|
i'm really mad about your left-aligned numbers but awesome job on everything else
|
#
?
Jul 23, 2017 21:11
|
|
|
|
| # ? May 18, 2024 04:25 |
|
|
median filters are cool as heck
|
|
#
?
Jul 24, 2017 01:02
|
|
|
oh and yes i am going to sync the grid movement with the wheel speed. already have that functionality written, just not active.
|
|
#
?
Jul 24, 2017 01:15
|
|
|
What's the best PCB design software for idiot hell fuckers like me
|
|
#
?
Jul 24, 2017 12:45
|
|
|
Malcolm XML posted:What's the best PCB design software for idiot hell fuckers like me geda/PCB isn't that bad if you don't mind open source, we used it in my robotics class for building the sensor system board
|
|
#
?
Jul 24, 2017 13:54
|
|
|
Bloody posted:you should be able to get muscle flexion pretty easily with any surface electrode pair, like if you stick one at the base of your thumb at the wrist and then up towards the top of that bigass thumb muscle you should expect to see lotsa bigass EMG signals in the 1s to 10s of mV range. i dont remember the frequency ranges off the top of my head but 50-3k sounds kinda high. i know you get tons of muscle artifacts in EEG, which is more like 1-50 Hz you're right, i was reading the wrong data i wrote down a year ago, amp i'm using is a TL072 and I've tried an LM318 as well most likely i'm just a doofus and had set it up for the wrong signal i might just get a free sample AD620 instead, a few things i've looked at recommended those but they're $25 per chip
|
|
#
?
Jul 24, 2017 14:00
|
|
|
Malcolm XML posted:What's the best PCB design software for idiot hell fuckers like me Altium
|
|
#
?
Jul 24, 2017 14:38
|
|
|
Malcolm XML posted:What's the best PCB design software for idiot hell fuckers like me kicad if ya poor, altium if ya not i'm poor and I use kicad - it's worked for me!
|
|
#
?
Jul 24, 2017 16:01
|
|
|
install diptrace
|
|
#
?
Jul 24, 2017 19:08
|
|
|
Apocadall posted:you're right, i was reading the wrong data i wrote down a year ago, amp i'm using is a TL072 and I've tried an LM318 as well AD620 is definitely the best of those. what kinda topology are you using and how much gain? are you just dumping that into an oscilloscope for now or do you have an ADC involved? i'm coming from a background where we either made our own parts or used things like http://intantech.com/products_RHA2000.html so im probably not gonna have cost-sensitive suggestions for parts lol
|
|
#
?
Jul 24, 2017 19:11
|
|
|
Bloody posted:AD620 is definitely the best of those. what kinda topology are you using and how much gain? are you just dumping that into an oscilloscope for now or do you have an ADC involved? i'm coming from a background where we either made our own parts or used things like http://intantech.com/products_RHA2000.html so im probably not gonna have cost-sensitive suggestions for parts lol no ADC right now, just working on getting a good signal before i try to convert it and do something with it, just dumping to the oscilloscope for now any suggestions on an ADC? i'm aiming to use muscle movements to control motors, but figure i'll start with just getting them to control some LEDs and go from there eventually will use a pi zero to handle that part i think the gain was somewhere between 100k and 1M, but i'll have to go back through and clean up any information i currently have and consolidate into a more coherent mess
|
|
#
?
Jul 24, 2017 19:21
|
|
|
best i have for ADC's is some MCP3004 and MCP3008
|
|
#
?
Jul 24, 2017 19:27
|
|
|
ADS1115 is a pretty ok cheap 16-bit 4 channel I2C ADC that you can buy as a little PCB on eBay, 16-bit with 4 channels and a PGA. for a really precise application you might want something like the ultra fancy 24+ bit delta-sigma converters that TI/Analog/Linear make for this sort of application. IIRC they also come with 50/60 Hz notch filters as a selectable option.
|
|
#
?
Jul 24, 2017 21:18
|
|
|
ratbert90 posted:Altium not gonna drop 7.5k on a dumbass home automation project
|
|
#
?
Jul 25, 2017 01:00
|
|
|
longview posted:ADS1115 is a pretty ok cheap 16-bit 4 channel I2C ADC that you can buy as a little PCB on eBay, 16-bit with 4 channels and a PGA. my recco in the  category is ltc2378 its real good and nice
|
|
#
?
Jul 25, 2017 04:29
|
|
|
longview posted:ADS1115 is a pretty ok cheap 16-bit 4 channel I2C ADC that you can buy as a little PCB on eBay, 16-bit with 4 channels and a PGA. this is probably too slow of an adc
|
|
#
?
Jul 25, 2017 04:30
|
|
|
Apocadall posted:best i have for ADC's is some MCP3004 and MCP3008 10 bits kinda blows rear end but for EMG it might work lol
|
|
#
?
Jul 25, 2017 04:31
|
|
|
Apocadall posted:no ADC right now, just working on getting a good signal before i try to convert it and do something with it, just dumping to the oscilloscope for now that is an alarming amount of gain, post teh circuit
|
|
#
?
Jul 25, 2017 04:31
|
|
|
better yet see if you can just get a sample of a TI ADS1298R 
|
|
#
?
Jul 25, 2017 04:53
|
|
|
isn't TI weird about samples? or have I been listening too much to all the distys they "hosed over"
|
|
#
?
Jul 25, 2017 06:15
|
|
|
Malcolm XML posted:not gonna drop 7.5k on a dumbass home automation project Altium circuit maker is Free Software. I think the only real restriction is that you have to use their  and cannot use local storage
|
|
#
?
Jul 25, 2017 07:27
|
|
|
eagle, kicad, and dipstrace are the big three for free pcb design tools. eagle is a weird program with a terrible UI, but has a lot of traction in the hobby scene so you can get help with it easily. kicad and dipstrace are distant seconds in terms of popularity, but are nicer to use.
|
|
#
?
Jul 25, 2017 13:11
|
|
|
Spatial posted:because of this an m3 can pull off a division in just a few cycles if the operands are small. however it still can't compete with a shift, which will likely be combined with another operation for free thanks to the whole flexible operand2 thing in the isa. Except the Cortex series mostly supports Thumb1 (with a couple of Thumb2 instructions where absolutely necessarily), where the integrated barrel shifter thing doesn't apply. Edit: Wait, I guess that's just Cortex M0/1, derp.
|
|
#
?
Jul 25, 2017 14:18
|
|
|
Bloody posted:that is an alarming amount of gain, post teh circuit i'm going to redo the circuit tonight, or maybe on my lunch break going to recheck all my numbers and see if i can't find where i made an error before
|
|
#
?
Jul 25, 2017 14:25
|
|
|
Sweevo posted:eagle, kicad, and dipstrace are the big three for free pcb design tools. eagle is a weird program with a terrible UI, but has a lot of traction in the hobby scene so you can get help with it easily. kicad and dipstrace are distant seconds in terms of popularity, but are nicer to use. i've tried all three and diptrace is the only one that feels like something written in the last ten years
|
|
#
?
Jul 25, 2017 20:45
|
|
|
Sagebrush posted:i've tried all three and diptrace is the only one that feels like something written in the last ten years but do either eagle or kicad feel like good software written in the 1990s-2000s
|
|
#
?
Jul 26, 2017 08:00
|
|
|
eschaton posted:but do either eagle or kicad feel like good software written in the 1990s-2000s they're pcb cad programs what does your heart tell you
|
|
#
?
Jul 26, 2017 08:16
|
|
|
They both feel like exactly what they are Eagle, a CAD program from like 1988 that hasn't had a single usability update since the initial release KiCAD, the same thing but also made by Linux users
|
|
#
?
Jul 26, 2017 08:27
|
|
|
Apocadall posted:i'm going to redo the circuit tonight, or maybe on my lunch break yeah remember that gain bandwidth product is a thing that exists and youre using real cheap opamps
|
|
#
?
Jul 26, 2017 09:30
|
|
|
"arduino" means one thing again and we're also going to burn down two's complement and go back to signed-magnitude because bitflips in values cause more power burn than literally anything else, apparently
|
|
#
?
Jul 28, 2017 21:48
|
|
|
well, i mean, obviously bitflips is the only thing that actually by necessity uses power. i assume we are not at a point where that really is the reality we are up against granted, but, hey, interesting to think about
|
|
#
?
Jul 28, 2017 22:48
|
|
|
Cybernetic Vermin posted:well, i mean, obviously bitflips is the only thing that actually by necessity uses power. i assume we are not at a point where that really is the reality we are up against granted, but, hey, interesting to think about I thought leakage current was the major power draw these days I wonder if nano relays will make it for tiny rear end sensors. Zero leakage! Anal leakage
|
|
#
?
Jul 28, 2017 23:27
|
|
|
right, but the actual physical hard boundary of computation possible to perform is the number of bitflips needed to realize the computation as noted i highly doubt it has a ton of practical use in present day, but it is sufficiently interesting and fundamental that it bears thinking about a bit
|
|
#
?
Jul 28, 2017 23:42
|
|
|
the ALU is only a small part of an MCU, switching the type of arithmetic used wouldn't have any practical effect i think. besides, it's not so much bits flipping as it is transistors switching and that happens everywhere logic exists in the MCU, not just the values that get placed in registers. so the picture the guy paints with values switching less often is probably not terribly informative. in all the low-power / energy harvesting stuff i've worked on the difference in power consumption between workloads was negligible, even for stuff like infinite nops vs generating random number sequences. from a software point of view we treated the dynamic power as a constant for a given frequency and voltage and the big deal was always the MCU active time. Malcolm XML posted:Anal leakage Spatial fucked around with this message at 23:52 on Jul 28, 2017 |
|
#
?
Jul 28, 2017 23:50
|
|
|
Cybernetic Vermin posted:right, but the actual physical hard boundary of computation possible to perform is the number of bitflips needed to realize the computation  i know someone at work who has studied this subject directly and published, maybe i'll ask him about it next week.
|
|
#
?
Jul 28, 2017 23:51
|
|
|
I mean the self discharge rate of your battery will swamp the idle consumption but it's fun to think about.
|
|
#
?
Jul 29, 2017 00:52
|
|
|
ya we always focused on minimizing operating time in 99.n% duty cycle applications
|
|
#
?
Jul 29, 2017 00:53
|
|
|
Malcolm XML posted:I mean the self discharge rate of your battery will swamp the idle consumption but it's fun to think about. yeah i work with iot devices and the occasional "10 years battery life!" you see advertised is always funny
|
|
#
?
Jul 29, 2017 00:55
|
|
|
Cybernetic Vermin posted:right, but the actual physical hard boundary of computation possible to perform is the number of bitflips needed to realize the computation No, flipping a bit is reversible. The actual limiting operation is erasing a bit. (This is just the theory, though. As far as I know, nobody has built reversible-logic circuits yet.)
|
|
#
?
Jul 29, 2017 14:44
|
|
|
|
| # ? May 18, 2024 04:25 |
|
|
Zopotantor posted:No, flipping a bit is reversible. The actual limiting operation is erasing a bit. ah, right you are. i should know this, as reversible automata is a recurring theme at conferences i go to (the interpretation is of course very simple in the finite state case) one is rather staring even deeper into the theory abyss at that point though, clearly it is rather correct that any operation which is required, reversible or not, will in most imaginable practical circumstances expend some energy. there has to, after all, be some b to be gotten to from a, even if they stand in simple relation to each other
|
|
#
?
Jul 29, 2017 14:54
|
|