|
BlackTie posted:I suppose US plugs are also rated for a that many amps, so we get double the power. The typical US outlet is 15A, with fairly common 20A versions available.
|
#
?
Dec 2, 2013 20:12
|
|
|
|
| # ? May 9, 2024 00:40 |
|
|
BlackTie posted:Probably yes, since we use 240V instead of 120. Going by Wikipedia, Schuko plugs can handle 16A, which is about 3600W. I suppose US plugs are also rated for a that many amps, so we get double the power. Residential wiring in the US often tops out at 20A unless it is a specialty circuit, but that still isn't enough to make up for the 2x voltage. It's pretty sweet to be able to deliver more power on thinner wires 
|
|
#
?
Dec 2, 2013 20:15
|
|
|
In older housing it's common for the fuses or circuit breakers to trip at even lower currents
|
|
#
?
Dec 2, 2013 20:17
|
|
|
Otto Skorzeny posted:In older housing it's common for the fuses or circuit breakers to trip at even lower currents Or in the case of a mobile home, a hilariously bad residential fire 
|
|
#
?
Dec 2, 2013 21:05
|
|
|
Posted in the headphones thread, but if anybody is looking for a great DIY project, Glass Jar Audio is having a 15% black friday sale with a coupon code: http://glassjaraudio.com/ The AMB kits are some of the best hifi equipment out there, if not the best. Worth looking into if you're serious about audio and love tinkering.
|
|
#
?
Dec 2, 2013 23:04
|
|
|
Anyone have any manufacturers to recommend for high-reliability / industrial-grade microSD cards?
|
|
#
?
Dec 3, 2013 02:16
|
|
|
movax posted:Anyone have any manufacturers to recommend for high-reliability / industrial-grade microSD cards? Zynq corruption? I can relate. Got the highly recommended Sandisks and they still flaked out after a day or so. I ended up making a read only partition for the rootfs. If it's not that, Sandisks are good I hear for ~most~ stuff 
|
|
#
?
Dec 3, 2013 02:23
|
|
|
priznat posted:Zynq corruption? I can relate. Got the highly recommended Sandisks and they still flaked out after a day or so. I ended up making a read only partition for the rootfs. Corruption?  nah, in this case just evaluating different options / suppliers. I generally run with a read-only /boot and rootfs, with a third rw partition for user stuff.In a completely unrelated note, anyone have good app notes for 24-bit RGB interfaces? Have a goofy LCD and pin placement to deal with; pretty sure pixel clock is the key driver in terms of lengths / mismatches / etc, just want to sanity check against a vendor reference.
|
|
#
?
Dec 3, 2013 08:21
|
|
|
I'm absolutely new to all this and am trying to make an internet-enabled infrared transmitter (I want it to control some infrared curtains and also the TV) - so far I've managed to set up an IR receiver (with one pin going to ground, one to 3V, and one to p23) - this works fine. I've also managed to get it transmitting - with one pin connected to p24, the other through a 330ohm resister to ground - however it only transmits over very short distances (It works when less than 30cm away and doesn't work when around a meter away, I can't test inbetween due to plug socket positions). I guess this is more complicated than I assumed it would be! how would I go about making it work over a longer distance?
|
|
#
?
Dec 3, 2013 13:40
|
|
|
I'd like to build a benchtop variable DC power supply, and maybe a simple 'scope that I can hook to a PC. Any recommendations for cheap and cheerful starting points? I don't necessarily need a kit, although I wouldn't rule one out if that's the best option.
|
|
#
?
Dec 3, 2013 16:30
|
|
|
saint gerald posted:I'd like to build a benchtop variable DC power supply, and maybe a simple 'scope that I can hook to a PC. Any recommendations for cheap and cheerful starting points? I don't necessarily need a kit, although I wouldn't rule one out if that's the best option. Max/min voltages required? Amperage requirements?
|
|
#
?
Dec 3, 2013 17:00
|
|
|
Motronic posted:Max/min voltages required? Amperage requirements? I guess about 1-2V up to maybe 12 or 24? No more than a couple of amps.
|
|
#
?
Dec 3, 2013 17:05
|
|
|
saint gerald posted:I guess about 1-2V up to maybe 12 or 24? No more than a couple of amps. If that's all you need it's pretty hard to beat one of those $60 0-30v ones you can get on eBay. You may be able to build it cheaper, but that would probably depend on what you can scavenge rather than buy. Something like this: http://www.ebay.com/itm/30V-5A-110V...=item460dfedbce
|
|
#
?
Dec 3, 2013 17:51
|
|
|
saint gerald posted:I'd like to build a benchtop variable DC power supply, and maybe a simple 'scope that I can hook to a PC. Any recommendations for cheap and cheerful starting points? I don't necessarily need a kit, although I wouldn't rule one out if that's the best option. The cheap benchtop supplies may be a good fit, but also check out the computer power supply -> desktop supply converters. If you have an extra power supply kicking around, it's not a bad way to get yourself up and running. https://www.sparkfun.com/products/9774 Heisenberg1276 posted:I'm absolutely new to all this and am trying to make an internet-enabled infrared transmitter (I want it to control some infrared curtains and also the TV) - so far I've managed to set up an IR receiver (with one pin going to ground, one to 3V, and one to p23) - this works fine. Not being familiar with your hardware at all (what transmitter, what receiver? P23 and p24 of what?) I'm guessing you're not bright enough. You may have to juice the poo poo out of the LEDs, make a big bank of them, I don't know. Look at a few of the tv-b-gone projects for inspiration - they are trying to do something similar as far as distance goes.
|
|
#
?
Dec 3, 2013 18:19
|
|
|
Delta-Wye posted:Not being familiar with your hardware at all (what transmitter, what receiver? P23 and p24 of what?) I'm guessing you're not bright enough. You may have to juice the poo poo out of the LEDs, make a big bank of them, I don't know. Look at a few of the tv-b-gone projects for inspiration - they are trying to do something similar as far as distance goes. Sorry yeah that was terribly explained. P23 and P24 of the GPIO on a Raspberry PI. I'm just using an IR LED (I'm not sure exactly which). Reading around it seems the Raspberry Pi won't be able to do it because it doesn't have enough current - and I'll need a transistor http://upverter.com/alexbain/f24516375cfae8b9/Open-Source-Universal-Remote/embed_img/13702922880000/ This is a schema for what I need to do apparently. I guess the transistor has one pin to ground, one to GPIO, and one to the LEDs with 5v on the other side. Am I reading that correctly?
|
|
#
?
Dec 3, 2013 18:38
|
|
|
It would probably help to have an IR camera around. Fortunately, that often boils down to "what's the oldest cell phone you have" and possibly stripping a filter off.Delta-Wye posted:I'm guessing you're not bright enough.
|
|
#
?
Dec 3, 2013 18:38
|
|
|
movax posted:Corruption? Good call. I found on the zedboard it was fine just leaving rootfs rw but on the developed board things would go bad fast. Signal integrity for the sd card signals I suppose? I miss Zynq stuff now, mostly just writing C test programs on multi core devices at a new job. Thinking of getting a microzed to mess around with at home!
|
|
#
?
Dec 3, 2013 18:50
|
|
|
Delta-Wye posted:I'm guessing you're not bright enough. rude  priznat posted:Good call. I found on the zedboard it was fine just leaving rootfs rw but on the developed board things would go bad fast. Signal integrity for the sd card signals I suppose? Maybe; speed is limited to ~25MHz if you route it through EMIO, but I think in general, removable SD cards just don't seem suited for this type of use (eMMC seems to be the better choice). Trying to remove as much risk / issue(s) as possible though by picking a "nice" SD Card. Jawn I know you're lurking this thread somewhere, I feel like you may have done stuff with 24-bit parallel RGB.
|
|
#
?
Dec 3, 2013 18:53
|
|
|
movax posted:Jawn I know you're lurking this thread somewhere, I feel like you may have done stuff with 24-bit parallel RGB. Nothing generic enough for your question. But I'd be happier cutting a check to someone like Ramtex and getting along with my application than reverse engineering an undocumented pin interface.
|
|
#
?
Dec 3, 2013 19:07
|
|
|
JawnV6 posted:Nothing generic enough for your question. But I'd be happier cutting a check to someone like Ramtex and getting along with my application than reverse engineering an undocumented pin interface. I've got full documentation for the screen and the display controller, their documentation is just lovely-enough (TM) that it doesn't provide useful information re: layout tips. Right now the physical layout of the system has pins divided up R-G-B immediately over a connector that is B-G-R, and byte order is swapped 
|
|
#
?
Dec 3, 2013 19:11
|
|
|
That's is the point where I ask one of my counterparts working in Shenzhen to call up the factory and get a reference design or something. Cheer up though, sounds like there's only a hundred or so legit configurations you have to physically wire up and test! You can totally get that done by the end of the year. It would probably be more fun to design a big enough mux and give yourself physical switches to change each of RGB to the other channels and swap the endianness of each byte. Collapses the space of possibilities and simplifies the test plan to switch configurations instead of resoldering every time. Even flex connectors are a pain to swap around a few dozen times.
|
|
#
?
Dec 3, 2013 19:29
|
|
|
movax posted:I've got full documentation for the screen and the display controller, their documentation is just lovely-enough (TM) that it doesn't provide useful information re: layout tips. Right now the physical layout of the system has pins divided up R-G-B immediately over a connector that is B-G-R, and byte order is swapped What exactly are you doing/need help with? Have you got a link to the docs for the screen and controller?
|
|
#
?
Dec 3, 2013 19:57
|
|
|
JawnV6 posted:It would probably help to have an IR camera around. Fortunately, that often boils down to "what's the oldest cell phone you have" and possibly stripping a filter off. Turns out the iSight on a Macbook shows IR, so thanks for that. It makes it much easier to know if something is happening or not. With the IR LED connected the way I had it I can't even see anything using the camera (though it does turn the TV on from very close range). I have a couple of baby questions if anyone is willing to answer them. I'm trying to put together the bottom part of http://upverter.com/alexbain/f24516375cfae8b9/Open-Source-Universal-Remote/embed_img/13702922880000/ To the best of my understanding: GPIO 22 goes to the transistor B pin The transistor E pin goes to GND Now I'm confused what I connect the transistor's C pin to. My first guess was to the LED + and to 5v (with the LED - going to GND) but that will just mean the LEDs get 5v constantly won't it? I'm all kinds of confused... can anyone point me in the right direction?
|
|
#
?
Dec 3, 2013 19:58
|
|
|
Heisenberg1276 posted:My first guess was to the LED + and to 5v (with the LED - going to GND) but that will just mean the LEDs get 5v constantly won't it? You're really close. Think about the transistor as a gate. Without voltage on the 'B' pin, it won't connect C to E. So you have a LED with 5v on one pin and the other end floating. Nothing will happen. Until, of course, you put voltage on B, then C and E are connected and your circuit is complete. edit: oh, LED '-' pin goes to C, sorry i missed that
|
|
#
?
Dec 3, 2013 20:09
|
|
|
priznat posted:Zynq corruption? I can relate. Got the highly recommended Sandisks and they still flaked out after a day or so. I ended up making a read only partition for the rootfs. Can you talk about the kind of corruption you mean? We use SD cards in our products here, and occasionally they get irreparably corrupted, and in fact sometimes even trying to fix them using low-level SD card tools on a PC doesn't work (stuff like changing the reported capacity and stuff). Various measures have been taken to fix issues, including changing SD card suppliers, but I'm wondering if you (or anyone else) knows anything about what you can actually mess up when writing to a card.
|
|
#
?
Dec 3, 2013 20:12
|
|
|
Slanderer posted:Can you talk about the kind of corruption you mean? We use SD cards in our products here, and occasionally they get irreparably corrupted, and in fact sometimes even trying to fix them using low-level SD card tools on a PC doesn't work (stuff like changing the reported capacity and stuff). Various measures have been taken to fix issues, including changing SD card suppliers, but I'm wondering if you (or anyone else) knows anything about what you can actually mess up when writing to a card. It starts to spew ext4-FS errors but after doing a fsck on it in a PC it is usually fine (until the next time). I did have one card that got to be unrecoverable but it was a cheap no name and I didn't spend any time root causing or anything, the card just went in the trash. I had tried to reduce the possibility of writes by disabling swap on the partition, trying with journaling both off and on, and nothing seemed to fix it 100%. Really the best possible solution is to mount the FS via nfs if it's available and/or keep a backup on the SD in RO mode with a rw user partition. Running the rootfs on a rw sd card is just a real hassle and gamble.
|
|
#
?
Dec 3, 2013 20:19
|
|
|
Heisenberg1276 posted:I'm absolutely new to all this and am trying to make an internet-enabled infrared transmitter (I want it to control some infrared curtains and also the TV) - so far I've managed to set up an IR receiver (with one pin going to ground, one to 3V, and one to p23) - this works fine. One general comment: IR LEDs, unlike normal LEDs, are often built not for sustained on/off usage, but for modulated transmissions. This means that instead of needing to be limited to 30 mA, they can run with little to no series resistance (ie. the inherent resistance of the diode can be enough). This works because the diode can't heat up to a dangerous temperature during the on-time off the modulated signal, and can then cool back down during the off-time. They probably change the geometry of the diode itself to help minimize heating, too. The circuit you linked to doesn't use any external resistor, but also has the voltage drop of 2 diodes and a BJT, which limits the maximum current through the diodes.
|
|
#
?
Dec 3, 2013 20:21
|
|
|
priznat posted:It starts to spew ext4-FS errors but after doing a fsck on it in a PC it is usually fine (until the next time). I did have one card that got to be unrecoverable but it was a cheap no name and I didn't spend any time root causing or anything, the card just went in the trash. For a new project using an actual OS, we might be mounting our RFS as read-only, and directing all of our log-files to a different partition (with journaling, i think...). Hopefully this would make a bad-pointer less likely to screw up something on the card compared to current stuff using custom embedded file system and SD card drivers...
|
|
#
?
Dec 3, 2013 20:25
|
|
|
Yeah as long as your rootfs is RO you should be fine. It was just a pain when developing but once it was stable it was fine to leave like that, with logs going to a separate rw partition.
|
|
#
?
Dec 3, 2013 20:33
|
|
|
FSMC posted:What exactly are you doing/need help with? Have you got a link to the docs for the screen and controller? This is the screen: http://www.newhavendisplay.com/specs/NHD-3.5-320240MF-ATXL-1.pdf Here's the controller: http://www.mouser.com/ds/2/20/gcc-2%20new-4948.pdf Got some interns working on putting it together, and I'm sanity checking their layout. Was hoping to find an appnote from someone like TI/Microchip/etc where they might have the checklist(s)/math handy for verifying net lengths / etc. Been hunting through TI DaVinci / other "media processor" docs to see if they've got anything.
|
|
#
?
Dec 3, 2013 20:41
|
|
|
JawnV6 posted:You're really close. Think about the transistor as a gate. Without voltage on the 'B' pin, it won't connect C to E. So you have a LED with 5v on one pin and the other end floating. Nothing will happen. Slanderer posted:One general comment: IR LEDs, unlike normal LEDs, are often built not for sustained on/off usage, but for modulated transmissions. This means that instead of needing to be limited to 30 mA, they can run with little to no series resistance (ie. the inherent resistance of the diode can be enough). This works because the diode can't heat up to a dangerous temperature during the on-time off the modulated signal, and can then cool back down during the off-time. They probably change the geometry of the diode itself to help minimize heating, too. Thank you very much for this. It works fine now from across the room, and I understand what the transistor is doing rather than blindly connecting it! Now just to package it up and my first electronics project will be complete
|
|
#
?
Dec 3, 2013 21:15
|
|
|
movax posted:This is the screen: http://www.newhavendisplay.com/specs/NHD-3.5-320240MF-ATXL-1.pdf "CONFIDENTIAL" haha
|
|
#
?
Dec 4, 2013 04:40
|
|
|
So the LCD clock panel on my oven is going out. It's audibly buzzing and doing the 60 Hz blink now. Display segments that shouldn't be lit up are being lit up, just not as bright. I haven't opened it up yet to check it out. What's the most likely culprit on that board?
|
|
#
?
Dec 4, 2013 05:30
|
|
|
kid sinister posted:So the LCD clock panel on my oven is going out. It's audibly buzzing and doing the 60 Hz blink now. Display segments that shouldn't be lit up are being lit up, just not as bright. I haven't opened it up yet to check it out. What's the most likely culprit on that board? If I had to totally guess, a diode failed in the rectifier and is now shorted.
|
|
#
?
Dec 4, 2013 07:33
|
|
|
movax posted:Anyone have any manufacturers to recommend for high-reliability / industrial-grade microSD cards? For my bachelor's thesis we used Cactus SD cards, they're SLC based so should last a long rear end time. Prices are ridiculous though, and maximum size is very small compared to modern MLC.
|
|
#
?
Dec 4, 2013 19:24
|
|
|
Greetings electronics goons. I have a failing in my skillset I want to rectify. How does one become smart with connectors? I am unaware of the impact of choosing .1" headers against molex, or SMA against BNC. Where does one get the knowledge to choose application appropriate connectors? If I'm being too vague, one of the cases I am working on is wiring up a set of thermoelectric coolers so they can be easily swapped between for a TEC.
|
|
#
?
Dec 5, 2013 04:30
|
|
|
motoh posted:Greetings electronics goons. I have a failing in my skillset I want to rectify. If anyone answering this question cares to tack on a little bit extra, I've been curious what options I should be looking into for PCB to PCB connections for standard microcontroller communication formats (I2C, SPI, etc.) over distances of 1 to 10 feet that need to be idiot resistant and reasonably durable.
|
|
#
?
Dec 5, 2013 04:38
|
|
|
poeticoddity posted:If anyone answering this question cares to tack on a little bit extra, I've been curious what options I should be looking into for PCB to PCB connections for standard microcontroller communication formats (I2C, SPI, etc.) over distances of 1 to 10 feet that need to be idiot resistant and reasonably durable. You can consider off the shelf cables like ethernet cables (which can be shielded). Just plan the twisted pairs. With I2C you could do SDA twisted with power and SCL twisted with GND for example. HDMI cables also work and offer more pins. Otherwise standard 0.1" ribbon is convenient. motoh posted:Greetings electronics goons. I have a failing in my skillset I want to rectify. Honestly, connectors suck. Beyond that it all depends on the application, and annoyingly, every application seems to demand entirely different connectors. The important factors are shielded/not, twisted/not, current capacity (gauge), voltage rating, number of conductors, size, shape, ribbon/crimp etc. Twisted pairs and shielding both go a long way to improving signal integrity with the next step being impedance controlled (usually for analog or very high speed). 0.1" can be twisted but aren't impedance controlled or shielded (generally). SMA and BNC are shielded and impedance controlled which is ideal for signal integrity but they only carry a single signal. My go-to for connectors is Samtec, their catalog and documentation is a step above the rest (cost is a bit higher too but for low volume or hobby this doesn't matter). asdf32 fucked around with this message at 05:07 on Dec 5, 2013 |
|
#
?
Dec 5, 2013 04:59
|
|
|
Connectors really, really suck. Whoever figures out how to make browsing connectors easy will be a bajillionaire. Honestly, I find paper catalogs useful for connectors. I've got an entire drawer full of 'em at work. You can thumb through a catalog, point at the connector that "looks" right, and try to find it on Digikey. Otherwise, cozy up to a few manufacturers and learn the lineup. I like Samtec, Molex, and JST for most things. Samtec offers free samples on just about everything, and I've never been able to order enough that they cut me off.
|
|
#
?
Dec 5, 2013 14:41
|
|
|
|
| # ? May 9, 2024 00:40 |
|
|
Connectors and cables are a field of their own. For the hobbyist, I'd recommend sticking to the ridiculously common commodity connectors. RJ11/RJ45, barrel power connectors, headphones/RCA plugs, D-sub, banana, Molex KK, .100 ribbon, etc. If it doesn't show up at your local big box store it probably isn't worth fooling with. Automotive connectors might be great if you need to survive constant vibration, heat, and grime but your project probably doesn't need it. You can't afford circular connectors. Tiny low-profile multipack connectors are too much trouble to actually install and mate.asdf32 posted:SMA and BNC are shielded and impedance controlled which is ideal for signal integrity but they only carry a single signal. I wouldn't recommend BNC for anything besides low frequency stuff, VHF and below. Your typical commercial BNC is an absolute piece of junk. I've had 50-ohm BNC connectors that were 75 ohms from immediately behind the interface all the way to the cable. A $50 BNC might be "good" to 8 GHz. Meanwhile, a $1 SMA from a fly-by-night factory in China will reach 12.4 GHz without issue.
|
|
#
?
Dec 6, 2013 00:01
|
|