|
Selling a house. no loving clue how i imagine i will have to pay lawyers money and it makes the bile come up
|
# ? Aug 20, 2013 19:12 |
|
|
# ? Apr 26, 2024 09:49 |
|
peepsalot posted:its like 175/mo oh that's kind of a lot
|
# ? Aug 20, 2013 19:13 |
|
Jonny 290 posted:Selling a house. craigslist
|
# ? Aug 20, 2013 19:17 |
|
Jonny 290 posted:Selling a house. you gotta get a realtor or look up how to do FSBO
|
# ? Aug 20, 2013 19:20 |
|
Jonny 290 posted:Selling a house. do you guys have those signs stapled to telephone poles everywhere that say I WILL BUY YOUR HOUSE SELL ME YOUR HOUSE PLZ why not just do that
|
# ? Aug 20, 2013 19:21 |
|
FSBO: BYOB OH (open house)
|
# ? Aug 20, 2013 19:23 |
|
yeah i'm going to make up a nice one-sheet about the house and then hit up every single property management company worth a poo poo in the area i suspect a rental type would want to snatch up a property that mortgages for 600 and can rent for 1k easy
|
# ? Aug 20, 2013 19:26 |
|
after getting pissed at how useless and frustrating SSIS is, i'm building a sql server bulk data copy tool in c# and winforms. v1 is complete and does a naive copy of a list of tables while disabling and re-enabling constraints. (by 'naive' i mean it assumes the column schema matches 100% and even an extra nullable column will crash it) future plans: v2 i will make it asynchronous so it can do parallel loads (i believe i can do up to three simultaneously) and hopefully get it to the point where it can nicely handle tables where the columns are out of order or there are extra (nullable) columns in source or destination. v3 i will save/load config files (right now i use it for one purpose so app.config is good enough). v4 i would like to drop and recreate indexes and constraints before and after the run so it will be faster and i can use table truncates. right now it's doing unbatched deletes and that's super slow. also i really should convert it to mvp and stuff all the goodies in a class library so i can reuse it in other places easier, right now it's all in the codebehind for the main form, heh
|
# ? Aug 20, 2013 19:33 |
|
01011001 posted:https://www.youtube.com/watch?v=zyiY_3cyEuo here's a robutt. only fell in once filming it, too that's a gr8 robort
|
# ? Aug 20, 2013 19:35 |
|
01011001 posted:https://www.youtube.com/watch?v=zyiY_3cyEuo here's a robutt. only fell in once filming it, too
|
# ? Aug 20, 2013 19:45 |
|
i've been doing research for my sous vide thing and holy gently caress ive fallen down a rabbit hole. every question brings more questions: do you want to build a sous vide machine? yes. ok. do you know how PID controllers work? no. do you want to know? sure. do you remember high school calc? no. okay, relearn all that poo poo, then research how a pid works. now, do you want to use a regular pid controller or an arduino? i dunno. okay, research them. okay. i guess i want to use an arduino because its more customisable. great! what kind of temperature sensor do you want to use? a pt100. cool. that's insanely and unnecessarily complex for this application, but i like your gusto! do you have an ic in mind to amplify the voltage? huh? okay, research that and decide. now, what ui do you want? an lcd and four buttons. you know you'll need a contrast pot for that, right? ugh. that sounds unnecessary! i'll just set it once and never use it again. can't i use a resistor instead? sure, but you'll need a voltage divider. what's that? research it. okay, now, here are the libraries you'll need for the pid software. you just need to write the code to get your pt100 to talk to it, and the 328p to talk to the ssr. got it? but i don't know python! learn it. now, that should all be done! you know how to wire it up, right? which resistors/caps/transistors to use? no. do you know anything about any of those things? no. okay, learn about them. now, what form factor do you want to use? well, my hackspace has a solder reflow oven, so maybe i could use that... do you know how smd and solder reflow works? no, i can barely solder. you know you're biting off more than you can chew, right? ...yes, but i'll do it dammit! oh christ...
|
# ? Aug 21, 2013 00:04 |
|
yospos i'm going insane. it's poo poo i don't know all the way down. like, what the gently caress is this?!? jonny, pls tell me that you were as naive as i was before you built raspberry high...
|
# ? Aug 21, 2013 00:05 |
|
WorkingPeer posted:i've been doing research for my sous vide thing and holy gently caress ive fallen down a rabbit hole. every question brings more questions: you forgot to put funney emotes in
|
# ? Aug 21, 2013 00:06 |
|
graph posted:you forgot to put funney emotes in isn't that bannable now or was i just having a wonderful dream
|
# ? Aug 21, 2013 00:09 |
|
WorkingPeer posted:yospos i'm going insane. it's poo poo i don't know all the way down. this is a great thing you're doing, jumping in at the deep end. you'll learn an unbelievable amount if you stick with it, but it's going to take a while and you can't hope to do it all in one go. what you need to do is come up with some subproblems that will be easier to solve. that'll stop you from burning out. in this case, they could be things like
what you definitely shouldn't do is worry about all these things at once. come up with a list of two or three small subproblems, and forget about everything that doesn't contribute to solving those problems. you can come back and make a full design later on, and it'll be a much better design for all that you've learnt from the subproblems.
|
# ? Aug 21, 2013 00:21 |
|
coffeetable posted:this is a great thing you're doing, jumping in at the deep end. you'll learn an unbelievable amount if you stick with it, but it's going to take a while and you can't hope to do it all in one go. phew, thanks. it's good to know that i'm not being too stupid for jumping into this head-first. you're right: i've got to eat the elephant one bite at a time, as it were. i guess my first task is to go out and buy an arduino and go from there. any recommendations of which model i should get?
|
# ? Aug 21, 2013 00:27 |
|
coffeetable posted:this is a great thing you're doing, jumping in at the deep end. you'll learn an unbelievable amount if you stick with it, but it's going to take a while and you can't hope to do it all in one go. yo could you recommend a good simulator? i keep finding lovely java ones but im convinced there's some nerdtool that is exactly what i'd like it to be (drag components onto a space and connect them and inspect voltage and see things happen etc)
|
# ? Aug 21, 2013 00:29 |
|
example: for my climate sim, my early subproblems are
coffeetable fucked around with this message at 05:12 on Aug 29, 2013 |
# ? Aug 21, 2013 00:29 |
|
WorkingPeer posted:i've been doing research for my sous vide thing and holy gently caress ive fallen down a rabbit hole. every question brings more questions: you could just do this http://learn.adafruit.com/sous-vide-powered-by-arduino-the-sous-viduino/sous-vide i mean yes you'll learn a shitload about the nitty details by rolling your own. but if what you want is a sous-vide that you can put together over a couple days/wees but still lets you hac all the pid stuff on your own time then this may be a way to go
|
# ? Aug 21, 2013 00:31 |
|
WorkingPeer posted:phew, thanks. it's good to know that i'm not being too stupid for jumping into this head-first. you're right: i've got to eat the elephant one bite at a time, as it were. Mido posted:yo could you recommend a good simulator? i haven't played with electronics in years, and my knowledge isn't near extensive enough to make any recommendations on either of these, sorry . if no-one else jumps in, you might wanna go try the electronics thread in the DIY forum.
|
# ? Aug 21, 2013 00:32 |
|
WorkingPeer posted:yospos i'm going insane. it's poo poo i don't know all the way down. PID is proportional, integral, differential the overall concept is not that complex. you have an error function e(t) which is just (target value - measured value) for any given time t. now you calculate three different components of error think of a line graph of your temperature on the y axis and time on the x axis proportional is just how far away are you right now from target, this is the height difference for the latest point on the graph = e(t) integral of the error with time tells you your history of cumulative error, aka the area under the curve = running sum of (e(t) * dt) differential of the error is the rate at which you are approaching the target, aka the slope of your graph = (e(t) - (previous e(t))) / dt dt is how often you take a temp sample in seconds or whatever (sample period) each component gets scaled by its own K constant(Kp, Ki, Kd). you have to tweak these values to best fit your specific system then you add up the three components and that gives you the value that you feed back into the system (heater pwm or whatever)
|
# ? Aug 21, 2013 00:43 |
|
Sagebrush posted:you could just do this yeah that's actually what i was initially looking at when i decided to try arduino rather than a normal off-the-rack pid controller. i will definitely base my design off of it to a large extent, and probably use their code. i don't actually want to integrate the full arduino into the final device, though.
|
# ? Aug 21, 2013 00:48 |
|
WorkingPeer posted:i don't actually want to integrate the full arduino into the final device, though. you'll need to explain this bit
|
# ? Aug 21, 2013 00:55 |
|
re: arduino purchase idk what you need specifically but usually your choice here is mostly how many inputs/outputs you need of each type (digital/analog in your case) and if you need some major built in feature (you dont). uno or mega 2560 (duemilanove or diecimila might do it too but those are p old now) are your best bets most likely due to being straightforward and less gimmicky than a lot of those offered my amphibious robot up there uses a mega and it works p well even after its been submerged more than a few times
|
# ? Aug 21, 2013 00:55 |
|
WorkingPeer posted:i don't actually want to integrate the full arduino into the final device, though. not 100% sure what you want to do then
|
# ? Aug 21, 2013 00:57 |
|
peepsalot posted:PID is proportional, integral, differential that's a very good concise explanation of it, which is surprisingly hard to find on the internet. i like this video because it explains it using a car analogy: https://www.youtube.com/watch?v=XfAt6hNV8XM
|
# ? Aug 21, 2013 00:57 |
|
Mido posted:yo could you recommend a good simulator? i keep finding lovely java ones but im convinced there's some nerdtool that is exactly what i'd like it to be (drag components onto a space and connect them and inspect voltage and see things happen etc) you could try this one, it looks kinda polished but i haven't really used it https://www.circuitlab.com/ i think there is also a fruit tablet version
|
# ? Aug 21, 2013 00:57 |
|
Dr. Honked posted:you'll need to explain this bit as in i want to use the arduino as a testbed and usb interface to program the microcontroller, then pop out the microcontroller and build a circuit around it from scratch. so i can just buy another 328 and reuse the arduino. like so: https://www.youtube.com/watch?v=NCGzKDTFBSQ&t=855s
|
# ? Aug 21, 2013 01:11 |
|
https://github.com/jonny290/yosvape/blob/master/yosvape.ino the core of yosvape is just a fuckin straight import of the arduino PID library. i didnt even change the constant names i was so lazy i never messed with the PID tuning library that is meant to work alongside the PID lib itself, but i heard it can help you get it pretty decently tight, tweak from there.
|
# ? Aug 21, 2013 01:22 |
|
01011001 posted:re: arduino purchase i tend to use teensys for all quick hacks nowadays cause they're cheap and light and small and can be usb peripherals too! also im pretty sure they have more flash than any of the regular arduinos and they definitely have more ram. to run them away from a computer you just need a 5v supply which can be as simple as a 7805 (though i bought a pile of $3 hobbyking switching ubecs that make a great alternative) it's like yeah you can program a dip mega328 and put it on a custom pcb with the other components and make it all pretty but you're probably only making one of the things anyway so i just solder a bunch of flying leads and dead bug stuff to the teensy and stick it in a box
|
# ? Aug 21, 2013 01:30 |
|
oh and for the really tiny things i use msp430s. you can program them with arduino code if you download energia and those chips are cheaaaap
|
# ? Aug 21, 2013 01:34 |
|
WorkingPeer posted:as in i want to use the arduino as a testbed and usb interface to program the microcontroller, then pop out the microcontroller and build a circuit around it from scratch. so i can just buy another 328 and reuse the arduino. like so: arrr i see
|
# ? Aug 21, 2013 01:45 |
|
Sagebrush posted:re: re: arduino good to know if you have more stomach than me for wiring, for sure. my setup is a drat mess as it is
|
# ? Aug 21, 2013 02:40 |
|
Mido posted:yo could you recommend a good simulator? i keep finding lovely java ones but im convinced there's some nerdtool that is exactly what i'd like it to be (drag components onto a space and connect them and inspect voltage and see things happen etc) didnt bother to look to see if this is answered but there's about a million spice variants that you could use. i usually use ltspice for no particular reason. its free and has the features you mentioned.
|
# ? Aug 21, 2013 02:45 |
|
Jonny 290 posted:https://github.com/jonny290/yosvape/blob/master/yosvape.ino did you autotune the yosvape, or tune it manually?
|
# ? Aug 21, 2013 09:18 |
|
CALCULUS N*GGA
|
# ? Aug 21, 2013 14:47 |
|
actually dealing with solving the differential equations for a tepid pool machine sounds completely overkill. stick in a large K, a medium D, and a small I and try it out.
|
# ? Aug 21, 2013 16:17 |
|
Bloody posted:actually dealing with solving the differential equations for a tepid pool machine sounds completely overkill. stick in a large K, a medium D, and a small I and try it out. uh i'll stick with my large d, thanks very much.
|
# ? Aug 21, 2013 16:23 |
|
WorkingPeer posted:did you autotune the yosvape, or tune it manually? nah i started with some generic constants and then tuned a _bit_ but not even much. then i just proportionally bumped those up for the less active modes. I really need to do an adaptive algorithm that starts out vague and gets more accurate over time, but it does have a bit of an oscillation problem if you time the hits wrong, so you can't get too aggressive with the PID constants
|
# ? Aug 21, 2013 16:23 |
|
|
# ? Apr 26, 2024 09:49 |
|
Jonny 290 posted:nah i started with some generic constants and then tuned a _bit_ but not even much. then i just proportionally bumped those up for the less active modes. who knew weed could be so complicated?!
|
# ? Aug 21, 2013 16:24 |