Electrical Engineering V2 - Developers

[QUOTE=No_Excuses;44221033]I'm curious as to what software you personally use to design PCBs[/QUOTE] I use Altium.

[QUOTE=No_Excuses;44221033]I'm curious as to what software you personally use to design PCBs[/QUOTE] I use EAGLE as well, its pretty decent once you figure out how it works. The board size limitation is rather annoying however. Here is a board I did in EAGLE: [url=http://i.imgur.com/3jpEkca.jpg][img]http://i.imgur.com/3jpEkcal.jpg[/img][/url]

I made 4 SR latches with transistor NOR gates for shits and giggles. [thumb]http://i.imgur.com/KSyaoNY.jpg[/thumb] The Arduino is just a 5v supply, I haven't gotten around to making a little one to sit in the breadboard yet. I put in an order on BangGood for some more resistors and transistors because I was running low on common values. Aaand some tactile buttons and a 4 digit 7 segment display because why not. I'll probably make some D flip flops for a shift register next since I have some free time soon.

[QUOTE=pentium;44220498] [...] Really fucking pissed off now. I can't afford to rebuild the optic block but either something with the circuit went funny or the damn 4093's blew up my phototransistors. There's no really fun way to reliably salvage this unless you could somehow find a way to define a LOW state as 1v or less and a HIGH state as 1.5v or more.[/QUOTE] Would an array of photocouplers work? You could potentially find a PNP type, and work a bit of voltage divider magic to define those specific states.

I used NPN IR phototransistors because they were cheap and small. To make them work I ahd to grind the sides off so they could sit side by side. Actually after a day of thinking, playing with resistor values and crying while eating raw cookie dough I managed to salvage it. None of what I did is directly related to the 4093 but it overall fixes some problems. First I was using nine IR LED's as a light source and for reasons out of my control the emitted light between them and the phototransistors was not uniform from one to the next. I had to add additional resistors to add or drop the voltage to each LED so they were all just bright enough for the task at hand. Too bright and even with a blocked channel the phototransistor would trigger. [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/IMG_5455.jpg[/IMG] The second problem was I was getting crosstalk from light leaking everywhere. My desk lamp is incandescent so it also fouled up the optic block by blasting the desk with IR light. I had to dismantle the optic block and replace a piece of clear plastic with a piece of black plastic so excess light wasn't falsely triggering the phototransistors. Finally I went over the pullup resistors for each phototransistor and changed them until I was getting a uniform 4v or so between logic states. NOW I have a somewhat healthy output and I was back at a 5v operating voltage again. [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/IMG_5484.jpg[/IMG] [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/IMG_5490.jpg[/IMG] This is where my success stops. So far it seems that no matter what logic I put on the phototransistors the voltage will get pulled down to a point where the high/low is too tight for the logic analyzer and the 4093's to detect states. I tried replacing the pullup with different values to counter this but that was becoming a nightmare. Either there was too much voltage in a low state or not enough in a high logic state. Even more confusing is that the drop will vary between phototransistors. It's not uniform at all. No, this was happening before I mix and matched resistors too. Actually, it was even worse then. [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/IMG_5497.jpg[/IMG] (yeah, it only got pulled to 3v here but on other channels it's down to 2v.) Anyways, I'm just barely able to get it working with the logic analyzer now but again, the high/low is too tight. I'm not sure how to deal with that. [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/IMG_5502.jpg[/IMG] [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/IMG_5508.jpg[/IMG]

I was about to say potentially buffering them with some Op-amps, but the 4093s already take only ~10uA. Potentially using some sort of AGC to level the phototransistors to optimal levels and then pipe it into logic, either that or replace the phototransistors with photodiodes to produce voltage instead of changing the gain (In assuming a voltage divider setup), that way you'll have more control over the gain ([URL="http://www.robotroom.com/ReversedLED.html"]I.e.[/URL])

Something is clearly loading the output, do you have a schematic ?

I think I was just tired. I tried the 4093 on a breadboard this morning and it worked. [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/IMG_5515.jpg[/IMG] Now to wire them in and move onto the microcontroller.

I really need to get myself a bench supply

There, the triggers are all in place. I had to add a 3904 and a 10K resistor because I have 8 data bits plus the timing sprocket hole. I need an order of capacitors to come in but we're otherwise ready to add in a microcontroller. I have one spare Pro Mini but I'm considering a PIC16C63. Really all we are doing is converting 8-bit patterns into 8-bit ASCII. [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/IMG_5547.jpg[/IMG] [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/IMG_5548.jpg[/IMG]

IEEE 754 (Single Point Float) is a strange mistress, I've figured it out by now for an assignment, however normalization is alot stranger than I would've expected (i.e. I thought it would only subtract the exponent by the number of bits from the left until the first 1 bit in the fraction, however I was wrong.) On a side note, our team has finally gotten most of the gremlins out of our navigation portion of our code, our image processing soon to follow.

Oh here's a weird problem. (on another project. Not related to the tape reader project above) Consider the schematic below. [IMG]http://i11.photobucket.com/albums/a166/ballsandy/asdsad.png[/IMG] The diode and phototransistor are part of an optical transducer. A blocked path generates a HIGH while an unblocked path is a LOW. That part is all correct but I couldn't figure out where an extra 1.2v was sneaking in and causing the 74LS14 to be stuck with a low output. (it won't switch high unless the input voltage is below .5v) Turns out it's coming frrm the 74LS14's input pin. I tried another 74LS14 and it was still there. What is going on here?

I am in no way qualified to be answering your question but could it have anything to do with the Schmitt trigger part of the 74LS14?

[QUOTE=pentium;44241990]Oh here's a weird problem. Consider the schematic below. [IMG]http://i11.photobucket.com/albums/a166/ballsandy/asdsad.png[/IMG] The diode and phototransistor are part of an optical transducer. A blocked path generates a HIGH while an unblocked path is a LOW. That part is all correct but I couldn't figure out where an extra 1.2v was sneaking in and causing the 74LS14 to be stuck with a low output. (it won't switch high unless the input voltage is below .5v) Turns out it's coming frrm the 74LS14's input pin. I tried another 74LS14 and it was still there. What is going on here?[/QUOTE] Could this just be because it's a TTL part? You're talking in terms of voltage but you also need to sink a good few milliamps of current to register as a low, right? Have you tried it with CMOS parts? [editline]blah[/editline] Do the waveforms look in spec if you take the 7414 away? If so you know the problem is caused by the introduction of the chip, more likely low impedance than a mystery voltage source!

[QUOTE=Leestons;44237296]I really need to get myself a bench supply[/QUOTE] What's more important, CV or CC? [editline]15th March 2014[/editline] I'm guessing CV but just want to check

[QUOTE=Leestons;44243223]What's more important, CV or CC? [editline]15th March 2014[/editline] I'm guessing CV but just want to check[/QUOTE] I have a powersupply that you can just set current limiting, it works perfectly for anything, if you want to control by voltage, just set the current up to max, if you want to control by current, set the voltage to max and slowly increase the current.

[QUOTE=pentium;44241990]Oh here's a weird problem. Consider the schematic below. [IMG]http://i11.photobucket.com/albums/a166/ballsandy/asdsad.png[/IMG] The diode and phototransistor are part of an optical transducer. A blocked path generates a HIGH while an unblocked path is a LOW. That part is all correct but I couldn't figure out where an extra 1.2v was sneaking in and causing the 74LS14 to be stuck with a low output. (it won't switch high unless the input voltage is below .5v) Turns out it's coming frrm the 74LS14's input pin. I tried another 74LS14 and it was still there. What is going on here?[/QUOTE] The input high current is 40uA which is equivalent to 125kOhm, so assuming the transistor is off your 5V will be attenuated to 1.05V which is below the HIGH threshold of the 74LS14, I suggest changing the 470k to 100k or lower. Not taking impedance in to account is the most common mistake, you always have to ask yourself "will this load my source" whenever you put components in series.

[QUOTE=Chryseus;44243466]The input high current is 40uA which is equivalent to 125kOhm, so assuming the transistor is off your 5V will be attenuated to 1.05V which is below the HIGH threshold of the 74LS14, I suggest changing the 470k to 100k or lower. Not taking impedance in to account is the most common mistake, you always have to ask yourself "will this load my source" whenever you put components in series.[/QUOTE] If I use a value lower than 470k the voltage on the low side of the signal rises (eg: on a 470k it's 1.2v and on a 220k it's about 2v). Now, I can use say a 1K resistor as a pulldown and that cancels out the voltage from the LS14 but then the high end is also dragged down. Okay, so replace the 470k to still pass a voltage, right? Well, it becomes a weird balancing game and after about two hours playing with values I still couldn't get it.

The transistor isn't pulling down hard enough due to the limited base drive, turn it into a darlington pair configuration. Or add a high impedance buffer before the Schmitt trigger so you can use a nice large resistor. Or use a CMOS Schmitt trigger.

Thanks Chryseus, your idea worked. [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/logic.jpg[/IMG] [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/IMG_5552.jpg[/IMG]

Does any one have any quadcopter resources? I don't know enough about motors or propellers and I'd like to change that.

[QUOTE=IpHa;44251664]Does any one have any quadcopter resources? I don't know enough about motors or propellers and I'd like to change that.[/QUOTE] Unless you want to create your own motor controller, you don't need to know much about them. When you buy a motor somewhere (Hobbyking for example), they will tell you the optimal propeller for this motor.

And if someone was to create their own motor controller, do you have any resources on that?

[QUOTE=Gulen;44254771]And if someone was to create their own motor controller, do you have any resources on that?[/QUOTE] Begin researching on the basics on motor control such as H-Bridges and the like, also research PWM motor control. And then you can look at such chips such as the L293 for starters or some smaller ICs that support both current sensing and PWM inputs.

[QUOTE=LoneWolf_Recon;44255004]Begin researching on the basics on motor control such as H-Bridges and the like, also research PWM motor control. And then you can look at such chips such as the L293 for starters or some smaller ICs that support both current sensing and PWM inputs.[/QUOTE] Don't use L293. They're ancient as hell, and there are many better drivers for the same money.

[QUOTE=DrDevil;44255438]Don't use L293. They're ancient as hell, and there are many better drivers for the same money.[/QUOTE] They're not the best for professional models, for that I'd recommend some of ST's [URL="http://www.st.com/web/catalog/sense_power/FM1965/SC1039/PF145473"]VNH2SP30[/URL] series of motor drivers, I just recommend the L293 as an education basis to learn the basics of motor. control

[img]http://u.cubeupload.com/Chryseus/MOqJJw.jpg[/img] It's time gentlemen, to venture in to the weird and wonderful world of switching power supplies. I can smell the magic smoke already.

-snipped because I'm a moron-

That IC is really close to those caps left of it. Just keep in mind that these components have size tolerances, and you might want to invest a couple more mm to give yourself a little room. About the inductor: I think your inductor isn't connected to vcc properly. Try moving the vcc symbol up by 1 grid.

I dunno about Eagle but you usually need to forward annotate the schematic changes to the PCB in order for it to properly deal with new parts or changes to the netlist. Your layout is also too cramped, there is a lot of wasted space at the bottom and top areas, there is no need to line your components up perfectly.