#qi-hardware IRC log for Saturday, 2016-03-12

DocScrutinizer05watch out for capacitor ESR02:02
wpwraki guess they rarely indicate that on the package. so he couldn't check that anyway. no need to worry :)02:26
DocScrutinizer05they sort of indicate it on the package, and you also can conclude about needed low ESR from position in circuit02:45
DocScrutinizer05replacing a capacitor that needs low ESR by a standard type usually results in a great working fix - for max 1 week02:46
DocScrutinizer05depending on circuit it also can go poof after just 10 seconds02:47
DocScrutinizer05and funny enough it's usually exactly those capacitors that need low ESR that pop in any appliance like TV, or computer PSU or whatever02:49
DocScrutinizer05http://www.low-esr.com/  http://www.illinoiscapacitor.com/pdf/Papers/low_ESR_fact_or_fiction.pdf etc02:53
whitequarkceramic capacitors are such a lie02:54
whitequarknot only massive variation of capacitance with temperature, but also voltage02:55
DocScrutinizer05very interesting: http://www.satcure.co.uk/accs/low_ESR_capacitors.htm#heating02:55
whitequarklol suggesting audiophile caps02:57
DocScrutinizer05there was an industry wide issue with capacitors a few years ago02:58
DocScrutinizer05billions of capacitors were faulty and had a way too short lifespan03:02
DocScrutinizer05whitequark: audiophile capacitors are one of the very few things that are not scam03:03
DocScrutinizer05IOW you actually can prove the effect with a meter03:03
DocScrutinizer05or scope or whatever03:04
DocScrutinizer05it's not like "audiophile caps" are the only brand/type of capacitor that yields decent audio results, but there are many non-audiophile caps that actually spoil it03:06
DocScrutinizer05whitequark: ((variation of capacitance with voltage)) yes, a known effect, and particularly nasty since it turns the capacitor into a non-linear component that introduces distortions03:23
DocScrutinizer05ideally when you need capacitors and there is no alternative to ceramic, you should use a capacitance *and* voltage way larger than what a conservative design would suggest03:24
DocScrutinizer05at least for signal paths03:24
DocScrutinizer05analog signal paths03:25
DocScrutinizer05whitequark: the capacitance can half when C is charged to nominal voltage. You know what happens to voltage of a (e.g. air+plates) capacitor when you decrease the capacitance (increase plate distance). When the capaictor is charged to Q at V/2, and the voltage rises to V, the capacity halves and Q stays constant03:55
wpwrakwhitequark: just use NP0 and you'll be fine ;-)07:26
eintopfi didn't check the esr07:26
wpwrakeintopf: it's a parameter you probably can't control without more preparation anyway, even if you can find out the design value (which is unlikely, given that the actual component you find may have an ESR that differs from the design)07:38
wpwrakoh, and "better" is almost always "lower", but there are exceptions07:38
eintopfi bought the parts twice07:43
eintopfso i am prepared for a next fix07:43
eintopfI have the feeling that the background light is better than before09:09
eintopfmaybe this was broken a long time ago09:09
DocScrutinizer05eintopf: when you actually need to do the fix again, I'd for sure not use the same capacitors, even when you got them twice. Rather get low-ESR types, they cost twice as much but possibly last 1000 times as long09:34
DocScrutinizer05 http://www.satcure.co.uk/accs/low_ESR_capacitors.htm has rather simple explanations about the whole problem, and their advice (though targeted at selling their stuff) is not completely bad09:40
eintopfwhere is the ESR value?09:43
DocScrutinizer05they seem to not state ESR, only max allowable ripple current which creates heat inside the component according to P=U^2*R  (where R == ESR)09:58
DocScrutinizer05err sorry09:58
DocScrutinizer05they seem to not state ESR, only max allowable ripple current which creates heat inside the component according to P=I^2*R  (where R == ESR)09:58
DocScrutinizer05and I=Ripplecurrent09:59
DocScrutinizer05watch out, they state ripple current as RMS and @ 120Hz10:00
DocScrutinizer05those are usually no low-ESR type, far from it10:01
DocScrutinizer05what's closest to ESR rating is their "Dissipation Factor" I guess10:02
DocScrutinizer05a switched power supply will charge and discharge the capacitor (over the "virtual resistor" ESR) a 500 to 20000 times more often per second than what they are rating their "max ripple current" for (120Hz)10:07
eintopfah, you mean I should buy parts which are marked as "low-ESR" only?10:07
DocScrutinizer05yes, they should be explicitly marked "Low ESR (x.xx Ohm)"10:08
DocScrutinizer05at least in datasheet10:08
eintopfbut now it works somehow10:09
DocScrutinizer05sure, it works for a few seconds to a few weeks always10:09
eintopfyou believe it will break in ~ few weeks?10:10
eintopfbut there is also a chance it will work ~years10:10
DocScrutinizer05first approach you could check if the new capacitors get hot - beware, remove mains voltage before you touch them, and keep your left hand in pocket!10:10
DocScrutinizer05actually remove all cables10:11
eintopfthen I need to disassemble everything again :(10:11
eintopfyes, sure10:11
DocScrutinizer05well, you can just wait and see how long it works until next failure, maybe you're lucky10:11
eintopfI will be back when it's broken again :-)10:12
DocScrutinizer05I fixed a PC PSU of a client when I wasn't aware of ESR yet. It worked great... for 2 weeks10:12
eintopfbut why they don't selling low-ESR only?10:12
DocScrutinizer05low ESR are way more expensive, 2 to 5 times10:13
eintopfah, ok.10:13
eintopfthis stupid cost-factor makes everything more breakable10:13
DocScrutinizer05nowadays almost everything appliance is not working longer than 3 years, due to poor component quality10:14
DocScrutinizer05manufs are happy with it10:14
DocScrutinizer05cheaper products, sold more often10:15
eintopfbut then you could buy broken ones at ebay10:15
eintopfrepair it10:15
eintopfthen sell it again10:15
DocScrutinizer05that's what a lot of repair shops do, commercially10:16
eintopfbut this makes you not lucky and rich10:16
eintopfwin-win situation for repair shops and manufactors10:18
DocScrutinizer05I once bought a refurbished washing machine. Of the two heating rods one was broken and they "fixed" it by simply cutting the wires for power supply to that rod.10:18
eintopfah, ok.10:18
eintopfI know some heating issues at coffee machines10:18
eintopfthey have two heating pipes10:18
DocScrutinizer05I literally can hear the service dummy: "this is beyond economic repair, we deliver the new machine and fetch the old one for free"10:19
whitequarkreminds me I want to make a tea machine that's like...10:19
eintopfI had two of them and repaired one10:19
whitequarkproduces tea instantly  at a drinkable temperature10:19
whitequarktank + heater + leaf cartridge + adjustable mixer + PID10:20
whitequarkcan't be too hard10:20
whitequarknice exercise in mechanical design too10:20
eintopfsounds like a good kickstarter project10:21
eintopfI want such teamachine10:21
eintopf"produces tea instantly  at a drinkable temperature"10:21
whitequarki'm not going to spend my time and mental health on that treadmill10:22
whitequarkin fact, if this goes viral on hackaday and some chinese company makes a clone, i would be very happy10:22
eintopfthings on hackaday will be cloned from chinese companies?10:22
whitequarkI mean, hypothetically10:23
whitequarkif I made a kickstarter, it will obviously get cloned at some point if it's successful10:23
whitequarksince i'm not going to get more out of a kickstarter campaign than i put in, then how about cutting out the middle step?10:23
whitequarkin qty 1 this thing is a breeze10:24
eintopfyou need to announce it on kickstarter when the product is already done, but this is not how kickstarter works10:24
whitequarkin qty 1000 already it's a complete nightmare10:24
whitequarkno, I'm not saying this is what will happen. I'm saying this is what I wish would happen because I sure as hell am not going to manufacture that10:25
whitequarkmost likely nothing happens at all and i'm ok with that too.10:25
whitequarkout of stock...10:35
whitequark"quite prone to defects" too10:36
DocScrutinizer05this is a really great design. Only one "downside": water is always boiling hot when it gets on the tea leafs10:37
whitequarkright, so it needs to have *two* mixing taps10:37
whitequarkI also wonder if you could make it automatically swap the leaves10:38
DocScrutinizer05it needs another electronic valve on top tank which heats the water, it now only has a inverted U pipe that makes water go down to tea when it boils10:38
DocScrutinizer05(auto swap leafs) hardly10:39
whitequarkwhy not?10:39
whitequarkyou could make a piston with two O-ring seals that can take three positions in a horizontally placed cylinder10:39
whitequarkleft: open from top, can add leaves10:39
whitequarkcenter: connected to hot water inlet and tea outlet10:39
whitequarkright: open from bottom, ditches content10:39
whitequarkand a servo or something to move it10:40
DocScrutinizer05hmm, maybe it's actually simpler to just pull out the middle container, turn it top down and spray some small amount of water in to remove the tea leafs sticking to walls, the turn top up and fill with new tea leafs by whatever means you like (you don't want to drink same tea all week long, do you?), then shove the thing in again10:47
whitequarkyou mean manually?10:47
whitequarkthat's exactly what i suggest to do automatically10:47
DocScrutinizer05no, you can do that by any servo10:48
DocScrutinizer05actually iirc you *need* to remove the middle container to disengage the electromagnetic valve at bottom10:50
DocScrutinizer05aaah, for the temperature of water: you can add a pressurizer air hose with valve (or a small air pump) to press the water out of top container. Instead of using steam pressure of boiling water10:52
whitequarkor just a pump10:53
whitequarklike peristaltic10:53
whitequarkor even piston pump10:53
whitequarkhow so?10:53
DocScrutinizer05boiling water, moving parts10:53
whitequarkthe pump acts on cold water exclusively10:53
whitequarki want to use an inline heater10:54
DocScrutinizer05ooh, well10:54
DocScrutinizer05that's not the best idea for tasty tea10:54
whitequarkwhy not?10:54
DocScrutinizer05tea should get boiling water purred over it, in short time10:55
whitequark4kw inline heater10:55
DocScrutinizer05well, that needs a special outlet10:55
whitequarkdoes it?10:55
DocScrutinizer05here yes10:55
DocScrutinizer053kW is abs max10:55
whitequarkahhh right10:56
whitequarkI meant 3kw, misremembered what the limit is10:56
whitequarkstill should be enough10:56
whitequarkit's not enough if you wanna take a shower, but to make some tea? sure10:56
DocScrutinizer053kW will take a minute or more to boil 2L of water 10:57
whitequarkwho said about 2L?10:57
whitequarkwhat I want is a device with a button. press a button, get a teacup. instantly.10:58
whitequarki have the attention span of a ferret on meth10:58
DocScrutinizer05ouch, one cup10:58
whitequarklike... you know why i want this device, right?10:58
whitequarkbecause if i have to leave tea to extract then I forget about it entirely10:58
whitequarksame about water in teapot10:59
whitequarkactually even better idea, direct line to mouth. take a straw, pull air from it, get some nice warm tea10:59
whitequarksince we're already doing *everything* inline that could just work10:59
DocScrutinizer05you're better off with http://www.amazon.de/dp/B00THB4VT4/?tag=glv-21&ascsubtag=4ea4350ada05458b2d468c3f13adaea110:59
whitequarkI don't think that has the cool down function11:00
DocScrutinizer05((if i have to leave tea to extract then I forget about it entirely. same about water in teapot)) that's the nice thing about the first one. You simply fill in cold water and tea, and a 5 min later you got a complete can of ready made and filtered tea11:01
DocScrutinizer05with extraction time to the second11:01
whitequarkbesides it uses the capsules, like keurig,which are insanely expensive and probably unavailable here in HK11:01
whitequark(5 min later) see this is the problem11:01
DocScrutinizer05you're in HK now?11:01
whitequarkI will forget about it 5 minutes later, 100% certainty,11:01
whitequarkyeah, I moved to HK11:01
DocScrutinizer05hmm, congrats?11:02
whitequarkehhh I guess11:02
whitequarkthere's not much difference frankly11:02
whitequarksomewhat faster post, less food that i like11:02
DocScrutinizer05and Mao instead Stalin11:03
whitequarkless extreme cold but no heating meaning that 15° is excruciating11:03
whitequarkyou wouldn't know about Mao if you just lived in HK as a foreigner11:04
whitequarkit's a very western place11:04
DocScrutinizer05I know, I think I've been there11:04
whitequarkhence: not much difference11:04
whitequarktaobao is fairly nice11:05
whitequarkalthough US amazon is still more cnvenient if you live near it11:05
DocScrutinizer05how's internet in HK, particularly on weekends?11:14
DocScrutinizer05in TPE it was abysmal11:14
whitequarkextremely shit11:14
DocScrutinizer05well, a 7 years ago11:14
whitequarkI have ADSL 8 down 0.5 up11:15
whitequarkat home11:15
whitequarkand fiber 10 down 10 up in the office11:15
whitequarkwhy do yo ueven bother with fiber if there's no plan over 20mbps??11:15
DocScrutinizer05well, as long as it *works*11:15
whitequarkit's stable11:15
whitequarkbut mind you11:15
whitequarki used to have 200/20011:15
DocScrutinizer05on weekends you always had almost no connection to EU and USA11:15
whitequarknot anymore11:16
DocScrutinizer05I could get a max 150 down here, on shared medium (cable), which is not even a guarantee for 10 when your neighbours also use the net. Or I get 'up to' 16 via ADSL11:18
DocScrutinizer05not even VDSL11:19
DocScrutinizer05could you be my "boots on the ground" in HK?11:20
whitequarkwhatcha need?11:21
DocScrutinizer05maybe components or devices, e.g. refurbished N90011:21
whitequarkpossible but keep in mind i dont speak chinese11:22
DocScrutinizer05I see11:22
whitequarkthis is a huge problem overall11:22
arossdotmewhitequark, thx17:20
--- Sun Mar 13 201600:00

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