|
ANIME AKBAR posted:Well you need to start somewhere. Do you want static or alternating fields? What strength and volume? 1. (unrelated) Your own NMR? That's some pretty goddamn serious physical chemistry. I've never heard of anyone doing that before. If you're willing to talk more about it, I'd love to know more. 2. I don't really know what range of field strength I'm going for, since I'm not actually trying to accomplish anything. Honestly, I just want to make a really strong electromagnet, possibly for an electromagnetic lock or something. The problem is, I've never seen any practical designs for electromagnets, so I don't really know much about them. So, I guess, some example questions: a) What might be the benefits or using AC vs DC, or the other way around? I'd imagine that an AC powered electromagnet might simplify my circuit somewhat, since I could just reduce it to a transformer, a fuse, and a relay (or a triac, if I wanted to control the power). How would I go about calculating losses/heating due to eddy currents and magnetic hysteresis? b) What sort of core materials do I want do use? Are combinations of materials ever used (ie, a ferrite core inside the coil, perpendicular to an iron plate/disc at the end of the core?)
|
#
?
Jun 28, 2011 19:39
|
|
|
|
| # ? May 8, 2024 22:22 |
|
|
Slanderer posted:1. (unrelated) Your own NMR? That's some pretty goddamn serious physical chemistry. I've never heard of anyone doing that before. If you're willing to talk more about it, I'd love to know more. quote:2. I don't really know what range of field strength I'm going for, since I'm not actually trying to accomplish anything. Honestly, I just want to make a really strong electromagnet, possibly for an electromagnetic lock or something. quote:a) What might be the benefits or using AC vs DC, or the other way around? I'd imagine that an AC powered electromagnet might simplify my circuit somewhat, since I could just reduce it to a transformer, a fuse, and a relay (or a triac, if I wanted to control the power). quote:How would I go about calculating losses/heating due to eddy currents and magnetic hysteresis? quote:b) What sort of core materials do I want do use? Are combinations of materials ever used (ie, a ferrite core inside the coil, perpendicular to an iron plate/disc at the end of the core?) Sometimes combinations of materials are used if you want to focus the field as much as possible. Like in that picture you posted, the tips of those cones (where the flux density is most concentrated) are probably made out of something like FeCo (with an extremely high saturation B), while the bars making up the outer flux path are probably made of something more typical like laminated steel. This is really only necessary if you want to play with fields more than 1.4T.
|
|
#
?
Jun 29, 2011 03:44
|
|
|
ANIME AKBAR posted:So I'm assuming that the area of interest is in between the tips of those two cones, right? Also all the posting about homemade NMR makes me drool at the possibilities that room temperature superconductors could provide. Pizer fucked around with this message at 06:09 on Jun 29, 2011 |
|
#
?
Jun 29, 2011 04:42
|
|
|
Thanks for all the help. I'm sorry that pic I posted was misleading--it was just a pic of the variable strength magnet we used for measuring the Bohr magneton using the zeeman effect. A cadmium lamp went in the middle, and with the power supply at the highest safe value the field strength was measured at something like 1.6T. I 'm not really interested in duplicating it, but it planted the seed of interest in my mind, since I was curious about why it was designed in such a way (sorta-hourglass shape), and why there was the iron ring all the way around it (something about the stray fields?). My prof. couldn't really give a good explanation for its design, since my drat questions were delaying our experiment enough as it was, and I've always remembered it as a neat design. Would a single coil around a core be good for a locking mechanism, or would two coils around the straight parts of a u-shaped core be ever better?
|
|
#
?
Jun 29, 2011 04:57
|
|
|
The coil image looks like a helmholtz coil slightly modified
Pizer fucked around with this message at 06:10 on Jun 29, 2011 |
|
#
?
Jun 29, 2011 05:07
|
|
|
Slanderer posted:Thanks for all the help. I'm sorry that pic I posted was misleading--it was just a pic of the variable strength magnet we used for measuring the Bohr magneton using the zeeman effect. A cadmium lamp went in the middle, and with the power supply at the highest safe value the field strength was measured at something like 1.6T. I 'm not really interested in duplicating it, but it planted the seed of interest in my mind, since I was curious about why it was designed in such a way (sorta-hourglass shape), and why there was the iron ring all the way around it (something about the stray fields?). My prof. couldn't really give a good explanation for its design, since my drat questions were delaying our experiment enough as it was, and I've always remembered it as a neat design. Magnetomotive force (MMF or just F) is roughly equivalent of your voltage source. It's measured in amp-turns, so it's easy to calculate. Reluctance (R) is the equivalent of resistance in magnetism. Reluctance can be found for a complete closed path (and magnetic fields always form closed paths) or a section of the magnetic path. Reluctance is given by R=(path length)/(crosssectional area * permeability), which is pretty much the same as the formula for resistance (except replace permeability with conductivity). Higher permeability gives lower reluctance Magnetic flux (Iota) is the equivalent of current. If you apply a given MMF to a path with a given R, then your total flux through the path will be MMF/R. Magnetic flux density (B) is analogous to current density. It's just the flux per unit area. So if you want to make a big B, then you want a lot of flux in a small cross sectional area. That requires a very low reluctance. So it's easy to generate stong magnetic field if you have just a big ring of iron with a coil around it. But that field will only exist inside the iron, so you can't play with it. You need to put a gap in the ring, in which you can put whatever you want. The trouble is, putting the gap there will greatly increase the overall reluctance of the path, decreasing your flux. So you want to make it as small as possible. So that's why that picture you posted has iron going all the way around. If it wasn't there, then most of your path length would be air and the reluctance would be huge. It would just suck. And the gap is tiny because the smaller it is the better field you can get. If you want to play around with this stuff, you should use FEMM, a free, simple, 2D FEM program. quote:Would a single coil around a core be good for a locking mechanism, or would two coils around the straight parts of a u-shaped core be ever better?
|
|
#
?
Jun 29, 2011 14:13
|
|
|
Pizer posted:Field guides for magnetic fields? What are those? quote:Also all the posting about homemade NMR makes me drool at the possibilities that room temperature superconductors could provide.
|
|
#
?
Jun 29, 2011 14:18
|
|
|
Hello Electronics thread (I knew there was a reason SA had forums other than Games), I have a few silly questions. I've started reading about Electronics a couple of times in my life, and I usually get all the way up to understanding how transistors work before losing interest (mainly because I'm unfocused like that). I've always been able to understand how individual components work and what they do, but never been able to create bigger things combining them. I mean, when one of you looks at this:Slanderer posted:Can you understand at a glance what is going on, and how modifying components and such would affect it? Are you able to create circuits like that from scratch? What books do you recommend? I know the "Lessons in Electric Circuits" from the OP are rather good, but any textbooks I should look into? Also, I'd love to hear from people who are into Electronics just as a hobby. How for can you get in the hobby without any formal education (ie studying electronic engineering at the university)? PS: Kudos to the guy building a DIY NMR, you are the dream of every aspiring Chemist.
|
|
#
?
Jun 29, 2011 15:51
|
|
|
Markovnikov posted:Can you understand at a glance what is going on, and how modifying components and such would affect it? Are you able to create circuits like that from scratch? Hah, you only understand how circuits at a glance if you've spent too much time working with 555/556 timer ICs. They are (or at least were) a tremendously prolific IC (maybe one of the most common ICs ever?). I'm not sure what to recommend for intro reading, but The Art of Electronics is still a standard reference text if you decide to get into more complicated stuff. It's outdated, and at times pretty dry, but it still covers a lot of material. Personally, I think a strong basis in the math is an important thing to start from. Learning all the analysis methods for linear circuits, for one, as well as learning complex numbers are they relate to impedance and AC signals. Getting a good grasp on Op-amps is especially important, IMO. You can get pretty far teaching yourself; however, access to equipment and parts is normally what can hinder people. There's a lot to be gained from puzzling over a circuit your partner broke in lab, with a signal generator on your left and an oscilloscope on your right, with a drawers of resistors and capacitors right behind you (and a guy to yell at you for not verifying their actual values with the LCR meter before sticking them into your breadboard). Other than that, there is a ton of stuff that comes from experience, since no one seems to write it down half the time.
|
|
#
?
Jun 29, 2011 18:01
|
|
|
Slanderer posted:I'm not sure what to recommend for intro reading, but The Art of Electronics is still a standard reference text if you decide to get into more complicated stuff. It's outdated, and at times pretty dry, but it still covers a lot of material. I would recommend AoE, even if it is a little old. I didn't find it to be too dry, though. It uses plain language and gets to the point quickly. Markovnikov posted:Also, I'd love to hear from people who are into Electronics just as a hobby. How for can you get in the hobby without any formal education (ie studying electronic engineering at the university)? Read read read. Read books, read design ideas, read white papers. Use simulation to test out the ideas you aren't sure about. (This part really lets you get a handle on what will happen if you do [whatever] in your circuit.) Then try to do it. It really helps if you come up with an idea that you want to do, then research all the things you will need to know about to do it. It will keep you motivated to learn more about whatever is involved in your project. taqueso fucked around with this message at 18:23 on Jun 29, 2011 |
|
#
?
Jun 29, 2011 18:16
|
|
|
Oh, a quick question to throw to ya'll about the best way to finish building a quick-and-dirty battery monitor: I was asked to quickly throw together a data logger for measuring the voltages of three ~3.3V batteries, along with a current. The 3 batteries are in series, and are gonna be running through a load at a current between 50 and 400 amps, depending on which load they attach. I've decided to go with a low-side current sensor, only because (a) it's easier, (b) I have some extra instrumentation amplifiers lying around and (c) because my monitor needs to be attached to the circuit's common, which mean's that my supply voltage is too low to make the high side measurements. For the sake of time, I just threw a data logger shield I used a while back onto an arduino, so I can log the voltages onto an SD card, with the RTC timestamp. However, here's the problem: In order to measure all three batteries while sharing a common with them, I need to divide down the voltages from each of them. However, because the atmega has a poo poo ADC, this gives me an effective resolution of like 10mV/bin if I divide the highest voltage battery in half. I'd like to increase it easily, so here are my potential solutions: 1. Oversample the ADC, since I only need 1 point of data a second, and the voltage will be changing at subhertz frequencies. If I do this, whats the best solution if there's not enough noise. Dithering? Adding a few feet of wire to the board? 2. I feed the outputs from the voltage divider into single op-amps configured as DC amplifiers with virtual grounds set at a little less than the (divided down) cutoff voltages of the batteries. Alternatively, I could run the op amps off of a new supply of like two 9V batteries in series (or just a 10+ volt supply), remove the voltage dividers, and set the virtual grounds as before. If I made the op-amps unity gain for either case, I could still increase my resolution by switching my ADC reference from 5V to the internal 1.1V 3. (the crazy solution that I'm only writing down for fun). As before, I use a higher voltage supply for my external circuitry, but I make my measurements by using this: http://www.linear.com/product/LTC1043 to switch a flying capacitor from across individual batteries to my ADC input. I only mention this because I've never been able to think of something neat to do with these ICs. They would be even better if they were able to tolerate voltages about VCC at the switch inputs. So what do ya'll think? Would oversampling be enough, or should I seriously consider throwing a generic TI quad op amp onto the board? I'm even more curious about things that I haven't thought of, because I'm mostly just talking out of my rear end here. So if there are any other common ways to deal with situations like this, let me know.
|
|
#
?
Jun 29, 2011 18:26
|
|
|
Slanderer posted:I'm not sure what to recommend for intro reading, but The Art of Electronics is still a standard reference text if you decide to get into more complicated stuff. It's outdated, and at times pretty dry, but it still covers a lot of material. Holy gently caress, that was one of the books I was  ).I'm not that worried about the math part as due to my career choice (Chemistry) I have a rather solid background in Math, and what I don't know I can learn/bug my mathematician friends to teach me. I'm fine with complex numbers as long as no one starts deriving/integrating with them (if the horror stories my math friends tell me are to be believed)  .taqueso posted:Read read read. Read books, read design ideas, read white papers. Use simulation to test out the ideas you aren't sure about. (This part really lets you get a handle on what will happen if you do [whatever] in your circuit.) Then try to do it. I remember going through a similar process when learning programming on my own, so I have part of that down already at least. Speaking of which, I remember there being simple, computer programmable IC's; how far up the hobby progress chain is that? Thanks for all the information guys.
|
|
#
?
Jun 29, 2011 20:26
|
|
|
Slanderer, you want an opamp. With a quad opamp you could take the three differential measurements pretty easily.
|
|
#
?
Jun 29, 2011 20:29
|
|
|
Markovnikov posted:Speaking of which, I remember there being simple, computer programmable IC's; how far up the hobby progress chain is that? I'm guessing you are referring to a CPLD or FPGA. I like Lattice's MachXO/XO2 because it is pretty powerful, has a PLL, and internal flash so you don't need to have a 2nd device to hold the configuration data. This thing will get you going quickly and cheaply: http://www.latticesemi.com/products/developmenthardware/developmentkits/machxo2picokit.cfm It has a USB-serial chip that is used to program the chip, so you don't need to buy/make a programming dongle. The software to compile and program is free. I'd recommend starting out learning Verilog over VHDL, but it is really personal preference. You can do whatever you need to with either.
|
|
#
?
Jun 29, 2011 20:42
|
|
|
Markovnikov posted:Hello Electronics thread (I knew there was a reason SA had forums other than Games), I have a few silly questions. I've started reading about Electronics a couple of times in my life, and I usually get all the way up to understanding how transistors work before losing interest (mainly because I'm unfocused like that). I've always been able to understand how individual components work and what they do, but never been able to create bigger things combining them. I mean, when one of you looks at this: I started with http://ocw.mit.edu/courses/electrical-engineering-and-computer-science/6-002-circuits-and-electronics-spring-2007/video-lectures/ which I found a really good start to the theoretical side of electronics. With regard to say the 555, I basically went through how it actually worked and derived the formulas myself. So spending allot of time thinking about how the 555 worked was really useful. In regards to circuits, often they look really complicated. I find one of the best ways to learn is to try and design circuits yourself. As you want to do more stuff you'll find you need to add bits in here or there and things start making sense.
|
|
#
?
Jun 29, 2011 20:54
|
|
|
Markovnikov posted:Speaking of which, I remember there being simple, computer programmable IC's; how far up the hobby progress chain is that? If you're talking about microcontrollers, take a look at the Arduino / Atmel AVR platform. Don't worry too much about the analog stuff, yeah you need to know some of the basics, but if you're not going to be designing power supplies or signal processing hardware you can probably avoid most of it. For that example with the 556, if you can program you can re-create that on a microcontroller more cheaply and with no external parts.
|
|
#
?
Jun 29, 2011 21:34
|
|
|
taqueso posted:I'm guessing you are referring to a CPLD or FPGA. Since you're a student, I'd also look into Digilent boards (http://www.digilentinc.com/). They use Xilinx FPGAs and are a little more expensive, but the FPGAs are much bigger than a Lattice part and they have far more peripherals included. You will also probably be able to get more support/help for a Xilinx part, since they are one of the top two FPGA companies and more people use their parts. I also like the Xilinx tool chain better than Lattice, but that's just me.
|
|
#
?
Jun 29, 2011 22:05
|
|
|
Markovnikov posted:Hello Electronics thread (I knew there was a reason SA had forums other than Games), I have a few silly questions. I've started reading about Electronics a couple of times in my life, and I usually get all the way up to understanding how transistors work before losing interest (mainly because I'm unfocused like that). I've always been able to understand how individual components work and what they do, but never been able to create bigger things combining them. I mean, when one of you looks at this: I actually wrote a post a little while ago explaining how that particular IC, the 555 timer, works. Take a look at the last page or two of my post history in this thread.
|
|
#
?
Jun 29, 2011 22:26
|
|
|
drat, I've been throwing questions at this thread every day! But, I gotta assemble some boards that a coworker designed, and I think there's a bit of an issue... It's all SMT, but the problem is that there are some sensor components that need to be soldered to the bottom side. Most of them are flat-leaded SMT, but one of them is a 6 lead(?) BGA package. When I soldered these to the previous version of the design, I used an aluminum foil heat shield around the rest of the board, and using a temperature-regulated hot air gun, sorta followed the reflow profile to attach it. It's an optical sensor chip, so I don't want to damage the clear package, the optical filters, or the chip itself by reflowing it too hot or too long, nor do I want to damage the BGA balls. What's the best way to solder something like this, when there are dozens of SMT parts on the other side of the PCB? The topside parts are being soldered using solder paste + stencil, reflowed on an enormous electric griddle. I don't have access to any fancy IR soldering stuff, only a regulated hot air gun, soldering irons, and the reflow griddle.
|
|
#
?
Jun 30, 2011 20:34
|
|
|
Slanderer posted:drat, I've been throwing questions at this thread every day! But, I gotta assemble some boards that a coworker designed, and I think there's a bit of an issue... If your board is in a fixture of some type so that it doesn't move, the surface tension should hold the opposite side parts on, as long as they're not too heavy. It's even easier if you're using hot air since you only have to worry about the parts that are close to the components your're soldering.
|
|
#
?
Jun 30, 2011 21:44
|
|
|
SnoPuppy posted:It's even easier if you're using hot air since you only have to worry about the parts that are close to the components your're soldering. There's like a dozen components directly underneath the one thing I'm soldering, however. And probably a ground fill too, just to make my life annoying.
|
|
#
?
Jul 1, 2011 00:11
|
|
|
Slanderer posted:There's like a dozen components directly underneath the one thing I'm soldering, however. And probably a ground fill too, just to make my life annoying. Why don't you just solder the BGA component along with the other devices that you're doing with paste on a griddle?
|
|
#
?
Jul 1, 2011 01:49
|
|
|
Krenzo posted:Why don't you just solder the BGA component along with the other devices that you're doing with paste on a griddle? ....because it's on the other side of the board? Like I said, I didn't do the layout.
|
|
#
?
Jul 1, 2011 01:54
|
|
|
We used an Altera Cyclone 3 Development board in my Digital Logic class to learn about FPGA's. I've been thinking about getting one for my own fun because I've got some cash to burn. Should I go for that guy, thats about $210 right now on Altera's website, or should I go for something from Xilinx like this: http://www.digilentinc.com/Products/Detail.cfm?NavPath=2,400,792&Prod=S3EBOARD Or should I move on up to something like this (apparently you can get Linux running on this bad boy with Xilinx's Microblaze dealio): http://www.digilentinc.com/Products/Detail.cfm?NavPath=2,400,836&Prod=ATLYS Any thoughts? Aluminum Record fucked around with this message at 02:08 on Jul 1, 2011 |
|
#
?
Jul 1, 2011 02:02
|
|
|
Slanderer posted:There's like a dozen components directly underneath the one thing I'm soldering, however. And probably a ground fill too, just to make my life annoying. If they're not too heavy, you're probably safe. Are they small passives? ICs? Here's how a double sided board is normally made: You build one side, flip it over, paste/place the other side, and pop it in the oven again. Done! Aluminum Record posted:We used an Altera Cyclone 3 Development board in my Digital Logic class to learn about FPGA's. I've been thinking about getting one for my own fun because I've got some cash to burn. Should I go for that guy, thats about $210 right now on Altera's website, or should I go for something from Xilinx like this: I normally use Xilinx FPGAs so I'm fairly biased towards them. I'm also less familiar with Altera. That aside, both digilent boards you listed should be a better FPGA than the Altera board, since the Spartan 6 is a generation (or more) ahead of the Cyclone III - it's more like the Cyclone IV or an Aria II FPGAs. The structure of the Spartan 6 will allow for more logic since it has a 6 input LUT with two flip flops per LUT. They also have DSP slices which contain a hard 18x18 multiplier, an adder, and an accumulator - the Altera parts only have the hard multiplier, as far as I know. I'd suggest the Atlys board since it has a much bigger FPGA, has a lot of peripherals, and is about the same price as your Altera board (assuming you're still a student). It also has DRAM and, as you mentioned, can run an embedded OS. The other digilent board would be able to run the Microblaze (I think), but without DRAM running an OS would be difficult. Digilent also has a ton of peripheral IO cards they call PMODs, so it has that going for it as well.
|
|
#
?
Jul 1, 2011 02:43
|
|
|
SnoPuppy posted:If they're not too heavy, you're probably safe. Are they small passives? ICs? Most components won't fall off. Only bulky stuff like connectors will have issues. If you do have anything you're worried about, just put a dab of glue on it after you reflow that side.
|
|
#
?
Jul 1, 2011 10:07
|
|
|
Slanderer posted:What's the best way to solder something like this, when there are dozens of SMT parts on the other side of the PCB? The topside parts are being soldered using solder paste + stencil, reflowed on an enormous electric griddle. I don't have access to any fancy IR soldering stuff, only a regulated hot air gun, soldering irons, and the reflow griddle. If you think you can get away with it, use leaded solder on the BGA. If the other components are soldered with lead free solder, they won't reflow when soldering the BGA (so long as you control the temperature well). In industry, using solders with different melting points is a very common way to do double sided SMT (though they wouldn't use leaded and lead free, but rather two types of lead free with some nickel added to one). Aluminum Record posted:We used an Altera Cyclone 3 Development board in my Digital Logic class to learn about FPGA's. I've been thinking about getting one for my own fun because I've got some cash to burn. Should I go for that guy, thats about $210 right now on Altera's website, or should I go for something from Xilinx like this: ANIME AKBAR fucked around with this message at 13:06 on Jul 1, 2011 |
|
#
?
Jul 1, 2011 13:03
|
|
|
ANIME AKBAR posted:If you think you can get away with it, use leaded solder on the BGA. If the other components are soldered with lead free solder, they won't reflow when soldering the BGA (so long as you control the temperature well). In industry, using solders with different melting points is a very common way to do double sided SMT (though they wouldn't use leaded and lead free, but rather two types of lead free with some nickel added to one). Actually, I believe most manufactures just use the same paste for both sides. Keeping track of two different pastes can be annoying - if somebody loads the wrong paste, bad things happen. The general rule of thumb is to make sure that the components that will be put on first (generally the back side) should have a mass/pad area < 30 g/in^2 ANIME AKBAR posted:The hardware offered by Altera and Xilinx is fairly similar, and prices aren't too different. If you're just starting out, I would base the decision on the development software (which is way more important IMO). Go with whatever you can afford a cheap license for. As far as I know, both companies offer pretty much the full FPGA design tools for free for their smaller parts. The exception being the Xilinx EDK, but you should be able to get a trial license for it if you need it. You might even be able to get a free license by asking nicely, since you're a student - it helps to have a project in mind. I imagine your school has a license for modelsim (I know mine did), but if not, the Xilinx tools come with a simulator. Xilinx and Altera make their money by selling parts - they like to give students discounted/free software because you're more likely to spec in something you already have experience with when you're in industry. SnoPuppy fucked around with this message at 15:21 on Jul 1, 2011 |
|
#
?
Jul 1, 2011 15:01
|
|
|
grams per square inch are you loving kidding me
|
|
#
?
Jul 2, 2011 15:19
|
|
|
I prefer newtons/hectare.
|
|
#
?
Jul 2, 2011 19:08
|
|
|
Pascals
|
|
#
?
Jul 2, 2011 20:43
|
|
|
Mill Town posted:grams per square inch are you loving kidding me Uh, well most pcb design is done using mils, and most components are spec'd in grams. So it's easier to just keep the units in separate systems.
|
|
#
?
Jul 2, 2011 22:00
|
|
|
Any engineer (professional or otherwise) doesn't give two shits what the unit of measure is. The ones who care go crazy and find a different job/hobby.
|
|
#
?
Jul 3, 2011 00:05
|
|
|
csammis posted:I was thinking about getting a logic analyzer at some point to help me out with continuing adventures in embedded development and have found http://www.saleae.com/logic/. Does anyone in this thread have any experiences good or bad with it? My needs are not many at this stage - I would like to be able to see PWM output on something other than an LED and a guess at the percentage of brightness, for example. I'm a bit confused is there any reason you can't just use an Osciloscope?
|
|
#
?
Jul 3, 2011 01:18
|
|
|
Zaxxon posted:I'm a bit confused is there any reason you can't just use an Osciloscope? Yes; I don't have one. e: or have any idea where to get one, how to use one, what all good it would be over a logic analyzer, etc. If you can enlighten me I'm all ears.
|
|
#
?
Jul 3, 2011 04:59
|
|
|
Don't logic analyzers tend to come with a lot more channels plus the ability to decode serial protocols like I2C and RS232? Of course you can't use them for much else. I mostly do power stuff so I wouldn't go anywhere near a logic analyzer. Speaking of which I just tried a new coilgun design that uses a pair of 12V lead-acid batteries instead of a capacitor bank. My initial tests where I manually shorted the leads instead of using a switching mechanism were encouraging so I picked out one of my MOSFETs that's good for 180 amp surges and it seems to be able to handle the currents without exploding. Next up is building a firing circuit with an optical sensor that cuts off the power when the projectile is closer to the centre of the coil. I was also thinking of doing multiple stages.
|
|
#
?
Jul 3, 2011 06:30
|
|
|
BattleMaster posted:Don't logic analyzers tend to come with a lot more channels plus the ability to decode serial protocols like I2C and RS232? The one he was looking at makes a trade off of having more channels, but the total sample time is divided up to give you those extra channels. So instead of one channel with a 200 MS/s, you get 8 with 25 MS/s. That's not very good if you're trying to look at any decently fast logic.
|
|
#
?
Jul 3, 2011 19:32
|
|
|
Krenzo posted:The one he was looking at makes a trade off of having more channels, but the total sample time is divided up to give you those extra channels. So instead of one channel with a 200 MS/s, you get 8 with 25 MS/s. That's not very good if you're trying to look at any decently fast logic. It's fast enough for acting as a bus analyzer, though, for anything except USB, really. It's basically the best product on the market for that price range, as it's software is usually pretty good at decoding stuff.
|
|
#
?
Jul 3, 2011 20:48
|
|
|
Random: I've noticed that on a lot of laser pointers, or in their instructions or labels or whatnot, they suggest you don't fire them continuously for more than a couple minutes. Is that just due to heat dissipation related issues? Such that if you had a laser with improved heat dissipation capabilities, such as some sort of active cooling, you could effectively fire the laser continuously for an indefinite period of time? And with that in mind, is it reasonable to control the brightness of a laser via, say, PWM? Or would such rapid cycling possibly damage it somehow?
|
|
#
?
Jul 5, 2011 20:51
|
|
|
|
| # ? May 8, 2024 22:22 |
|
|
Bad Munki posted:Random: For the first question: heating is indeed the problem, and will damage the diode over time. Continuously is still probably not possible for all diodes, though, since external heatsinking and cooling can only do so much if the internal structure doesn't transfer heat out very effectively. For the second, I don't see any reason why you couldn't do this, if you were just driving a bare diode without any regulating circuitry. I could be wrong, but it seems very doable. However, the current still needs to be limited. It's often possible to drive LEDs without a limiting resistor if you keep the pulse width super short, relying on the internal resistance of the diode to prevent mayhem. The behavior, I'd guess, would be nonlinear, since the actual semiconductor junction would be rapidly heating and increasing it's resistance, but I don't know to what extent. It's heavily frowned upon, but still viable as long as the parameters of the LEDs match well with each other (ie, internal resistance doesn't vary much). In fact, many IR LEDs can be driven like this, and even include specifications for driving them this way. However, I would never ever do this for a laser diode. If it's a low power diode, a limiting resistor will probably be fine (as long as you're okay with the power loss due to the resistor), and then I imagine you could drive the whole thing with a mosfet, which you could control with a TTL signal.
|
|
#
?
Jul 5, 2011 21:24
|
|