|
Zuph posted:I've ordered there before. The shipping was a little slow because I got the cheapest shipping option, and she's pretty far away, but otherwise fantastic. One more vote here; ladyada is one of the best things on the net for starting out in electronics.
|
#
?
Jan 12, 2009 19:39
|
|
|
|
| # ? May 21, 2024 13:08 |
|
|
Great thanks guys, gonna get some Arduino stuff from her then. Plus the bitchin tutorial.
|
|
#
?
Jan 13, 2009 02:21
|
|
|
I'm putting together a power supply with three different voltages out -- unimpeded (at 9v), 5v (brought down via a 7805 stage), and an adjustable (brought down by a fancy robotics adjustable regulator). It's coming from a 9v wall wart. I think I'm going to need to add bypass caps to this. My rough schematic is here: http://data.tumblr.com/hGNNUr9IPipr9n53RGG9ndRyo1_500.gif Where on this schematic should I add them? Right after the Vout on each of the different voltages, right?
|
|
#
?
Jan 14, 2009 17:32
|
|
|
Colin Ex Machina posted:I'm putting together a power supply with three different voltages out -- unimpeded (at 9v), 5v (brought down via a 7805 stage), and an adjustable (brought down by a fancy robotics adjustable regulator). It's coming from a 9v wall wart. just put a cap between vout and ground for each of them. That is what one normally does with bypass caps.
|
|
#
?
Jan 14, 2009 17:42
|
|
|
Okay, so it bridges the outs and the grounding of the outs? I couldn't figure if it was in series with the out or if it shorted the vo and gnd together. Thanks. Now I will add bypass caps to everything I own.
|
|
#
?
Jan 14, 2009 17:45
|
|
|
Keep in mind that wallwart supplies are almost never actually regulated. Their voltages depend heavily on the load. Make sure that with no load its voltage isn't more than the arduino can handle.
|
|
#
?
Jan 14, 2009 19:46
|
|
|
mtwieg posted:Keep in mind that wallwart supplies are almost never actually regulated. Their voltages depend heavily on the load. Make sure that with no load its voltage isn't more than the arduino can handle. the arduino has a regulator in it I believe, it is a 5v device and accepts a 9v wall wart.
|
|
#
?
Jan 14, 2009 20:30
|
|
|
Metajo Cum Dumpster posted:Has anyone ordered from Adafruit before? They have a nice Arduino starter kit bundle that goes with this Arduino beginner's tutorial. I got a usbtinyISP from her before Chriastmas, it came in <1 week and I soldered it up. It's a great deal if you want a programmer. I'm kinda wishing I went for futurelec's programmer because it has JTAG, though. 
|
|
#
?
Jan 15, 2009 03:37
|
|
|
mtwieg posted:here are some without the locking mechanism. Now that I think about it the locking mechanism might make it impossible to connect two of them next to each other, so you should go with these simple ones. just a word to everyone else as I had to figure it out the hard way. The way these things work is you solder the wire onto the end of the silver thing sticking out the back, then you bend in all the little parts that stick up and then cram the metal part into the rectangular housing until it clicks in, then you have a connector. I had to break a pair of them before I figured that out. The datasheet was no help.
|
|
#
?
Jan 15, 2009 08:12
|
|
|
I'm looking for some sort of robot that can walk/drive around the house, has some type of sensors connected to it, is programmable in C/C++ (or even runs Linux) and is extensible e.g. by adding new types of sensors/actuators. For a 2nd year computer science project we actually had to program such a robot and write device drivers for it (it ran Linux), so I should know what I'm doing software wise, but I have no clue about putting one together myself. So do these kind of pre-assembled (or simple DIY) robots exist? I'm mostly interested in programming it and making it do stuff. I guess I'm basically looking for a programmable board with wheels and sensors. Was looking to spend about $150-200.
|
|
#
?
Jan 15, 2009 16:11
|
|
|
Paul MaudDib posted:I got a usbtinyISP from her before Chriastmas, it came in <1 week and I soldered it up. It's a great deal if you want a programmer. I'm kinda wishing I went for futurelec's programmer because it has JTAG, though. The futurlec programmer is just an avr ice clone, which can only do JTAG programming/debugging on a limited range of chips (Granted, this includes the popular 16x series). You can build it out of parts for less than $15, and buy a kit for under $20. I think I'm going to end up buying the AVR Dragon. I'd like to have JTAG debugging for the atmega 324P that I'm using for an upcoming project: Designing a "smart" power outlet that monitors power usage, and can shut off an outlet based on time/date settings, or power draw. This will all interface to a PC wirelessly.
|
|
#
?
Jan 15, 2009 16:47
|
|
|
Zaxxon posted:the arduino has a regulator in it I believe, it is a 5v device and accepts a 9v wall wart. yes but the onboard regulator has a maximum input rating of 12V IIRC, and a 9V rated wall wart can easily be 14 volts if its rated for higher current than the arduino actually draws. Just a precaution.
|
|
#
?
Jan 15, 2009 18:18
|
|
|
amotea posted:I'm looking for some sort of robot that can walk/drive around the house, has some type of sensors connected to it, is programmable in C/C++ (or even runs Linux) and is extensible e.g. by adding new types of sensors/actuators. For a 2nd year computer science project we actually had to program such a robot and write device drivers for it (it ran Linux), so I should know what I'm doing software wise, but I have no clue about putting one together myself. iRobot Create. Roomba without the vacuum.
|
|
#
?
Jan 16, 2009 04:10
|
|
|
Zaxxon posted:just a word to everyone else as I had to figure it out the hard way. The way these things work is you solder the wire onto the end of the silver thing sticking out the back, then you bend in all the little parts that stick up and then cram the metal part into the rectangular housing until it clicks in, then you have a connector. I had to break a pair of them before I figured that out. The datasheet was no help. I'm pretty sure there is a crimping tool for those. Those fingers at the end crimp onto the insulation and the other bit onto the conductor, then you slide the whole thing in.
|
|
#
?
Jan 16, 2009 05:31
|
|
|
BattleMaster posted:The official PIC C compiler costs money and the third party ones also cost money or are incomplete. However, they are really easy to program with assembly language because all RAM and I/O peripherals are mapped as CPU registers so you don't spend time shuffling data around as you work with it. It takes more work than C, but you have much greater control over timing and the inner workings of the CPU. The PIC C compiler has a free student edition which doesn't offer the highest level of code optimization after 30 or 60 days of use, but if your learning how to program the micro controllers in C you really don't want optimizations. I personally stand that yes the Arduino is the cheapest to begin to learn but the PIC24 is probably the most versatile microchip on the market, but costs alot more due to having to either build your own dev board or use the explorer 16+ PICKIT 2 combo. On another note has anyone used any of these XMOS micro controllers I want to start playing with them and would like to know how good they are. Also does anyone know a good dsPIC or digital signal micro controller book?
|
|
#
?
Jan 16, 2009 07:11
|
|
|
turbo sex bat 4000 posted:I'm pretty sure there is a crimping tool for those. Those fingers at the end crimp onto the insulation and the other bit onto the conductor, then you slide the whole thing in. well it's all quite doable with solder and fingers, so I might stick with that.
|
|
#
?
Jan 16, 2009 21:45
|
|
|
Ham Session posted:has anyone used any of these XMOS micro controllers I want to start playing with them and would like to know how good they are. those seem Pretty cool. I'm curious about them myself.
|
|
#
?
Jan 16, 2009 21:50
|
|
|
babyeatingpsychopath posted:iRobot Create. Roomba without the vacuum.
|
|
#
?
Jan 16, 2009 23:15
|
|
|
I hope it's kosher to plug an SA-mart thread in here that some of you nerds might be interested in. If not let me know and I'll edit this out. I'm selling a nice FPGA on a development board and I was also considering selling...amotea posted:I'm looking for some sort of robot An iRobot Create Premium Development Package. This retails for $300 and is in brand new condition, but I'm not looking for that much. Come and see! Make offers! http://forums.somethingawful.com/showthread.php?threadid=3056171
|
|
#
?
Jan 17, 2009 20:26
|
|
|
Do any of you Arduino users know what the maximum sampling rate is for the analog inputs? The website says they are 10 bit, 0-5V inputs but doesn't give a settling time. If a 100+ kHz sampling rate is possible you could use it with some pre-amplifiers to do crazy stuff with audio signal processing.
|
|
#
?
Jan 18, 2009 01:13
|
|
|
no, you can get close to 80ksps, but your resolution will suffer. At max resolution the atmega168 datasheet says 15Ksps is possible, so it's no good for audio. But with any decent external ADC you could likely get sample rates that fast (assuming you can run the arduino at a fast clock speed).
|
|
#
?
Jan 18, 2009 01:51
|
|
|
Integrated uC ADCs are pretty crappy anyways, with a few notable exceptions (none on the arduinos that I'm aware of, some PICs and the Analog uC line have good ADCs). If you're sampling a depth sensor or doing something pretty innocuous then whatever, but if you're pushing the ADC you probably won't like the results. SAR(successive approximation)-based ADCs are cheap-as-water and pretty powerful nowadays. Jameco and Digikey have pretty extensive lists of ADCs up and around 100kHz. Remember that for sample rate you need double the highest frequency you will be sampling, and I'd actually make that triple to be on the safe side (to deal with filter rolloff and aliasing effects, go lower if you have a really sharp filter with little bouncing). Also, do not forget to put a lowpass filter in front of the ADC to block high-frequency signals, even if it's just a low-pass RC circuit. Rolloff should be so that you pick up minimal response at frequencies at or above half the sample rate. Note for audio I wouldn't go less than 12-bits unless you're trying to get a distorted sound. 16-bits is the gold standard, 14-bits will probably be alright. 12-bits (and maybe 10-bits) gives the same performance at very high sample rates when you resample, like 300kSPS or so (I can do the math if that helps anyone). clredwolf fucked around with this message at 03:28 on Jan 18, 2009 |
|
#
?
Jan 18, 2009 03:24
|
|
|
atmel's new line of uCs, the XMEGAs, have integrated 12 bit 1MSPS adc (along with 12 bit 1MSPS DAC) which is drat awesome. I've been wanting to toy around with them, but I would need to buy another programming kit.
|
|
#
?
Jan 18, 2009 04:04
|
|
|
Geez that's amazing. Wonder what the SNR/SFDR specs for that ADC is like. You could almost do some fancy IF sampling with that (or at least it's getting into that territory)! Some of the Analog uCs have ADCs in that range, but those tend to be pricier than normal uCs. Those are pretty much uCs with a proper ADC copy-pasted into it's silicon, so it's kind of cheating. clredwolf fucked around with this message at 04:45 on Jan 18, 2009 |
|
#
?
Jan 18, 2009 04:42
|
|
|
Yeah they're pretty nuts. They're also a lot faster (32MHz, along with the new DMA system which means data transfer is done independent of the cpu, and the hardware event system which means you can program your own hardware interrupt routines), lower power, at least five 16 bit counters, and lots of other neat stuff like built in encryption handlers and IR communication stuff. The amount of bells and whistles is has is insane. Here's the datasheet for the cheapest one, which I think are under five bucks on digikey (but aren't in stock yet  )http://atmel.com/dyn/resources/prod_documents/doc8077.pdf
|
|
#
?
Jan 18, 2009 05:13
|
|
|
Thanks for the responses, I figured it was too good to be true but the prospect of a $30 device that I could just plug into a USB port and program was mighty enticing. What I'm trying to do is build a cheap lock-in amplifier. They work by taking a pure reference sine wave in one channel and an unknown input signal on a second channel and digitizing them both. Then you multiply the two digitized values together and integrate the product. Sine waves of different frequencies are orthogonal so the integral of two waves at the same frequency is a non-zero value but the integral of two waves at different frequencies is zero over many cycles. This means that any components of your input signal at the reference frequency integrate to something measurable while noise at all other frequencies gets zeroed out. The noise rejection on these things is insanely good, too bad commercial ones cost thousands of dollars. I did some quickie modeling since my last post, the stuff I'm interested in is about 10 kHz so the Arduino won't hit the Nyquist point and would be aliased as hell. I'll have to check out the XMEGAs in the morning, hopefully they have a dev kit that isn't hideously expensive and badly documented.
|
|
#
?
Jan 18, 2009 06:00
|
|
|
Ahh, for that kind of app I honestly wouldn't even consider an internal ADC. There's a lot of noise pickup from the microcontroller, digital noise pickup (which is non-random!). By separating the ADC and the uC, you get some noise immunity, especially if you make liberal use of ground planes (recommended) and seperate the analog and digital power sources (probably not necessary). Honestly interfacing to an external ADC isn't very tough, esp. if you can find a pre-built board. Some of the really high-end ones have a SPI interface for control and that's about it. It is a step or two up from just hooking a uC to a USB port though.
|
|
#
?
Jan 18, 2009 07:26
|
|
|
yes, interfacing microcontrollers to external adcs is simple. Most midrange adcs use spi interface, which all atmel microcontrollers (including most arduinos) have. Just a matter of telling the loading the output data into a register, then hardware takes care of transferring the data, and you read the result. I definitely woudn't limit myself to internal adcs.
|
|
#
?
Jan 18, 2009 09:08
|
|
|
I saw this thread and figured it'd be a good place to ask... I am having a TERRIBLE time trying to fix two flex pcb cords with broken traces. I've tried using pencil graphite to bridge the broken connections (which works, but ends up bleeding onto other traces and getting signals mixed up), and using a silver trace pen which doesn't seem to work at all because the tips are too wide to use on these near-microscopic traces. I've even tried using needles dipped in the silver solution, which only makes a huge mess. I know that flex pcbs are prone to damage and as such are usually cheap to replace, but that's not the case with the mp3 player I'm trying to fix because the company doesn't even sell them and buying a broken one for parts on ebay actually costs more than I paid for the drat thing. That was long, but any suggestions would me MUCH appreciated!
|
|
#
?
Jan 18, 2009 20:55
|
|
|
If the PCB is cracked usually I'd just throw the thing out. If you want to though, best way I've found to fix them is to use solder and a lot of flux. Put down flux on the traces to fix. I like to use flux pens personally, you can precisely put down lots of flux on an area with them. Then put just a small dab of solder on the iron, and bridge the gap. If the PCB is just cracked, it shouldn't be a large gap. Then check to see if the circuit works. If it does, superglue it down. If not, keep at it until you either destroy the board or fix it. It's a royal pain in the rear end to do this though.
|
|
#
?
Jan 18, 2009 21:05
|
|
|
PDP-1 posted:Do any of you Arduino users know what the maximum sampling rate is for the analog inputs?
|
|
#
?
Jan 19, 2009 07:15
|
|
|
schnarf posted:I don't know, but you don't need a 100+ kHz sampling rate for audio. 48k is the highest useful rate; 96k is used but pretty much unnecessary. Human hearing is pretty much nonexistent above 20k, and the only reason you see significantly higher sampling rates is to give some wiggle room for the lowpass filters in the A/D converters. Look at the spectrum of a signal that is sampled near the nyquist and one that is sampled much higher (say 10x). Even looking at the waveform should show you how higher sampling rates will be useful - a 48 KHz sample rate ADC will only have 2.4 samples per cycle at 20 KHz. Even at 10 KHz you only have about 5 samples/cyc. In addition to providing less distortion, better sampling rates also tend to have better SNR and ENOB, although strictly speaking this is a function of how much jitter is on your clock circuit.
|
|
#
?
Jan 19, 2009 16:16
|
|
|
anyone here ever work with solar cells? I'm having a hell of a time trying to connect them together. Solder won't stick to them and it's hard as poo poo to get them to do anything without breaking.
|
|
#
?
Jan 19, 2009 17:41
|
|
|
schnarf posted:I don't know, but you don't need a 100+ kHz sampling rate for audio. 48k is the highest useful rate; 96k is used but pretty much unnecessary. Human hearing is pretty much nonexistent above 20k, and the only reason you see significantly higher sampling rates is to give some wiggle room for the lowpass filters in the A/D converters. They are also nice for software-defined-radios because the bandwidth of the SDR is proportional to the sample rate. But, that's not audio.
|
|
#
?
Jan 19, 2009 21:12
|
|
|
Has anyone here experience with metal detector circuits or capacitance sensors ? I am trying to add a sensor to my project in order to detect the presence of a metallic object above an electrode. The electrode should be ca. 3" x 3" and detection distance should be up to 1". The past two weeks I have been playing around with an Omron B6TS capacitance sensor but unfortunately could not get the SPI to work properly even though the design looked nice. Basically I am looking for a replacement with somewhat increased sensitivity. Any recommendations ?
|
|
#
?
Jan 19, 2009 21:28
|
|
|
I've used eddy current sensors to do this, but they were quite expensive. I think the sensor/driver pair we were using came in at around $1500. The install I worked on had around 100 - I'm glad I didn't have to foot the bill. Maybe there are cheaper options, though. The model we used was specified by the end user so we didn't price out anything else. Here is the driver we used: http://www.lionprecision.com/inductive-sensors/ecl100.html e: Looks like it can't quite meet your range reqs anyway. v  v
|
|
#
?
Jan 19, 2009 21:40
|
|
|
Cross_ posted:Has anyone here experience with metal detector circuits or capacitance sensors ? Do you need a capacitance sensor specifically? That is, could you build the electronics yourself? I would think you could build an oscillator based on the self-capacitance (and whatever stray capacitances are around it) of the electrode itself and run this through a comparator tied to a known frequency. With some callibration you should be able to correlate the output level of the comparator to the shift in capacitance induced by your metallic object. This would all take a bit of experimentation, but its a nice "quick and dirty" method.
|
|
#
?
Jan 20, 2009 03:43
|
|
|
Nope, it doesn't have to be capacitance- all I want to know is whether my metallic object is within sensor range. However, following along the same idea I tried out the CapSense example found here: http://www.arduino.cc/playground/Main/CapSense It's just timing how long it takes for the output voltage across the improvised capacitor to reach a reference value. While I did get some clear spikes they were limited to a distance of no more than 0.25". And since I was using a non-grounded laptop the values greatly fluctuated depending on whether I was touching the keyboard at the time :-( Even though this looks promising I don't know how reliable it would be. Will the system go haywire if a user touches the board enclosure? Can this be tweaked to get more sensitivity and thus increased range? With this out of the way I am curious how your approach would work. Are you suggesting an RC/LC oscillator and then measuring the phase/frequency shift induced in the capacitor or coil? Would that provide improved resolution compared to the timing method above?
|
|
#
?
Jan 20, 2009 21:59
|
|
|
Cross_ posted:With this out of the way I am curious how your approach would work. Are you suggesting an RC/LC oscillator and then measuring the phase/frequency shift induced in the capacitor or coil? Would that provide improved resolution compared to the timing method above? Yes, an LC oscillator. Measure the frequency shift by comparing it to a local oscillator set to the unloaded resonant frequency (it doesn't have to be super precise). You'll get a DC level output that varies according to magnitude of the frequency difference. Monitor this output with a scope as you bring your metallic object near and take note of what it reads when you get to 1". Provide this voltage to level comparator (something that spits out a 1 or 0 depending on a voltage threshold - I can't think of the proper name). You could put potentiometers on DC sources for adjustment/calibration.
|
|
#
?
Jan 22, 2009 02:11
|
|
|
|
| # ? May 21, 2024 13:08 |
|
|
A scheme like what Cyril Sneer is describing is likely your best bet. Meauring capacitance can be tricky business depending on the properties of the capacitor under test (such as size, absorption, dielectric, parasitic resistance, etc). Unless you're measuring things with near ideal properties (like an actual capacitor), using timing methods probably won't work. But whatever properties a cap has, if you put it into a resonant circuit, any change in its properties will result in a change in frequency, and even if the capacitance is very low and frequency is very high you can easily downmix that signal to something lower, as opposed to using a timing method in which a small capacitance might charge up in a few nanoseconds, requiring very fast circuitry to measure it. Theramins actually work on the same principal. One's body acts as a capacitor, and by changing that capacitance you can modulate a very high frequency signal into the audible range. Your circutry could be as simple as the tank oscillator with its output connected to a tuned filter with its center frequency tuned to the oscillator's natural frequency (that is, it's frequency with no metal present). The amplitude at the output of the filter would change depending on the frequency of the oscillator. By measuring the amplitude (via a simple rectifier circuit) you could approximate capacitance.
|
|
#
?
Jan 22, 2009 17:41
|
|