arduino board for prototyping or something. At some point you need to graduate from using a newbie kit and learn how to do the stuff yourself.
|
|
#
?
Jan 10, 2010 21:19
|
|
|
It looks like sparkfun has atmega328's that are preloaded. The only problem is that I suspect I'll end up wanting to tinker with it a few months down the road (say, throw on some more sensors or something akin to the sparkfun logger shield) and at that point, I'd imagine that I'd need to put on interface stuff to work on programming and I'd really just be building another arduino. It's got to cost somewhere around $30 to do that (guessing here).
|
|
#
?
Jan 11, 2010 00:29
|
|
|
Take a look at your Arduino. The IC is seated in a little chip holder. You can pry it out and swap microcontrollers around without any soldering or anything. Use the same thing on your project, and that's all you'll need for future expansion. Just pop it off and plug it back into your Arduino to tinker around with your code. The preloaded atmegas sound like a neat solution if you don't to mess with bootloaders. There are advantages to using a full Arduino in your project. It has both a clock and a power source without you doing anything. I'm trying to discourage you from doing that, in case you couldn't tell. It's cheaper and your final project will look better and more professional.
|
|
#
?
Jan 11, 2010 01:05
|
|
|
Anyone know of a cheap uC that possesses 2 UARTs, 2 PWMs, and does not require 3.6v or less? I am working on a project that requires one, everything is designed at 5V already and I don't want to go back and remake everything.
|
|
#
?
Jan 11, 2010 06:41
|
|
|
Twerpling posted:Anyone know of a cheap uC that possesses 2 UARTs, 2 PWMs, and does not require 3.6v or less? I am working on a project that requires one, everything is designed at 5V already and I don't want to go back and remake everything. The Atmega164PA seems to be the cheapest AVR that has what you need. 2 UARTs, 6 PWM channels, about $4.40 in single quantities from digikey. e: also 2.7v-5.5v supply. You can probably find an equivalent PIC for the same or cheaper. e2: but that PIC might be 3.3v only so who knows? Finding a PIC is left as an exercise to the reader. For future reference, I used the table here to pick this out: http://www.atmel.com/dyn/products/param_table.asp?family_id=607&OrderBy=1262&Direction=DESC#760 Mill Town fucked around with this message at 09:18 on Jan 11, 2010 |
|
#
?
Jan 11, 2010 09:14
|
|
|
I think the only PIC18s with two UARTs are in the PIC18 J-series, which are 3.3 volt only. None of the lesser PICs have two and many of the 16- and 32-bit ones do but are also 3.3 volt only.
|
|
#
?
Jan 11, 2010 13:26
|
|
|
BattleMaster posted:I think the only PIC18s with two UARTs are in the PIC18 J-series, which are 3.3 volt only. None of the lesser PICs have two and many of the 16- and 32-bit ones do but are also 3.3 volt only. MAPS lists a fair pile of pic18Fs, 2-5.5V with 3PWMs and 2 A/E/USARTs. It lists individual parts for about $3-4 or so. Some even come in 28-pin SPDIP (.070 pitch DIP).
|
|
#
?
Jan 12, 2010 05:57
|
|
|
babyeatingpsychopath posted:MAPS lists a fair pile of pic18Fs, 2-5.5V with 3PWMs and 2 A/E/USARTs. It lists individual parts for about $3-4 or so. Some even come in 28-pin SPDIP (.070 pitch DIP). Oh I know why I didn't realize those parts existed, because all of them have 64+ pins except for a 28/40 pin K-series part that isn't actually available yet. None of the other 40 or less pin parts have 2 UARTs except some of the J-series parts. BattleMaster fucked around with this message at 06:09 on Jan 12, 2010 |
|
#
?
Jan 12, 2010 06:07
|
|
|
Hypnolobster posted:It looks like sparkfun has atmega328's that are preloaded. The only problem is that I suspect I'll end up wanting to tinker with it a few months down the road (say, throw on some more sensors or something akin to the sparkfun logger shield) and at that point, I'd imagine that I'd need to put on interface stuff to work on programming and I'd really just be building another arduino. It's got to cost somewhere around $30 to do that (guessing here). The Arduino isn't much more than an oscillator (16Mhz) and an ATmega 328. what takes up the space on the board is the FTDI chip for USB->Serial, the regulated power supply, etc. It would probably cost you $8 to make your own "non-fancy" Arduino. isagoon fucked around with this message at 18:26 on Jan 12, 2010 |
|
#
?
Jan 12, 2010 18:23
|
|
|
I decided to smoosh two PIC18F2420s together via I2C to create the interface I needed. This was probably the better idea since I can now access the flash without having to go through the main uC and suspend sensor data flow.
|
|
#
?
Jan 12, 2010 18:33
|
|
|
Ok, heres my problem: I want to build a device that has heating power of about 150 watts. And a fan. I want it to run off 120v/60hz grid power. So far, I have an AC fan and I'm planning to buy nichrome wire to make the heating element. quote:1. Gauge 32 AWG 0.008" (0.20 mm), Kanthal Nikrothal 80 Plus, Length 33 ft (10 m), Resistance 10.66 Ohm/ft (35.0 Ohm/m), Temp. max 2190 F (1200 C), Surface clean shining silver Now here is where everything I learned in highschool comes to a crash. What resistance do I need the resistor to be? How the hell do I calculate this? Does it depend on what amperage I want to give it? I also bought a light dimmer to regulate the power of the heating element. However, the dimmer only affects voltage, not amperage. Is there something I'm missing? How do I prevent my element from drawing infinite amps and tripping the breaker? How do I regulate amperage? Additional resistors? Where do I find something powerful enough to dissipate up to 14 amps in the air? I have the feeling I'm missing something about working with power straight out of the wall socket, I understand grounding and I also understand that I don't need a transformer since all my stuff runs off AC.
|
|
#
?
Jan 12, 2010 22:13
|
|
|
All you need here are anyone who works with electronics two favorite equations: V=IR and P=IV. You want 150W max off of a 120V source, so 150W=I*120V, and I is 1.25A. To get 1.25A from a 120V source: 120V=1.25A*R, and R works out to 96 ohms. From your list, you would need 9 feet of the 32 AWG wire (@ 10.66 ohm/ft) to make 96 ohms. If you turn down the voltage, the resistance remains constant, so the current also drops. Say you turn it halfway down to 60V: 60V/96ohms = 0.625A 60V*0.625A=37.5W So you can see that taking away half the voltage reduces the power by 3/4 due to combining these two equations into P=I^2R. Hope that helps!
|
|
#
?
Jan 12, 2010 22:24
|
|
|
Delta-Wye posted:You probably ought to use a voltage regulator, but I can't see why this wouldn't work if the drop is pretty close. My SparkFun order came (they hardly screwed it up at all), so now I can play with microcontrollers. Maybe someday I'll figure enough out to be helpful.
|
|
#
?
Jan 12, 2010 23:21
|
|
|
Hillridge posted:All you need here are anyone who works with electronics two favorite equations: V=IR and P=IV. This is helpful, I know those equations but drat if I would have arrived to those figures. 9 feet of wire is a lot tho, but it should pan out. Thanks!
|
|
#
?
Jan 12, 2010 23:30
|
|
|
you can cut down the length of wire by adding power resistors in series with it. A 10 ohm resistor rated for 20W (or more) will reduce the wire needed by about a foot, but almost 17W of your power/heat is going to come from the resistor instead of the wire.
|
|
#
?
Jan 13, 2010 01:10
|
|
|
|
| # ? Jun 3, 2024 17:22 |
|
|
Hillridge posted:you can cut down the length of wire by adding power resistors in series with it. A 10 ohm resistor rated for 20W (or more) will reduce the wire needed by about a foot, but almost 17W of your power/heat is going to come from the resistor instead of the wire. Alternatively, just run it in a zigzag or spiral or coil pattern so you take up less space.
|
|
#
?
Jan 13, 2010 03:44
|
|