It works but you're driving the input beyond the maximum rated operation, so it'll just slowly die. At least put a current limiting resistor there, resistor divider would probably be even more ideal.
Or just build a level-shifter, that's even better.
I did some more testing, and I can get the LED to dim, but not turn off entirely. If I follow Lonewolf's schematic and have a direct connection between emitter and ground, nothing happens though, but it dims with a 100ohm resistor.
Anyone have a good formula for calculating flyback energy of an inductive load? I'm not really wanting to add a flyback diode so I want to verify that my MOSFET can tolerate it repeatedly (I'm assuming that particular parameter is avalanche energy or maybe just max VDS).
[QUOTE=Gulen;49467190]I did some more testing, and I can get the LED to dim, but not turn off entirely. If I follow Lonewolf's schematic and have a direct connection between emitter and ground, nothing happens though, but it dims with a 100ohm resistor.[/QUOTE]
From the schematic that I posted, is R your 100ohm resistor or is it R2? What is the other value of the resistor (If R = 100ohms then what is R2)?
For the LED to fully turn off the voltage across it (ergo the voltage across the transistor's collector/emitter, C-E, terminals) must be less than the LED's turn-on/forward voltage ([URL="https://en.wikipedia.org/wiki/Light-emitting_diode#Colors_and_materials"]List of forward voltages vs LED colors[/URL]). For example, your red LED has an approximate forward voltage of 1.7V, to have it fully turn off we want the voltage across it to be much less than that due to the [URL="https://en.wikipedia.org/wiki/Diode#Shockley_diode_equation"]Diode Equation[/URL].
In order for the C-E voltage to be lower, the transistor must be able to sink the current that would normally go through the LED. Based upon the equations Chryseus provided, compute the collector current required to turn off the LED.
e.g. Lets say we want the C-E voltage to be near zero (How about 0.1V) with R2 = 1kOhms on a 9V battery. Thus the voltage drop across R2 would be 9V - 0.1V, with a current flow of 8.9mA. In order to sink that current using a [URL="https://www.fairchildsemi.com/datasheets/BC/BC547.pdf"]BC547[/URL] transistor (which has a minimum current gain, HFE, Beta_F of 110) the base current must be atleast 80.9uA. Common B-E voltage drops on transistors are around 0.7V, thus (9V - 0.7V) / (80.9uA) = R = 102kOhms maximum.
These values can vary (most notably the B-E voltage drop) which is due to several things but you don't have to worry about it now. (However a great DC model for a BJT transistor is the [URL="https://en.wikipedia.org/wiki/Bipolar_junction_transistor#Ebers.E2.80.93Moll_model"]Eber-Moll's Model[/URL]).
[QUOTE=LoneWolf_Recon;49469326]Anyone have a good formula for calculating flyback energy of an inductive load? I'm not really wanting to add a flyback diode so I want to verify that my MOSFET can tolerate it repeatedly (I'm assuming that particular parameter is avalanche energy or maybe just max VDS).[/quote]
The voltage is V = L * dI/dt
The energy is e = 0.5 * L * I^2
In practice I just usually increase the supply voltage while monitoring Vds and apply a snubbing network as needed.
[QUOTE=Chryseus;49469519]The voltage is V = L * dI/dt
The energy is e = 0.5 * L * I^2
In practice I just usually increase the supply voltage while monitoring Vds and apply a snubbing network as needed.[/QUOTE]
Perfect, thanks, thought it was alot more complicated due to inductor parasitic resistive losses. But worst case scenario is welcome. I typically assume that if the max energy/amperage is less than a 1/10 of the maximum then it can be repetitive. (Famous last words)
with UART, isn't it RX-TX to TX-RX from one device to another?
My ESP module and Arduino are wired RX > RX and TX > TX. It does work though.
[editline]6th January 2016[/editline]
Nevermind, has to do with me bypassing the ATmega328 again. Wiring displayed on the pcb is for that one. It results in RX/TX being switched around when the ATmega16u2 is doing stuff.
[editline]6th January 2016[/editline]
With voltage divider (R1 = 200 ohm, R2 = 390 ohm) the ESP module didn't receive data anymore. Unsoldered R2 to GND and it works again. Atleast I have a current limiting resistor there now.
Isn't the telltale signs for failed capacitors that they are bulging at either or both of the ends and that they're optionally leaking, yet they can still fail without any visual indicator?
Because I've currently disassembled my Asus monitor which had flickering that slowly got worse over ½ year period.
But the capacitor all LOOK fine, even those on the logic board and not just those on the power supply board.
Also the only branding I can see on the capacitors on the power supply board are KY on some of them and KMG on others, I got no idea if they belong on the capacitor "naughty list"
If it's not Nichicon or Rubycon, it's probably shit. I'd replace them.
[QUOTE=Van-man;49492533]Isn't the telltale signs for failed capacitors that they are bulging at either or both of the ends and that they're optionally leaking, yet they can still fail without any visual indicator?
Because I've currently disassembled my Asus monitor which had flickering that slowly got worse over ½ year period.
But the capacitor all LOOK fine, even those on the logic board and not just those on the power supply board.
Also the only branding I can see on the capacitors on the power supply board are KY on some of them and KMG on others, I got no idea if they belong on the capacitor "naughty list"[/QUOTE]
yeah and another great way to tell your caps are prolly garbage is if they're only rated for like 80-85c or whatever the lower one is. obviously there are some cases where that doesnt matter but ideally they should pretty much all be 105c rated. even if they dont get anywhere near that got it's really an indicator of a long operating life generally speaking
Only problem regarding replacing them is that the main "pre-stage" capacitor is a 400volt 150uf bastard that's hard to find a equivalent of cheap, without being some shitty Chinese capacitor that's probably even worse than the current one.
Also being rated for only 400v yet placed right after the 1:1 isolation transformer and diode bridge means it's exposed to approx 320v peak, which is a little bit too close for my liking.
Should probably look for something 500volt 150uf for replacing that.
You most likely aren't going to find a cheap replacement for that; there's a reason the name brand caps cost that much and it's because they're going to last.
Not sure what size you need in particular but here's a [URL="http://www.mouser.com/ProductDetail/Nichicon/LKX2W151MESA35/?qs=sGAEpiMZZMvwFf0viD3Y3QOh4aH3eNb8JnRn7npZgOU%3d"]150uf 450V Nichicon[/URL] for a good price.
[editline]9th January 2016[/editline]
Just noticed that's a snap-in package but through hole caps should be a similar price.
[QUOTE=Van-man;49492984]Only problem regarding replacing them is that the main "pre-stage" capacitor is a 400volt 150uf bastard that's hard to find a equivalent of cheap, without being some shitty Chinese capacitor that's probably even worse than the current one.
Also being rated for only 400v yet placed right after the 1:1 isolation transformer and diode bridge means it's exposed to approx 320v peak, which is a little bit too close for my liking.
Should probably look for something 500volt 150uf for replacing that.[/QUOTE]
The main high voltage filter cap right after the rectifier almost never fails in those kind of supplies fyi.
[QUOTE=Van-man;49492984]Only problem regarding replacing them is that the main "pre-stage" capacitor is a 400volt 150uf bastard that's hard to find a equivalent of cheap, without being some shitty Chinese capacitor that's probably even worse than the current one.
Also being rated for only 400v yet placed right after the 1:1 isolation transformer and diode bridge means it's exposed to approx 320v peak, which is a little bit too close for my liking.
Should probably look for something 500volt 150uf for replacing that.[/QUOTE]
When you look for a replacement, make keen note of the ESR and try to lower it as much as possible while still being in your price range. Caps heating from joule ESR heating (Typically not that much, but in power supplies or rectifier backends it matters alot more) can lead to shorter lifespans.
[QUOTE=ddrl46;49493559]The main high voltage filter cap right after the rectifier almost never fails in those kind of supplies fyi.[/QUOTE]
That's both what I'm hoping and also suspecting, since the smaller "final stage" caps are sandwiched between some TO-220's bolted to heatsinks, which means those caps have always been pretty toasty.
While that main cap is located far away from any components developing any significant amount of heat.
[QUOTE=LoneWolf_Recon;49494588]When you look for a replacement, make keen note of the ESR and try to lower it as much as possible while still being in your price range. Caps heating from joule ESR heating (Typically not that much, but in power supplies or rectifier backends it matters alot more) can lead to shorter lifespans.[/QUOTE]
Keep ESR as low as possible, while temperature and voltage rating ideally higher than original, got it.
Guess I don't need to replace the capacitors after all.
Turns out the TO220 packages surrounding the capacitors are Schottky diodes, but the one responsible for the 5v line were barely even attached to its heatsink.
Desoldered both diode and heatsink & then unscrewed the diode from the heatsink, and it became apparent why it wasn't attached well to its heatsink.
There were some significant burrs around the screw-hole in the heatsink, so I degreased both diode and heatsink, removed burrs, put some good thermal paste on the contact area between diode and heatsink and screwed it in properly and soldered it up again.
Apparently that did the trick, because I haven't seen any flicker after doing that & attaching the power supply for a test-run.
Weird, but it has saved me from buying capacitors (for now).
Do you guys know what's inside those fluke k type wire adapters?
[QUOTE=POLOPOZOZO;49507131]Do you guys know what's inside those fluke k type wire adapters?[/QUOTE]
[URL="https://en.wikipedia.org/wiki/Thermocouple#Type_K"]Chromel & Alumel[/URL]
[QUOTE=LoneWolf_Recon;49507281][URL="https://en.wikipedia.org/wiki/Thermocouple#Type_K"]Chromel & Alumel[/URL][/QUOTE]
I can't find it one there, I'm referring to the little block that the wire feeds into, for fluke it's called 80ak-a, it doesn't seem like the leads are soldered directly to the banana plugs, otherwise I could make my own. Measuring the resistance across the plugs gives ~5 ohms so something is definitely in there and I can't figure out what.
So I took apart this LED TV today. There's 60 in series split in two sections giving about 110V @ 400mA per string, according to LG's datasheet. Being in series, if one fails, they all fail. I gotta give it to them for making the backlight so accessible but clearly they've fucked the dog if it fails after 2.5 years. Never seen inside so I thought it was kind of neat looking. Strips of SMD LEDs with a neat looking lense on each one.
[thumb]http://i.imgur.com/8FfhDl7.jpg[/thumb]
[thumb]http://i.imgur.com/pYCvLjt.jpg[/thumb]
I found the replacement LEDs on ebay so I might take a crack at it. The strips are fairly expensive but the LED's are reasonable.
I have a schematic for an ADSR envelope generator that I want to put together as part of a project. I've only ever moved simple schematics to stripboard, but I think the circuit is too complex to try and fit on a stripboard (and too much potential for mistakes).
I'm considering having a pcb fabricated (I'll need 3/4 of them made), but I have no experience using layout software. Are there any useful techniques for going from a schematic to a layout? I'd probably try and keep it simple with a 2 layer board, with just a few jumper connections if need be.
[QUOTE=Mattz333;49530686]I have a schematic for an ADSR envelope generator that I want to put together as part of a project. I've only ever moved simple schematics to stripboard, but I think the circuit is too complex to try and fit on a stripboard (and too much potential for mistakes).
I'm considering having a pcb fabricated (I'll need 3/4 of them made), but I have no experience using layout software. Are there any useful techniques for going from a schematic to a layout? I'd probably try and keep it simple with a 2 layer board, with just a few jumper connections if need be.[/QUOTE]
OSHPark is honestly your best bet with $5USD/in^2 (cheap and high quality boards with wonderful silkscreen) and they give you 3 boards per order.
I also recommend DipTrace for Schematic/PCB software as they typically have a full library of components that tie their schematic symbols with the physical part design that would be incorporated into the PCB layout. EagleCAD and KiCAD are also other options.
For something as simple as ADSR, you can rely upon the autorouter/autoplacer of the software that you're using for the PCB layout (after setting proper [URL="http://cryoarchive.net/tutorials/diptrace-tutorials/diptrace-design-your-pcb-for-oshpark-com/"]design rules from OSHPark for DipTrace[/URL] and the board size you'd like to make). After that you simply export the design as a set o Gerber and Drill files, combine them into a zip file, upload to OSHPark, review/verify and order.
The link for the design rules is honestly the best tutorial for using DipTrace with OSHPark.
These fuckers were hiding INSIDE a sealed AC line filter. I had to dremel the box open to replace them.
[IMG]http://i11.photobucket.com/albums/a166/ballsandy/IMG_2268.jpg[/IMG]
[QUOTE=LoneWolf_Recon;49531130]OSHPark is honestly your best bet with $5USD/in^2 (cheap and high quality boards with wonderful silkscreen) and they give you 3 boards per order.
I also recommend DipTrace for Schematic/PCB software as they typically have a full library of components that tie their schematic symbols with the physical part design that would be incorporated into the PCB layout. EagleCAD and KiCAD are also other options.
For something as simple as ADSR, you can rely upon the autorouter/autoplacer of the software that you're using for the PCB layout (after setting proper [URL="http://cryoarchive.net/tutorials/diptrace-tutorials/diptrace-design-your-pcb-for-oshpark-com/"]design rules from OSHPark for DipTrace[/URL] and the board size you'd like to make). After that you simply export the design as a set o Gerber and Drill files, combine them into a zip file, upload to OSHPark, review/verify and order.
The link for the design rules is honestly the best tutorial for using DipTrace with OSHPark.[/QUOTE]
Any thoughts on CircuitMaker? I use Eagle sparsely and I usually find it pretty clunky but I've never used anything else up this point.
My main problem with Eagle is that making anything but the most basic custom component footprint is a huge pain. Like the mechanical CAD part of it is just atrocious.
Check out KiCAD, I recently switched to it from EAGLE and it's pretty nice so far.
[QUOTE=No_Excuses;49537732]Any thoughts on CircuitMaker? I use Eagle sparsely and I usually find it pretty clunky but I've never used anything else up this point.
My main problem with Eagle is that making anything but the most basic custom component footprint is a huge pain. Like the mechanical CAD part of it is just atrocious.[/QUOTE]
Circuit maker is shit, don't poke it even with a 3m pole. Your schematics should never be stored in a cloud only, that's utter shit.
[QUOTE=paindoc;49537292][url]http://www.elab-hackerspace.org/2015/09/18/ion-thruster/[/url]
:o[/QUOTE]
Holy fuck, I always thought it wasn't possible in normal air without really low thrust, but that's seemingly a strong bit of thrust. I really want to try that out with Xenon or some other gas in a vacuum chamber now.
[QUOTE=No_Excuses;49537732]Any thoughts on CircuitMaker? I use Eagle sparsely and I usually find it pretty clunky but I've never used anything else up this point.
My main problem with Eagle is that making anything but the most basic custom component footprint is a huge pain. Like the mechanical CAD part of it is just atrocious.[/QUOTE]
I haven't tried CircuitMaker (first time that I've heard of it), I dislike Eagle quite alot because the interface is a clusterfuck with little organization IMO yet it has a sort of de-facto standard behind it as almost alot of manufactures/companies explicitly work in their proprietary brd format. (Thankfully DipTrace can import EagleCAD schematics.)
Also I didn't like the workflow of Eagle, it didn't flow as well as DipTrace. DipTrace is seamless comparatively with the Component Design -> Schematic Capture -> PCB Layout process.
I hate all PCB tools, they're either too bloated, lacking features, missing common parts or a pain in the arse to use.
That pretty much sums up every piece of software ever made.
[url]http://fritzing.org/home/[/url] has done me well from simple stuff. Check it out and i think you can add custom parts also.
Sorry, you need to Log In to post a reply to this thread.
Electrical Engineering V3
