| 01:00.27 | *** part/#edev mozmck (~moses@client-67.210.159.209.dfwtx.partnershipbroadband.com) |
| 02:00.51 | *** join/#edev Belliash (~belliash@asiotec/constitutive/belliash) |
| 03:02.49 | *** join/#edev orhan89 (~orhan89@103.10.171.14) |
| 04:27.04 | *** join/#edev [7] (~quassel@rockbox/developer/TheSeven) |
| 05:31.43 | *** join/#edev mozmck (~moses@67.210.159.209) |
| 05:47.14 | *** join/#edev ntfreak (~ntfreak@unaffiliated/ntfreak) |
| 06:06.55 | *** join/#edev kuldeepdhaka (~kuldeepdh@117.254.218.251) |
| 06:15.14 | *** join/#edev kuldeepdhaka_ (~kuldeepdh@117.254.222.99) |
| 06:43.39 | *** join/#edev kuldeepdhaka (~kuldeepdh@117.254.220.97) |
| 07:25.11 | *** join/#edev kuldeepdhaka (~kuldeepdh@117.254.220.97) |
| 09:26.05 | *** join/#edev mbuf (~user@14.97.51.81) |
| 11:37.29 | *** join/#edev g1powermac (~g1powerma@unaffiliated/g1powermac) |
| 11:50.09 | *** join/#edev RampantOwl (~RampantOw@22.53.214.193.static.cust.telenor.com) |
| 12:06.06 | *** join/#edev kuldeepdhaka (~kuldeepdh@117.254.217.25) |
| 12:48.51 | *** join/#edev sndcrb (~sndcrb@dsl-hkibrasgw1-58c385-172.dhcp.inet.fi) |
| 14:30.19 | *** join/#edev landley (~landley@70.114.134.56) |
| 14:35.33 | *** join/#edev KiriKari (KiriKari@m90-129-68-81.cust.tele2.se) |
| 14:38.13 | KiriKari | If I have variable StepMotorPosition which is shared between an interrupt and the main-loop. when i write a new value to StepMotorPosition, should I disable interrupt and then enable again afterwards? disable_interrupts(); StepMotorPosition = 1000; enable_interrupts(); And in the interrupt I dont have to care because it will always finish before it does something else? If I dont disable interrupts |
| 14:38.13 | KiriKari | when writing to the variable it could cause problems? or does an interrupt wait until current instruction is finished? |
| 14:55.53 | *** join/#edev gustavoz (~gustavoz@host197.190-137-239.telecom.net.ar) |
| 15:52.07 | MonMotha | KiriKari: depends on whether the type of the variable in question is atomic on your architecture. If it's not (e.g. because its type exceeds the word size of your architecture or exceeds the memory bus size and your bus controller can be interrupted while splitting up the write), then yes, you need a lock of some sort. |
| 15:52.58 | MonMotha | also, if this is C or something similar, it's quite likely you need to qualify said variable as "volatile" to ensure the compiler does not generate an instruction stream which could use a stale copy of the data |
| 16:11.07 | *** join/#edev kuldeepdhaka (~kuldeepdh@117.254.217.150) |
| 16:12.26 | KiriKari | OK,ty. It is just an ordinary variable, not an IO or anything though. Not sure about the wordsize. |
| 18:20.48 | *** join/#edev Kasreyn (~i@unaffiliated/inquisitor) |
| 18:28.04 | *** join/#edev blueness (~blueness@cpe-74-77-145-97.buffalo.res.rr.com) |
| 18:28.05 | *** join/#edev blueness (~blueness@gentoo/developer/blueness) |
| 18:46.07 | *** join/#edev djerome (~djerome@68.98.228.41) |
| 18:49.48 | *** join/#edev Kasreyn_ (~i@h5ef52150.selulun.dyn.perspektivbredband.net) |
| 18:49.48 | *** join/#edev phonkane_ (phonkane@gateway/shell/tkk.fi/x-ikqieloaiudbkyuj) |
| 18:49.54 | *** join/#edev ccole_ (~cole@coledd.com) |
| 19:08.42 | *** join/#edev florian (~fuchs@Maemo/community/contributor/florian) |
| 19:26.47 | *** join/#edev blueness (~blueness@gentoo/developer/blueness) |
| 19:29.43 | *** join/#edev kuldeepdhaka (~kuldeepdh@117.254.217.150) |
| 20:12.36 | *** join/#edev Guest62783 (~kuldeepdh@117.254.220.249) |
| 20:13.15 | *** join/#edev kuldeepdhaka_ (~kuldeepdh@117.254.220.249) |
| 20:36.55 | *** join/#edev Lars_G_ (~Lars@unaffiliated/lars-g/x-000001) |
| 20:45.51 | *** join/#edev orbarron (~orbarron@nat/ti/x-kgezvychzormbwfz) |
| 21:55.03 | g1powermac | anyone know off hand what kind of capacitor is 'low leakage'? |
| 21:57.35 | devilsadvocate | g1powermac: poly |
| 22:02.34 | g1powermac | devilsadvocate, hmm, what value is used to measure the leakage of these caps, ESR? |
| 22:07.10 | Lars_G_ | g1powermac: Why the hell are you here too? |
| 22:09.07 | g1powermac | cause I am :-) |
| 22:19.24 | SpeedEvil | Context dependant. |
| 22:19.53 | SpeedEvil | A low leakage cap may be femptoamps, with glass standoffs in vacuum, or amps. |
| 22:20.05 | SpeedEvil | In the case of a warehouse sized cap bank |
| 22:37.10 | g1powermac | SpeedEvil, I'm looking at using this chip: http://www.ti.com/lit/ds/symlink/bq25504.pdf |
| 22:37.30 | g1powermac | it requires 'low leakage capacitors', but that's pretty much all the datasheet says about them |
| 22:37.45 | SpeedEvil | No, it's not. |
| 22:38.22 | SpeedEvil | It says that leakage increases quiescant current, and decreases efficiency. |
| 22:38.59 | SpeedEvil | If you have a 100uA input, and your input capacitor leaks 50uA at the input voltage, then what's your maximum input current? |
| 22:39.40 | SpeedEvil | Similarly with the output cap |
| 22:39.43 | SpeedEvil | Also - what are you trying to power it from - there are often better ways |
| 22:39.45 | g1powermac | solar panel |
| 22:40.04 | SpeedEvil | You mean a solar cell? |
| 22:40.26 | g1powermac | yea |
| 22:40.37 | SpeedEvil | What for? |
| 22:41.07 | g1powermac | to charge NiMH batteries for wireless sensor net I'm working on |
| 22:43.48 | SpeedEvil | Why NiMH? |
| 22:43.57 | SpeedEvil | At least - are you using eneloop or similar? |
| 22:44.13 | SpeedEvil | Other ones can have serveral percent a week charge loss. |
| 22:45.06 | SpeedEvil | For NiMH, you can cheat, and just use a panel slightly over their charged voltage, sized so as to not exceed the maximum trickle, and get >80% charge efficiency |
| 22:46.02 | g1powermac | was planning NiMH mainly because I have a batch of them already |
| 22:47.32 | g1powermac | interesting though, so if I have lets say 3.6v of NiMH cells in series, I could go for a 4v solar cell with no controller? |
| 22:52.54 | g1powermac | these are the batteries I'm using: http://www.eneloop.info/fileadmin/web_data/Data-Sheets/HR-3U_2500_.pdf |
| 22:56.24 | SpeedEvil | Basically, yes. |
| 22:56.24 | SpeedEvil | You need a shottky diode |
| 22:56.57 | SpeedEvil | Those ones don't specify a maximum continuous charge current though |
| 22:58.25 | g1powermac | yea, always need a diode |
| 23:06.33 | g1powermac | SpeedEvil, it does say this under the first note on that datasheet: Charge : 250mA ? 16h |
| 23:06.59 | SpeedEvil | That's normal, not trickle |
| 23:07.07 | g1powermac | hmm |
| 23:07.37 | g1powermac | wonder how I could figure out the trickle amount on these things |
| 23:12.58 | g1powermac | according to this page, 1/40th of the mAH is the trickle current for NiMH: http://www.talkingelectronics.com/projects/ChargingNiMH/ChargingNiMH.html |
| 23:16.35 | devilsadvocate | g1powermac: run a sample-and-hold kind of circuit and measure decay time |
| 23:18.52 | devilsadvocate | g1powermac: for that kind of application, i wouldn't worry so much.. Use low ESR, maybe one of those X5R caps or so |
| 23:34.24 | g1powermac | this forum thread seems to pretty much spell it out |
| 23:34.25 | g1powermac | http://forum.allaboutcircuits.com/showthread.php?t=83127 |
| 23:34.36 | g1powermac | seems incredibly simple compared to using that TI chip |
| 23:52.00 | *** join/#edev landley (~landley@70.114.134.56) |
| 23:52.07 | *** mode/#edev [+o landley] by ChanServ |