#robotics | Logs for 2016-11-30

Back
[00:09:14] <mbrumlow> Oh man that cheta looks awesome.
[01:19:16] <rue_house> I been reverse engineering the LM567, and I'm quite amused
[01:20:36] <rue_house> both phase detectors are XOR gates, but they run off oscillator schmitt triggers of different voltages
[01:20:49] <rue_house> which is apparently what makes one of them 'quadrature'
[01:22:06] <Jak_o_Shadows1> I think that I really should make a breakout board for this.
[01:22:25] <Jak_o_Shadows1> Haven't figured anything out enough to do so yet
[01:22:51] <rue_house> ... I bought a meter of silicone tubing, I dont know why
[01:23:01] <rue_house> I just got it, it was ordered oct 1
[01:23:12] <rue_house> I need to keep a journal of why I buy things
[01:23:35] <rue_house> too bad aliexpress dosn't let me put notes on purchases
[01:24:20] <rue_house> maybe its for airmuscles
[01:24:37] <rue_house> yea, I think it is
[01:24:53] <rue_house> I wanted to find an alternate to baloons that disintegrate
[01:25:13] <Jak_o_Shadows1> TBH, what else was tbing gonna be for
[01:25:21] <rue_house> this tubing is a lot of things, but not soft
[01:26:12] <rue_house> maybe I can make a systallic pump for a supper machine with it
[01:26:47] <rue_house> :)
[01:27:05] <rue_house> glad I didn't pull a rue and order 300 feet.
[01:28:00] <Jak_o_Shadows1> I wish I'd bought more female dupont crimp connectors :(
[01:28:01] <Jak_o_Shadows1> I've run out
[01:28:23] <rue_house> I got a roll of like 10000 for like, $13
[01:28:40] <rue_house> part of my "lifetime supply" program
[01:28:57] <rue_house> which is why I have like 300 74HC00 chips...
[01:29:05] <rue_house> and 74HC02
[01:29:12] <rue_house> or are they LS, anyhow
[01:29:26] <Jak_o_Shadows1> It's a bad tim e of year to be buying stuff though :(
[01:29:31] <rue_house> yea
[01:29:37] <Jak_o_Shadows1> I bought some from an australian reseller because I want them soon
[01:29:43] <rue_house> :)
[01:29:50] <Jak_o_Shadows1> (then I also bought a bunch much cheaper from hong kong and china)
[01:29:56] <rue_house> the shipping from oz is pretty good
[01:30:08] <Jak_o_Shadows1> I bought 2 lots of 200, from separate places, in the hope that one will arrive sooner
[01:30:12] <Jak_o_Shadows1> Within australia is good yeah
[01:30:21] <rue_house> :/ I'm catching myself doing that too
[01:30:34] <rue_house> but you have to change the quatites to know whihc seller it came from
[01:30:53] <rue_house> so, from china, I might buy 10 from one seller and 20 from the other
[01:30:58] <Jak_o_Shadows1> That's a good point
[01:31:01] <rue_house> if 15 arrive, I'm screwed
[01:31:05] <Jak_o_Shadows1> This is China and Hong knog though, so I'm fine.
[01:31:16] <rue_house> ...
[01:31:38] <rue_house> I'v not found any relation between the seller and the package origin
[01:31:45] <Jak_o_Shadows1> ahaha
[01:31:51] <Jak_o_Shadows1> I suppose I havent' been paying attention
[01:32:01] <rue_house> I think almost all the sellers are vendors for direct-ship warehouses
[01:32:19] <rue_house> which is why the images are all the same
[01:53:52] <rue_house> https://www.aliexpress.com/item/1-75-3mm-3D-printer-accessories-Kraken-E3D-4-multi-head-4-color-water-cooled-hot/32670721845.html
[01:53:54] <rue_house> 8-|
[01:53:56] <rue_house> its...
[01:54:04] <rue_house> a quad head with water cooling....
[01:54:07] <rue_house> wow
[01:56:00] <rue_house> why didn't I think to make a water cooled head?
[01:56:17] <rue_house> 8-]
[01:56:25] <rue_house> with the cnc machine, I can make a water cooler...
[02:06:30] <Jak_o_Shadows1> Is the temperature actually limiting you at the moment?
[02:06:56] <rue_house> no, but the fan hanging off the head is causing my grief
[02:07:05] <rue_house> esp if I'm gonna do ABS
[02:09:05] <rue_house> $10 for an 80mm ratiator,
[02:09:16] <rue_house> OR I go to the scrapper, get a heater core for ~$20?
[02:09:30] <rue_house> OR use one of the heater cores in the basement for $0 ...
[02:09:54] * rue_house cant get over how cute an 80mm radiator is
[02:28:46] <Jak_o_Shadows1> aha
[10:56:08] * z64555 looks at the water cooler head
[10:56:58] <z64555> huh, for the coolant pipe, they drilled two holes for the inlet and outlet, and then drilled another hole through the side to connect them, and just capped it off to use as a cleanout
[11:03:47] <SorcererX> I wish digital servos would make less noise
[11:06:52] <branjb> anyone familiar with the stm32 line?
[11:11:38] <anniepoo> .
[11:11:53] <anniepoo> I've fiddled with it a bit
[11:12:00] <anniepoo> but am far from an expert
[11:12:09] <branjb> you used pwm with them?
[11:13:38] <anniepoo> no
[11:15:34] <branjb> their libraries have come a long way but is not documented wonderfully
[11:16:55] <z64555> the chips themselves, right? You could look at their datasheet and see what registers control the PWM generation
[11:17:35] <branjb> as far as I can tell for stm32's you have to set up timers manually to generate pwms
[11:18:10] <z64555> that's if they don't have a pwm modeul, which generally sucks
[11:18:16] <branjb> i'm not 100% sure but they have a bunch of DACs on the chip i am using, i think that's supposed to take the place of PWMs
[11:18:19] <branjb> yeah no pwm module
[11:19:00] <branjb> honestly the biggest downside in using stm chips
[11:19:36] <z64555> ah, it has two PWM timers
[11:19:43] <branjb> yeah
[11:20:27] <branjb> never set pwm's up manually with timers, only ever did it via registers like on avr's. so this is new
[11:21:23] <z64555> its the same principle, just different sequence
[11:21:42] <z64555> I'd suggest making an outline of the steps to make on paper, so you don't forget them
[11:22:12] <z64555> "Advanced-control
[11:22:12] <z64555> timers (TIM1, TIM8) The advanced-control timers (TIM1, TIM8) can be seen as three-phase PWM generators multiplexed on 6 channels."
[11:22:26] <z64555> o rly
[11:23:14] <branjb> from what i understand, one timer can control 1 pwm on multiple GPIOs right? so if I wanted 3 GPIOs one with a 25% duty cycle, one with a 50% duty cycle, and one with a 75% duty cycle, I can't do that with only 2 PWM timers right?
[11:23:37] <branjb> but I could probably do some funky stuff with other timers and prescalers to synth new duty cycles
[11:23:40] <z64555> If that means what I think it does, it has the capacity to run a pair of BLDC's given the proper switches
[11:24:11] <z64555> branjb: that's what I'm looking for, as well
[11:24:28] <branjb> I'm only controlling two motors so I probably won't mess with it too much right now
[11:27:35] <branjb> but the general concept is, to control duty cycle I control how many pulses per period, and then the duty cycle is a function of picking which pulse transitions from high to low right?
[11:27:52] <branjb> ie, Period = 1200, Pulse = 600, 50% DC?
[11:30:55] <z64555> DC = pulse / period. yes
[11:34:46] <z64555> branjb: http://www.st.com/content/ccc/resource/technical/document/reference_manual/59/b9/ba/7f/11/af/43/d5/CD00171190.pdf/files/CD00171190.pdf/jcr:content/translations/en.CD00171190.pdf
[11:35:40] <z64555> Looks like they have independant channels per timer
[11:36:19] <branjb> yeah looking at that now, and you can scale each channel individually
[11:36:36] <z64555> so they'll have the same period across the channels, and you can define a seperate pulse per channel
[11:37:25] <branjb> moving from a mega2560 to stm32 is daunting, so many more features and options
[11:39:47] <branjb> good learning experience though
[11:39:49] <z64555> it doesn't help that they're made by different companies :)
[11:40:26] <branjb> i've dedicated december to "learn some new microcontrollers" month
[11:40:59] <branjb> going to port my robot to the stm32, also got a few cypress psocs to experiment with
[11:45:28] <deshipu> stm32 is pain
[11:46:16] <deshipu> I'm actually learning AVR now, and I can't understand why people even use things like Arduino -- the bare AVR stuff is so simple!
[11:46:42] <veverak> :D
[11:47:21] <branjb> arduino is great for hobbiests, a bane for industry; kids come out of CpE programs now not knowing how to use or configure any kind of microcontroller outside of arduino + importing other peoples libraries
[11:48:17] <veverak> it's calld "unexperienced programmer"
[11:48:19] <veverak> happens evrywhere
[11:48:37] <branjb> haha
[11:49:54] <deshipu> arduino is nice if you just want to get stuff done fast
[11:50:09] <deshipu> and don't much care what it actually runs on
[11:50:14] <branjb> stm32cubemx pin configuration is crap
[11:50:31] <deshipu> branjb: doing it manually is not pleasant either
[11:50:37] <branjb> yeah
[11:50:41] <deshipu> branjb: especially with the bugs in silicone
[11:50:52] <deshipu> silicon
[11:51:05] <deshipu> in the chips
[11:51:06] <branjb> with the pin configurator if you try to enable stuff from the left, it rearranges your already configured pins
[11:51:07] <branjb> annoying
[11:52:18] <branjb> ah, just got to be smarter than the software. you can lock pins in
[11:58:10] <deshipu> no idea, I don't even have Windows to run this crap
[16:53:19] <mbrumlow> Not sure if this is the right channel for this. But I got a small toy RC car that I want to control with a micro controller. From what I gather I need to get a SDR to analyze the signal is that correct? If can anybody point me in the right direction for a good setup with Linux?
[16:55:11] <z64555> software defined radio?
[16:55:49] <z64555> I mean, I guess you could use that as an oscilloscope/spectrometer
[16:56:42] <z64555> maybe try looking on some amateur radio (HAM) websites to see if they did anything like that
[16:58:50] <mbrumlow> Well if there is a better way I am all ears. But you think I can do it with a oscilloscope?
[16:59:39] <z64555> What exactly are you trying to measure?
[17:00:48] <mbrumlow> I thought I saw somebody do this before. But, I want to see what the signal looks like so I can reproduce it.
[17:01:11] <mbrumlow> I did not think they were using a oscilloscope though, I thought they were using a SDR.
[17:01:48] <z64555> the SDR will take care of the modulation from the radio
[17:02:07] <z64555> so if you want to see what data is being sent, the SDR is what you'd use
[17:02:22] <mbrumlow> Ok good, I was not too far off.
[17:02:43] <z64555> somewhat cheaper to do is to hook up a reciever to an oscilloscope
[17:02:51] <z64555> and see what it's putting out
[17:03:18] <z64555> the RC recievers I know just put out PWM's to the various servos
[17:04:11] <z64555> They have at least one servo connected to a PWM channel, and sometimes they'll slave multiple servos to the same channel
[17:04:23] <mbrumlow> I want to see the data, then I can buy the hardware / the right transceiver that I can use with a rpi3, or arduino.
[17:07:51] <z64555> I don't know if they sell the radios seperate from the controller
[17:09:05] <mbrumlow> I am thinking something like a RF24L01
[17:15:19] <z64555> won't work
[17:15:36] <z64555> It has to have the same freq hop algorithm as the reciever you're trying to talk to
[17:16:45] <mbrumlow> I doubt the thing I am working with has a freq hop, Its a $10 toy that has "27 MHz" pasted to the side of it.
[17:17:25] <mbrumlow> It also has a sticker that says "FULL FUNCTION"
[17:17:32] <mbrumlow> but there is no on or off swtich on the remote :p
[17:17:56] <z64555> ah, you might have a chance then, that's an FM radio
[17:19:10] <mbrumlow> So, I need a SDR to figure out what the protocol looks like. Then maybe I can use the RF24L01 or something like that to reproduce the signals.
[17:19:13] <mbrumlow> ?
[17:19:24] <mbrumlow> Any recomendations on a SDR?
[17:19:47] <z64555> sadly no, I haven't gotten into that stuff
[17:20:11] <z64555> (the SDR stuff, that is)
[17:21:54] <mbrumlow> Well thanks, you have been a help :)
[19:19:35] <MrCurious> Spee Devil!
[19:33:12] <Snert_> sounds like a german ww2 uboat.