Back
[01:01:09] <rue_house> ah!
[01:01:16] <rue_house> well I'm glad you asked
[01:01:38] <rue_house> the tiny26, is, by FAR the cheapest 10 bit, 11 channel adc there is
[01:01:59] <rue_house> but, you need to write some software to turn it into a generic adc
[01:02:04] <rue_house> but I dont want any generic adc
[01:02:16] <rue_house> I want a steaming, time sensitive adc
[01:02:26] <rue_house> cause I'm using it in a servo feedback loop
[01:02:48] <rue_house> so I mauled spi
[01:03:03] <rue_house> I have 3.5 lines,
[01:03:13] <rue_house> miso, sck, and sync
[01:03:22] <rue_house> the mosi is used to chain devices
[01:03:51] <rue_house> its operated by pulsing the sync pin, which makes all the devices write their values to the shift registers,
[01:04:00] <rue_house> then shifting out all of them at once
[01:04:18] <rue_house> you never have to send an address
[01:04:37] <rue_house> each device sends a 2 byte packet that has the address and the adc value
[01:05:00] <rue_house> and it appears to be working
[01:05:36] <rue_house> so now I have to write code for the mega8 that can take in the data and send it to a serial port for a PC
[01:06:15] <rue_house> if this works, I'll have a board that I can use modified servos with to make a force feedback servo set
[01:06:26] <rue_house> its 8 channel
[01:10:23] <rue_house> because, for testing, I'm not taking over the reset line, I will need the mega8 to do 6 adc channels on its own, which looks like it'll work
[01:11:24] <rue_house> matter a fact, I bet I could maul the isp on the mega8 to send the data to the pwm controller at the same time its loading the data from the adc's
[01:12:24] <rue_house> I'm outputting 16, 16bit values, and inputting 20 8 bit numbers
[01:12:51] <rue_house> no
[01:12:57] <rue_house> its worse than that jim
[01:13:07] <rue_house> I'm outputting 16, 12 bit values, and inputting 20, 8 bit numbers
[01:13:30] <orlock> typical IT response would be
[01:13:35] <orlock> "Did you log a ticket?"
[01:13:49] <orlock> doh wrong channel
[01:14:01] <rue_house> ticket number 2!, AFTER 6 YEARS WHY HASN'T IT BEEN ADDRESSED YET!
[01:14:19] <rue_house> IT MUST BE MARRING YOUR PERFORMANCE REPORT!!!!
[01:14:40] * rue_house was in IT too.
[01:15:48] <orlock> I should have the test-fit part of my robot done with by tomorrow
[01:15:49] <rue_house> this is good, the pins dont conflict, I have the isp, serial and adc all available on the mega8
[01:16:08] <orlock> then its printing the "strong" parts, and software
[01:16:42] <rue_house> ITS PRINTING THE SOFTWARE!! DAMNIT, why didn't I think of that!
[01:17:24] * rue_house chants to himself 'make the MACHINES do the work'
[01:18:28] <rue_house> I wonder if I'v ever coded for an atmega8 before
[01:19:16] <rue_house> ... a light blinker and a pwm -> adc converter... ugh
[01:22:38] <orlock> i had to increase the diameter of my gears by .7mm each so they would mesh
[01:22:45] <orlock> i just hope i didnt increase it too much
[01:33:36] <rue_house> hahah I just realized I can work out the shutterspeed of the camera by photographing a scope trace
[01:33:53] <rue_house> use the length of the line with the time/div
[01:33:54] <rue_house> hahah
[01:35:47] <z64555> er, what about offsets
[01:36:25] <z64555> oh, wait, nevermind
[01:37:00] <rue_house> looking over the photos of what I did this last week while on vaccation, all in all, not much, ARG
[01:37:46] <orlock> That's usually the point of a vacation, isnt it?
[01:38:13] <rue_house> most of my time went into putting th new roof on the greenhouse, 2 days
[01:38:27] <rue_house> I got some important code written tho
[01:38:31] <orlock> rue_house: WOuldent that depend on the phosphourous too? (sp)
[01:38:43] <orlock> re: taking pictures of CRO with camera
[01:38:56] <rue_house> no, the phosphor lasts longer than the ... it does
[01:39:20] <rue_house> I caught part of a wave, it was a relativly slow trace
[01:39:27] <rue_house> crisp ends
[01:39:43] <orlock> ahh
[01:39:48] <rue_house> the phospor lasts less than the trace time
[01:39:54] <orlock> i've done similar things, but it was because i didnt have a DSO at the time
[01:40:07] <rue_house> :)
[01:40:22] <rue_house> I got a slow one, mainly cause it was a kit and i wanted to have some fun assembling it
[01:40:33] <rue_house> 500Khz sample iirc
[01:40:36] <orlock> You should read "Spycatcher"
[01:40:50] <rue_house> no, reading stories is a waste of time for me
[01:40:51] <orlock> just for the interesting bits of tech they talk about
[01:41:06] <rue_house> I need to get a new platform build to load my mind onto before I die
[01:41:19] <orlock> Like wiring the output of microphones into a bank of CRO/CRT's
[01:41:34] <rue_house> mmhmmm
[01:41:36] <orlock> and then capturing the waveforms on video
[01:41:59] <orlock> The microphones were being used to monitor the noises/vibtrations from encryption wheels being set
[01:42:19] <orlock> enigma style crypto machines
[01:42:44] <rue_house> :) gotta wake at 6:15am tommoro, gnight
[01:45:02] <orlock> so that they could reconstruct the state of the wheels from the recorded audio fingerprints
[01:45:29] <orlock> 'night
[01:45:36] <orlock> .. 6:15am?
[01:45:43] <orlock> you mean a SLEEP IN?
[01:45:51] <orlock> ahh, kids
[01:59:57] <z64555> I actually think this would be a good experiment to try
[02:08:00] <z64555> hook up a battery to a uC and have it log voltage readings over time
[02:08:40] <z64555> as the battery discharges through two resistors
[02:09:07] <z64555> one connecteed across the pack terminals, and the other across a single cell
[02:10:26] <z64555> theory being that the rate of discharge of a cell is dependant on the maximum current going through it
[02:27:44] * z64555 fusses with kicad
[03:52:49] <z64555> anonnumberanon: 10 units of a SMPS for $8? not bad
[03:53:37] <z64555> this, actually makes me a bit suspicious because I see similar units that are $8 for just 1
[04:09:14] <z64555> oh, they use only one side of the board
[04:21:49] <Snert> anonnumberanon: I have seen people tap off at the balance tap and run a regulator with it.
[04:22:03] <Snert> just no huge boatloads of current is all.
[04:22:26] <Snert> like... a 6s lipo...tapped at the 2nd cell.
[04:22:45] <Snert> to get 5v regulated.
[04:24:20] <Snert> the balance tap wires can actually handle quite abit of current. That being said, it's all relative. 2-3Amps maybe more wouldn't be out of the question.
[04:24:47] <Snert> the actual balance tap wire gauge can handle alot more than that.
[04:25:49] <Snert> but one depletes one or 2 cells in the pack faster than the others. No big deal. Just gotta watch it is all.
[04:26:52] <Snert> Usually they just run the Rx and servos from it's own battery pack instead of tapping off the balance tap.
[04:28:08] <z64555> yeah, can get a longer flight time if you do that
[04:28:13] <z64555> presumably
[04:28:19] * z64555 hasn't flown anything yet
[04:28:22] * z64555 is ashamed
[04:28:59] <Snert> I gave up on 450 and 500 size helis.
[04:29:17] <z64555> why?
[04:29:21] <Snert> not enuff payback twixt crashes.
[04:29:30] <z64555> oh. lol
[04:29:31] <Snert> after months all I could do was barely hover.
[04:29:45] <z64555> didn't have a gyro?
[04:29:49] <Snert> so fuggit...rather have a quad where I can just fly.
[04:30:09] <Snert> yea it had a gyro and all. Still, it's alot of maintenance and crash downtime.
[04:30:28] <Snert> no real continuity of learning curve.
[04:30:42] <z64555> hm, must've been a crummy gyro
[04:30:51] <Snert> barely get it fixed again after a week wait for parts...then another crash...another week downtime.
[04:30:58] <z64555> Quad's by default must have a decent control mechanism
[04:31:30] <z64555> model heli's for the longest time got by without gyros
[04:31:37] <Snert> lol I have a $900.00 gyrobot that can do a hands off inverted hover.
[04:32:03] <Snert> but fuggit. lost the intrest.
[04:32:21] <Snert> too much crash downtime is all.
[04:32:41] <z64555> what kept breaking?
[04:33:04] <Snert> it's always everything. Different thing each time.
[04:34:01] <Snert> it was alot of fun though.
[04:34:54] <Snert> and then drones hit the scene where one can just fly for hours and not crash.
[04:35:12] <Snert> no torque rotation to compensate for.
[04:35:29] <Snert> well, relatively speaking, that is.
[04:37:17] <Snert> http://imgur.com/ZM9x7s5
[04:37:36] <Snert> http://imgur.com/PJHpo3t
[04:37:51] <z64555> yeah, the flight controller does a good job of making an otherwise unstable platform flyable
[04:38:35] <Snert> http://imgur.com/fWClKTZ
[04:44:08] <Snert> bust up one of these and it hurts.
[04:44:12] <Snert> http://www.rc-tek.com/store/i2rc-fbl500a-pro-500-class-high-performance-flybarless-rotor-head-p2214.html
[04:44:34] <Snert> I2 is the best there is.
[08:43:40] <z64555> well, I can put "subtractor" circuits on each cell
[08:44:20] <z64555> that would bring down the voltage of a cell to be in reference to common ground
[08:44:48] <z64555> I can't share resistors, though. which is going to be quite a few resistors
[08:54:14] <z64555> ooh!
[09:00:39] <z64555> hm, no. that won't work
[09:17:48] <rue_house> z64555, its called a zener diode
[14:12:58] <z64555> hm, can't seem to remember any zener diode circuits that would be of use. rue_house what did you have in mind?
[14:14:34] <z64555> the voltage shifter? but that would dissapate power through the resistor
[18:49:12] <anonnumberanon> connect a cell to a resistor to the adc of your uC to check the voltage
[18:49:26] <anonnumberanon> z64555,
[18:49:55] <orlock> yay
[18:50:01] <orlock> "high strength" gear printed
[18:50:10] <orlock> the issues i was having must have been the filament
[18:50:20] <orlock> this is from a different vendor
[19:29:01] <z64555> anonnumberanon: well, no, I cant just do that. the voltage will be referenced to ground, and if the resistor is a pull down (meaning it's connected to ground at one end, the cell being measured at the other) then current will pass through it
[19:29:49] <z64555> tweaking the resistor's value would then scale the seen voltage, instead of shifting it
[19:31:07] <z64555> the voltage shifter circuit, which is a zener diode and a resistor, is essentially a voltage divider circuit with one of the resistors replaced by the zener diode
[19:33:54] <anonnumberanon> z64555,
https://www.youtube.com/watch?v=2pHdO8m6T7c&t=3m11s
[19:38:32] * z64555 grumbles as he has to wait for a clearer format loads up
[19:42:09] <z64555> um. I need to see if my uC can handle the higher voltages on its ADC. I think the motorola HC9S12 could do that, but I forget its limits
[19:44:06] <z64555> that schematic isn't particularly helpful
[19:46:09] <z64555> omfg.
[19:46:20] <z64555> "If you don't have a soldering iron, what can you do?"
[19:46:27] <z64555> "may you can try glue for connection."
[19:49:30] <anonnumberanon> z64555, how is it not?
[19:51:26] <z64555> it simply says "battery input"
[19:51:40] <z64555> is that a cell, or is that indeed the entire battery?
[19:51:48] <z64555> how many cells are in the battery?
[19:52:20] <anonnumberanon> oh by the way that uC board in the middle is pretty lenient with input voltages
[19:52:33] <anonnumberanon> worse case senario should be 14.2 V
[19:52:41] <anonnumberanon> diode takes .7 off of it
[19:53:01] <anonnumberanon> that's the entire battery
[19:53:16] <z64555> ok. It pulls 5V off of the "reciever battery" as well
[19:53:35] <z64555> I'm not familiar with that chip
[19:53:50] <z64555> or board, for that matter
[19:54:17] <anonnumberanon> No it gives a regulated 5V to the receiver to power it.
[19:54:54] <z64555> mkay, so it has a regulator on the board
[19:54:59] <anonnumberanon> yes
[19:55:18] <z64555> that's nice
[19:55:56] <anonnumberanon> I'm not familiar with the way power is regulated on the Arduino board though.
[19:58:40] <anonnumberanon> this part is the voltage regulator, not sure if it's just a standard regulator or if it switches, i don't think it switches
[20:00:20] <anonnumberanon> Input Voltage (recommended) 7-12V
[20:01:01] <z64555> if it's a chip with three pins on it, it's most likely a standard regulator
[20:01:16] <anonnumberanon> max of like a 3cell lipo is 3 * 4.2 V = 12.6, diode removes .7 so max you'll ever see will be 11.9 wich is weirdly perfect
[20:01:17] <z64555> the SMPS chips have like, 5 or 6 pins on them
[20:01:28] <anonnumberanon> yeah
[20:01:59] <anonnumberanon> does your uC have a regulator?
[20:02:17] <z64555> think it does. I don't have the parts list for the board (yet)
[20:03:11] <anonnumberanon> so you have the housing but no chip to go on it yet?
[20:03:22] <anonnumberanon> trying to find the name of it again
[20:03:28] <anonnumberanon> i opened it 30 min ago fuck
[20:04:14] <anonnumberanon> okay cant find it pls link
[20:05:10] <z64555> no the chip's on there
[20:05:19] <anonnumberanon> ah cool
[20:05:21] <z64555> http://www.ti.com/tool/EK-LM4F120XL
[20:05:30] <anonnumberanon> btw this is a more useful diagram
[20:05:36] <anonnumberanon> https://www.arduino.cc/en/uploads/Main/Arduino_Uno_Rev3-schematic.pdf
[20:07:33] <z64555> gah, damn these companies obscure naming schemes
[20:08:57] <anonnumberanon> yeah it feels like it
[20:08:58] <anonnumberanon> http://www.mouser.fr/ProductDetail/Texas-Instruments/LP2985-33DBVR/?qs=paYhMW8qfiswyUdMjD%2FxlQ%3D%3D
[20:09:09] <anonnumberanon> im trying to find the power supply of yours
[20:09:18] <anonnumberanon> it maybe just takes usb 5V only
[20:09:24] <z64555> found it on the bill of materials, it's a 3.3V regulator
[20:09:52] <anonnumberanon> oh that' sright, so your entire board runs at 3,3?
[20:09:57] <z64555> uh. and there's a "voltage supervisor?"
[20:10:05] <z64555> TPS73633DRBT <- the regulator
[20:10:21] <anonnumberanon> do they provide schematics at all?
[20:11:11] <z64555> there's a schematic in the user manual
[20:11:20] <z64555> http://www.ti.com/lit/ug/spmu289c/spmu289c.pdf
[20:13:35] <z64555> mkay, so the voltage supervisor ensures the supply voltage doesn't dip below its set voltage
[20:13:54] <z64555> so if Vdd is 4.9 volts, the output of the supervisoer is 0
[20:14:20] <z64555> basically a comparator
[20:15:28] <anonnumberanon> im looking at the one on the arduino and i feel stupid but i thinking it has very high efficiency for a voltage regulator
[20:15:50] <anonnumberanon> kinda mind blowing and no wonder the board never ever gets hots
[20:16:30] <anonnumberanon> lm7805 is ridiculous compared to it (keeping in mind you should not draw more than.150 A)
[20:17:11] <z64555> which is not gonna happen unless you're using the chip as a motor driver
[20:18:17] <anonnumberanon> yours is better though if it can supply 0.4A
[20:19:50] <z64555> it's quite likely that your wall wort outputs close to the 5V to begin with. the closer the input voltage is to the output, the cooler it is
[20:19:59] <anonnumberanon> to be honest it's pretty good to have that power because then that allows you to power at least one servo directly with the board only, without an external power supply, for like trying stuff on the fly
[20:20:31] <anonnumberanon> yeah
[20:21:45] <anonnumberanon> so what does the supervisor chip do ?
[20:22:06] <z64555> it's just a comparator
[20:22:42] <Casper> Anyone here used 100% co2 for mig welding?
[20:22:59] <z64555> Vout = 0 if Vin < 5V; or Vout = Vin if Vin >= 5V
[20:23:37] <z64555> It prevents the circuits attached to it from going into an undefined state because of a too-low voltage supply
[20:23:59] <Tom_itx> Casper it's a cheap alternate
[20:24:33] <Casper> yes, that I know, but... from what I see, it is not better than fluxcore... why so many people push for C100?
[20:25:08] <Tom_itx> why not just try it
[20:25:21] <anonnumberanon> z64555, for some reason I see that as a flaw, for handling voltage spikes for example, guess capacitors would fix those
[20:25:40] <Casper> 180$ bottle, 45$ gas, 20$ adapter. That's why
[20:26:11] <z64555> anonnumberanon: that's my layman's understanding of it by just skimming over the chip's description
[20:26:14] <Tom_itx> not to try the flux core
[20:26:25] <Casper> fluxcore I used it
[20:26:33] <Tom_itx> and get a small bottle to try silly
[20:26:44] <Tom_itx> you can refund the bottle when you're done
[20:26:55] <anonnumberanon> z64555, it's just an "extra neat thing to have"
[20:27:04] <Casper> 180 is the smallest: 20lbs... or rent the bottle, 90$/year
[20:27:13] <Tom_itx> i use small bottles since i don't do that much anymore
[20:27:32] <Tom_itx> i've got 2 in case i empty one halfway thru
[20:29:25] <Casper> do you know of a good ressource that show the differences "unbiased" between C25, C100 and FC?
[20:29:52] <Casper> the main issue I see with FC is multiple pass, where you will get slag inclusion...
[20:31:06] <Tom_itx> https://www.reddit.com/r/Welding/comments/3ce8w4/welding_hot_rolled_steel_with_straight_co%C2%B2/
[20:31:10] <Tom_itx> haven't read it
[20:31:55] <Casper> already read it... and meh...
[20:32:39] <Casper> one issue with all what I find is that they do not quantify anything beside "more" and "less" not even with something like "alot"
[20:33:29] <Tom_itx> gotta run..
[20:33:30] <Casper> like "less cleanup" "less splatter" ... but nobody seems to be able or want to say by how much...
[20:33:34] <Casper> ok see ya
[20:36:47] <z64555> ew.
[20:37:03] <z64555> anonnumberanon: that voltage regulator has a real small input voltage window
[20:37:21] <z64555> but it will take in 5V
[20:37:26] <anonnumberanon> yeah 0 to 7?
[20:37:40] <anonnumberanon> use one of the BECs then
[20:38:20] <z64555> 1.7 V to 5.5 V
[20:38:32] <anonnumberanon> but you can still make a voltage divider with the battery to the uC's ADC so that you can always know the voltage
[20:40:13] <anonnumberanon> ah, corre ct
[20:40:16] <z64555> well the good thing is that I can ditch both of the regulators that I have on the quad now
[20:40:42] <z64555> pull 5V off of the BEC's, and then pull 3.3V off of the uC
[20:40:58] <anonnumberanon> yeah at 400mA that's way enoufh to power any sister boards like the imu and a few other things
[20:41:19] <anonnumberanon> question, what's the power rating of the bec on one of those ESCs?
[20:41:27] <z64555> heck if I know
[20:41:39] * z64555 scrounges for the parts list
[20:43:59] <anonnumberanon> On average ... It is very common for speed controls to have BECs rated at 2 or 3A.
[20:44:07] <anonnumberanon> ^I'm quoting
[20:45:01] <z64555> this one says 0.5A
[20:45:15] <z64555> for an 18A ESC
[20:46:00] <anonnumberanon> yeah that's perfect
[20:48:50] <z64555> blah, I need to get a better IMU, or at least one that has a compass and maybe also a baro on it
[20:50:04] <z64555> oh, I was looking for the current draw the chip itself pulls
[20:51:33] <anonnumberanon> what is your current imu?
[20:52:32] <z64555> some old sparkfun piece of junk
[20:52:37] <z64555> 6DOF "Razor"
[20:52:59] <z64555> lol, the gyros consume more current than the accelerometers
[20:54:51] <z64555> 6.8mA for the 2 axis gyro, 5mA for the 1 axis gyro (seriously?) and 350uA for the 3-axis accelerometer
[20:57:21] <anonnumberanon> man it's a bit hard to find code for it
[21:00:52] <anonnumberanon> I cutnpasted my imu's code but I'd really like to take the time to do it myself some day. Feels like an achievement I could be proud of and talk about lol.
[21:01:47] <z64555> for what, the gyro?
[21:01:51] <z64555> er, the IMU?
[21:03:00] <z64555> that's probably the easiest thing to code
[21:03:51] <anonnumberanon> for getting all the data out
[21:04:42] <anonnumberanon> (into the uC and in usable form)
[21:09:11] <z64555> yeah, that's the easiest part
[21:09:36] <z64555> the IMU uses analog values, so it's no wonder you couldn't find code for it because you have to use the ADC
[21:09:50] <anonnumberanon> WTF
[21:10:14] <z64555> did I mention it was a piece of junk?
[21:11:07] <anonnumberanon> i dunno it does look pretty good in pictures
[21:12:09] <z64555> I works well, but there's so many other boards out there that work better
[21:20:59] <anonnumberanon> last comment on this, hilarious and also so true
[21:21:01] <anonnumberanon> https://www.reddit.com/r/arduino/comments/3d433u/mpu6500_and_mpu6050_differences/
[21:22:22] <z64555> RTFM is a meme for a reason. :)
[21:23:38] * anonnumberanon estimates 70 pages for one of those
[21:23:42] * anonnumberanon checks..
[21:24:29] <anonnumberanon> meh 48
[21:24:32] <anonnumberanon> https://store.invensense.com/datasheets/invensense/MPU_6500_Rev1.0.pdf
[21:30:20] <anonnumberanon> if you get one of those ill get back into it
[21:30:31] <anonnumberanon> gnihjt
[21:32:19] <z64555> hm? you mean get a MPU6500?
[21:32:57] <z64555> oh right, that's an IMU
[21:44:05] <z64555> blah, this side cutters are dangerous. I keep playing with them near the power cords
[21:44:08] <z64555> *these
[23:03:18] <z64555> oh now this looks nice:
http://www.dfrobot.com/index.php?route=product/product&product_id=987#.V7vDKDUpwwI
[23:16:11] <z64555> hey rue_house, are you around?
[23:21:31] <rue_house> WHO YOU CALLING ROUND!!!!
[23:21:42] * rue_house pokes his belly
[23:21:50] <rue_house> I'm no rounder than I'm supposed to be
[23:22:08] <rue_house> THE SPI SYSTEM ON THE ATMEGA8 IS SO MUCH BETTER THAN ON THE TINY26!
[23:22:31] <Casper> those tiny has been butchered quite often
[23:23:33] <rue_house> I'm all excited, all I have to do is write to a register and wait for the transfer to finish
[23:23:37] <rue_house> *or not*
[23:52:52] <z64555> ah, there you are. I was wondering if regular servos will take a 3.3V signal
[23:54:06] <z64555> of course I could suppose I could write a short program to test it with my ESC
[23:54:32] <z64555> yeah, I should probably do that
[23:54:43] * z64555 ends up not doing it for another day