#qi-hardware IRC log for Monday, 2014-06-02

whitequarkwhat the fuck.06:51
whitequarksilicone didn't work, and I have just learned why06:51
whitequarkbecause the water chamber is not actually welded shut, it's snap-on, like the motor chamber06:52
whitequark*of course* that shit leaks06:52
whitequarkwhee, 2nd issue of c't hacks12:49
whitequarkalso I've put a lot of the silicone over the latch of the water chamber and then closed it. it worked, it no longer leaks12:50
wpwrakmore silicone helps more. it's just a question of quantity until nothing could possibly leak anymore.12:52
whitequarkalternatively, encase it in a solid block of epoxy12:53
whitequarkno joke; that's how soviet military electronics was actually processed12:53
whitequarkthey even developed a special kind of epoxy which negligibly shrinks during curing, so that it wouldn't tear components off12:54
wpwrakmust be hell to change those tubes ...12:54
wpwrakor do they still use relays ?12:54
whitequarkmeh12:54
whitequarksoviet integrated circuits. biggest integrated circuits in the world12:55
whitequarkusually encased in ceramic & with a shitload of gold on the pins12:55
whitequarkyou could literally sell a few ICs and buy a nice car12:55
whitequarkAu, Pt, Pd, Ir, sometimes even more rare stuff12:55
whitequarklike this: http://files.radioscanner.ru/uploader/2009/119un2.jpg12:56
whitequark"one IC contains 19,4737mg of Au"12:56
whitequarkor this, for soviet PDP-type machine: http://patlah.ru/etm/etm-01/dom-promsl/uvelir/zoloto%20othod/zoloto-othod-08-04.gif12:58
whitequarkthat's in *grams*12:58
sb0they indicate the gold content in the datasheet?12:59
sb0or is this not a datasheet?12:59
whitequarkit's a datasheet12:59
whitequarka part of it12:59
sb0would be fun to visit the fabs :)13:00
whitequarkwell, you'd have to get a time machine first13:01
wpwrakDocScrutinizer05: you asked about raster's whereabouts. i guess this may explain it: www.heise.de/newsticker/meldung/Samsung-Z-Das-erste-Smartphone-mit-Tizen-statt-Android-ist-fertig-2213903.html13:24
wpwrakand this https://www.tizen.org/events/presentations/tizen-native-display-layer-efl-architecture-and-usage13:24
DocScrutinizer05eugh13:24
wpwrak(in german, "tizen" immediately draws your mind towards "tits". there are two words, "titten" and "zitzen", both meaning that)13:25
DocScrutinizer05hehe13:26
DocScrutinizer05I always grab my ear when I read "Tizen"13:26
DocScrutinizer05also not exactly a lucky association13:27
DocScrutinizer05http://www.google.de/search?hl=en-DE&source=hp&q=mike+tyson+ear+bit13:28
DocScrutinizer05http://www.youtube.com/watch?v=Wkx3NCv2XKg13:28
wpwrakouch !13:32
sb0whitequark, I suspect some are still around :)13:32
whitequarkbut I can say for a fact that *no* precious metals have survived for more than like13:33
whitequarkten minutes after the fall of ussr13:33
sb0I know a fab near Warsaw that still uses a lot of USSR-era equipment13:33
sb0including large oil diffusion pumps etc.13:33
sb0maybe mercury, even :)13:33
whitequarkmmm, pumps13:33
whitequarkmercury isn't really precious13:34
sb0(it's been on my todo-list for a while, but I'm never in Poland at the right time)13:34
sb0and I don't mean for the precious metals, just for the vacuum/electronic porn13:36
whitequarkwhat's a good way to make a constant-current regulator for 1A?15:29
whitequarklm317?15:29
whitequarkhm, no, that'd dissipate at least 3W15:33
ysionneauSTCS1A ?15:45
whitequarkysionneau: oh great17:24
whitequarkysionneau: oh actually not, I can't find anywhere to buy it in RU17:37
whitequarknot even "not in stock"17:38
whitequarkthough nevermind, I just realized I have digikey covered17:46
whitequarkand DK has it in stock17:46
sb0what do you need this regulator for?17:48
sb0just curious17:49
whitequarksb0: I have 24V input and I want to power a water (coolant) pump for my CNC mill17:49
whitequarkat about 1A it works best17:50
sb0constant voltage isn't good?17:50
whitequarkI dunno actually17:51
whitequarkthey're probably more or less equivalent17:52
whitequarkin this case17:52
sb0and what happens at max voltage?17:52
whitequarkyou mean at 24V?17:53
sb0the problem with driving a DC motor with constant current is that if you reduce the load (torque), the voltage and speed will go up17:53
sb0yes17:53
whitequarkoh17:53
whitequarkoh you're right.17:53
sb0if you run out of water, won't it make the motor spin too fast?17:53
whitequarkI need constant voltage here17:53
whitequarkit will, and it'll burn17:53
wpwrakthat should be easy to fix. even an arduino could do that ;-)17:54
sb0the problem with pure constant voltage is the current spike at startup, when the motor isn't already spinning17:56
sb0but in some cases, the circuit's resistance is enough to keep that current within a reasonable value17:57
sb0of course, if you block the motor then, the current will stay high17:57
wpwrakso have a controller. monitor voltage. on start, allow for inrush current. then limit voltage (probably needs a low-pass filter). if it gets too high, uncommanded, emergency shutdown.17:57
wpwrakand yes, maybe also if the voltage gets too low :)17:58
sb0motors are inductive too, which helps limit the inrush current17:59
sb0so in a lot of cases, constant-voltage is the best simple solution :)18:00
sb0you can also add a slow-acting fuse ...18:00
wpwraki guess you want to control both - current and voltage. regulate one, monitor the other. then you can detect any problems and just cut off. at least that's the theory :)18:02
whitequarkso contrived18:03
wpwrakin russia, it's normal for motors to burn18:03
DocScrutinizer05use a constant voltage with current limiting, and possibly even softstart. All available in very common cheap and simple regulator chips21:08
DocScrutinizer05usually with 3 pins21:08
DocScrutinizer05maybe for this usecase you prefer a 5pin version, that allows separate control for current limit and softstart21:09
DocScrutinizer05the 3 pin versions do no softstart and current limit is usually "hardcoded" and foldback characteristics which you don't really want21:10
DocScrutinizer05but actually you might get away with very convenient PWM on 50%, which doesn't even need any heatsink21:12
DocScrutinizer05since the motor is inductive, you only need a oscillator driving a power mosfet, and a clamp diode21:13
DocScrutinizer05http://www.ti.com/lsds/ti/power-management/pwm-resonant-controller-products.page might give you some ideas21:21
DocScrutinizer05http://www.ti.com/general/docs/datasheetdiagram.tsp?genericPartNumber=UCC28722&diagramId=SLUSBL7A21:25
DocScrutinizer05of course you also could use a NE555 and a FET ;-) (sounds familiar, eh?)21:30
--- Tue Jun 3 201400:00

Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!