#qi-hardware IRC log for Tuesday, 2014-04-22

DocScrutinizer05actually for *electrons* (physical direction pragma) it would be push and pull swapped. At least that's what I consider "push" and "pull"00:00
DocScrutinizer05(seen from pin3)00:01
nicksydneyso if i change "Holes are push and pull around" to "electrons push and pull around" makes more sense ?00:01
DocScrutinizer05yes :-)00:02
nicksydneycool..will do that :)00:02
nicksydneyDocScrutinizer05: http://postimg.org/image/9smy5wjzh/ better ?00:04
DocScrutinizer05unless you write from a C3 point of view. then you could say C3 pushes electrons to V+ and pulls them from ground. But usually it's the active component (NE555) for which the activities are defined00:04
DocScrutinizer05the one's push is the other's pull ;-)00:04
DocScrutinizer05whe I push you, you may say you pulled me00:05
nicksydneyyeah i'm more inclined to view it from the NE555 point00:05
DocScrutinizer05ooh, and actually not holes but electrons are "moving from ground to C3"00:06
DocScrutinizer05holes move from (V)+ to -/GND00:07
DocScrutinizer05electrons do the opposite00:07
nicksydneyahh ok00:07
nicksydneyDocScrutinizer05: like this http://postimg.org/image/dszgzmqx3/ ?00:09
DocScrutinizer05again broken, needs click to "section 251"00:10
DocScrutinizer05could this be my adblock?00:10
nicksydneytry this http://postimg.org/image/pcog1kkvh/fba772c5/00:11
DocScrutinizer05anyway looks good00:12
nicksydneywill be a useful reference for myself down the track as will not be able to remember everything i learned :)00:14
nicksydneythe whole exercise of learning this circuit is because i'm thinking of making RA8875 driver board...and one of thing that really confuse the hell out of me was producing the different voltage required for the Backlight00:18
DocScrutinizer05aaah yes00:19
nicksydneythe board will have RA8875 for interfacing with 4.3inch screen + BLE (for communication to Android) + temperature + barometer  and for MCU was thinking M3 or M0 00:20
DocScrutinizer05best you forget about electrons (and even holes) and just think "current" (keeping in mind this is technical current pragma, + -> - )00:20
DocScrutinizer05current always flows from + to minus, or from higher to lower voltage, whatever fits your model best00:21
nicksydneythis is the schematic for the power supply i found on the net http://postimg.org/image/o370v7dvz/f3d8104a/ ... trying to replicate that by using TPS61040 00:23
DocScrutinizer05only for electron beams in CRT ans the like, you need to know the electrons actually come from minus and move to plus00:23
nicksydneyDocScrutinizer05: ok got it...will think about it as just normal current going from + to lower voltage 00:23
DocScrutinizer05but even experienced EE usually don't bother about electrons00:24
DocScrutinizer05actually often I tend to forget when I better should remember00:24
DocScrutinizer05e.g. when pondering highly obscure stuff like electromigration (electrons bouncing metal atoms along the wire, literally!)00:25
nicksydneybecause this schematics http://postimg.org/image/o370v7dvz/f3d8104a/ shows that there is VLED+ and VLED- which makes me think need to understand about producing negative voltage and that's where i learned the termed 'Charge Pump' so it eventually brought me to learn about using 555 to produce negative voltage to understand the basic00:25
DocScrutinizer05seems to me like VLED- is pretty standard connected to GND00:26
DocScrutinizer05no need for negative voltage in there00:27
DocScrutinizer05*usually* GND = V-00:27
nicksydneyahhhh.....right....that's why when i was looking at that schematics again and trying to put the pieces together i realised that there is no diode present for the negative voltage00:27
DocScrutinizer05there's just a resistor for sensing the current 00:28
nicksydneywhat is "sensing resistor" used for actually ?00:29
nicksydneyDocScrutinizer05:  so "*usually* GND = V-" does that mean that when the backlight is connect the connection will be the VLED+ for BL+ and GND for BL- ?00:31
DocScrutinizer05L4 and D2 create a high (relative to input) voltage to feed backlight VLED+ with it. On  R8 there's a small voltage (some 0.x Volt) proportional to current through LEDs, and the chip U4 is sensing that via pin3/FeedBack00:31
DocScrutinizer05nicksydney: the connections are like shown in schematics00:32
DocScrutinizer05U4 is specialized chip to drive backlight00:33
nicksydneyyeah i was thinking of replacing U4 with TPS61040 as that AIC1896 is not easy to get ... but again i could be wrong00:34
nicksydneyR5 and R8 is a voltage divider correct ?00:35
nicksydneywhile R9 is to limit and pull current to GND ?00:35
DocScrutinizer05pin1/LX pulls current through L4 to store energy in that inductivity. Then U4 switches pin1 to open state, so the inductivity creates high voltage pushing through D2 into C23/2100:35
DocScrutinizer05yes, basically that's a voltage divider. But R8 and R9 (if populated) are parallel and can be considered one00:36
DocScrutinizer05(voltage divider) here it doesn't serve that purpose. The purpose of R5 is to make any current going through Zener D4 (when voltage > maybe 13V) immediately pull up U4:3 to stop the oscillating of that chip and thus reduce output voltage it creates 00:39
DocScrutinizer05so basically R8 and R8 pull current to GND00:57
nicksydneyahhh ok let me see if i get this right.....when current flow from FB and since R5 and R8 is connected to GND ...the current is "pulled" towards GND by the resistor  which leaves less current to be "pulled" via D4 ? 00:57
DocScrutinizer05FB is an input00:57
DocScrutinizer05D4 is normaly highimpedance aka "open"00:58
DocScrutinizer05simply forget it for now00:58
DocScrutinizer05voltahe upper end of R8/9 goes higher the higher the current00:58
DocScrutinizer05this voltage goes to FB input via R500:59
DocScrutinizer05when voltage goes too high, U4 reduces frequency to regulate the current to LEDs00:59
DocScrutinizer05that's normal operation00:59
DocScrutinizer05now think something strange happens and voltage on VLED+ (and top of D4) rises above any sane limit to maybe 13V or more01:00
DocScrutinizer05D4 is a Zener that "closes"/connects the path for electric current when voltage along D4 goes >12V (that's why it's written there)01:02
DocScrutinizer05current flows from VLED+ resp C23/D2 to D4, through D4 and directly into FB 01:02
DocScrutinizer05FB sees very high voltage and immediately stops activity of U401:03
nicksydney"current flows from VLED+ resp C23/D2 to D4, through D4 and directly into FB" so it means that the Zener Diode "REverse breakdown voltage" can be reached ?01:04
DocScrutinizer05yes, exactly01:05
DocScrutinizer05on >12V plus whatever voltage is at lower end of D401:05
nicksydneynow this is interesting.....as learning something new again :)01:05
DocScrutinizer05afk o/01:06
nicksydneyso the current 'push' and 'pull' between D4 and the FB uses that to determine when to 'supply' more current or not01:06
DocScrutinizer05there's no push/pull between D4 and FB. FB is an imput01:07
DocScrutinizer051/LX is the active output of U401:07
nicksydneyok let me go this step-by-step....01:07
nicksydney1/LX will provide current and the current flows 'through' D2 to supply to VLED+01:08
DocScrutinizer05LX connects to GND to increase current through L4. Then it does stop the current (high impedance) and thanks to "inertia" the current in L4 has to go through D201:09
DocScrutinizer05so LX pulls current through L2, and L2 then pushes current through D201:09
DocScrutinizer05this cycle happens like 1 mio times per second01:11
DocScrutinizer05and less often when FB goes high enough, so current through LEDs gets reduced when it's going too high01:11
DocScrutinizer05(actually it might adjust time of LX pulling, instead of frequency LX is pulling)01:12
DocScrutinizer05effect is the same: current reduction01:13
nicksydneyso when this happens --> "and less often when FB goes high enough, so current through LEDs gets reduced when it's going too high" .... that's the time when the voltage will be higher than 12V and reverse voltage of D4 happens  and this makes current "flow" through D4 but the R5 and R8 resistor limit the current for the FB pin..is that correct how my thinking is ?01:27
DocScrutinizer05without R5 the current through D4 would go straight to ground via very low resistance R8/903:36
DocScrutinizer05thanks to R5 the FB input goes high level very easily as soon as D4 allows a small current to flow03:39
nicksydneyDocScrutinizer05: so R5 is there to make sure "some current" does not flow through directly to GND so as can be picked up the signal by FB04:13
DocScrutinizer05we want volatge from R8/9 on FB, for current feedback, but we also want allow D4 to immediately and efficiently "shut down" the device when overvoltage detected on VLED+04:17
DocScrutinizer05R5 allows voltage from R8/9 to go to FB with little attenuation (FB is very high resistance input), but doesn't allow all of current through D4 (when overvoltage) vanish directly to GND via R8/904:19
nicksydneyI understand the part that current flows through D4 and R5 "stops" some current from goingt o GND to be picked up by FB but I'm lost when you say "R5 allows voltage from R8/9 to go to FB " 04:24
nicksydneyi think the question that my head is ringing with "Where does the voltage from R8/R9 comes from as current are trying to push itself all to GND" ? :)04:25
nicksydneysorry if the question sound stupid :)04:26
DocScrutinizer05when current flows along a resistor, it creates a volatge along the resistor. This means for 100mA via LEDs there will be +1.2V (vs GND) on upper pin of R8/904:31
DocScrutinizer05those 1.2V go straight to FB, via R504:31
DocScrutinizer05it's not like current has a direction and needs to break to make it through the curve towards R504:32
nicksydneyi see....so current has a destination and voltage does not have any ..methamorphically speaking04:33
DocScrutinizer05hmmm, I have to ponder this, it#s not easy for me to understand :-)04:33
DocScrutinizer05it might be correct but I haven't understood it completely yet04:34
DocScrutinizer05may it helps when you compare a electric circuit with your heating pipes and radiators and the pump for the water04:35
DocScrutinizer05some radiators may be parallel to each other, others may be in series 04:35
DocScrutinizer05the battery or power supply is the pump creating some pressure difference04:36
DocScrutinizer05the on/off-valves are resistors04:37
nicksydney"pressure difference" that's the word i was looking for...yes that makes sense ...because FB has less pressure the voltage "move" to that pin ?04:37
DocScrutinizer05also the radiators are rsistors04:37
DocScrutinizer05yes, some electricity moves there, but since it's a dead-end pipe basically that omly has a pressure gauge mounted to it (FB input), not much electricity is flowing through the R5 resistor to FB, and at FB you already have same pressure like at top pin of R804:39
DocScrutinizer05for the pressure to propagate, not much volume has to move around, so any valve/resistor that's in there will probably not matter much since not a whole liter of water flows through it. It's more like one ml per minute that gets lost somewhere in pressure gauge04:41
DocScrutinizer05basically zero current is flowing04:41
DocScrutinizer05via R504:41
DocScrutinizer05when D4 eventually "opens the valve", *some* water will flow via R5 down to the huge GND pipe and back to the pump, but the pressure on upper end of R5 will be hig, compared to lower end of same R504:43
DocScrutinizer05if R5 wasn't there, all water would go straight to R8 and GND and not much pressure would be seen at FB04:44
DocScrutinizer05I mean if R5 was a straight pipe (aka wire)04:45
DocScrutinizer05without any connection otoh the pressure at R8 couldn't get to FB at all04:46
nicksydneyso with connection that is there (R5) the pressure was slow down so it can reach FB04:46
DocScrutinizer05*theoretically* we could also use an ideal diode instead of R5, that would only let current flow from R8 to FB but not the other way round04:47
nicksydneyso that's why PIN3 is connected between D4 and R504:47
DocScrutinizer05but there are no ideal diodes, they always bite a bit out of pressure/voltage even in forward direction of current. Usually ~0.7V04:48
DocScrutinizer05and diodes are more expensive than resistors04:48
DocScrutinizer05and then there's also some non-ideal parts in D4 behaviour, it might not be completely "tight" even on voltages lower than Zener voltage04:49
nicksydneywhat will happen if PIN3 was connected between R5 and R8 ?04:50
DocScrutinizer05as long as this leak is negligible compared to the flow of current that R5 allows to drain to GND, it doesn't matter04:50
DocScrutinizer05then overvoltage protection would not work at all04:51
nicksydneymy best guess is it will not be picked up by FB 04:51
DocScrutinizer05when PIN3 was connected between R5 and R8, then overvoltage protection would not work at all04:51
nicksydneybecause the voltage between R5 and R8 is very high for FB ?04:53
nicksydney"<DocScrutinizer05> those 1.2V go straight to FB, via R5" ... is that the 1.2V ?04:53
DocScrutinizer05current via R8 would increase *marginally* when D4 opens up completely, since R5 limits the current to insignifican amount. Thus voltage across R8 wouldn't increase any noticeable amount, and thus FB doesn't see any change04:53
nicksydneyahhh so i was correct :) ... the FB does not detect any changes and like you say the overvoltage protection would not work04:54
nicksydneygot it !04:54
DocScrutinizer05so D4 and R5 are useless when FB  is connected to R804:54
nicksydneyok ok now i see04:54
Action: nicksydney jumping like a rabbit as he can see how things works now04:55
nicksydneyit's complicated but yet easy to understand !04:55
DocScrutinizer05yeah, this actually is a nice educational circuit04:55
DocScrutinizer05it shows a few concepts04:56
nicksydneyfor newbie like me all circuit are educational circuit :)04:56
nicksydneyso that means there is no point having R9 as R8 (changing the value) will be more than enough ?04:57
nicksydneyin terms of selecting the correct R value normally how do you go about determining it ?04:58
DocScrutinizer05as long as the value is correct to create the right FB volatge when current through LEDs is nominal, it doesn't matter if it's one or two parallel resistors04:58
nicksydneyi mean if you have to start designing this kind of circuit how to determine what is the correct resistor value to use ?05:00
DocScrutinizer05determining R8: find datasheet for U4. It has a formula that goes like: FB-voltage must be 1.3V, so for 100mA you need 13 Ohm, according to I*R=U rsp R=U/I05:01
DocScrutinizer05U=1,3V, I=0.1A, -> R=13Ohm05:03
nicksydneyso normally in the datasheet it does tells you that if say for example you use L4 with 10uH you will get 100mA and then you can use that current value to calculate the voltage required to get the resistor right ?05:03
DocScrutinizer05when FB volatge must be 1.1V (datasheet needs to tell you), then your R is 5.5 Ohm for 200mA05:03
DocScrutinizer05L4 is fixed by design of U4 basically. They will tell you what L4 type to use05:04
DocScrutinizer05it's not really related to current you want to get for your LEDs05:05
DocScrutinizer05only R8 determines the current05:05
nicksydneyDocScrutinizer05: something like this http://postimg.org/image/4asyduuf7/ba081bb3/ ?05:06
DocScrutinizer05and the current you get from datasheet of your display, it tells you how much current to use for the backlight05:06
DocScrutinizer05yes, exactly05:06
DocScrutinizer05my first estimation been already to the point05:06
DocScrutinizer05actually 1.233V05:07
nicksydneyfrom the datasheet the first page shows this circuit http://postimg.org/image/ufqga6udf/bd07bdb6/ and you were right as it specify the use of 10uH05:11
nicksydneybut hang on noticed in one of the page the circuit that produce negative and positive output the L1 is different  http://postimg.org/image/v3y6yhuvf/09634334/05:14
DocScrutinizer05that circuit is for regulated output voltage, for backlight you need regulated output *current*05:15
nicksydneyahhh we need "regulated output current" to control the amount of current that are available for the LED as we don't want too much current as that will create problem...got it !05:16
DocScrutinizer05and the 6.8uH is for very low output current, the diagram says 10V@15mA05:17
DocScrutinizer05you need to study the complete datasheet, it wil have proper instructions how to use the chip and which external components to get05:18
nicksydneyok cool....going to write down what i learned today in my notes otherwise will forget again05:20
DocScrutinizer05in http://postimg.org/image/v3y6yhuvf/09634334/ you will find the chage pump again05:21
DocScrutinizer05for the -10V section05:22
DocScrutinizer05it's exactly identical05:22
nicksydneyyes that's right...that's what we were discussing :)05:22
nicksydneyalways makes me nervous playing with high voltage stuff as not experienced enough :)05:23
nicksydneythat's why i always read and read more to understand first before doing any high voltage stuff :)05:23
DocScrutinizer05even C3 and C4 are named identical, haha05:23
DocScrutinizer05err, are you planning to operate that stuff at 100V instead of 10V?05:24
DocScrutinizer0510V is not exactly any high voltage05:24
DocScrutinizer05or 12V or even 18V which might be needed for your backlight05:25
nicksydneyno not 100V i'm already worried with 20V :)05:26
nicksydneylet me check what is the backlight voltage05:26
DocScrutinizer05hmm, not sure anymore about C4, was it actually C4 in that other Ne55 based circuit?05:26
nicksydneyyes C4 was used like you said as buffer cap in the NE555 circuit05:27
nicksydneyit's labelled C5 in the NE55505:27
nicksydneythis one http://postimg.org/image/j019ckor9/7cfa6149/05:28
DocScrutinizer05right, C4 is C405:28
DocScrutinizer05and C3 is C305:29
DocScrutinizer05btw 20V are nothing to worry about, actually not even with wet feet :-)05:30
DocScrutinizer05your landline phone has ~40V and is considered safe05:30
nicksydneyyikes ! didn't know that :)05:31
DocScrutinizer05vans have 24V batteries05:31
nicksydneyi'm worried to blow up something :)05:31
DocScrutinizer05your laptop has a 22V powersupply with exposed plug05:31
nicksydneyi blew up a regulator on my Arduino UNO already when was playing around with a TFT screen :(05:31
nicksydney*pop* and white smoke :)05:32
DocScrutinizer05yeah, you can blow up stuff anyway. even on 1.8V05:32
nicksydneytrue 05:32
DocScrutinizer05nevertheless it helps *a lot* to understand what you're doing05:33
DocScrutinizer05ooops, fuse of laptop blown ;-P05:33
DocScrutinizer05or panic shutdown due to dangerous PSU? 05:34
nicksydneyhahahaha......nah USB not working need 05:34
DocScrutinizer05have fun! I'm afk05:35
nicksydneythanks DocScrutinizer0505:37
whitequarkguys, have you seen PSHDL? http://pshdl.org/11:02
whitequarklarsc: ^11:02
whitequarkwhat do you think about it?11:03
larsctake the worst parts of verilog and vhdl and mix them with java ;)11:03
larscI don't like it, but that might be personal taste11:07
larscit's a bit like php11:08
ysionneauSo, is the charge absolute or not? damn it! :p11:12
Action: ysionneau leaving -> |_11:12
larscysionneau: it's relatively absolute11:14
ysionneauit's absolutely relative11:15
ysionneauor the other way round :)11:16
nicksydneyinteresting site - http://solderpad.com/12:50
nicksydneyuhhh MilkyMist One solderpad http://solderpad.com/solderpad/milkymist-one/12:51
whitequarktheir circuit viewer causes nausea12:53
nicksydneywhitequark: think of it as a 3d cinema :)13:05
wpwrak3D crimes ? :)13:31
larsc3d crimera?13:41
nicksydneyhttps://www.youtube.com/watch?v=HJQa1GlIi_Q SMD production line porn :)13:48
eintopfnicksydney: cool17:18
Action: whitequark ponders at "powdered alcohol" that recently appeared on the net22:41
DocScrutinizer05whitequark: congrats, your moscow streets get nicer now. Don't dare to think otherwise! The project manager clearly stated that those who don't like this project are not of a sane state of mind22:42
whitequarkapparently, ethyl alcohol trapped in cyclodextrin crystals22:42
whitequarkDocScrutinizer05: what?22:42
DocScrutinizer05"official" graffiti22:43
whitequarkwhat the hell are you talking about22:43
DocScrutinizer05"Putin" painting walls with graffiti showing Krim in russian colors, and of course the logo i´of his party next to it22:44
whitequarkyou clearly are informed much better than me22:44
DocScrutinizer05see? German TV news are not all that bad22:44
DocScrutinizer05I seem to recall they said "2000 walls"22:45
whitequarkI do not watch TV of any country22:45
DocScrutinizer05I can well understand that22:45
whitequarkhmmm, so, powdered alcohol. I'd need ±-cyclodextrin22:45
whitequarksigma-aldrich is selling it for like 300EUR/kg22:46
DocScrutinizer05seems this http://themoscownews.com/arts/20121210/190981821.html is an initial message preparing the public for what's rolling out now22:48
whitequarkapparently, I can make it myself with starch and some enzymes. ±-amylase is going for $20/120ml, but their price list is weird22:48
whitequarknow, I'd also need ±-CGTase. not sure where to get that yet, hmmmm22:48
whitequarkgrmbl. no vendors here. alibaba otoh has some22:50
whitequarkno, I don't need a *bucket*22:50
whitequark(o°¡°o5 ;;22:53
whitequarkwhy can't I buy a completely trivial purified enzyme in reasonable quantities, is it 2014 or so22:53
whitequarkwhy the fuck are we, as a society, focused on fucking rectangles with fucking displays22:54
whitequarkwon't work.22:55
whitequarkfascinating: apparently, a pinhole camera performs a spatial Fourier transform on the incoming light22:58
whitequarkI'm not sure I understand how that works22:58
nicksydneyDocScrutinizer05: just put a sample oder from TI for the TPS chip let's see if they approve it for the project :)23:05
nicksydneygoing to design the pcb to test it on when it arrive 23:05
DocScrutinizer05good luck!23:06
nicksydneygoing to divide and conquer...once the power part is working than i will work on the RA8875 chip 23:06
nicksydneythis is the block diagram for the board http://postimg.org/image/4xfx9luyr/3ebfbd85/ .. not my design ..this is from RA8875 demo board ... not going to use ROM in my design23:08
nicksydneyand this is the RA8875 chip schematics that i got from the same doc http://postimg.org/image/f0t7p7foj/bff6eb78/23:09
DocScrutinizer05whitequark: why do you want to build cyclodextrin (if that's the right name, I dunno). Seems that stuff is widely available for relatively low cost23:10
DocScrutinizer05and used in roundabout everything nowadays, like "Febreze air freshener"23:10
DocScrutinizer05also washing powder that claims to keep laundry fresh (read smelling) for 2 weeks23:11
DocScrutinizer05literally everybody and his dog is using that stuff for smart as well as nonsensical purposes23:12
DocScrutinizer05>>I'd need ±-cyclodextrin<< ooops, they use beta23:16
DocScrutinizer05and the most important question: what the heck do you need powdered alcohol for?23:18
DocScrutinizer05lol, http://www.sigmaaldrich.com/catalog/search?interface=Product%20Name&term=Alpha+Cyclodextrin&N=0+&mode=mode%20matchpartialmax&focus=product&lang=en&region=US  those guys deserve a price for the most fucked up webpage since... 5 days. When you reject their cookies, the page never stops reloading23:24
wpwrakthe world is full of this sort of pages ...23:25
DocScrutinizer05isn't there any way to instruct browsers to not accept cookies for $this-url ?23:39
DocScrutinizer05I'd love to send to /. something like "hey! those guys give away Alcohol for free! just fill the form :-)  http://www.sigmaaldrich.com/catalog/search?interface=Product%20Name&term=free-alcohol&ACCEPTCOOKIES=NO"23:41
DocScrutinizer05after all, when they already offer their own DDoS exploit, then way deny and reject it?23:42
DocScrutinizer05s/way /why /23:42
DocScrutinizer05never waste an opportunity for an exploit borne from ignorant use of deprecated technologies23:45
DocScrutinizer05hmm, maybe I just need to tackle this a tiny bit smarter: "hey, have you recently checked the vulnerability of your FireFox? To check, set 'accept cookies: never' and then visit this pake: $URL. When after 60min the page hasn't detected the vulnerability in your FF, then it's probably safe to simply abort the test. Nevertheless redo every other week, new vulnerabilities could sneek in to FF with updates that get installed 23:49
DocScrutinizer05page, even23:49
DocScrutinizer05might fly23:50
Action: DocScrutinizer05 wonders what's best channel to start this game23:54
DocScrutinizer05media channel23:54
DocScrutinizer05/.? reddit? some other silly fora? Twitter?23:55
DocScrutinizer05BOFH is quite inspiring sometimes23:57
DocScrutinizer05>>never waste a wrong connection<<23:58
DocScrutinizer05>>don't say 'sorry, wrong number', say 'Cindy? Cindy isn't available for you anymore! Cindy is completely pissed of you and I shall tell you to fuck off<<23:59
--- Wed Apr 23 201400:00

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