|
Can you take a few close up pictures of the actual board? Not promising we can help you but we might see something that you don't.
|
#
?
Aug 2, 2013 20:00
|
|
|
|
| # ? Jun 4, 2024 07:09 |
|
|
armorer posted:Can you take a few close up pictures of the actual board? Not promising we can help you but we might see something that you don't. Sure thing! Not sure if you can make anything from this picture, but here:  Thanks!
|
|
#
?
Aug 2, 2013 20:30
|
|
|
Check the polarity of C1 & C2. If they're backwards they're probably bad now, too. Also use a fresh 9V battery.
|
|
#
?
Aug 2, 2013 21:01
|
|
|
Phone screen is tiny and all, but I see a lot of questionable joints. It's low-density through-hole, don't be afraid to use more solder (and perhaps some flux)
|
|
#
?
Aug 2, 2013 21:34
|
|
|
Hah, and here I was worried I'd been using too much solder. Thanks for the tip, I'll try to re-solder some of the shoddier looking joints. The 9V is definitely brand new, and does give a loud tone and two bright LEDs. So that's not the issue. I'll double check the caps polarity, but I'm pretty sure I double-checked before soldering. But mistakes do happen...
|
|
#
?
Aug 2, 2013 21:39
|
|
|
Zaphod42 posted:I have a schematic, but its on paper. Here ya go. Never mind. In regards to current you can measure current of a resistor simply by measuring its voltage and calculating ohms law. This also tells you the current going through anything directly in series with it. asdf32 fucked around with this message at 00:17 on Aug 3, 2013 |
|
#
?
Aug 3, 2013 00:12
|
|
|
Zaphod42 posted:Hah, and here I was worried I'd been using too much solder. Thanks for the tip, I'll try to re-solder some of the shoddier looking joints. Keep in mind that a constant tone on your speaker still means that there's some oscillation going on, it's just faster than you expected. You also seem to be deflecting or don't understand the comments about how the human eye sees LED brightness. My money's on a problem with the connections to the 555, either bad solder joints, blown caps, or wrong value resistors.
|
|
#
?
Aug 3, 2013 00:50
|
|
|
So, like a few others on this board, I'm working on a Daft Punk helmet. I'm doing the Tron version of the Guy Manuel helmet. The helmet is basically done at this point, but I need some help on the electronics end. I've got a chaser circuit, which does a variety of patterns with 8 LEDs. LEDs in the visor are 6 rows of 24 LEDs each. I can either do horizontal lines, skipping the first and last pins on the chaser, or I can do vertical groups of 18 LEDs (3 columns per group). Either way, I need to drive a lot of LEDs off one pin. What's my best way of doing this? I also want to run it off my handy little USB battery pack (outputs 5V over USB, runs for 10+ hours per charge). This of course means my voltage is limited to 5V. I'm also using a PIC controller, rather than arduino, just because that's what I have.
|
|
#
?
Aug 3, 2013 01:51
|
|
|
ante posted:Keep in mind that a constant tone on your speaker still means that there's some oscillation going on, it's just faster than you expected. You also seem to be deflecting or don't understand the comments about how the human eye sees LED brightness. I don't have time now to walk through the whole schematic, but it appears as if the LED oscillation is independent of the 555 oscillations. Sort of, but here is my guess. The 555 is running in astable mode, with the control line being used to superimpose another signal - this is where the euro siren comes in. The 555 creates the overtone, the led oscillators creates the fade in and fade out. See http://www.talkingelectronics.com/projects/555/Page3-555.html 555 amplifier, the previous pages also have good info about the 555 in general. The LED section is a pretty straightforward monovibrator (  ) circuit, and I'm guessing that is what is failing to oscillate. Check values, solder jobs, transistor polarity (triple check that, in fact).Can you measure the voltage of the base, emitter, and collector of the two BJTs? I think those voltages would be useful if it's not oscillating at all. TwystNeko posted:So, like a few others on this board, I'm working on a Daft Punk helmet. I'm doing the Tron version of the Guy Manuel helmet. The helmet is basically done at this point, but I need some help on the electronics end.  If your circuit isn't a purpose made chaser, and is a software program on a microcontroller, it may be worth looking into shift registers. It would be worth doing a back of the envelope calculation to see if you could get pixel-level control at a decent framerate. You're only controlling 144 LEDs, so it's not all that much data to pound out of a GPIO pin. Pregenerate some animations, store them in a big buffer, and just dump them out to the shift registers. More complicated, but I suspect you can do some cool stuff that way. TwystNeko posted:I'm also using a PIC controller, rather than arduino, just because that's what I have. 
Delta-Wye fucked around with this message at 06:05 on Aug 3, 2013 |
|
#
?
Aug 3, 2013 01:55
|
|
|
ante posted:Keep in mind that a constant tone on your speaker still means that there's some oscillation going on, it's just faster than you expected. You also seem to be deflecting or don't understand the comments about how the human eye sees LED brightness. What the hell man? That was stated like 5 times and I never disagreed with it a single time. How am I deflecting? Both LEDs should not be on at the same time. It should go visibly on and off with a human perceptible delay. I understand LEDs turn on and off, but this circuit is designed to do more than just feed a constant frequency to the LED. Look at the schematic. It is designed to emulate a European police siren, varying between two different LEDs, one off when the other is on, and two different tones. The caps all have the right polarity. I'll try re-soldering. I guess I'll also check to see if voltage is 0 across anything. Zaphod42 fucked around with this message at 05:52 on Aug 3, 2013 |
|
#
?
Aug 3, 2013 05:40
|
|
|
Zaphod42 posted:The caps all have the right polarity. I'll try re-soldering. I guess I'll also check to see if voltage is 0 across anything. Well, one thing I noticed is that your PCB is labeled Rev D, while your schematic is Rev C. Any chance you might be working with something out of date/out of sync? I'm also suspicious about the addition of the LEDs - even if the capacitors are fully discharged, I think the LED forward voltage will be enough to cause the transistors to be constantly conducting. Have you tried temporarily shorting out the LEDs to see if that fixes the audio?
|
|
#
?
Aug 3, 2013 08:20
|
|
|
Is there seriously no simple solution for charging three li-ion cells? I'm working on an integrated APRS tracker/digipeater and I found a 11.1V Li-ion pack + a balancing/protection PCB, but I seem completely unable to find a decent Li-Ion charging module. Only thing I found that seemed like a good bet was the Silvertel AG112-3 (http://www.semiconductorstore.com/cart/pc/viewPrd.asp?idproduct=44732) and that has a minimum order of 1000. Having an outboard charger is a big drawback, so I was really hoping to find a PCB module that I could integrate.
|
|
#
?
Aug 3, 2013 11:00
|
|
|
longview posted:Is there seriously no simple solution for charging three li-ion cells? I'm working on an integrated APRS tracker/digipeater and I found a 11.1V Li-ion pack + a balancing/protection PCB, but I seem completely unable to find a decent Li-Ion charging module. Lithium charging modules in a PCB form factor are pretty rare apparently, probably because they require somewhat complicated control schemes they don't want the user to be able to tamper with. Heating is also a frequent problem in the charging circuitry. If you want it on-board then you may have to DIY it, or get a dev board from one of the major chip manufacturers (TI, Linear, etc).
|
|
#
?
Aug 3, 2013 12:40
|
|
|
Zaphod42 posted:Sure thing! Not sure if you can make anything from this picture, but here: Folks mentioned double checking the polarity of the caps, but also double check the polarity of the LED that isn't lighting up. (Sounds from your description like only one LED lights up, and stays on?) Your circuit appears to be like the one titled POLICE SIREN on the link Delta-Wye provided, except that the leftmost 555 chip is replaced by the LED section. If you want a graphical idea of what should be happening in the LED section, check out this widget. Look at this widget to get an idea how the 555 section should be working. In your circuit the oscillating of the LED section drives the control pin on the 555 timer. This is what should be making the tone change. If you are getting a constant tone, then that control signal is stuck high or stuck low. You can probably backtrack through your circuit from that point and make sure everything is soldered correctly.
|
|
#
?
Aug 3, 2013 16:00
|
|
|
Delta-Wye posted:Easiest way is to use the output of the chaser circuit to drive some transistors to turn on your LEDs. Do you have room for eight transistors and a resistor for each and each led? If you use a NPN BJT, each of the 8 outputs drive a transistor and however many led/resistor pairs are connected to each transistor like so: I don't really have a whole lot of room in there. It's going to be tight as it is. If i was using SMD versions of stuff, on a flexible PCB, that would be ideal. In fact, I'm going to look into that now. The circuit is just a PIC, but it's a pretty small one. It's basically this: http://picprojects.org.uk/projects/ledchaserpwm/ The other idea I just came up with, that's simple, and perhaps a little ghetto, is to use some of the EL Tape I have.
|
|
#
?
Aug 3, 2013 18:04
|
|
|
armorer posted:Folks mentioned double checking the polarity of the caps, but also double check the polarity of the LED that isn't lighting up. (Sounds from your description like only one LED lights up, and stays on?) Actually both LEDs are fully lit up the whole time. Thanks for that widget, that explains a lot about about the circuit is supposed to be switching.
|
|
#
?
Aug 3, 2013 20:19
|
|
|
Zaphod42 posted:Actually both LEDs are fully lit up the whole time. So, I know you don't want to hear this anymore, but if the frequency of everything is ramped up, it may well appear that both LEDs are on the whole time and it may sound like the speaker is only playing one tone. If the LEDs alternate back and forth at, say, 1000 times per second, technically only one will be on at any moment, but they will appear to your crude human eyes to be on constantly. Same thing for the speaker: if the speaker is alternating back and forth between, say, 1000hz and 2000hz a thousand times a second, your lovely human ears can easily interpret it as one sound. As an example, open up two tabs here: http://onlinetonegenerator.com and on one tab, set it to 1000hz and on the other, 2000hz. Play them both at the same time. Depending on your brain, you will probably hear one tone. If the frequencies are closer or further apart the effect may be more or less pronounced. 400hz/500hz is also a good example. Bad Munki fucked around with this message at 20:32 on Aug 3, 2013 |
|
#
?
Aug 3, 2013 20:30
|
|
|
Zaphod42 posted:Actually both LEDs are fully lit up the whole time. Thanks for that widget, that explains a lot about about the circuit is supposed to be switching. Also see here. http://en.wikipedia.org/wiki/Multivibrator So a few suggestions. First, I suspect the left half of the circuit is what's broken and perhaps the right half is fine. You're hearing a constant tone so it probably is. To verify this further I would take the signal that goes to pin 5 of the 555 and short it to 9V. This is the state it's in when the right side LED is off, which is the state you're not getting into. On the left side I'd check the connections the best you can with the ohm meter. I'd also check the caps with the capacitance measurement on the meter. If you actually have polarized caps (do you?) it's possible you blew them without knowing. They are critical for the operation. Is there any chance you have the 1k and 56k value resistors reversed? The ohm meter will tell you. Also measure voltages. Given that both LED's are on I'd expect to see about 2V across the LED, nearly zero across the BJT's and therefore about 6V across the 1k/2x470 resistors. The bases of the BJT's should be at just about 0.7V. Any deviation from this might give you clue. If I had to guess I'd say you have something off with the caps. They transmit quick changes in voltages which create a positive feedback cycle. I also question the choice to put the LED between the caps and BJT instead of above where the caps connect. Reduction in current through one transistor is supposed to increase its CE voltage which transmits through the cap and turns on the other BJT harder, which reduces its voltage which transmits through the cap reduces base voltage at the first transistor which further cuts off current at the first transistor and so on. However in the middle of this path are the LED's - when a transistor reduces current it's voltage goes up, but the LED's voltage will decrease. This dampens the positive feedback which is required for oscillation. So anyway, if you don't get anywhere I'd suggest yanking out the LEDs and shorting across their footprint to create a stock multi-vibrator circuit. It should be slow enough that you'll see if this works with the multi-meter and/or you'll hear it on the buzzer.
|
|
#
?
Aug 4, 2013 01:49
|
|
|
Have you checked the transistors and led polarity yet? If not, do that before scratching your head further.
|
|
#
?
Aug 4, 2013 02:29
|
|
|
Bad Munki posted:So, I know you don't want to hear this anymore, but if the frequency of everything is ramped up, it may well appear that both LEDs are on the whole time and it may sound like the speaker is only playing one tone. Delta-Wye posted:Have you checked the transistors and led polarity yet? If not, do that before scratching your head further. asdf32 posted:useful
|
|
#
?
Aug 4, 2013 02:52
|
|
|
Can someone tell me if this is an okay regulator by which to power an atmega328p (and the associated bits to make a barebones arduino, i.e. a few caps, a resistor, and a crystal)? http://www.mouser.com/ProductDetail/STMicroelectronics/L78M05CDT-TR/?qs=sGAEpiMZZMug9GoBKXZ759dUb0oa4gSJfZtsjLnQjx0%3d It looks like it can take 7V-35V, and can provide as much as 500mA output, which should provide more than enough juice for the atmega and its bits, as well as a healthy smattering of accessories powered off the same regulator. And it looks like the recommended usage has a couple caps on it, anything else I need to consider?
|
|
#
?
Aug 5, 2013 03:01
|
|
|
Bad Munki posted:Can someone tell me if this is an okay regulator by which to power an atmega328p (and the associated bits to make a barebones arduino, i.e. a few caps, a resistor, and a crystal)? Looks good to me. Voltage regulators are pretty straightforward. The only other thing I'd take into consideration is that if you're planning on powering it on the higher end of that and using a lot of current it might get pretty warm, but you should be fine if you're powering it off a 9V or whatever.
|
|
#
?
Aug 5, 2013 04:34
|
|
|
Super. With that done, then, would anyone mind taking a look at this eagle project and offering some basic critique? I'm pretty green here, so any guidance is welcome. The project is just supposed to be a bare bones arduino as per this product/guide with an added on-board voltage regulator and holes/pads for header pins so it will mate with a breadboard. Here's the project, I think everything that's needed is included: http://gregshort.net/Barebones_Arduino.zip  e: I should point out that a couple parts on there are mis-labeled like the voltage regulator and the atmega, but that's just because I didn't have a library handy with the exact part, so I just picked a very close one with the same pinout/package. Bad Munki fucked around with this message at 16:06 on Aug 5, 2013 |
|
#
?
Aug 5, 2013 06:11
|
|
|
Might also want to look at something like this project if you're looking for a breadboardable barebones Arduino. I've got the parts for a couple of those sitting on my desk at home waiting for some free time. I don't have Eagle or anything installed so I can't look at your project files though.
|
|
#
?
Aug 5, 2013 21:21
|
|
|
n0tqu1tesane posted:Might also want to look at something like this project if you're looking for a breadboardable barebones Arduino. Well, I am, but I'm also looking to practice with eagle and get past this hump of not actually completing so many projects because I'm afraid to actually make/order PCBs for some inscrutable reason. The barebones arduino is just a good jumping off point and would apply to numerous projects. 
|
|
#
?
Aug 5, 2013 21:29
|
|
|
Either way, it's a good example circuit of using a voltage regulator similar to the one you're looking at in conjunction with a barebones arduino. It has a few extra features that the Bareduino doesn't, such as the voltage regulation, and a power/activity LED.
|
|
#
?
Aug 5, 2013 22:20
|
|
|
Zaphod42 posted:Hah, and here I was worried I'd been using too much solder. Thanks for the tip, I'll try to re-solder some of the shoddier looking joints. Cover up as much of the copper as you can, solder joints shouldn't have holes/pits in them like that. Let the solder creep all the way to the silk screen if you aren't super limited on it. And yes, get some flux, it makes all the difference when it comes to soldering, especially if you do anything beyond through-hole.
|
|
#
?
Aug 6, 2013 00:09
|
|
|
Arcsech posted:Looks good to me. Voltage regulators are pretty straightforward. The only other thing I'd take into consideration is that if you're planning on powering it on the higher end of that and using a lot of current it might get pretty warm, but you should be fine if you're powering it off a 9V or whatever. "Might"? No it would burn up immidiately. Never go by the front page power/current rating on a power component. Power components are typically designed to be heatsunk and/or the power dissipation varies widely over it's operating ranges. In the case of a regulator, peak current and peak voltage never go together, usually not even close. In this case at 500mA at 35V in it's dissipating 15W. Digging deeper you'll see a thermal coefficient of 50C/W (ja=Junction to Abmbient) which means it goes up 50 degrees C for every watt it's dissipating without a heatsink. The maximum junction temperature is 150C which means you're only good for about 2W (assuming an ambient of 25-50C). Note that 1-2W is a good rough number to remember for a TO-220. So run your numbers, even at 9V in 500mA will push you right up to the 2W limit. Final note, there is increasingly less and less reason to use linear regulators, at least from a cost perspective. Drop in switching regulators for these applications have come down to less than $3 at quantity 1. A bit bulkier, perhaps a bit of switching noise (rarely an issue), but switchers can (usually) hit their numbers with no nonsense concerning heat. http://www.digikey.com/product-detail/en/R-78E5.0-0.5/945-1648-5-ND/2834904 When you actually have need for high voltage inputs drop in one of these, then use linear regulators from 5 to 3.3/1.8/1.2 etc for nice clean busses if you'd like.
|
|
#
?
Aug 6, 2013 00:16
|
|
|
In this case, the power supply would probably be either a 9V battery for low-current installations, or a closer-to-target (6V?) wall-wart for higher-current/longer-term installations. At $3 a pop for a switching regulator, that roughly doubles the cost of the device.
|
|
#
?
Aug 6, 2013 00:49
|
|
|
Do consider that a 9v battery has something like a 200mAh capacity already, and burning up nearly half of that as waste heat will mean incredibly lovely battery life. Is it worth the $3 to not have to replace batteries all the time? e: looked it up and typical power consumption for an ATMega328 running code is roughly 12mA, so add that to whatever else you'd be driving off the board (a single 5mm LED: 15-30mA) and I think you can see the problem. Sagebrush fucked around with this message at 01:32 on Aug 6, 2013 |
|
#
?
Aug 6, 2013 01:26
|
|
|
Bad Munki posted:In this case, the power supply would probably be either a 9V battery for low-current installations, or a closer-to-target (6V?) wall-wart for higher-current/longer-term installations. 78xx regulators really do require a few volts overhead to operate (probably 7V to 8V minimum for 5V output). There are other low-dropout regulators, however, if you really want to power off a 6V wall wart rather than a 9V wart. For a 9V battery you have relatively small capacity, linear regulation from 9V down to 5V you are literally wasting 45% of the power and correspondingly more battery life. An efficient switcher will get most of that back. For example, if your circuit requires 100mA and you have a linear regulator, you'll draw 100mA the source, but if you use a switcher, you'll draw 60mA-ish from a 9V source. If it's a 500mAh battery, you've gone from 5 hours of life to over 8. eta: and you can get boost converters if you'd rather use a couple AAs or a single lithium cell.
|
|
#
?
Aug 6, 2013 02:08
|
|
|
All right, thanks for the power tutorial, all of you. It's a shame that a switching regulator would multiply my cost that way, but it sounds like it'd be a far better general-purpose solution, so for a non-application-specific board, it's probably the best bet.
|
|
#
?
Aug 6, 2013 02:59
|
|
|
Sorry to double-post, but is something like this the right answer? http://www.mouser.com/ProductDetail/Seiko-Instruments/S-8354A50MC-JRJT2G/?qs=sGAEpiMZZMtitjHzVIkrqaV7sk7EKPTuwwgsWXc7YT4%3d What does it mean when it says output current is "fixed?" I still expect to see a value there, so I'm confused. Sorry, power regulation is a new field for me. e: Or is this a better answer? http://www.mouser.com/ProductDetail/Texas-Instruments/TL2575-05IKTTR/?qs=sGAEpiMZZMtitjHzVIkrqZAkPplH0S%2fkfBD4JgyCXn4%3d Bad Munki fucked around with this message at 03:31 on Aug 6, 2013 |
|
#
?
Aug 6, 2013 03:27
|
|
|
Bad Munki posted:Sorry to double-post, but is something like this the right answer? http://www.mouser.com/ProductDetail/Seiko-Instruments/S-8354A50MC-JRJT2G/?qs=sGAEpiMZZMtitjHzVIkrqaV7sk7EKPTuwwgsWXc7YT4%3d Neither, both require external inductors and diodes which add cost (probably another ~$1+) and complexity. Unless they actually specify the diode/inductor part numbers I'd rule these out. E: Although a quick search confirms what I said below, pre-packaged boosts are hard to come by, so if you go the boost route something like this isn't a bad option. Note that when you're looking at "controllers", which I classify as anything which doesn't have the inductor, "synchronous" is a nice option because it removes the diode. On digikey stay in under the power supply category not the IC category to get fully integrated DC-DC's. http://www.digikey.com/product-search/en/power-supplies-board-mount/dc-dc-converters/ I like the boost idea in theory but I've found boost converters harder to find and more expensive as a result. That Recom is pretty good, I recently designed it into a board. If you end up finding something better let us know. Note that if your input is 5.5V or less a whole other class of smaller/cheaper switchers is available to go to 3.3 and below. asdf32 fucked around with this message at 04:11 on Aug 6, 2013 |
|
#
?
Aug 6, 2013 03:53
|
|
|
When I need a little buck converter for an electronic project, I tend to use these: http://hobbyking.com/hobbyking/store/__15212__HobbyKing_Micro_UBEC_3A_5v.html They're designed for radio control usage and so are pretty small and light. Not a great package if you're trying to make a PCB, obviously, but for hackery stuff where you're just going to cram it in an Altoids tin they're fine. Cheap, stable and pretty efficient (70-90% depending on draw and input voltage). Also note that that unit will source up to 3A vs. that TO-220 drop-in replacement's 0.5A -- much better if you're going to be running little motors, strings of high-powered LEDs, etc. As asdf32 says, boost converters are a little harder to find, but here's one designed to take 3.2-4.2v from a single lithium-polymer cell and boost it to 5v for your Arduino or whatever. Same sort of package as the buck above but still nice and compact. Very nice for when you don't have a whole lot of power requirement but need to squeeze as much energy from your battery as you can. http://hobbyking.com/hobbyking/store/__11784__TURNIGY_Voltage_Booster_for_Servo_Rx_1S_to_5v_1A_.html Sagebrush fucked around with this message at 04:31 on Aug 6, 2013 |
|
#
?
Aug 6, 2013 04:29
|
|
|
If you're trying to keep costs down, it's hard to beat the sketchy direct-from-china
|
|
#
?
Aug 6, 2013 04:34
|
|
|
I'm having a hard time finding consistent-seeming package names for crystals. So you've got something like this: And usually it's called HC-49/US. Sometimes USX, I think HC-49/U (without the S) is similar but taller? And then there's an SMD variant, basically just the same thing but with an insulating sheet underneath and the pins bent sideways to attach to pads:  But that's also called HC-49/US? Or sometimes they'll just name the package by the outer dimensions. There are also some that are just flat with little pads underneath, and I'm wondering what the most normal/common SMD crystal package is, or if there is one.
|
|
#
?
Aug 6, 2013 22:59
|
|
|
Honestly, you don't need a switcher for an arduino clone, though. That's just adding unnecessary cost and complexity, since you're unlikely to need very much power.
|
|
#
?
Aug 7, 2013 00:11
|
|
|
Okay, but just to be clear, the question(s) re: crystals are completely aside from the earlier regulator chat. 
|
|
#
?
Aug 7, 2013 01:36
|
|
|
|
| # ? Jun 4, 2024 07:09 |
|
|
Does anyone have any suggestions for organizing parts bins for an electronics shop? Our building is getting renovated at work so we have to clear out twenty-odd years of accumulated junk from the electronics room. We are also getting new shelving units when we move back in, each filled with wide trays containing round-bottom partition bins. This means we have a once-in-a-decade chance to try to set up a system that is better than the rows of cheap plastic bins that we have now, where poo poo is labeled only by part number (e.g. XF-381) and nothing is 'discoverable' unless you already know what that part number means. Off the top of my head I was thinking of trying to group stuff by function, something like Caps->Electrolytic->50v->1uF, or Connectors->0.1" Headers->6pin->socket, and then leave about 20% of the bins in each tray empty for future expansion. That way you could start off with a vague notion like "I need a differential amplifier", then go to the OpAmp shelf, pull out to the differential input tray and only have to check a few part numbers of items you already know might work to find one that fits your specs. Nothing is set in stone yet, so if there's better options lemme know. Also - inventory management. Today we just try to remember what parts we noticed were running low whenever we put in an order for some project, and it can be hit-or-miss at times. I was thinking of hanging a couple of clipboards around where folks could write down stuff we are low/out of, and we'd scratch off items whenever we ordered replacements. Any other/better suggestions?
|
|
#
?
Aug 7, 2013 01:50
|
|
