Back
[00:00:30] <orlock> And, for example - a dark frame - if it's taken at a different temperature to your actual exposures will cause artifacts
[00:01:21] <Jak_o_Shadows1> orlock, how would I predict, based on camera focal length and imager plane size, etc, how big/how many pixels across a star is?
[00:01:32] <orlock> http://somepicturesof.space/Antares.jpg - this one has some weird faint coloured streaks thrugh it
[00:01:45] <orlock> That's because the dark frames were taken inside at a different temperature
[00:02:37] <orlock> maths.. i forget the details though, but it's not terrribly hard
[00:03:17] <Jak_o_Shadows1> What's it based on, on the star side of things. HOw bright the star is? Or is that sort of thing just in a database somewhere?
[00:04:45] <z64555> oh, birghtness would make a difference on its captured size...
[00:04:51] <orlock> yeah
[00:04:55] <orlock> overflowing buckets
[00:05:05] * z64555 tries to smack down the dyslexia of tiredness
[00:05:51] <orlock> https://en.wikipedia.org/wiki/Angular_diameter
[00:06:27] <Jak_o_Shadows1> hmm. Thanks
[05:15:13] <Jak_o_Shadows> I think I am bad at putting the IDC connectors in
[05:27:32] <Snert> 1st day on the job I ripped up some pcb traces on some to220 power transistors.
[05:27:44] <Tom_L> awesome!
[05:28:03] <Snert> the engineer was pretty pissed lol.
[05:28:12] <Tom_L> pack your shit and get out...
[05:28:18] <Snert> told me to make traces out of buss wire.
[05:28:52] <Snert> so I did...and epoxied the wires onto the board. It worked out fine.
[05:29:01] <Tom_L> that or solderwick if you need heavy ones
[05:29:35] <Snert> Ya. Needless to say my soldering and my use of wick improved greatly after that day :)
[05:29:55] <Snert> I had never even seen wick yet, at the time.
[05:30:21] <Triffid_Hunter> wick is magic, just drinks the solder away when it gets hot enough
[05:30:30] <Tom_L> and they handed you a soldering iron the first day?
[05:30:40] <Triffid_Hunter> need a big tip though, like 4-6mm chisel for best results
[05:31:02] <Snert> no choice. I had to correct the production fuckups on the crystal ovens of 10 radios.
[05:31:14] <Tom_L> i sometimes add a bit of liquid flux to it as well
[05:31:47] <Tom_L> helps transfer the heat quicker
[05:32:17] <Jak_o_Shadows> One of the IDC connectors I put on is
[05:32:19] <Jak_o_Shadows> is fine*
[05:32:32] <Jak_o_Shadows> The other two (one which is mid-way in the wire) is buggered. Only 1-2 pins actully connect
[05:41:47] <Snert> got any porn on that?
[05:42:20] <Jak_o_Shadows> Photos? Nah, can't be assed.
[05:42:21] <Snert> shint solder joints really do it for me lol
[05:42:33] <Jak_o_Shadows> It's not working, so I'll have to redo it, and get it right htis time
[07:57:59] <orlock> i just had an idea
[07:58:07] <orlock> i think it probably already exists
[07:58:19] <orlock> but i'm unsure of how to put it into a concise google search
[12:43:03] <codepython777> has anyone used AS5047P ?
[12:51:42] <deshipu> no, they are manufacturing thousands of it for nothing
[12:58:45] <Tom_L> it's just to keep the chinese kids employed
[12:59:27] <codepython777> :)
[12:59:42] <codepython777> is there a better optical one to use?
[13:35:47] <delinquentme> how do I make my arduino sketches upload faster?
[13:37:58] <deshipu> write less code
[13:40:23] <codepython777> :)
[14:37:23] <Tom_L> codepython777, use a faster programmer
[21:23:43] <pokmo> do 3d printers have some sort of rotary encoders onboard?
[21:23:45] <pokmo> most of them use steppers, right
[21:23:53] <orlock> yes and maybe
[21:24:01] <orlock> sorry, yes they use steppers
[21:24:04] <orlock> and th encoder is a maybe
[21:24:08] <orlock> but none that i have seen
[21:24:33] <pokmo> orlock, then how do they control positioning?
[21:24:46] <orlock> Step by step
[21:24:49] <z64555> dead reckoning
[21:24:58] <orlock> If that fucks up, there's something wrong anyway
[21:25:09] <z64555> yup
[21:25:10] <pokmo> orlock, what if the motor somehow misses some steps
[21:25:20] <orlock> Then built your printer right
[21:25:26] <z64555> then something's wrong with your motor
[21:25:33] <orlock> "somehow" == "something fucked up"
[21:25:45] <pokmo> hmm
[21:26:01] <z64555> unlike normal motors, steppers are pretty darn accurate
[21:26:29] <pokmo> right
[21:26:35] <z64555> the only major issue you'd face with 3d printing is from something else
[21:26:40] <z64555> Like gear backlash
[21:26:57] <pokmo> yeah
[21:27:24] <z64555> solution: don't use gears
[21:27:28] <orlock> they dont use gears usually
[21:27:41] <orlock> it's all belts or screws
[21:28:11] <z64555> screws can have backlash, but the one I've seen gets around that by having the screw only do the Z direction
[21:28:38] <z64555> which has no backlash at all
[21:28:53] <pokmo> is it why servos aren't used?
[21:29:25] <orlock> z64555: Slop is still a problem - Depends if you are use leadscrews of fastner-screws
[21:29:26] <z64555> yes, because most servo's use gears, or a continous motor
[21:29:40] <orlock> pokmo: Servo is an imprecise definition
[21:30:08] <z64555> I'm assuming the type you mean are the same type used in RC stuff
[21:30:12] <orlock> and a servo could very well be a stepper plus an encoder, with a control board
[21:30:59] <pokmo> yeah
[21:31:05] <pokmo> rc servos
[21:31:33] <z64555> if you want to be technical, a stepper can be considered a servo
[21:31:37] <orlock> ~190 degree field of movement....
[21:31:56] <orlock> on most RC hobby servos - wont get you far
[21:32:33] <z64555> since it has a control circuit in it to insure the shaft rotates to the next pole (or half-pole) on a pulse
[21:32:56] <orlock> But there's no feedback if it doesnt
[21:33:28] <z64555> RC servo's don't have feedback, either, but they work continously
[21:33:53] <z64555> *feedback to the controller issuing the pulse or position signal
[21:33:54] <orlock> They have feedback built into their control circuits - steppers dont
[21:33:58] <orlock> yeah
[21:34:36] <orlock> rc servos will keep fighting to go to the specificed position, steppers will just miss and settle
[21:34:59] <z64555> RC servo's usually have a PID or similar analog controller to adjust the shaft to the desired angle
[21:35:09] <z64555> yes
[21:35:19] <rue_bed> not PID, not PI, not PD, just P
[21:36:11] <rue_bed> http://ruemohr.org/~ircjunk/tutorials/elex/hobbyservo/servo101.html
[21:36:32] <pokmo> how accurate are Nema 17s usually?
[21:36:43] <orlock> 1.8 or 0.9 degree per step
[21:37:08] <pokmo> that's pretty much like rc servos
[21:37:23] <orlock> nah
[21:37:34] <orlock> rc servos have way more positions available
[21:38:19] <pokmo> orlock, but you could do half steps with steppers, right? if not finer
[21:38:37] <orlock> yes but accuract/repeatability goes down
[21:38:38] <z64555> you can't do quater steps
[21:38:43] <orlock> even though you increase smoothness
[21:38:44] <z64555> *quarter
[21:39:08] <z64555> ok, there is definitely some sort of gunk under my 'r' key
[21:40:10] <pokmo> what's the accuracy of rc servos usually?
[21:40:21] <Tom_L> not very
[21:40:21] <pokmo> 0.1?
[21:40:46] <orlock> i didnt bother measuring, but they usually have a PWM range
[21:40:50] <rue_bed> 1/22000 of 300 degrees
[21:41:06] <rue_bed> for my librarys output pulses anyhow
[21:41:43] <orlock> pokmo: also,
https://www.ncsu.edu/labwrite/Experimental%20Design/accuracyprecision.htm
[21:42:39] <z64555> that would depend on what chip they use, how accurate it's registers it is
[21:43:03] <rue_bed> m328
[21:44:59] <orlock> rue: oh - openssh server running on windows, authenticating against AD
[21:45:06] <orlock> this is the one that Microsoft's been working on
[21:45:08] <orlock> it's great
[21:46:28] <z64555> atmega 328, 8-bit processor, seems reasonable
[21:46:44] <z64555> Position is probably incoded as a uint16_t
[21:47:59] <z64555> 0.005deg is likely the best accuracy you'll ever get
[21:48:08] <z64555> *maybe*
[21:49:36] <Snert> orlock: 'sbout damned time microsoft came up with a built in ssh server. Been waiting for that for years.
[21:49:38] <z64555> 0.010deg is more likely, but again at best
[21:53:40] <z64555> ty rue_bed
[22:13:48] <orlock> Snert: It's not quite builtin - It's hosted on github, but it "works" without needing anything else
[22:14:17] <Snert> even better :)
[22:15:08] <orlock> and it's an actual microsoft project
[22:16:34] <Snert> yea, all that seemed to change when ballmer stepped aside.
[22:17:00] <Snert> time for the "poison linux" world think to go.
[22:46:11] <rue_shop4> after all the forign anti-linux ads?
[22:53:04] <Snert> I just saw the tone of what ballmer was flapping his jaws about linux and the oss world.
[22:53:57] <Snert> it'll take eons for newthink to pervade microsoft, if it ever does.
[22:57:51] <rue_shop4> sure sure, thats why windows just happens to have an etc directory for its networking
[22:58:46] <Snert> I stay suspicious, so far. I need to see a big contribution, not just heisting oss code for their own purposes.
[22:59:21] <Snert> even if they do make those purposes available according to oss licensing.
[23:00:24] <Snert> are they just trying to look "with it" or are they really gonna help us out.
