--- Log opened Wed Apr 26 00:00:42 2017 01:48 #navit: < kcroot> morning 01:48 #navit: < kcroot> kazer: ping 02:03 -!- xenos1984 [~xenos1984@2001:bb8:2002:200:6651:6ff:fe53:a120] has joined #navit 02:18 -!- magruder [redurgam@gateway/shell/panicbnc/x-qtqjbdqtekvpcmia] has quit [Ping timeout: 264 seconds] 02:26 #navit: < kcroot> query 03:21 -!- jjelen [jjelen@nat/redhat/x-nvzfqzjvhgzqufje] has joined #navit 05:05 -!- redurgam [redurgam@gateway/shell/panicbnc/x-tfgmufojxliakcpq] has joined #navit 07:59 -!- pini [~pini@bou-fi.pustule.org] has joined #navit 10:16 -!- xenos1984 [~xenos1984@2001:bb8:2002:200:6651:6ff:fe53:a120] has quit [Quit: Leaving.] 12:01 #navit: < kazer> h ithere 12:01 #navit: < kazer> hi kcroot 12:16 -!- jjelen [jjelen@nat/redhat/x-nvzfqzjvhgzqufje] has quit [Ping timeout: 240 seconds] 12:46 -!- robertp [5849cf38@gateway/web/freenode/ip.88.73.207.56] has joined #navit 12:49 #navit: < kazer> hey robertp 12:52 #navit: < robertp> hallo 12:54 #navit: < robertp> Ive been trying and trying to get this fcuk uninstall target to work.... If noone cares I am about to drop this feature. A handmade shell script will do much better.... 12:54 #navit: < kazer> haha 12:54 #navit: < kazer> well i care too, it would be nice to have it 12:55 #navit: < kazer> where are you currently stuck? 12:58 #navit: < kazer> btw, not as pretty as your board but I made some tangible progress on my electronics : https://cloud.kazer.org/index.php/s/avzShPM0vMqPBhJ 13:15 #navit: < robertp> I am stuck at deleting empty folders and checking if they are empty. The docs seem to be some kind of outdated. 13:17 #navit: < robertp> I am using cmake 3.0.2 on Ubuntu 16.04 at my vm 13:21 #navit: < robertp> your electronics look good. how did you make the pcb, kazer? 13:23 #navit: < robertp> the print looks a bit misaligned o.O 13:34 #navit: < kazer> i used a cnc engraving machine to mill the pcb, and a laser cutter to etch the component placement. The print is indeed a bit misaligned, it was my first attempt :) 13:35 #navit: < robertp> do you also have a pic of the copper side? 13:35 #navit: < kazer> you can see that i really fucked up my settings because the shapes are not consistent :) my corner speed was too high and i had some jitter 13:35 #navit: < kazer> sure hang on 13:37 #navit: < kazer> https://cloud.kazer.org/index.php/s/H7gzZKkmD5gVO5x 13:38 #navit: < kazer> i used a v-shaped carving tool for the pcb. Pro : much more difficult to break. Con : end result is a bit less pleasant, aestetically 13:38 #navit: < kazer> last time i used a small routing bit instead. Tracks edges were sharper, but these break really, really easily 13:40 #navit: < robertp> I'd solder all pins of the arduino for mechanical stability. If you mount that in your car, that will break soon. 13:40 #navit: < kazer> thanks, that's useful advice 13:40 #navit: < kazer> any comment welcome btw, i'm still new to this 13:40 #navit: < kazer> also the arduino is mounted on a riser on this board, as it's the first prototype 13:42 #navit: < kazer> https://cloud.kazer.org/index.php/s/D0PTMkBPKXC10A3 13:43 #navit: < kazer> btw : next i am going to need a board to switch between audio sources. Have you done that before? 13:47 -!- xenos1984 [~xenos1984@22-164-191-90.dyn.estpak.ee] has joined #navit 13:47 #navit: < robertp> what kind of source? for analog audio signals use relais, for (digital) i2c audio i2c-audio-ics 13:48 #navit: < kazer> analog yeah. I am going to need 4 sources (raspberry, bluetooth, FM, line in) 13:53 #navit: < robertp> I'd use a on/off relay for every channel and take care that only one is active at a time to prevent output latchup of the source (outputs should be short-circuit-safe but I wouldnt rely on that assumption) 13:54 #navit: < robertp> There were some analog-switch-IC for analog TVs in the past but they are very rare and horribly expensive 14:10 #navit: < kazer> thanks! 14:16 #navit: < kazer> also looks like i'd want non-latching relay, correct? 14:28 #navit: < robertp> No, latchup or latch-up is a reaction of a cmos output to inductive load. This leads to direct connection of the power and ground of the cmos-output which lets the silicon melt. and smoke 14:29 #navit: < robertp> have to google what latching relay are.... 14:29 #navit: < kazer> sounds bad :) but this is different from latching / non-latching relays right? 14:30 #navit: < kazer> from what i gather a latching relay will hold its current position until toggled, where a non latching relay will revert to the NC state when the coil looses power 14:30 #navit: < robertp> yes, that latch-up has nothing to do with relays 14:30 #navit: < robertp> that depends on your relay driver technique 14:31 #navit: < kazer> i was thinking of driving the relays with an arduino 14:31 #navit: < kazer> like i do currently to enable the car amp 14:48 #navit: < kazer> how about these? http://pccomponents.com/datasheets/PANA-TQ2.PDF the tq2-5v 14:58 #navit: < robertp> So cheap standard automotive relays will do 14:58 #navit: < robertp> do you have schematics of your pcb? 14:59 #navit: < robertp> you should.... would you mind sharing? 15:00 #navit: < kazer> for the power board? yeah sure. For the audio not yet, will work on it in the next days 15:00 #navit: < kazer> just ordered a bunch of tq2-5v, will try them 15:00 #navit: < kazer> do you use kicad? 15:09 #navit: < robertp> No I use Eagle and Altium (CircuitMaker) 15:09 #navit: < kazer> ok 15:09 #navit: < kazer> i'll push the pcb and schematics to github today 15:11 #navit: < kazer> in the meantime : https://cloud.kazer.org/index.php/s/NbraDEwpsS07t2m 15:11 #navit: < kazer> any comments are welcome, so shoot :) 15:13 #navit: < robertp> So better hide savely. O.o 15:13 #navit: < robertp> What values do the resistors at the Ardiuni IOs have? 15:19 #navit: < kazer> r1 = 33k, r2 = 10k 15:19 #navit: < kazer> the voltage divider on BATT is meant to monitor battery level, and shutdown the pi if battery goes low 15:20 #navit: < kazer> the voltage divider on IGNITION is meant to power on the pi when starting, and giving it some time to cleanly shut down when turning the engine off 15:22 #navit: < robertp> I'd recommend R1 47k, R2 22k and R3 (from R1/R2 in series to the arduino io) also 47k and a 22nF capacitor parallel to R2. The protection diodes on the arduino can be massacred with up to 1mA continously 15:23 #navit: < robertp> So the series resistor from ~12V to the Arduino is about 100k. That protects the IC from voltages up to 1000V. 15:24 #navit: < robertp> In noisy automotive environments voltages to 1000V can occur. 15:25 #navit: < robertp> The capacitor takes care about needle pikes 15:26 #navit: < kazer> thanks! 15:27 #navit: < kazer> mm, 47k/22k gives 5.1V at 16V input. Isn't it a bit much? 15:28 #navit: < robertp> the arduino has diodes from IO to VCC 15:28 #navit: < kazer> so it's safe to provide, say, 5.5V on an input, we will just read it as 5v? 15:29 #navit: < robertp> so everything above 5.6V will be lead to vcc 15:29 #navit: < kazer> ok thanks! 15:29 #navit: < robertp> yes 15:29 #navit: < robertp> but take care about the current. giving the protection diodes more than 1mA might harm the arduino 15:30 #navit: < robertp> thats why i recommend 2x 47k in series 15:31 #navit: < robertp> for adc use reduce the R2 of the divider to adjust the operating range 15:31 #navit: < kazer> adc? 15:32 #navit: < robertp> analog digital converter 15:35 #navit: < kazer> ah, thanks 15:35 #navit: < robertp> You read the analog voltages, don't you? 15:36 #navit: < kazer> yes correct 15:36 #navit: < kazer> so something like this? https://cloud.kazer.org/index.php/s/EJgoLRYtwqSFIuV 15:37 #navit: < robertp> almost 15:37 #navit: < robertp> use ceramic capacitors or foil caps 15:37 #navit: < kazer> so, unpolarized? 15:37 #navit: < robertp> electrolytic caps are way too slow 15:39 #navit: < robertp> yes 15:40 #navit: < kazer> https://cloud.kazer.org/index.php/s/PTCYvephe9UEVDq ? 15:46 #navit: < robertp> ok. Next: You are also missing a diode antiparallel to the relay coil 15:46 #navit: < kazer> indeed thanks 15:46 #navit: < kazer> i noticed it while looking into relays earlier 15:46 #navit: < robertp> the induction spark of the coil will latch-up ( ;) ) your MOSFET 15:48 #navit: < kazer> ha yeah :) 15:49 #navit: < kazer> like this? https://cloud.kazer.org/index.php/s/ZG3l0LA4OmizcRW 15:49 #navit: < kazer> any advice on the part itself? 15:49 #navit: < robertp> where is your voltage regulator? It is very important, too 15:50 #navit: < kazer> i am using a ubec for that, it's the 4 pins on the top right corner (j9-j11) 15:50 #navit: < kazer> i will have to add a part to my library 15:52 #navit: < robertp> So obviously no over voltage and short circuit protection? 15:53 #navit: < kazer> i *believe* that it is handled by the ubec 15:56 #navit: < robertp> can you link a schematic? 15:57 #navit: < kazer> sure 16:02 #navit: < kazer> mmm. they are hard to find it seems. But the specs for the ones i have say that it has over current and reversed pins protection. Nothing about over voltage or short circuit though 16:04 #navit: < kazer> this seems close : https://www.monolithicpower.com/Products/Product-Detail/238/Step-down(Buck)/SwitchingRegulators/DC-DCPowerConversion/Converters/MP2307 16:04 #navit: < robertp> short circuit protection can be done with a simple fuse 16:05 #navit: < kazer> adding a fuse sounds like a good idea indeed 16:05 #navit: < kazer> you mentioned ceramic capacitors earlier. Should I use -|(- instead of -||- in my schematics? 16:07 #navit: < robertp> The schematic symbol SHOULD match the footprint. So use -|(- for electrolytics and -||- for ceramics or foil or.... 16:07 #navit: < kazer> yep i want them to match :) 16:07 #navit: < kazer> in the schematics i shared they use -|(- for ceramics 16:08 #navit: < kazer> ( C1 ) 16:08 #navit: < robertp> seems wrong ;) 16:08 #navit: < kazer> ok cool 16:11 #navit: < robertp> for over voltage protection use a fuse (we mentioned already) and a TVS (i use P1,5KE18A in automotive envs) from the fuse to ground 16:12 #navit: < kazer> will add that, thanks! 16:12 #navit: < robertp> http://www.rfwireless-world.com/ApplicationNotes/TVS-diode.html 16:13 #navit: < kazer> wait, aren't fuse protecting from over current rather than over voltage? 16:13 #navit: < robertp> Luke, the TVS-site reading you must.... 16:14 #navit: < kazer> so the fuse and the tvs diode would go from batt to ground, not 5v to ground, correct? 16:14 #navit: < kazer> haha ok 16:14 #navit: < robertp> the tvs leads current to ground if voltage exceeds the value printed on the plastic 16:15 #navit: < robertp> if there was a fuse, the fuse would now break and stop the current. 16:15 #navit: < kazer> so i should protect the 5v circuit, right? because the batt voltage can fluctuate 16:16 #navit: < kazer> mmm.. but the ubec should already guarantee a 5v output 16:16 #navit: < robertp> If there wasnt a fuse the TVS woul break and directly connect 12V to GND. If there is a thin cable from the battery to the circuit, it will play fuse 16:16 #navit: < robertp> is ubec designed for use in automotive environments? 16:17 #navit: < robertp> If so: lean back and enjoy 16:17 #navit: < robertp> if not, back to fuse and TVS 16:18 #navit: < kazer> it's mostly used for rc, like quadcopters where you power the electronics via some powerful batteries. I have used them with 4S, 5000mAh batteries for example 16:20 #navit: < robertp> Ok. But in your car is a lot of electrical noise. The relays, the alternator and electric motors spread a bunch of high voltage peaks over the 12V line. You cant compare that to LIPO cells of a RC thing 16:20 -!- xenos1984 [~xenos1984@22-164-191-90.dyn.estpak.ee] has quit [Ping timeout: 260 seconds] 16:21 #navit: < kazer> so to be safe, i should protect the circuit with fuse and tvs between bat and ground, with a tvs rated for .. 14V? 16:23 #navit: < robertp> That depend on the VinMax of your voltage regulator 16:23 #navit: < robertp> Most discrete Vregs can be used with up to 30V 16:23 #navit: < robertp> So there I use a TVS with 18V 16:24 #navit: < kazer> the ubec should accept up to 26V 16:24 #navit: < robertp> You should use the TVS which is only used if such an evil peak occurs 16:25 #navit: < robertp> in my car 14,8V are OK, so 15V would be too close. 16:25 #navit: < robertp> that would harm the TVS 16:26 #navit: < kazer> makes sense. 18V then 16:27 #navit: < kazer> like this? https://cloud.kazer.org/index.php/s/nocDecheFBPSinW 16:28 #navit: < robertp> No. 16:28 #navit: < kazer> doh 16:29 #navit: < kazer> re-reading the tvs page 16:29 #navit: < robertp> The fuse in series and then the TVS parallel 16:29 #navit: < robertp> like a voltage divider 16:34 #navit: < kazer> updated the schematics, same link as before. Correct? 16:37 #navit: < robertp> Almost 16:38 #navit: < robertp> Use a unidirectional TVS 16:38 #navit: < robertp> Cathode pointing to the fuse 16:38 #navit: < kazer> ok thanks 16:42 #navit: < kazer> updated 16:46 #navit: < robertp> very good. 16:47 #navit: < kazer> sweet. Anything else you see? 16:48 #navit: < robertp> What does J5 do? 16:49 #navit: < kazer> it's the connector to the pi. pin 1 is GND, pin 2 is 5V, pin 3 is a signal from the pi to power on the car amp 16:50 #navit: < robertp> Ah... the relay switches the AMP? 16:50 #navit: < kazer> yes correct 16:51 #navit: < kazer> i need to do a few tests, as the pi might be a pull down instead of a pullup (or the opposite, i'm always confused) 16:51 #navit: < kazer> to switch the amp on, i just need to provide power to the cable labelled "antenna" 16:52 #navit: < kazer> (that's its name on the wiring harness) 16:52 #navit: < robertp> I've seen another thing: The LED connected to 12V has the same series resistor like the one at 5V. 16:52 #navit: < kazer> i have to admit that i loosely spec'ed the resistors here :) 16:53 #navit: < robertp> haha. Its always the same. Everywhere. :D 16:53 #navit: < kazer> haha 16:53 #navit: < kazer> glad to know that i'm not alone :) 16:53 #navit: < kazer> does it matter much? 16:53 #navit: < robertp> Ive designed a board where I used DIP swicthes. 16:54 #navit: < robertp> http://excelcell.manufacturer.globalsources.com/si/6008800045588/pdtl/DIP-digital/1059992751/Through-Hole-Dip-Switch.htm?state_filt=CN_CNFUJ?? 16:54 #navit: < robertp> like this one 16:55 #navit: < robertp> I placed them at the end of the pcb to be able to access the through holes in the housing 16:55 #navit: < robertp> But a fellow soldered them and I realized that i rotated them 16:55 #navit: < robertp> to I couldnt access them through the holes 16:56 #navit: < robertp> To your LEDs 16:56 #navit: < robertp> they have a standard forward current and the resistor should be used to adjust the current and the brightness 16:57 #navit: < kazer> they are merely troubleshooting indicators in my case. I do not want them to be visible when the box is closed, too much distraction 16:57 #navit: < robertp> If you exceed the forward current the led will shine very bright but probably very short 16:57 #navit: < robertp> :D 16:57 #navit: < kazer> so i don't need them to be bright, and should just bump the R value? 16:57 #navit: < robertp> So dont solder them 16:58 #navit: < robertp> I have to go 16:58 -!- robertp [5849cf38@gateway/web/freenode/ip.88.73.207.56] has quit [Quit: Bye] 18:37 -!- xenos1984 [~xenos1984@22-164-191-90.dyn.estpak.ee] has joined #navit 19:22 -!- Horwitz [~mich1@p200300800E100000022268FFFE64E7C4.dip0.t-ipconnect.de] has quit [Ping timeout: 260 seconds] 19:34 -!- xenos1984 [~xenos1984@22-164-191-90.dyn.estpak.ee] has quit [Quit: Leaving.] 19:35 -!- Horwitz [~mich1@p200300800E7BEB00022268FFFE64E7C4.dip0.t-ipconnect.de] has joined #navit 19:35 -!- mode/#navit [+o Horwitz] by ChanServ 21:05 -!- noradtux [~noradtux@2a02:2028:60c:bc01:3d43:be8f:a687:b91e] has quit [Ping timeout: 260 seconds] 21:09 -!- noradtux [~noradtux@2a02:2028:569:6b01:187d:338c:deb4:6977] has joined #navit --- Log closed Thu Apr 27 00:00:44 2017