|
Stabby McDamage posted:On the decoupling caps - I have 10uF on both the 12V in and the regulated 5V; these are at the end of the breadboard. I've seen recommendations for microcontrollers to also put something like 0.1uF right next to the chip. Is that what you'd recommend? If so, what is it about the proximity/capacity that would lead you to having both? Most people don't bother with the 100nF 'close' decoupling caps on breadboarding, largely because you really can't get close enough to the pins for it to matter anyway, and also because on a temporary breadboard there is some expectation of occasional reboots or weirdness from occasional power rail glitches, which is the purpose of the 'close' decoupling cap, to clean out any minor noise that may get picked up by PCB traces.
|
#
?
Apr 14, 2015 05:56
|
|
|
|
| # ? Jun 4, 2024 21:56 |
|
|
Stabby McDamage posted:I was able solve it by finding a way to force the PWM frequency up to 32kHz, so I can tell the noise is still there, but it's high/faint enough to not care about. The idea behind decoupling capacitors is that they create a low pass filter with the power input. The R in the RC circuit comes from the conductor on the power side itself, even if it is ridiculously small. The center frequency is 1/2πRC, and we're aiming to limit anything above 0 Hz, so it works out. The reason why they usually suggest ceramics at around 0.1 uF is because they're better at handling high-frequencies, which is typical of smaller capacitors. You can combine this with a electrolytic with more C to be certain, but you don't want to rely on an electrolytic alone. This is how it was explained to me, and wiki will tell you the same. Larger capacitance from the electrolytic can help avoid dropout when you're dealing with power applications and/or an anemic power supply. The proximity is simply because you don't want any long conductors (traces, leads, wires) in between your clean power supply and your component where more noise can be introduced. EDIT: I looked at a bunch of those amp modules on eBay, and it looks like they generally already have a decoupling cap on the board, so I don't think you'll get much out of fussing over it. If shortening the lead doesn't help, then just put a low pass filter on the signal going into the amp, and go with the higher frequency PWM on the LEDs. Effective-Disorder fucked around with this message at 06:25 on Apr 14, 2015 |
|
#
?
Apr 14, 2015 06:09
|
|
|
Stabby McDamage posted:I'm building an Arduino-based system to adorn a charity auction basket. It plays WAV music from an SD card through an LM386-based amplifier to a little speaker, while also pulsing an LED strip by means PWM on a MOSFET. 1) put regulators and bypass caps in the middle, then you're two seperate circuits on either side 2) add a gate resistor to slow the fet down, perhaps 1k 3) add a bypass cap, like 1uf right at the led's and right at the speaker circuit Edit: it looks like you did some good things by having your audio pull power very close to the regulator. The closer the better. You could also add bypassing to the audio chip itself, or create a dedicated audio power island for this purpose. Wait is the audio chip digital or getting an analog signal from the Arduino? If it's getting an analog signal then your layout with the Arduino all the way on the left is a major problem, and you really should re-layout with power and Arduino in the middle. The main goal is to "isolate" the two circuits so current from one doesn't effect the other. Physical layout is key here. Second is to minimize the disturbance the switching causes. Slowing down the switching and having proper bypass caps is the key for that. asdf32 fucked around with this message at 14:56 on Apr 14, 2015 |
|
#
?
Apr 14, 2015 14:47
|
|
|
So I think it's analog? Here is a good way to think about layout - like a structure. Power and ground are the supports. So here is what I think you have (and if I'm wrong it's still a good approach): A balcony supported on the right (regulators on the right), with a fat guy jumping up and down in the middle (mosfet) and a guy (Arduino) all the way on the left using a long stick (wire) to play a piano (amp) that's all the way back on the right. Needless to say the fat guy jumping up and down messes up the piano. Putting power and the Arduino in the middle means that what happens on the left and right of that won't interfere with each other. You can also see the importance of a good ground plane. It's like the strength of the structure. More copper means less bounce in the whole thing.
|
|
#
?
Apr 14, 2015 15:15
|
|
|
Stabby McDamage posted:I was able solve it by finding a way to force the PWM frequency up to 32kHz, so I can tell the noise is still there, but it's high/faint enough to not care about. Any chance you could explain how you got to the 32kHz? I've been trying to get higher frequency PWM to avoid harmonics with low (<=100Hz) frequency signals and I've only seen good methods for getting it up to about 7.8kHz unless you're using an external chip.
|
|
#
?
Apr 14, 2015 16:11
|
|
|
poeticoddity posted:Any chance you could explain how you got to the 32kHz? I've been trying to get higher frequency PWM to avoid harmonics with low (<=100Hz) frequency signals and I've only seen good methods for getting it up to about 7.8kHz unless you're using an external chip. The formula for figuring out the frequency is 16,000,000 / DIVISOR / TOP / PHASECORRECT = Hz, where TOP is 256 for Timer0 and Timer2, and 65536 for Timer1. PHASECORRECT is either 1 or 2, Timer1 and Timer2 are 2 (because they are phase-correct PWM) and Timer0 is 1 (because it is fast PWM). The highest practical frequency is 62.5kHz.
|
|
#
?
Apr 14, 2015 20:00
|
|
|
I used this setPwmFrequency code snippet, which sets the clock divider for any of the three timers (the actual function is behind the "get code" link at the bottom).
|
|
#
?
Apr 14, 2015 20:00
|
|
|
asdf32 posted:So I think it's analog? Here is a good way to think about layout - like a structure. Power and ground are the supports. So here is what I think you have (and if I'm wrong it's still a good approach): I'd like to pretend that I'm going try improving it, but it's good enough now, and my 3D printer just came in the mail, sooooooooooo....
|
|
#
?
Apr 14, 2015 20:01
|
|
|
SybilVimes posted:Most people don't bother with the 100nF 'close' decoupling caps on breadboarding, largely because you really can't get close enough to the pins for it to matter anyway, and also because on a temporary breadboard there is some expectation of occasional reboots or weirdness from occasional power rail glitches, which is the purpose of the 'close' decoupling cap, to clean out any minor noise that may get picked up by PCB traces. Decoupling caps We had a discussion about this a while back (2013, holy poo poo). It starts at this post and continues over 2 pages or so: http://forums.somethingawful.com/showthread.php?threadid=2734977&pagenumber=225&perpage=40#post420141964 In short: Decoupling is more ritual than engineering for most hobbyists (and many engineers). There may be reasons for doing it one way that sound completely reasonable and plausible, but often they are not backed up by actual math. Better decoupling strategies start from more accurate modeling of your components, PCB, and environment, along with rigorous EMC testing. However, all that is out of the reach of hobbyists, so cargo cult decoupling is the dominant philosophy.
|
|
#
?
Apr 14, 2015 20:37
|
|
|
asdf32 posted:So I think it's analog? Here is a good way to think about layout - like a structure. Power and ground are the supports. So here is what I think you have (and if I'm wrong it's still a good approach): I ended up needing to rebuild it anyway, so I moved the audio stuff to one side, and bam, noise gone. Thanks! Effective-Disorder posted:The idea behind decoupling capacitors is that they create a low pass filter with the power input. The R in the RC circuit comes from the conductor on the power side itself, even if it is ridiculously small. The center frequency is 1/2πRC, and we're aiming to limit anything above 0 Hz, so it works out. This made it clear for me -- it's close and a low C because the R comes from lead distance. The bigger C is for coarse-grained stuff like sudden loads. EDIT: Video: https://www.youtube.com/watch?v=aYUdSVaslV0 Stabby McDamage fucked around with this message at 02:31 on Apr 15, 2015 |
|
#
?
Apr 15, 2015 02:27
|
|
|
Slanderer posted:Decoupling caps Good reads I missed. Thank you for disabusing me of my ritualistic tendencies. Those PDFs you linked to back there are useful.
|
|
#
?
Apr 15, 2015 06:23
|
|
|
Hey, I'm starting to get into Electronics from a software engineering background. I've started my first project to create a modular, battery powered & wirelessly controllable roller blind and I'm kind of stumped when it comes down to choosing the right motor to use. To give myself a good feel of what might be needed I went ahead and purchased a 6V 30RPM 2.5kg.cm geared motor but as expected it's not cutting it. I want to keep this battery operated so I'm reluctant to go toward larger, 12V motors. But was wondering if perhaps coupling two 6V motors together would be a better option or choosing a lower RPM motor, and what methods can be used to calculate/predict the required strength of a motor or is it just a gut feeling sort of thing you pick up over time. So far I've been reading through many sources on choosing the right DC motor, but I was hoping for some advice based on this particular application. I haven't had much experience applying the laws of physics 
|
|
#
?
Apr 15, 2015 16:19
|
|
|
Addz posted:Hey, I'm starting to get into Electronics from a software engineering background. I've started my first project to create a modular, battery powered & wirelessly controllable roller blind and I'm kind of stumped when it comes down to choosing the right motor to use. To give myself a good feel of what might be needed I went ahead and purchased a 6V 30RPM 2.5kg.cm geared motor but as expected it's not cutting it. The thing you're looking for is torque and power, really. If your blinds don't go up with 2.5kgcm, then you need more torque. Once you find out how much torque that is, then you figure out how fast you want them to go up. If you need (say) 30rpm at 10kgcm, then that's like a 3W motor. 3W @ 6V is probably going to be about 1A draw after everything, so that affects battery selection, etc. If you're OK with the blinds going up slower, then use gearing to lower your RPM and raise your torque. Gears and pulleys both work for this. You can also counterweight the blinds so it takes less torque to raise them. This isn't strictly an electronics problem, but a general engineering one.
|
|
#
?
Apr 15, 2015 16:31
|
|
|
I work in mfg automation and we spec motors to replace people power all the time. I seriously just go to the lever or cord and hang weights or pull with a scale until the thing moves. Then I estimate the moment arm with a tape measure. This gives me a rough torque minimum so I'll add another 50-300% to get an output torque requirement. Then I'll open up a vendor catalog and find a motor with gear head that does something close to the speed I need and just choose the closest match off the shelf. For blinds I'm sure you'll need a pretty good reduction. Maybe you can use a pulley on the original chain so the motor is more accessible and you keep manual control. Also they may already have a counter weight to keep them from falling closed all the time that you could remove.
|
|
#
?
Apr 15, 2015 17:21
|
|
|
If you search Google/eBay for "tubular motor" you might find that's easier then building your own if you're more interested in the software/control part than the hardware part.
|
|
#
?
Apr 16, 2015 00:02
|
|
|
Here's a possible wild goose chase: http://www.amazon.com/12VDC-designed-operate-Chrysler-product/dp/B003XKVNUA/ref=sr_1_1?ie=UTF8&qid=1314994170&sr=8-1?tag=citofgamonlco-20 I have a handful of these motors, they have a huge gear reduction and IIRC a worm gear so they hold position. I cannot stop the shaft with my bare hands. They were all over the place at surplus houses but the supply has gotten spotty, apparently. I bought a few with a thought to motorize my blinds, but never did it.
|
|
#
?
Apr 16, 2015 21:53
|
|
|
Thanks for all the replies. Sorry if it doesn't adhere to the threads primary purpose, it's just that in this case being able to decide the best electrical components is something thats strongly tied with my electronics learning process. Which is also a reason I wont be considering tubular motors, gotta try to learn some hardware/electronics basics myself. I think I'm going to try two coupled 6V 6RPM motors to see if that has enough torque, then move onto low RPM 12V motors if that doesn't work. I will mess around with counterweights because yeah there is one on the roller & blind. Hopefully I'll have something cool to display in the projects thread one day for a cheap DIY option.
|
|
#
?
Apr 17, 2015 01:37
|
|
|
Addz posted:Sorry if it doesn't adhere to the threads primary purpose well half the time we're blowing our own dicks off with 400,000V modules that we bought for three bucks on ebay so i think this topic fits within the thread
|
|
#
?
Apr 17, 2015 06:34
|
|
|
Had a  moment recently some of you might appreciate:One of my dissertation committee members asked me to replace a dammaged SMB connector (which I'd never seen before) on some equipment he uses. After I replaced it, it wouldn't fit back in the case until I'd filed down all of the pins (even after I carefully cleaned my iron and flattened out all of the solder joints). The original connector broke because the company charging four figures for this device didn't mount it flush against the board because it wouldn't fit in their custom machined housing, otherwise. Unrelated: I got my first voltage booster circuit up and running yesterday without any of the magic blue smoke escaping. 
|
|
#
?
Apr 17, 2015 15:53
|
|
|
poeticoddity posted:Had a I guarantee you a hallways-reverberating gently caress was emitted after spending 100k on custom PCB manufacture and a second 100k on injection molding tooling and they assembled the first prototype. I almost want to be a fly on the wall for one of those.
|
|
#
?
Apr 17, 2015 16:13
|
|
|
longview posted:Has anyone here worked with ECL much? I've left this post open for like the past 3 weeks because I wanted to look at it for you; where are the power nets for U3 and the others / what schematic capture tool is that? Is it just masking them to show it from a more logical POV? I think your swapping of Q looks OK. Is this for some of your radio stuff? That's a pretty long conversion chain to go 15MHz->10MHz, but there's a good reason for it I'm sure.
|
|
#
?
Apr 17, 2015 18:47
|
|
|
insta posted:I guarantee you a hallways-reverberating gently caress was emitted after spending 100k on custom PCB manufacture and a second 100k on injection molding tooling and they assembled the first prototype. I almost want to be a fly on the wall for one of those. The housing was actually machined from aluminum (and would have been really nice if it'd been about 2mm higher). That same professor actually stopped me on my way into the office today becuase it had started shorting and randomly signaling, so I ended up filing it down further (which just feels wrong) and putting electrical tape inside the housing. Blegh. I had a "hallways-reverberating gently caress" moment a few weeks back when, after 8 hours of work, I'd finally gotten a tricky double-sided board etched in my lab and I realized that the bottom layer was rotated 180 degrees because I am a moron. I taught another graduate student the basics of KiCAD today and got her to lay out her own break-out board, though, so I feel like that balances it out a bit.
|
|
#
?
Apr 17, 2015 21:45
|
|
|
A DC motor I'm using has this PTC in it. I'm trying to maximize motor power for 20 second bursts, so I want to nail down how the PTC will affect things. I have the free and stall currents of the motor. Will current draw increase linearly with more load on the motor? My big question is about the dissipation power in "tripped mode", given on the data sheet. I was under the impression that resistance would increase exponentially past a certain temperature. Is that dissipation value OK to use and if so how doesn't it depend on voltage?
|
|
#
?
Apr 19, 2015 00:16
|
|
|
movax posted:I've left this post open for like the past 3 weeks because I wanted to look at it for you; where are the power nets for U3 and the others / what schematic capture tool is that? Is it just masking them to show it from a more logical POV? I think your swapping of Q looks OK. Is this for some of your radio stuff? That's a pretty long conversion chain to go 15MHz->10MHz, but there's a good reason for it I'm sure. No worries, I've been busy building a prototype of an RF prescaler for my HP 5335A. As I said in the post above the picture it's a 5V PECL design, I drew the logic symbols with hidden power and ground nets, but I'll probably redraw them with visible pins since I might want to use them in a NECL design later. The intent was to get a low jitter conversion and use the 10 MHz output to a) drive a general distribution amp (already built with 54LS logic), and b) provide a single clean 10 MHz output for instruments/radios. The signal source is a Lucent Z3810AS GPS timing system, which has two 15 MHz sine outputs (redundant, can be combined directly with some loss). Schematic capture is Proteus 8, it's pretty good. PCB layout is kind of like Allegro except missing a few features and without the terrible command line hack for the UI.
|
|
#
?
Apr 19, 2015 09:43
|
|
|
We have a few part number wizards here so I thought I would ask - I'm looking for a pcb or panel mount USB connector and I can't find poo poo. I need a male USB micro that I can set my phone on. I want to build a charging cradle but am stuck on finding the connector. I can't find anything on digikey or molex's sites (par for the course there I suppose) and the connectors I got off of amazon aren't USB micro male, they just look a lot like it. Has any one seen anything like this? I could also go try some other qi chargers and see if I can find something that works with my phone. The last ones were hilarious garbage with my phone but worked for a friend so I suspect it's my phone and not the charger. I gave up on a slick wireless solution and figured a stand with a physical connector was an easy plan b.  When the charger works it is pretty boss. I think I can pretty easily design a charging pcb and mount it at the bottom where the phone sits, if I could only find the drat connector I want. I haven't come up with a very good cable solution yet either but I thought you can buy android boomboxes with these connectors so I'm sure they are out there, somewhere.
|
|
#
?
Apr 19, 2015 17:27
|
|
|
I would recommend something like this: http://www.aliexpress.com/item/New-...2032118130.html
|
|
#
?
Apr 19, 2015 17:42
|
|
|
ante posted:I would recommend something like this: gently caress, the automatic spanish translation in ali-express describes that literally as a "macho for the woman".  "2015 de la promoción Stock Cables Usb Mini Usb Smartphone nuevo Micro Usb macho a para mujer extensión del Cable adaptador de ángulo conector Cable"
|
|
#
?
Apr 19, 2015 17:50
|
|
|
That Nokia Qi charger is tits, but you have to center the charger's coil over the phone's charging coil or you're going to have problems. If you're absolutely positive they're centered, the coil is probably damaged on either device, I have three of the Nokia models and they work great.
|
|
#
?
Apr 19, 2015 19:16
|
|
|
Hadlock posted:That Nokia Qi charger is tits, but you have to center the charger's coil over the phone's charging coil or you're going to have problems. If you're absolutely positive they're centered, the coil is probably damaged on either device, I have three of the Nokia models and they work great. I had two Nexus 4s to play with, and they both did the same thing. I did a screen swap on mine and /could/ have damanged the antenna but the other is stock. They would start charging, charge for a few minutes, and then pop back off. All night it was beep beep beep beep as it cycled and in the morning you would wake up to a discharged phone. My friend's Nexus 5 and whatever his wife has worked great  ante posted:I would recommend something like this: I could probably box that connector in enough that it will stay put. Does aliexpress ship to WA state now?
|
|
#
?
Apr 19, 2015 19:53
|
|
|
I had the N4 and N5, the N4 had loving awful Qi reception, if you could even get the coil centered correctly (I eventually traced the outline of the phone on the charger in pencil so it would charge reliably) The N5 I had just loving worked, like your friends. I have a Moto X now and I sort of want to go back to the N5 simply for the Qi charging. I don't know why we can't get flagship phones with Qi, it's the best feature ever. Qi charging on the N4 is mostly a gimmick, good luck sir.
|
|
#
?
Apr 19, 2015 20:04
|
|
|
How's this look for my first PCB layout? I already had the boards made and they work, which is nice. Seeed Studio prices blow my mind($10 for 10 boards up to 5cm x 5cm, crazy). I'm still not sure about the capacitors. I thought I was just using the reference design from the vreg's datasheet but realized it was supposed to be .33uf. I'm not sure how much these really matter in the grand scheme of things. The bottom of the board is ground.
|
|
#
?
Apr 26, 2015 00:26
|
|
|
You should be able to find the right cap in the same form factor and use that, if you care. Assuming it's a regulator, the value could be important for output stability, but if it's working, ehh, who cares For the board itself, I'd rearrange the components a little and route the traces so that your lines are straight or at 45 degrees only. Also the trace on the LED pin could be going to the other trace to make it shorter/cleaner. That's nitpicky stuff, though, it's a simple board and if it works, who cares 
|
|
#
?
Apr 26, 2015 00:51
|
|
|
What software you guys using to draw up your gerber files?
|
|
#
?
Apr 26, 2015 10:33
|
|
|
Sizone posted:What software you guys using to draw up your gerber files? I've been using Altium because I have access to a license, but for less money/free Eagle and Kicad are popular. Eagle is full-featured and has a free version that limits how big the design can be. I don't know too much about Kicad, but it seems to be gaining popularity.
|
|
#
?
Apr 26, 2015 16:46
|
|
|
Sizone posted:What software you guys using to draw up your gerber files? First rule of EDA is to understand that they all suck, just because Altium is $18k, doesn't make it not suck, it just sucks expensively. Also, don't ever trust any PCB footprints from anyone other than yourself, you might start off using eagle or kicad's built-in footprints, but sooner or later they WILL burn you with a footprint that doesn't match reality, and from that day forth you will want to ensure all your footprints - and probably your schematic libraries too - are done by yourself.
|
|
#
?
Apr 26, 2015 17:41
|
|
|
Thinking about using AVRs instead of PICs for my future projects, and I'm trying to figure out which debugger interface to get. I'm thinking about using something like the ATxmega256A3U for sophisticated things and ATtiny167 for general purpose, the first one it seems I can use with a standard JTAG and a cheap-rear end debugger like this: http://www.aliexpress.com/item/AVR-...1703487574.html But it looks like that won't work with the ATtiny, which cheap chinese debugger tool should I get?
|
|
#
?
Apr 26, 2015 19:53
|
|
|
longview posted:Thinking about using AVRs instead of PICs for my future projects, and I'm trying to figure out which debugger interface to get. I use an Arduino and ArduinoISP for stuff that doesn't fit in my big boy programmer and it works great. You just have to put together a few parts you probably already have (like a crystal, capacitor, etc) and wire it up and you're done. I made a little shield on some protoboard so I can just snap it in and program.
|
|
#
?
Apr 26, 2015 20:18
|
|
|
longview posted:Thinking about using AVRs instead of PICs for my future projects, and I'm trying to figure out which debugger interface to get. Just buy a JTAGICE3 like everyone else, it's compatible with just about everything and is fast. Not sure if it supports HV programming though. edit: whoops look like Atmel stopped producing the JTAGICE3 in favor of the ATMEL-ICE, looks like the same thing except it supports their newer ARM chips too. ANIME AKBAR fucked around with this message at 21:07 on Apr 26, 2015 |
|
#
?
Apr 26, 2015 21:02
|
|
|
Lt Moose fucked around with this message at 05:45 on Apr 4, 2016 |
|
#
?
Apr 28, 2015 04:01
|
|
|
|
| # ? Jun 4, 2024 21:56 |
|
|
Seems like this might be the right thread - I want to connect this DVB tuner to my car's stock antenna. Supposedly I need two adapters: BNC female to MCX male and BNC to Motorola to get this done. Since I'm not in the US I'm just ordering from Ebay and I got the first adapter here but not the BNC/Motorola one. The seller has a shitload of various BNC adapters, but I can't find the one I need - could someone have a look if the seller (or someone else) has it under a different name? I just can't figure out all the connector types that could make this work.
|
|
#
?
Apr 28, 2015 11:13
|
|