Electrical Engineering V3 - Developers

When I started using my synths and such it was weird for me to realize I was actually in a situation where wire quality could be a thing, even slightly. I'm trying to figure out if a mega 2560 has enough power to perform orbital Mechanics calculations. An idea I have for a side project is to write something akin to guidance computer firmware for a satellite, since I need a boost for my SpaceX internship app. I know the arduino has no proper FPU, but it should work, yes? Most orbital mechanics math is just vector calculus, parametrization, and differential equations afaik. [editline]21st October 2015[/editline] Microwave antenna is kaput because I can't afford to buy any supplies and I have 3 spare megas sitting in my desk

[QUOTE=No_Excuses;48950006]How many boards and what do they do? I'm about to pick up this receiver that has some issues with the left channel. When I looked up the model on google I came across this [URL="https://docs.sony.com/release/ES_STR_05_Final.pdf"]technical brochure from Sony[/URL]. There's some great reading in it such as: [img_thumb]http://i.imgur.com/48CPUYk.jpg[/img_thumb] (Yet More Audiophile Shit) [/QUOTE] Oh boy time to dig this out again: [IMG]http://oi60.tinypic.com/2lx86ma.jpg[/IMG] I mean atleast the inductors could have some legitimacy to them, cylindrical inductors are more prone to exernal B fields thus can pickup noise moreso than torodial (all B fields confined to the toroid). But even that is stretching it.

[QUOTE=paindoc;48950572]When I started using my synths and such it was weird for me to realize I was actually in a situation where wire quality could be a thing, even slightly. I'm trying to figure out if a mega 2560 has enough power to perform orbital Mechanics calculations. An idea I have for a side project is to write something akin to guidance computer firmware for a satellite, since I need a boost for my SpaceX internship app. I know the arduino has no proper FPU, but it should work, yes? Most orbital mechanics math is just vector calculus, parametrization, and differential equations afaik. [editline]21st October 2015[/editline] Microwave antenna is kaput because I can't afford to buy any supplies and I have 3 spare megas sitting in my desk[/QUOTE] I'd say it depends more on how much stuff you have to monitor and control, because the 2560 has a fair amount of juice behind it, but it can still only effectively do one thing at a time. But as far as calculating the math behind orbital mechanics (rather than actually maintaining an orbit of some sort), should be fine, just a lot slower than a modern CPU would do.

[QUOTE=No_Excuses;48950517]$30k for that power cable...is that site actually serious? I've seen things like this before but the site layout is so amateur and the name, Coconut Audio.[/QUOTE] I got that website when the video of some idiots setup got posted in CIPWTTKT a while back, don't have the video link because it's been deleted. Couldn't see shit because he was crawling around with a fucking torch to not disrupt his precious power lines. All his cables were those silver ones, sparkly rocks on everything. Giant tubes on USB cables and everything.

[QUOTE=Zero-Point;48951585]I'd say it depends more on how much stuff you have to monitor and control, because the 2560 has a fair amount of juice behind it, but it can still only effectively do one thing at a time. But as far as calculating the math behind orbital mechanics (rather than actually maintaining an orbit of some sort), should be fine, just a lot slower than a modern CPU would do.[/QUOTE] Thats what I figured. The constraints are part of the challenge for me, I may construct look up tables for the trigonometric functions since those seem to be a bottle neck from what I've seen so far. Part of my idea is that the use of microcontrollers like this is closer to the hardware in spacecraft as is. I might get really crazy and make a fancy box for it with buttons and switches and lights and stuff. I haven't programmed in ages, but the microcontroller programming was always my strong point. I only have a basic understand of the mathematics of orbital mechanics now, but one of the things I most want on my birthday list is a hardcover orbital mechanics textbook. I've had a ball reading the pdf online and doing those practice problems, and I love love love physical copies of textbooks to study and read from. Although I guess this would be trying to maybe "maintain" an orbit of sorts. I'm trying to figure out a way to simulate this firmware actually doing something, and I've found a few free libraries of stuff that could work but I also know some people have integrated arduino into KSP and used those as hardware controllers. This would be effectively the same, just sending hardware commands involving translation/rotation and such. I'm really interested by this problem now, damn. My differential equations class has gotten me really interested in applications and retaking vector calc has made me really like it this time around. We're supposed to start learning the Laplace transform soon and I"m unbelievably hyped about that. [editline]21st October 2015[/editline] Forgive me for my verbosity, I'm just a bit excited this morning. Might actually do well academically this quarter and I'm starting to take more fun math/science classes, and next year (if i get into aero/astro) I start the real shit like Spacecraft Design, Spacecraft Power systems, Spacecraft Controls Integration, Introduction to Aerospace Plasmas, Satellite Communication, etc etc. I'm hoping to get my PhD in aerospace plasmas or controls through my dept and the dream is getting real close now after 13 years of boring school.

[QUOTE=paindoc;48953095] [b]Forgive me for my verbosity, I'm just a bit excited this morning.[/b] Might actually do well academically this quarter and I'm starting to take more fun math/science classes, and next year (if i get into aero/astro) I start the real shit like Spacecraft Design, Spacecraft Power systems, Spacecraft Controls Integration, Introduction to Aerospace Plasmas, Satellite Communication, etc etc. I'm hoping to get my PhD in aerospace plasmas or controls through my dept and the dream is getting real close now after 13 years of boring school.[/QUOTE] Nonsense, I shall do no such thing. They say necessity is the mother of invention, but I'd say more accurately that it's passion. It's good that you're that enthusiastic about it!

It's fun when you have a radio that needs one certain connector for power that can only be found in about three places and retails for $90-170 per connector. Probably going to end up soldering wires directly to the radio.

[QUOTE=paindoc;48953095]We're supposed to start learning the Laplace transform soon and I"m unbelievably hyped about that.[/QUOTE] Laplace transforms are pretty baller. What's that? An awkward differo-integral equation? Nah, I'll turn it into a polynomial and solve that instead!

[QUOTE=Falcqn;48953590]Laplace transforms are pretty baller. What's that? An awkward differo-integral equation? Nah, I'll turn it into a polynomial and solve that instead![/QUOTE] And then do lots of partial fractions and look at it for a long time to figure out what it is in t domain again :wow: I'm kidding, they are incredibly baller. [editline]vlah[/editline] Slightly less baller if your diffy Q is not linear We've had one question where a differential equation in t becomes another differential equation in s, which you can quickly get a series solution for, which gives you another series solution in t. Mental.

Use the shit out of Laplace & Fourier transforms. Potentially you could throw the required equations into matlab, by-hand and just transform them down to easier polynomials that are computationally easy. But either way its dubious to do orbital math using pure integers, however the AVR math.h library is already double floating point based so you could do some tests on [URL="http://people.ece.cornell.edu/land/courses/ece4760/Math/Floating_point/"]their performance[/URL]. Some of the 32-bit AVRs (The UC3 variety) have built in FPUs while still being supported by WinAVR IIRC. Also Atmel has direct rad-hard designs that can be transitioned too easily from their commercial counterparts. Good luck with your studies & the internship! I honestly wish my Uni had those classes and keep us posted!

[QUOTE=LoneWolf_Recon;48954872]Use the shit out of Laplace & Fourier transforms. Potentially you could throw the required equations into matlab, by-hand and just transform them down to easier polynomials that are computationally easy. But either way its dubious to do orbital math using pure integers, however the AVR math.h library is already double floating point based so you could do some tests on [URL="http://people.ece.cornell.edu/land/courses/ece4760/Math/Floating_point/"]their performance[/URL]. Some of the 32-bit AVRs (The UC3 variety) have built in FPUs while still being supported by WinAVR IIRC. Also Atmel has direct rad-hard designs that can be transitioned too easily from their commercial counterparts. Good luck with your studies & the internship! I honestly wish my Uni had those classes and keep us posted![/QUOTE] Thanks! Yeah, the FPU problem and precision of the floats is what I see. My primary extracurricular work right now is with the 3D printing club, and one of our members (applied math PhD) is a fellow EE mad scientist, and his first big paper was on optimizing the kinematics equations to run faster. He actually rewrote the entire core of grbl's kinematics system, then ported it to run on the Arduino Due's chipset core, or something. Been a while since I talked to him. Made his own PCB I think for just 3D printers. He's head of a makerspace now :v: has a pick and place, pcb scribe/mill, reflow oven, the whole works. I'll probably talk to him again and get some perspective and tips from him, we get on quite well. Also the rad-hard versions of the atmel chips and how common those are is the majority of the reason I want to use them. If this gets advanced enough, down the road I make look at redundancy for bit-flipping. Should at least include it in my technical writeup. Do the atmel chipsets have ECC for that stuff? [editline]21st October 2015[/editline] I can't use Matlab yet, but I'm taking an entire course on it next quarter, starting in January so that option will be explored too. Due to classes I have to take this slowly, but I'm documenting stuff and gathering research when I can already. [editline]21st October 2015[/editline] So many of my friends don't understand why this list gets me so hyped [url]https://www.aa.washington.edu/courses/courseResult.php?search=ALL[/url] Some of you may catch why I'm hyped. Seems I mixed courses up or changed names around, heh. The Satellite comms was one of the "special topics" classes last year afaik

[QUOTE=paindoc;48953095]Thats what I figured. The constraints are part of the challenge for me, I may construct look up tables for the trigonometric functions since those seem to be a bottle neck from what I've seen so far. Part of my idea is that the use of microcontrollers like this is closer to the hardware in spacecraft as is. I might get really crazy and make a fancy box for it with buttons and switches and lights and stuff. I haven't programmed in ages, but the microcontroller programming was always my strong point. I only have a basic understand of the mathematics of orbital mechanics now, but one of the things I most want on my birthday list is a hardcover orbital mechanics textbook. I've had a ball reading the pdf online and doing those practice problems, and I love love love physical copies of textbooks to study and read from. Although I guess this would be trying to maybe "maintain" an orbit of sorts. I'm trying to figure out a way to simulate this firmware actually doing something, and I've found a few free libraries of stuff that could work but I also know some people have integrated arduino into KSP and used those as hardware controllers. This would be effectively the same, just sending hardware commands involving translation/rotation and such. I'm really interested by this problem now, damn. My differential equations class has gotten me really interested in applications and retaking vector calc has made me really like it this time around. We're supposed to start learning the Laplace transform soon and I"m unbelievably hyped about that. [editline]21st October 2015[/editline] Forgive me for my verbosity, I'm just a bit excited this morning. Might actually do well academically this quarter and I'm starting to take more fun math/science classes, and next year (if i get into aero/astro) I start the real shit like Spacecraft Design, Spacecraft Power systems, Spacecraft Controls Integration, Introduction to Aerospace Plasmas, Satellite Communication, etc etc. I'm hoping to get my PhD in aerospace plasmas or controls through my dept and the dream is getting real close now after 13 years of boring school.[/QUOTE] I've tried writing an orbital simulation on my computer once, and that already took an enormous computational time. Orbital calculations rarely are done on the spacecraft themselves, but on the ground. The reason for that is the enormous amount of data required to describe the non-uniform gravitational field of the earth, and the heavy processing power required to turn that information into a force you can apply to your spacecraft. And if you don't take the non-uniformity into account, your prediction will only be good for less than one orbit. That can already be useful when you try to approach another object for example, but only if the calculation itself doesn't take half an orbit.

[QUOTE=DrDevil;48958259]I've tried writing an orbital simulation on my computer once, and that already took an enormous computational time. Orbital calculations rarely are done on the spacecraft themselves, but on the ground. The reason for that is the enormous amount of data required to describe the non-uniform gravitational field of the earth, and the heavy processing power required to turn that information into a force you can apply to your spacecraft. And if you don't take the non-uniformity into account, your prediction will only be good for less than one orbit. That can already be useful when you try to approach another object for example, but only if the calculation itself doesn't take half an orbit.[/QUOTE] Shit, I totally forgot about the non-uniformity of the gravitational field. Can't believe I forgot that too, it was a big deal and reveal by our physics professor last year that we had fun with. Okay, consolidating to making the "onboard" computer capable of doing orbital maneuvers and transfer, and calculating maneuver nodes to rotate the craft towards. I'm going to have to do more research on this, I found a pdf of that orbital mechanics book but I'm trying to not read it and wait for the hardcover on my birthday [editline]edited[/editline] As much as I love my fellow ME's and MsE's, they are bad at taking care of the Arduino's strapped to their 3D printers here. I have a stack of 3 Arduino's, 2 of which have completely fried 5v regulators (not a problem for me) and one of which has a bootloader that needs to be reflashed with ISP. Powering the first two from a supply rail thats already regulated, or over usb, works. The latter one I'm reflashing with another Arduino used as an ISP. Yay, free Arduino's! Because my plan just got a lot more mad scientist-y. And crazy. And unnecessarily complex but hey in the case of a fun project I will embrace this complexity.

Yes. [b]YES![/b] :excited: [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/IMG_1344.jpg[/IMG] [media]https://www.youtube.com/watch?v=xnGbBbd-RLw[/media] Once I disabled a darlington pair by removing a resistor I was able to supply an external voltage. From that I added a 100 ohm resistor to stabilize the voltage further due to output distortion and suddenly both on the bench and when clipped to the equipment things started working. If it passes a few more tests I'll finish it off by installing the bell ringer and fabricating the new perfboard.

I wrote the start of a vectormath library last night. So far, just have the functions for dot and cross products, and then the ability to save to non-volatile flash and load from flash into the volatile SRAM on the Mega. That needs more work, since i don't understand the avr progmem library, but my arduino cookbook (reference manual with wonderful examples) is under my 80lb tv. Plans are to find that and improve memory tracking and be able to not forget where i saved vectors, and I'll have to add a clean function but that might have to go in a more general memory management class. Also want to get the unit vector, vector projection, and scalar projection. Not sure how to optimize for performance, though. also I did all of this to avoid panicking about my diff eq exam in about 30 mins. I studied pretty well already and studying last minute late at night or this morning probably wouldnt have done much good... [editline]23rd October 2015[/editline] I forgot how much I loved C and hated the arduino IDE. Had to set up a custom run command so I could use N++ and then just compile in arduino haha

[b]YES! YES! Sweet christ after two and a half years of work it's finally finished!![/b] Testing the ringer. The DTMF adapter means that we drop the anti-tinkle resistor (technically it's still recommended but it's unlikely to be needed these days) and all you need is the capacitor to only current to flow in AC mode. [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/CGS_1113.jpg[/IMG] The redesigned hookswitch. Still two relays and the capacitor but the pin headers are now for phone/power in, bell out and passthrough to the DTMF adapter. [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/CGS_1114.jpg[/IMG] The DTMF adapter is mounted on the backside of the transistor board. It draws +5 from the power rail while the pin headers are the phone line and input form the rotary dial. [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/CGS_1115.jpg[/IMG] The finished CCU. All components fitted and ready for final testing. [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/CGS_1116.jpg[/IMG] This thing now also consists of 12 pages of hand drawn diagrams, 32 lines of Nikomo's Arduino code and countless hours bickering between Chryseus, Ddrl46 and a few other people on other forums.

So I made a LED tester for my high power LEDs and I managed to sear reddish orangish spots into my visual cortex. The good news is, all of my LEDs are functioning. The bad(cool?) news is, everything I'm writing here seems to be highlighted with a yellow/orange spots.

So I'm taking another junior/senior level "research" next quarter class where I'm going to be learning about GPU and voxel computing, and cluster computing+parallel processing etc. Its graded based on how you apply the project and demonstrate your learning with an academic journal, as the professor doesn't believe midterms to be a fair or effective way of demonstrating knowledge. And he likes seeing applications. I think he'll like some of my ideas :D Also ordering a XMS432P401R launchpad from TI- Cortex M4F so I get a real FPU plus actual hardware divide/multiple and ridiculously low power. and easily clustered up. Slow going on the arduino work too. I forgot all the weird formatting shit arduino does, and I'm investigating a couple workarounds to get higher precision out of my hardware now. Sorta hesitant to work too hard if I have new hardware on the way though, since its so much more powerful and adaptable. Before I do order, has anyone used the C2000 series? They seem really cheap and able to do what I want so I could order 2-3 of them and play more with parallel processing. Or if anyone else has ideas for powerful MCU's for <$20 since I am actually that broke right now.

Made one little modification just to make this thing totally modular. The motor control relay was moved from the motor fuse to just before the plug block for the motor. This can now be dropped into any model 33 teletype machine and no wires need to be routed or added to the existing machine. [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/CGS_1117.jpg[/IMG] Finally, for reference the original UCC6 standard teletype power supply and console control. Below that the new UCC6[b]H[/b] Hayes modem interface built from the same assembly. [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/UCC6.jpg[/IMG] [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/CGS_1118.jpg[/IMG] [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/CGS_1119.jpg[/IMG] [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/CGS_1122.jpg[/IMG] [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/CGS_1123.jpg[/IMG] [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/CGS_1121.jpg[/IMG] [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/CGS_1120.jpg[/IMG] There, I'm done talking about it. I need to take a vacation. :v:

[QUOTE=pentium;48972702]Made one little modification just to make this thing totally modular. The motor control relay was moved from the motor fuse to just before the plug block for the motor. This can now be dropped into any model 33 teletype machine and no wires need to be routed or added to the existing machine. [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/CGS_1117.jpg[/IMG] Finally, for reference the original UCC6 standard teletype power supply and console control. Below that the new UCC6[b]H[/b] Hayes modem interface built from the same assembly. [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/UCC6.jpg[/IMG] [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/CGS_1118.jpg[/IMG] [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/CGS_1119.jpg[/IMG] [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/CGS_1122.jpg[/IMG] [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/CGS_1123.jpg[/IMG] [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/CGS_1121.jpg[/IMG] [IMG]http://i11.photobucket.com/albums/a166/ballsandy/Computer%20related/CGS_1120.jpg[/IMG] There, I'm done talking about it. I need to take a vacation. :v:[/QUOTE] Okay so I'm missing a bunch of background here, to be fair, but what is this? It looks really really neat (i see the arduino nano hiding out btw) but I can't quite figure out what exactly it is or what it does. (not that it needs some special purpose, projects for the hell of it are still worthy :D)

[url=https://facepunch.com/showthread.php?t=1250138&p=43433374&viewfull=1#post43433374]Context from when I was just exiting the early drafting stage which took fucking forever.[/url] Between V2 and V3 there's a fair number of progress updates and LOTS of pictures. Also some hipster shithead with glasses who took my work and half-assed my design to flail around Youtube before I could.

[QUOTE=pentium;48976011][url=https://facepunch.com/showthread.php?t=1250138&p=43433374&viewfull=1#post43433374]Context from when I was just exiting the early drafting stage which took fucking forever.[/url] Between V2 and V3 there's a fair number of progress updates and LOTS of pictures. Also some hipster shithead with glasses who took my work and half-assed my design to flail around Youtube before I could.[/QUOTE] Thats really cool, I wish I could understand half of what you did .

Hey resident RF wizards, I want to get a nice stable BFO to complement my DDS VFO. I figured I could just slap in another TCXO at 20MHz but I didn't realize that 90% of them output Clipped Sinewave which is not good for an RF mixer. How would I go about getting a sinewave from the output of one? Or is there another way to get a good stable base frequency to mix with? I've tried googling but I don't really know what I'm looking for specifically. If you're wondering I'm working on a SSB Transceiver for 40 & 20 meters.

[QUOTE=papkee;48982851]Hey resident RF wizards, I want to get a nice stable BFO to complement my DDS VFO. I figured I could just slap in another TCXO at 20MHz but I didn't realize that 90% of them output Clipped Sinewave which is not good for an RF mixer. How would I go about getting a sinewave from the output of one? Or is there another way to get a good stable base frequency to mix with? I've tried googling but I don't really know what I'm looking for specifically. If you're wondering I'm working on a SSB Transceiver for 40 & 20 meters.[/QUOTE] You could use filters to get rid of all the harmonics. This limits you to a certain frequency range though.

[QUOTE=papkee;48982851]Hey resident RF wizards, I want to get a nice stable BFO to complement my DDS VFO. I figured I could just slap in another TCXO at 20MHz but I didn't realize that 90% of them output Clipped Sinewave which is not good for an RF mixer. How would I go about getting a sinewave from the output of one? Or is there another way to get a good stable base frequency to mix with? I've tried googling but I don't really know what I'm looking for specifically. If you're wondering I'm working on a SSB Transceiver for 40 & 20 meters.[/QUOTE] What IFs are you using?

[QUOTE=Cakebatyr;48983493]What IFs are you using?[/QUOTE] The plan was to use an IF of around 20MHz, and the VFO would be tuned so that either 7MHz or 14MHz would be mixed up to 20MHz and product detected from there. I'm working on the receiver first to get my feet wet in RF design. TX will be later.

I should not be allowed to shop on ebay for chinese clones of mcu hardware while drunk i ordered things [editline]26th October 2015[/editline] false alarm i didn't but oh god i hadn't checked ebay for some reason yet and now all my birthday money is spoken for

Just ordered some raw PCB's from AliExpress, gonna try and use our CNC mill to see if it's accurate enough for PCB prototyping :cool:

[QUOTE=kokonut;49000683]Just ordered some raw PCB's from AliExpress, gonna try and use our CNC mill to see if it's accurate enough for PCB prototyping :cool:[/QUOTE] Cool, I've also been trying out some PCB milling on my milling machine lately, just have to get the hang of board leveling and proper alignment for double sided boards. [url=http://i.imgur.com/xPJ3Ivs.jpg][img]http://i.imgur.com/xPJ3Ivsl.jpg[/img][/url] [url=http://i.imgur.com/xH3LFBd.jpg][img]http://i.imgur.com/xH3LFBdl.jpg[/img][/url]

I'm currently trying to build a stirling engine and an alternator with the hopes of being able to construct the two as a single unit, thus heat->stirling engine-> electricity. However I'm not very good at it ;_; I can't figure out how to make a crankshaft for the life of me. I've tried using a metal wire hangar, but it was too curvy. My current attempt is having fashioned one from modelling clay. My goal is to do it with as many DIY materials as possible. Furthermore, I don't entirely understand the basic principle and concepts of an alternator, just that they exist and function. Can anyone explain them or link me to somewhere where I can figure it out better? Thanks!