|
Well, I got it to work. I tested the switch in the original remote to see what pins were connected at what stages of press. With that, I tried hooking it up using transistors to emulate the same short circuits. Couldn't get it to go. Swapped out the transistors for relays, and it works fine. Pretty drat sure I had the transistors hooked up properly, but I also have to assume that's probably where the problem was. Anyhow, works great now! My camera doesn't have a built-in time lapse feature, so this way I can do time lapses without having to lug my laptop along with. Just the camera, and this little box hanging off the bottom. Also, this is the first step in my hope to make a single-axis time lapse panning deal, using pretty much the same design for the panning axis as my cnc mill. Woo! Thanks guys. 
|
#
?
Jul 14, 2011 19:59
|
|
|
|
| # ? May 30, 2024 16:45 |
|
|
IAmCorbin posted:With the 1381 circuit it just reaches the peak and never discharges Have you tried using the 1381 alone, out of circuit? I would hook up the 1381 by itself and feed it with an adjustable voltage source to see if it ever triggers at all.
|
|
#
?
Jul 14, 2011 23:35
|
|
|
Is there a "rule of thumb" for pricing contractual work? I just placed my first bid ever and I'm curious how reasonable/outrageous it was. It was for a pretty simple microcontroller-based prototype for one of my coworker's pet projects.
|
|
#
?
Jul 15, 2011 00:05
|
|
|
Do you have a value decided on for an hour of engineering work? estimated hours * hourly rate * 2 The *2 part is to account for the estimate of how much time it will take always being low, but if you are awesome at estimating then maybe you don't need it or could make it 1.5. The hourly rate can vary hugely depending on location, experience, your overhead, etc. taqueso fucked around with this message at 00:15 on Jul 15, 2011 |
|
#
?
Jul 15, 2011 00:13
|
|
|
I picked up this black box "laser power supply" on the cheap. 9V in, 2400V out, according to the sign. Was the last one. Anything cool I can do this with? I'm thinking it's a charge pump design but I don't really know that much about lasers
|
|
#
?
Jul 15, 2011 04:12
|
|
|
Corla Plankun posted:Is there a "rule of thumb" for pricing contractual work? I just placed my first bid ever and I'm curious how reasonable/outrageous it was. It was for a pretty simple microcontroller-based prototype for one of my coworker's pet projects. Assuming you're out of school and have held a full-time engineering job before, start by figuring out what your employer paid to have you around and convert it to an equivalent per-hour rate. I don't mean your take home salary here, but what your employer paid - including insurance, vacation, retirement, etc. That's the amount that someone was willing to shell out to keep you around. Then for consulting/contracting work you multiply that rate by 3x for shorter term projects that you have really strong expertise in, or 2x for long term (several month) projects or for things that you can do better than the person who wants to hire you but aren't truly an A-list expert in. You can shade the multiplier between those values depending on a lot of other factors, but that's the general range. The result might surprise you. For a reasonably experienced engineer $100-$200 per hour is not uncommon. Just keep in mind that you are running a business here and (1) You have to cover all your own insurance, equipment costs, etc., (2) You are working on 'plumber time' in the sense that you make a lot while you work but then can have long stretches of downtime where you still need to pay the bills, and (3) You are insulating the person hiring you from the risks of taking on a full-time employee that they might not need for the long run. Like some other professions that could be named, people don't pay consultants to work, they pay us to go away after the job is done.
|
|
#
?
Jul 15, 2011 14:37
|
|
|
Pizer posted:I picked up this black box "laser power supply" on the cheap. 9V in, 2400V out, according to the sign. Was the last one. It's not a charge pump, but instead converts the DC to some degree of AC (even if its just a pulsed waveform), then through a transformer, then probably rectifies + smooths the voltage. It may be high even voltage for certain Crookes tubes or cool things like that, but since it depends on how much current it can actually supply. Someone is probably more knowledgeable than me on the practicalities.
|
|
#
?
Jul 15, 2011 15:15
|
|
|
csammis posted:Out of curiosity I took apart one of those cheap solar LED lamps you'd find in garden supply stores and found a grand total of five components: a small NiCad battery, the solar panel, an 8kohm resistor, an LED, and a component I don't recognize. In the two different devices I got, they both have different chips. One had a SOT footprint and the other had a more TO-92 shaped 4 legged transistor looking thing. In both cases, the thing that looks like a resistor is actually an inductor. The whole thing is actually a boost converter of some sort in that one of the legs is turning the inductor on and off at a somewhat high frequency. I thought they were just standard voltage boost converters but they are actually constant-current power supplies of some sort. If you remove the LED load you'll see the voltage goes way up (to over 7V on my system from 2.7V with the LED in place). The solar panel "charges" the battery directly through a standard schottkey diode. I'm sort of disappointed because I was hoping they were 2.7V or 3.3V boost converters. I can't really find any through-hole boost chips so I thought the solar lights would be a good way to get a solar panel and boost for $5. Is there any other cheap easily cannibalizable device with a 2.7V, 3.3V, or 5V boost converter in it?
|
|
#
?
Jul 15, 2011 19:38
|
|
|
CapnBry posted:Is there any other cheap easily cannibalizable device with a 2.7V, 3.3V, or 5V boost converter in it? I'd look for some kind of cheap battery operated toy. You said you couldn't find TH boost chip, maybe LT1302 will work? edit: I'm pretty sure the mintyboost uses a DIP packaged converter. Yep, MAX756 taqueso fucked around with this message at 20:12 on Jul 15, 2011 |
|
#
?
Jul 15, 2011 20:01
|
|
|
Cyril Sneer posted:Have you tried using the 1381 alone, out of circuit? Hooking up the 1381 (pin 1 = OUT, pin 2 = V+, pin 3 = GND) to my adjustable 1.25V-11V power supply feeding pins 2 & 3 and oscilloscope probe connected to pins 1 & 3 looks like it is reading properly off until a certain voltage where it triggers, but the lowest voltage I can get it to trigger on at is like 2.88V. The circuit it only charging the capacitor up to 2.7V...so I think I just found my problem as I am writing this post. Yep, I just got out a brighter lamp and it triggers just like it should, no wonder I was so stumped I had to wiring right, I was just working with a weak lamp. The next 1381 chip (1381J) is supposed to trigger at 3V. Are these chips just not that accurate or did I have the wrong chips sent to me? Thanks for the help, I'm new to these forums and I really appreciate the quick suggestions and assistance offered here. I have been trying to teach myself electronics with several texts and watching videos but the thing I've really been looking for is a place to get questions answered by people who know what they are doing. I am still having an issue with the BEAM bot I made out of the FLED circuit (here is another video I made of it today out in the sun - http://www.youtube.com/watch?v=5eIZFsHIfhA) and also just noticed the same issue with the circuit I just got working. This issue is that occasionally (actually quite often) the motor will not trigger and instead make a high pitch sound that will change pitch with the amount of light received (if you move the lamp closer it will never give it enough power to actually trigger and move, instead it will just make more noise. But if you remove the light source for a second (or just cover the solar panel with shadow (as you can see in the video above) then it pretty much always fires properly the 2nd time after dropping the voltage and allowing it to increase again. Anyone else out there ever had a similar issue with BEAM bots or DC motors in general? [EDIT] I found some info here (http://junkbots.solarbotics.com/sample_chapter/sample_chapter.htm) that said if a high pitch squeal is heard then the resistor value should be lowered. So I hooked up my multimeter to read the amps going through the collector on the NPN transistor (since it is not supposed to exceed 200mA and I started lowering the resistor value, I lowered it in increments all the way down to 180 ohms, testing it hooked up to the oscilloscope the entire time watching the patterns. It didn't seem to make much of a difference. One thing I did notice was that if I held the light really close to the solar panel it would stall more often, it seems to possibly be related to the heat of the solar panel? IAmCorbin fucked around with this message at 07:07 on Jul 16, 2011 |
|
#
?
Jul 16, 2011 06:31
|
|
|
Does anyone have any advise on what software to use for PCB design. Eagle Cad: Free version seems to be limited but it's a popular piece of software gEDA: Has lots of stuff. Would need to install a virtualbox and linux to run it though. It has spice support though. KiCad: Free open source. gEDA and KiCAD are free open source stuff. This is ok but I'm guessing I'm going to have to spend load of time to get things to work? I ideally just want stuff to work. I gave KiCAD a quick go but the component library felt really limited so I would have to search for libraries or make the components myself. Essentially I just want something that will work where I won't have to put excessive time and effort getting the program to work or make components, etc. The circuit spice simulation in gEDA seems really attractive so I'm not too opposed to putting some a little extra effort in if it's worth it.
|
|
#
?
Jul 16, 2011 14:46
|
|
|
Unparagoned posted:Does anyone have any advise on what software to use for PCB design. I love Eagle. I've used it a bunch, so I have a library built up and have quite a bit of familiarity with it. I buy a lot of parts from Sparkfun, so their Eagle library is a huge pro too. I've tried using KiCAD and I found the interface impossible. Gimp x 10; maybe if you really got to love it, it would be okay, but it felt clunky as hell to a beginner. Plus, it's primarily developed in French which is kind of a pain in the rear end. Been watching some guys at the local hackerspace design boards in gEDA (they 'require' free software) and it looks alright. It's different than Eagle and I haven't felt like playing with it yet, but it looks functional. Plus, if you're doing for-profit work, Eagle can be pretty expensive.
|
|
#
?
Jul 16, 2011 14:56
|
|
|
Unparagoned posted:Does anyone have any advise on what software to use for PCB design. There's also Design Spark PCB (http://www.designspark.com/theme/designspark-pcb). I've never really used it much (only installed it to play around), but it doesn't have really any restrictions and it can import Eagle libraries. It's made/licensed by RS Components, so the idea is that they give it away for free, but make it super easy to generate part orders from your design. There's nothing stopping you from just buying from digikey or mouser or anywhere else though. edit: I don't really see SPICE integration as much of a benefit either. I think it's much better to simulate sub circuits independently and then enter them into the schematic capture. That way you can set up various sweeps and analysis on the sub circuit without having to apply it to the whole design. It also lets you use "ideal" components (like current sources) without mucking up the schematic. You also don't have to worry about not having a proper SPICE model for something that you don't really care about simulating. SnoPuppy fucked around with this message at 15:37 on Jul 16, 2011 |
|
#
?
Jul 16, 2011 15:32
|
|
|
Can someone recommend me a decent autoranging multimeter with capacitance measurement? I'm too poor for a Fluke meter, so is there anything reasonably decent for under $100?
|
|
#
?
Jul 17, 2011 18:38
|
|
|
nobody- posted:Can someone recommend me a decent autoranging multimeter with capacitance measurement? I'm too poor for a Fluke meter, so is there anything reasonably decent for under $100? http://www.eevblog.com/2010/06/04/eevblog-91-50-multimeter-shootout/ I think he has a bunch of other meter reviews scattered through the blog. I just bought an EX330 a couple months back and it is pretty decent, especially for the price. I've even used the temp probe a few times.
|
|
#
?
Jul 17, 2011 18:43
|
|
|
taqueso posted:http://www.eevblog.com/2010/06/04/eevblog-91-50-multimeter-shootout/ Thanks for the link! In one of the comments on that site, someone mentions the Fluke 17B and 18B multimeters. Anyone used one of these? A $100 Fluke meter just sounds too good to be true.
|
|
#
?
Jul 17, 2011 21:04
|
|
|
nobody- posted:Thanks for the link! I've had one for almost a year now, it's alright, auto ranging is slow especially for resistance but otherwise nothing in particular wrong with it as far as I can tell.
|
|
#
?
Jul 18, 2011 15:42
|
|
|
So I'm building a homemade ionization chamber for fun, and a lot of the different guides I've read recommend using soup cans. As luck would have it I bought some soup last week and still have a couple of the cans, but they seem to have a nonconductive coating on the inside. Is this going to be a problem? I'd imagine it is, but none of the guides mention such a thing at all.
|
|
#
?
Jul 19, 2011 02:31
|
|
|
BattleMaster posted:So I'm building a homemade ionization chamber for fun, and a lot of the different guides I've read recommend using soup cans. As luck would have it I bought some soup last week and still have a couple of the cans, but they seem to have a nonconductive coating on the inside. Is this going to be a problem? I'd imagine it is, but none of the guides mention such a thing at all. So if people say it works, then I guess I'd believe them, but they probably wouldn't be able to tell if it weren't working well. Just take some sandpaper and rub the insulating layer off. You probably don't have to be thorough at all.
|
|
#
?
Jul 19, 2011 03:22
|
|
|
ANIME AKBAR posted:Hard to tell... with an insulator, the ions will still move and you will still sense some displacement current, but I think if the ions don't actually meet a conductor, then they will pile up at the wall of the can until they build up a field that opposes the one you're imposing on the chamber, causing it to work. But it's likely that the ionization will find a way to neutralize anyways... ions are just like that in air. Cool, thanks. I'll go hunting for some sandpaper. I'm guessing the coating is epoxy or something like that?
|
|
#
?
Jul 19, 2011 04:41
|
|
|
I'm pretty sure the inner coating of tin cans is always BPA.
|
|
#
?
Jul 19, 2011 16:53
|
|
|
I got a few ATMega32u2 Micropendous boards dumped on me to mess around with. I have Flip for loading Hex over the usb interface, and I have a copy of AVR Studio 5 installed, now how do I output a hex file from AVR studio 5 to load with Flip?
|
|
#
?
Jul 19, 2011 20:39
|
|
|
Hillridge posted:I got a few ATMega32u2 Micropendous boards dumped on me to mess around with. I have Flip for loading Hex over the usb interface, and I have a copy of AVR Studio 5 installed, now how do I output a hex file from AVR studio 5 to load with Flip? Is one generated when you do a build?
|
|
#
?
Jul 19, 2011 21:31
|
|
|
BattleMaster posted:Is one generated when you do a build? Maybe I've asked this before, but does anyone have any suggestions on good literature for improving my C coding style in embedded software? Either textbooks or online class notes or whatever. I just want to be better, damnit.
|
|
#
?
Jul 20, 2011 03:53
|
|
|
Slanderer posted:Maybe I've asked this before, but does anyone have any suggestions on good literature for improving my C coding style in embedded software? Either textbooks or online class notes or whatever. I just want to be better, damnit. Check out back issues of Embedded Systems Programming magazine if you can find them, there is a lot of good stuff in there. Michael Barr has some great article series on embedded software basics IIRC. Aha - found it: http://www.eetimes.com/electronics-blogs/4216468/Embedded-Systems-Design-magazine-archive It says it is the ESD archive, but far enough back and it turns to ESP. (ESP was better IMHO) eetimes and some of the other magazine portals have pretty good collections of white papers, including embedded programming stuff. Also, you can get a free subscription to ESD: http://www.eetimes.com/electronics-subscriptions taqueso fucked around with this message at 19:53 on Jul 20, 2011 |
|
#
?
Jul 20, 2011 19:47
|
|
|
I have a bank of dip switches, 8 of them. On the one side, they're hooked up to the +5V source on my arduino. On the other side, they're hooked up to 8 digital input pins on the same arduino. At first, the reading was...unstable. Sometimes, I'd get a HIGH when the switch was off. Okay, I figured, must need pull-down resistors. I just got a bunch of poo poo from digikey, among the pile of loot an assortment of resistors. I grabbed the biggest ones I had, 1MΩ, and connected one resistor per pin to ground, on the far side (opposite the voltage source). I'm still occasionally getting weird readings, though. Am I doing something silly? Should I use a smaller resistor? I just assumed bigger would be better. Should I somehow isolate them from each other, instead of just connecting them all to the same ground? Is there likely something else I'm misunderstanding? I can post a diagram if needed, but the application seems pretty simple: read the value, reliably, of these 8 switches connected up to my arduino. Basically, for one switch: code:
|
|
#
?
Jul 21, 2011 00:35
|
|
|
There is no problem with using the same ground for all the resistors. 1M might be too weak a pulldown, try 10k-100k. I pulled this pic off GIS:  Where it says "buffer" are your inputs. From your description, I think you have exactly this setup right now. It should work.
|
|
#
?
Jul 21, 2011 00:45
|
|
|
I'm no expert on digital logic but I think that value might be too big. I often see 10k resistors used for pull-downs. I don't know how you're supposed to select values though. I'm fairly certain that it's not advisable to leave input pins unconnected on a CMOS IC because stray charge can build up on the gates of the transistors. Maybe if the resistor value is too high that can happen as well.
|
|
#
?
Jul 21, 2011 00:49
|
|
|
All right, I'll try it with 10kΩ resistors. I, too, saw examples using that value, but I didn't see anything saying one particular value is better than another, or as you said, how to select the proper resistance. Oh well!
|
|
#
?
Jul 21, 2011 01:07
|
|
|
Too low (in value) and you waste power, too large and you become susceptible to noise. http://www.ganssle.com/articles/aresist.htm
|
|
#
?
Jul 21, 2011 01:10
|
|
|
Well, the 10k resistors seemed to do the trick. I guess if I wanted, I could figure out an optimal resistance, but for this application, I don't really care. I won't usually be powering it by battery, anyhow, unless I find I get really crazy good battery life (like ten days continuous running.)  It's a time lapse control module for my camera. Unfortunately, there was no time lapse feature built into my camera, so I made this instead. The switches let me set the delay from 1 to 256 seconds between pictures. The memory card in my camera will hold 48600 images. I really like making time lapses, so I think I'm going to have a blast with this. Previously, I had to lug my laptop around. Now, it's just this little module that hangs nicely off the bottom of my tripod. Here's a video I made with rev1, which you guys helped me with last week: http://www.youtube.com/watch?v=NgQNfxjnZjY. I'm looking forward to playing with rev2 now that I've got it working. I have a trip coming up this weekend that I hope to test it out on. If it works out as expected, I'll soon drop the arduino and just build a board to hold it all proper (still powered by the atmega, of course, 'cause that's what I know, but why waste a $30 arduino when a $5 atmega would do just as well.) And if it works REAL well, maybe I'll have a PCB printed up, mostly because I've always wanted to do that but haven't been able to bring myself to do it for the several projects which are currently stalled at that phase.  Thanks y'all!  e: is it okay to embed pictures from picasa/google+? If not, I'll put them on imgur, but I think google's terms allow it.
|
|
#
?
Jul 21, 2011 05:21
|
|
|
So I was messing around with making boards using the photo developer method. I made an owen splitter which splits a signal into two and has 20 dB of isolation between the two output ports. That worked perfectly fine. However, I then tried to make an amplifier, and it's not working very well. Here's the eval board I have Here's the board I made Here's what the output of the amplifier should look like Here's what the output of my board looks like I think the problem is with my vias that connect the top and bottom ground planes together. I just drilled holes without any real pattern and then put wires through the holes and soldered them to connect the top and bottom holes. The weird thing is that if I stick my finger against the board in certain areas, it will attenuate that noise on the oscilloscope. For example, if I stick my finger on the resister or the amp transistor, the noise goes away. Any ideas?
|
|
#
?
Jul 21, 2011 22:42
|
|
|
Krenzo posted:So I was messing around with making boards using the photo developer method. I made an owen splitter which splits a signal into two and has 20 dB of isolation between the two output ports. That worked perfectly fine. However, I then tried to make an amplifier, and it's not working very well. Also when I make vias, I always drill two adjacent holes (like 70mil apart), then thread them with a U-shaped piece of bare wire and solder it in place. That way I get a lower impedance connection in a small space.
|
|
#
?
Jul 22, 2011 04:06
|
|
|
Krenzo posted:So I was messing around with making boards using the photo developer method. I made an owen splitter which splits a signal into two and has 20 dB of isolation between the two output ports. That worked perfectly fine. However, I then tried to make an amplifier, and it's not working very well. Without being able to see your circuit, my guess is that you're capacitively loading your amplifier feedback input with your ground plane setup. That extra capacitance slows down the feedback signal so the amplifier can't keep up with changes in its own output, causing the output signal to swing around wildly. By sticking your finger on the board you're introducing just enough bypass capacitance to stabilize the oscillator again. In addition to what ANIME AKBAR said regarding vias, take a look at the positioning of C3 and C4 on the original board - they are tied to a section of ground plane that is not connected directly to the transistor. On your board they are connected to the same ground plane segment as the transistor and if that plane is high-ish impedance they could be coupling signal into the amplifier, or capacitavely loading down the inputs.
|
|
#
?
Jul 22, 2011 14:03
|
|
|
Simple (maybe) question. What is the application of a fusible resistor? When would it be chosen over a fuse or PTC resettable fuse?
|
|
#
?
Jul 22, 2011 15:13
|
|
|
They're cheap, take forever to blow, and until they blow they act as a PTC resistance--limiting current.
|
|
#
?
Jul 22, 2011 18:15
|
|
|
PDP-1 posted:In addition to what ANIME AKBAR said regarding vias, take a look at the positioning of C3 and C4 on the original board - they are tied to a section of ground plane that is not connected directly to the transistor. On your board they are connected to the same ground plane segment as the transistor and if that plane is high-ish impedance they could be coupling signal into the amplifier, or capacitavely loading down the inputs. I made another attempt at the board. I cut down the size of the top ground planes, and I moved C3 and C4 over to go to the top right ground plane. It didn't help. I think I'm going to try moving the components on the right as close together as possible and see if that helps. Topics on self resonance imply that trace length plays a role. Edit: Putting the components closer did not fix the issue. 
Krenzo fucked around with this message at 20:50 on Jul 22, 2011 |
|
#
?
Jul 22, 2011 18:41
|
|
|
sixide posted:They're cheap, take forever to blow, and until they blow they act as a PTC resistance--limiting current. Sure, but when is an appropriate time to use one over the other options? When costs are a concern? I see that many of them are flameproof, maybe that is a reason? PTCs can fail short in some cases (overvoltage), are fusible resistors immune to that? Are they just oldschool?
|
|
#
?
Jul 22, 2011 19:39
|
|
|
Krenzo posted:I made another attempt at the board. I cut down the size of the top ground planes, and I moved C3 and C4 over to go to the top right ground plane. It didn't help. L1 might be an issue if its self resonant frequency isn't high enough. Same goes for the DC blocking capacitors. As for trace length, that shouldn't matter at all so long as they have the correct characteristic impedance.
|
|
#
?
Jul 22, 2011 22:02
|
|
|
|
| # ? May 30, 2024 16:45 |
|
|
taqueso posted:Sure, but when is an appropriate time to use one over the other options? When costs are a concern? I see that many of them are flameproof, maybe that is a reason? PTCs can fail short in some cases (overvoltage), are fusible resistors immune to that? Are they just oldschool? Best I can tell they're used primarily when you need to replace a fuse+input resistor with a single part and/or when inrush currents are a big concern.
|
|
#
?
Jul 22, 2011 22:49
|
|