Electrical Engineering V2 - Developers

I've been experimenting with some cheap RF transmitters and receivers today, I bought them for £3 each a week or so ago and only just had time to mess around with them They are hilariously temperamental, and very sensitive to noise, but I actually found a way to make them work reasonably reliably: [img]https://dl.dropbox.com/u/10518681/Screenshots/2013-03-16_23-27-56.png[/img] The noise between the messages can be easily removed, the key point is that the string of bytes we care about is coherent and correct. Purely through an informed guess, I got it to work this reliably by sending a byte of data ("\") every 10ms if it wasn't transmitting any other data. Incredibly inefficient but it seemed to work (and I still can't figure out why, even prepending a string of bytes before sending my data didn't help!) If anyone wants to try them, it's these little buggers: [img]http://winavr.scienceprog.com/sites/default/files/images/stories/2008/rx_tx_433/RX433_TX433.jpg[/img] [url]http://dlnmh9ip6v2uc.cloudfront.net/datasheets/Wireless/General/RWS-374-3_315MHz_ASK_RF_Receiver_Module_Data_Sheet.pdf[/url] [url]http://dlnmh9ip6v2uc.cloudfront.net/datasheets/Wireless/General/TWS-BS-6_315MHz_ASK_RF_Transmitter_Module_Data_Sheet.pdf[/url] I might try making some kind of encoding such as Manchster, like XBees do (though I'm not sure they actually use Manchester specifically) but if anyone has any better ideas I'll gladly try them

I figured out a way I can get the mouse click working on my keyboard using the joystick click, despite already having no free pins. I'll just wire it up as part of the matrix and map a special function for that "key" to work as the left mouse button. Problem is, I came up with the following design and it does nothing as far as I can tell. [img]https://dl.dropbox.com/u/62766551/Norseboard/mouse_click.png[/img] Does anyone have any advice on the situation? Is there something fundamental that I'm missing here?

[QUOTE=HiddenMyst;39941621]I figured out a way I can get the mouse click working on my keyboard using the joystick click, despite already having no free pins. I'll just wire it up as part of the matrix and map a special function for that "key" to work as the left mouse button. Problem is, I came up with the following design and it does nothing as far as I can tell. [IMG]https://dl.dropbox.com/u/62766551/Norseboard/mouse_click.png[/IMG] Does anyone have any advice on the situation? Is there something fundamental that I'm missing here?[/QUOTE] Not sure I understand your switch diagram, is it active high or is it high pull up with an active low? NPNs are only really good for pulling "down" on a high signal, as you need a positive voltage of > ~0.6v from base to emitter to enable current flow. They always sink current down to the emitter - when your row gets scanned you're actually going to be detecting the small base-emitter current, so the key will always appear to be on! Ideally you would have a PNP with the switch attached as a sinking sensor: [IMG]http://i.imgur.com/0gU1Vfj.png[/IMG] You're not really at liberty to do that, but you [I]can[/I] invert the switch's output with an NPN and then pass this signal to a PNP. Even better would be to solder wires directly to the switch on the board so it behaves just like the other contact switches on the keyboard.

[QUOTE=DrLuckyLuke;39937862]Got a whole tape of mega1284p, though it's the MU package. [thumb]http://i.imgur.com/XCUFWBR.jpg[/thumb][/QUOTE] Not breadboard or idiot friendly.

[QUOTE=r0b0tsquid;39943501]Not sure I understand your switch diagram, is it active high or is it high pull up with an active low? NPNs are only really good for pulling "down" on a high signal, as you need a positive voltage of > ~0.6v from base to emitter to enable current flow. They always sink current down to the emitter - when your row gets scanned you're actually going to be detecting the small base-emitter current, so the key will always appear to be on! Ideally you would have a PNP with the switch attached as a sinking sensor: [IMG]http://i.imgur.com/0gU1Vfj.png[/IMG] You're not really at liberty to do that, but you [I]can[/I] invert the switch's output with an NPN and then pass this signal to a PNP. Even better would be to solder wires directly to the switch on the board so it behaves just like the other contact switches on the keyboard.[/QUOTE] It turned out I had the transistor the wrong way around from the get go, but I sorted that out. Then I encountered the issue that you described, where it was passing HIGH until the button was pressed, which is seemingly an issue, until it hit me that I don't care about the state, just as long as I can detect the change, because at the end of the day I can just change the code the the mouse "key" to work in the opposite way to the others.

[QUOTE=Van-man;39943833]Not breadboard or idiot friendly.[/QUOTE] You'd be surprised how easy it is to solder them onto a pcb. They pretty much snap into place once the solder is liquid.

You just need hot air gun right? Because I can't imagine doing this with iron.

[QUOTE=HeatPipe;39944481]You just need hot air gun right? Because I can't imagine doing this with iron.[/QUOTE] If you get the TQFP version you can solder it with an iron, if you get the QFN part you need a hot air gun.

[QUOTE=HeatPipe;39944481]You just need hot air gun right? Because I can't imagine doing this with iron.[/QUOTE] You can also use a cooking plate to solder. That's how I solder all my boards so far.

[QUOTE=DrLuckyLuke;39944623]You can also use a cooking plate to solder. That's how I solder all my boards so far.[/QUOTE] Yeah, I can imagine that. Doesn't that gets messy?

[QUOTE=DrLuckyLuke;39944429]You'd be surprised how easy it is to solder them onto a pcb. They pretty much snap into place once the solder is liquid.[/QUOTE] Are you really implying a beginner could design and make their own PCBs? I like my DIP packages, I just put them into my breadboard/protoboard, solder them if needed, bam it's ready.

[QUOTE=DrLuckyLuke;39944429]You'd be surprised how easy it is to solder them onto a pcb. They pretty much snap into place once the solder is liquid.[/QUOTE] [I]"If you make something idiot proof, someone will just make a better idiot"[/I] Besides I [I]JUST[/I] barely managed to solder a 5 pin sot23 package chip. suddenly having to deal with 44 pins that are even closer to each other would make me a nerve wreck. Also, breadboards and perfboards, waiting for cheap pcb's to arrive from Asia or paying out of the ass for quick made & delivered one's isn't something I fancy. And I'd rather spend time on [URL="http://filesmelt.com/dl/DSC004172.JPG"]bodging up a franken UV laser plotter then[/URL]. I've had nothing but failures with toner transfer methods, and there's no good places nearby where I can get something copied over to "overhead projector" film. for UV box use.

[QUOTE=Van-man;39947118][I]"If you make something idiot proof, someone will just make a better idiot"[/I] Besides I [I]JUST[/I] barely managed to solder a 5 pin sot23 package chip. suddenly having to deal with 44 pins that are even closer to each other would make me a nerve wreck. Also, breadboards and perfboards, waiting for cheap pcb's to arrive from Asia or paying out of the ass for quick made & delivered one's isn't something I fancy. And I'd rather spend time on [URL="http://filesmelt.com/dl/DSC004172.JPG"]bodging up a franken UV laser plotter then[/URL]. I've had nothing but failures with toner transfer methods, and there's no good places nearby where I can get something copied over to "overhead projector" film. for UV box use.[/QUOTE] I've become a fan of strip board recently, it combines the flexibility and rapid production of breadboards with the permanent nature of a PCB, without the messy wiring required for a perfboard Just pick up a track breaker and you're set

I tend to make traces on pad board. [url=http://i.imgur.com/zR90o9O.jpg][img]http://i.imgur.com/zR90o9Ol.jpg[/img][/url] Poor man's circuit board I guess.

[QUOTE=ddrl46;39947479]I tend to make traces on pad board. [url=http://i.imgur.com/zR90o9O.jpg][img]http://i.imgur.com/zR90o9Ol.jpg[/img][/url] Poor man's circuit board I guess.[/QUOTE] looks neat but doesn't that waste a lot of solder?

[QUOTE=Trumple;39947494]looks neat but doesn't that waste a lot of solder?[/QUOTE] Not really but I'm not sure how strong the traces are yet, regular wire or the leads of components soldered down to the pads are probably a better way to go. Edit: Better example on square pads. [url=http://i.imgur.com/BFc6EdZ.jpg][img]http://i.imgur.com/BFc6EdZl.jpg[/img][/url]

[img]http://i.imgur.com/Rs7Sk.png[/img] Try not to vomit.

[QUOTE=ddrl46;39947527]Not really but I'm not sure how strong the traces are yet, regular wire or the leads of components soldered down to the pads are probably a better way to go. Edit: Better example on square pads. [URL="http://i.imgur.com/BFc6EdZ.jpg"][IMG]http://i.imgur.com/BFc6EdZl.jpg[/IMG][/URL][/QUOTE] I've been looking for cheap & large (board size) square pad perdboards all over the internet. It's almost like finding a leprechauns pot o' gold, damn impossible. [QUOTE=Chryseus;39947625]http://i.imgur.com/Rs7Sk.png Try not to vomit.[/QUOTE] Beats the time I hot-glued components upside-down on a piece of cardboard and sloppily soldered on wires between component legs.

[QUOTE=Van-man;39947707]I've been looking for cheap & large (board size) square pad perdboards all over the internet. It's almost like finding a leprechauns pot o' gold, damn impossible. Beats the time I hot-glued components upside-down on a piece of cardboard and sloppily soldered on wires between component legs.[/QUOTE] I get mine from [url]http://bitsbox.co.uk[/url], they sell a pretty large square pad board.

[QUOTE=VistaPOWA;39946951]Are you really implying a beginner could design and make their own PCBs? I like my DIP packages, I just put them into my breadboard/protoboard, solder them if needed, bam it's ready.[/QUOTE] Making PCBs really isn't that hard. The first PCB I made was dips only, and the second PCB I ever made directly jumped to 0603 and QFN44 (!!!) packages. I even handapplied solderpaste, which is a bit troublesome (you get too much very easily), so I switched to stencils instead. Mind you that this is a commercial project so I get everything paid, which is why I understand why you wouldn't want to use stencils. But the fact is, that tinning the pads and cooking your PCB on a cooking plate works a treat, just apply a lot of flux (make sure it no-clean flux, because you won't be able to clean it up from under the IC). The hardest step in all of this process is to have the balls and try it out, and risk wasting a few bucks at first. But in the long run you end up with being able to produce higher quality electronics than any breadboard or perfboard could ever give you. Also it looks really beautiful: [thumb]http://i.imgur.com/wVP6ofT.jpg[/thumb] [thumb]http://i.imgur.com/jTeOQJ2.jpg[/thumb]

[QUOTE=DrLuckyLuke;39947989]Making PCBs really isn't that hard. The first PCB I made was dips only, and the second PCB I ever made directly jumped to 0603 and QFN44 (!!!) packages. I even handapplied solderpaste, which is a bit troublesome (you get too much very easily), so I switched to stencils instead. Mind you that this is a commercial project so I get everything paid, which is why I understand why you wouldn't want to use stencils. But the fact is, that tinning the pads and cooking your PCB on a cooking plate works a treat, just apply a lot of flux (make sure it no-clean flux, because you won't be able to clean it up from under the IC). The hardest step in all of this process is to have the balls and try it out, and risk wasting a few bucks at first. But in the long run you end up with being able to produce higher quality electronics than any breadboard or perfboard could ever give you. Also it looks really beautiful: [/QUOTE] You can make cheap-ass stencils out of an aluminum soda can and my beloved hydrochloric acid using the toner transfer method.

[QUOTE=ddrl46;39947819]I get mine from [url]http://bitsbox.co.uk[/url], they sell a pretty large square pad board.[/QUOTE] Either that's a incredibly shittily designed site, or I can only find large boards with round pads.

[QUOTE=Van-man;39948817]Either that's a incredibly shittily designed site, or I can only find large boards with round pads.[/QUOTE] He must be confused I don't remember them ever selling square pads.

[QUOTE=Chryseus;39947625][img]http://i.imgur.com/Rs7Sk.png[/img] Try not to vomit.[/QUOTE] I am intrigued, what is it? Poor man's pcb club unite! Fuck fancy pcb's anyway. [editline]18th March 2013[/editline] Nevermind, I must learn to read.

Just finished my Keypad Security Lock project, which I've been working on and off with a break in the middle for the past few weeks now. Held together with sticky pads and hookup wire, I present my latest creation: [media]http://www.youtube.com/watch?v=YtwQJi1Ke7s[/media] (excuse the shoddy camera angles, it's tricky to hold the camera and type when the keypad is on the other side of the door!) Components: ATmega 328p Strip board ABS housing Full rotation servo (why did I get a full rotation one...) Reed switch and magnet pair (to accommodate for the stupid decision above) 3x4 matrixed keypad (Internal circuit: [url]http://www.rapidonline.com/pdf/78-0305.pdf[/url]) 240v -> 0.7A 5v mains transformer ribbon cable, hookup wire, resistors, etc. So, the software side is actually pretty straightforward. I'm using a matrix keypad, so I just use a scanning method to set the columns high, and check if any input is high, and infer which key is pressed depending on which output/input pair is high. You can see that in the source code Beyond that, it's just some PWM to change the speed and direction of the servo, an array to store the password, a bit of serial debugging using a serial class I made for an earlier project, and some reed switch feedback. The overall program is essentially "check the sequence of input buttons is correct, if so turn the servo for 170ms" while it's unlocked and "do the same but only turn the servo the other way until the reed switch is active, then stop" when it's in the locked state. I do plan to add on-the-fly password changing so when my housemates figure out the password I can change it without taking it apart. But I have to take it apart to add that feature because I didn't make a programming port :v: The hardware side was an absolute pain in the ass. The ABS box needed to house the servo, so I had to cut out a hole using just a strip board track cutter because I left all my tools at home (I'm currently at my house at Uni). I ended up cutting a rough rectangle by making numerous holes, then sanding it down to make it look acceptable. The hardest part was hooking up to the pre-existing lock, which I could not modify because I'm renting this house and didn't really want to fix it back up when I moved out/got tired of this lock. So, I opted for hookup wire (I know...) and looped it around the grub screw on the shaft of the lock. It actually worked surprisingly well, and I could thread it through the servos arm holes. On top of this, I needed to ensure the lock didn't turn too much (because the lock eventually goes over a "click" which requires too much force than the servo can handle to go back over), and that the servo could turn the lock enough wherever it was positioned. Oh and the reed switch needed to be positioned enough so that the timed reversal/unlocking of the door was just right (trial and error!). I really should have bought a standard 270 degrees servo, so I could choose the angle and not have to worry about feedback. Surprisingly, the switches were happy without debouncing. Perhaps that's done in the keypad, it isn't really specified in the datasheet. So anyway, source code for those interested: [url]http://pastebin.com/tvfxuV92[/url] I wonder how long til' it falls apart and I get locked outside!

[QUOTE=Van-man;39948817]Either that's a incredibly shittily designed site, or I can only find large boards with round pads.[/QUOTE] [url]http://www.bitsbox.co.uk/hardware.html[/url] Scroll down, the large and medium boards with round pads actually have square pads (at least when I last ordered them).

[QUOTE=DrLuckyLuke;39947989]Making PCBs really isn't that hard. The first PCB I made was dips only, and the second PCB I ever made directly jumped to 0603 and QFN44 (!!!) packages. I even handapplied solderpaste, which is a bit troublesome (you get too much very easily), so I switched to stencils instead. Mind you that this is a commercial project so I get everything paid, which is why I understand why you wouldn't want to use stencils. But the fact is, that tinning the pads and cooking your PCB on a cooking plate works a treat, just apply a lot of flux (make sure it no-clean flux, because you won't be able to clean it up from under the IC). The hardest step in all of this process is to have the balls and try it out, and risk wasting a few bucks at first. But in the long run you end up with being able to produce higher quality electronics than any breadboard or perfboard could ever give you. Also it looks really beautiful: [thumb]http://i.imgur.com/wVP6ofT.jpg[/thumb] [thumb]http://i.imgur.com/jTeOQJ2.jpg[/thumb][/QUOTE] Are there any non-chemical methods for manufacturing PCBs? My govt. cracked down really hard on drug manufacturers, and it's very hard to get chemicals, unless you have a license for them.

[QUOTE=VistaPOWA;39957035]Are there any non-chemical methods for manufacturing PCBs? My govt. cracked down really hard on drug manufacturers, and it's very hard to get chemicals, unless you have a license for them.[/QUOTE] I'm just getting mine manufactured in china: [url=https://docs.google.com/document/d/1p6FH25ltGpzJQ5_8fbflDukqEKghiEcpuhJpngth2Is/edit]Hackvana[/url] But you can also use a CNC machine to route a PCB out, though you most likely won't be able to go down to QFN sizes.

[QUOTE=VistaPOWA;39957035]Are there any non-chemical methods for manufacturing PCBs? My govt. cracked down really hard on drug manufacturers, and it's very hard to get chemicals, unless you have a license for them.[/QUOTE] You could try ordering ferric chloride or ammonium persulfate powder/crystals on ebay ( or just look in your local electronics shop ) and hope they slip through customs, although i believe there won't be any problems. Have you tried looking for hydrochloric acid in hardware stores? I usually find it in the paint section near all the solvents and spirits. Trust me, making your own PCB's ( low quantities, not product manufacturing ) is so much cheaper and faster than ordering them from China. Also: where do you live that you can't get basic chemicals because someone might make meth out of them?

[QUOTE=DrLuckyLuke;39957131]I'm just getting mine manufactured in china: [url=https://docs.google.com/document/d/1p6FH25ltGpzJQ5_8fbflDukqEKghiEcpuhJpngth2Is/edit]Hackvana[/url] But you can also use a CNC machine to route a PCB out, though you most likely won't be able to go down to QFN sizes.[/QUOTE] I don't have access to a CNC machine, I can't even design PCBs that are meant to be manufactured. I am really just a beginner.