#qi-hardware IRC log for Friday, 2016-11-11

planasbofftopic, but maybe anyone know how write emulator main loops ?13:00
planasbhttp://stackoverflow.com/questions/40548409/slow-downing-emulator-speed13:01
pcercueidon't13:02
pcercueigenerally emulators synchronize on the audio output or the video output13:03
planasbthanks!13:03
planasbso i will need video dot clock .. I was thinking about. But was searching for easier solution.13:04
pcercueibasically you will run the CPU as fast as possible until one full video frame is rendered13:05
pcercueiit's when you call SDL_Flip() (or similar) that your program will sleep13:06
pcercueiso that you end up with a beautiful 60fps frame rate13:06
eintopfwow - pcercuei is an emulator expert :-)13:30
eintopfI think writing emulators makes fun13:30
pcercueieintopf: well, I've worked on some13:36
eintopfsome gaming console emulator?13:37
pcercueinothing from scratch13:38
pcercueiI'm working on a dynarec for a PSX emulator these days13:38
pcercueihttps://github.com/pcercuei/lightrec13:38
eintopfmhhh13:40
eintopfI know a PSX emulator13:40
eintopfwas very popular13:40
eintopfepsxe13:40
eintopfor something like that13:40
eintopfbut the PSX Bios isn't free yet or?13:40
pcercueiit can be HLE'd13:40
pcercueisimulated13:40
eintopfah, ok.13:42
DocScrutinizer05arossdotme-planb: (( looks like there saying the ltc3895 doesnt need those extra components(?).)) yes, but the point is the plot on left side, which shows spikes to iirc 150V18:44
DocScrutinizer05so that's what a car amp is *supposed* to be able to cope with18:45
arossdotme-planbyea sorry, i had a good at the plot, i see what you mean... 150v! though only for a peak.... still surprising18:47
arossdotme-planb(for me)18:47
arossdotme-planbnot used to the idea for lowish voltage componets being exposed to such high peaks.18:48
arossdotme-planbhad any experience with (not)openups?18:48
arossdotme-planbhttp://www.mini-box.com/OpenUPS18:49
arossdotme-planbtrying to get the dam thing to charge balanced li-ion S418:49
arossdotme-planbthe doc is awful18:49
arossdotme-planbno diagrams18:49
arossdotme-planbfor wiring it up18:49
arossdotme-planbive check again and again and again18:50
arossdotme-planbhttp://resources.mini-box.com/online/PWR-OpenUPS/PWR-OpenUPS-hardware-manual.pdf18:51
arossdotme-planbknow of any usb hid linux nut ups software, plug and play thx to usb hid standard for ups, usb voltage current monitor? for diy computer batt monitoring for diy ups?18:51
arossdotme-planbor a usb serial voltage & current meter that a diy script can read do actions18:53
arossdotme-planbaccordingly18:53
DocScrutinizer05I just have fun with my own (Fortron) cheesy UPS, the batteries are dead from beginning18:53
arossdotme-planbaww18:53
arossdotme-planblead? not cycled for months?18:53
arossdotme-planbhmm i should cycle mine come to think of it18:53
DocScrutinizer05lead gel, recieved a replacement device today, battery low (from 100%) after 68s18:54
arossdotme-planbdarn18:54
arossdotme-planblots of computer watts heh ;)18:54
DocScrutinizer05lead doesn't need cycling18:54
arossdotme-planbi tought if ti was floated for most of its life/months at a time that would kill them?18:55
DocScrutinizer05meh, that UPS is 1500VA/900W, I tested it with 700W and result see above18:55
arossdotme-planb700w heh18:55
arossdotme-planb(load)18:55
DocScrutinizer05that's NiCd18:55
arossdotme-planbyea18:55
DocScrutinizer05Pb want steady float maintenance basically18:56
arossdotme-planbhmm looks like i need to re-read up18:56
DocScrutinizer05what they do NOT like is deep discharge and storage at that, just like LiIon18:56
arossdotme-planbyea18:56
DocScrutinizer05what they also don't like is overcharging (keeping afloat at >14V4) since they lose water then18:57
arossdotme-planbhmm i guess when my prev ups batt got full discharge but not recharged for 12hours untill power came back on i guess that damaged it18:57
DocScrutinizer0512h should be harmless18:58
DocScrutinizer05a month, not so much18:58
arossdotme-planbcus the batts are from the same time but some have not been kept float charging yet the ups batt which was kept floating 24/7 and died a lot last year or 2 while the other batts are still fine..... hmm :/18:59
DocScrutinizer05yeah, weird19:00
DocScrutinizer05I admit I don't know for sure19:00
arossdotme-planbthough i should bear in mind the other batts have loads like 3amp 12 led light vs 500w+ via ups :)19:00
DocScrutinizer05anyway an UPS has no means for scheduled discharge, so...19:00
arossdotme-planbso hard to compared biased on time19:00
arossdotme-planbi though no scheduled discharge was something they skimped on which came at a long term cost.... but ok so they dont need it anyway heh19:01
DocScrutinizer05and I *never* heard of scheduled discharge in e.g. battery for emergency light in cinema, or for subway (yes, they do Pb UPS and that's *damn huge* battery)19:02
arossdotme-planbi guess its a case of silly me19:04
arossdotme-planbdam :|19:04
arossdotme-planbwhat might a serial voltage meter be called? im trying to search on aliexpress19:04
DocScrutinizer05*damn huge* here means  http://wstaw.org/m/2016/11/11/plasma-desktopUy2244.png - the three round lids are for acid refuel19:06
arossdotme-planbflip19:07
arossdotme-planbfound this https://www.aliexpress.com/item/CS5460A-Serial-Single-phase-Bi-directional-Power-Energy-Metering-ADC-Module-AD/32746735855.html?ws_ab_test=searchweb0_0,searchweb201602_0_10037_10091_10090_10032_10088_10089_10077,searchweb201603_1&btsid=c3bbf60f-823d-4194-bc1b-bd9acec46f6719:09
DocScrutinizer05https://www.google.com/maps/@49.4487956,11.0620215,3a,75y,15.59h,103.89t/data=!3m6!1e1!3m4!1sz1_ha_HoCDemUA9zWirhRg!2e0!7i13312!8i6656?hl=de19:14
DocScrutinizer05note that the info about batteries in there is not confirmed19:17
DocScrutinizer05but it's plausible by a number of aspects19:18
arossdotme-planbo hmm19:19
DocScrutinizer05in cinemas however they use e.g. 4* 24V 400Ah looking like car batteries, and nothing in any manual suggests to discharge those periodically during the supposedly 20 years of usage19:20
arossdotme-planbi think ill just sell one this notopenups got it 2rd hand... so its not so bad19:20
arossdotme-planboh,19:20
arossdotme-planbi really need to re educate my self19:21
arossdotme-planbre lead batts19:21
DocScrutinizer05same in alarm center (house alarm) where they use small (12V 3Ah) lead battery for embedded 'UPS' and those of course never must get discharged19:22
DocScrutinizer05aka the must not get discharged ever19:23
arossdotme-planbthere was a emp proof solar controllor/charger that i think discharged the batts a bit come to think of it19:23
DocScrutinizer05since that would open up a vulnerability window for burglars etc19:23
arossdotme-planbwatched a vid about it the other day19:23
arossdotme-planbyea dont want your key batt dischaged when you need it most19:23
arossdotme-planbhmm then again if i instead find a bluetooth volt + amp meter then it can wirelessly connect to laptop so then i only have power cable to laptop19:24
DocScrutinizer05I'm honestly interested where from you got that concept of Pb maintenance needs discharge. A dude at seler told me same (I answered that IF that's needed then the UPS is fucked up ny design since it doesn't allow to do that without risking to shut down hard your IT from power failure, and also it has an incorrect user manual then since it doesn'T tell about that)19:27
DocScrutinizer05at UPS seller*19:28
arossdotme-planbIDK where i got the concept from19:43
arossdotme-planbether was my own thought or i read it somewhere19:43
arossdotme-planbor heard it from someone19:43
arossdotme-planbi would not be supprised if it was the case, of course manufactures will skip features that would prolong the batts cus then they sell more batts19:44
arossdotme-planblike with laptops and how they rechage the batt each time you unplug and replug, even if only been a few mins, i suspect this uses up the cycle life of the battery. i have a artical about li-ion cylces i was about to read in my web browser tab backlog19:45
arossdotme-planb*recharge19:45
arossdotme-planb*cycle count of batt19:46
arossdotme-planbso this is only my suspision, i dont know for sure19:46
DocScrutinizer05http://batteryuniversity.com/learn/article/how_to_restore_and_prolong_lead_acid_batteries  http://batteryuniversity.com/learn/article/corrosion_shedding_and_internal_short  http://batteryuniversity.com/learn/article/water_loss_acid_stratification_and_surface_charge19:49
DocScrutinizer05>>**Always keep lead acid charged. Avoid storage below 2.07V/cell or at a specific gravity level below 1.190.**<<19:50
arossdotme-planbheh19:55
arossdotme-planbi should charge my up some more then....19:55
arossdotme-planbhmm well idk, ill check the cap monitor on it later19:56
DocScrutinizer05http://batteryuniversity.com/learn/article/charging_the_lead_acid_battery19:57
DocScrutinizer05my UPS claims the battery has 27.2V and allegedly 100% SOC now19:59
DocScrutinizer05this makes for a cell voltage of 2.27V20:00
DocScrutinizer05>>The correct setting of the charge voltage limit is critical and ranges from 2.30V to 2.45V per cell. Setting the voltage threshold is a compromise and battery experts refer to this as dancing on the head of a needle. On one hand, the battery wants to be fully charged to get maximum capacity and avoid sulfation on the negative plate; on the other hand, over-saturation by not switching to float charge causes grid corrosion on 20:00
DocScrutinizer05the positive plate. This also leads to gassing and water-loss.<<20:00
DocScrutinizer05hmmm  >>The recommended float voltage of most flooded lead acid batteries is 2.25V to 2.27V/cell.<<20:03
DocScrutinizer05honestly, http://batteryuniversity.com/learn/article/charging_the_lead_acid_battery *HIGHLY* recommended, also to FORTRON engineers20:07
DocScrutinizer05>>Most stationary batteries are kept on float charge and this works reasonably well<<20:08
DocScrutinizer05>>Another method is the hysteresis charge that disconnects the float current when the battery goes to standby mode. The battery is essentially put in storage and is only borrowed from time to time to apply a topping-charge<<20:09
Action: DocScrutinizer05 wishes he could tweak/improve the firmware of that PSU controller20:10
DocScrutinizer05unplugging the UPS from mains causes battery voltage to drop from 27.2 to 25.8 in just 60s, for the maybe 80W of consumption the UPS itseöf eats from battery20:28
DocScrutinizer05lol, plugging in the 700W load to the UPS in battery mode made it shut down completely20:34
DocScrutinizer05tried again and this time it stayed on and powered the load for ~120s until "LOW BATTERY" and then another 45s until I felt with the poor battery at 20.3V and allegedly 11%, so I plugged in mains20:35
DocScrutinizer05now battery charging, it's at 25.2V again, and allegedly 50% SOC20:36
DocScrutinizer05anyway I *think* this is better that first test, so the cells might actually need training to live up to full performance20:37
DocScrutinizer05aka "Formatting", see http://batteryuniversity.com/learn/article/how_to_restore_and_prolong_lead_acid_batteries20:37
DocScrutinizer05next few test I woll "go easy" as suggested there, and discharge the battery from UPS internal usage only, should serve nicely for pretty accurate time values til shut down, in the 3 digit minutes range20:40
DocScrutinizer05and duuude that toxic smell this thing emits for the first days of operation, unbearable20:44
wpwrakthey only gave you that one so that you'd detox it for them ;-))20:47
DocScrutinizer05like the other one? U dunno if they like the nice cig smoke smell better that I replaced then chemical smell with21:04
DocScrutinizer05*I*21:04
DocScrutinizer05that smell actually *tastes* a little minty on the tongue :-o21:05
DocScrutinizer05damn Rigol, slowest X timebase: 50s/div21:06
DocScrutinizer05I'd rather like 50min/div21:07
DocScrutinizer05I could smaple the waveforms and stitch them together though21:13
DocScrutinizer05http://maemo.cloud-7.de/maemo5/usr/local/bin/rigol-waveform21:13
DocScrutinizer05maybe not with 24Msamples memory depth though ;-)21:14
wpwrakat 50 min/div, you use a DVM ;-)21:16
DocScrutinizer05I'd do if I had 2 of them21:16
DocScrutinizer05or a mux21:16
DocScrutinizer05oooh that's an idea, I got that relay cards21:16
DocScrutinizer05so http://maemo.cloud-7.de/maemo5/usr/local/bin/u1233-wip and the relay controller script, frankensteined21:17
DocScrutinizer05but maybe long term the Rigol hack is just more useful21:38
DocScrutinizer05I can set MDEPTH as low as 6000 for 2chan, even 3000 for 4chan21:40
DocScrutinizer05this is a humble 120 samples/s or even 6021:40
DocScrutinizer05it's a pity not being able to exploit the whole 12M depth, but meh!21:41
DocScrutinizer05oops nope, this is mdepth per *screen*, not per div21:42
DocScrutinizer05>> The relationship among the memory depth, sample rate and waveform length satisfies the following equation:21:44
DocScrutinizer05Memory Depth = Sample Rate × Waveform Length21:44
DocScrutinizer05Wherein, the Waveform Length is the product of the horizontal timebase (set by the :TIMebase[:MAIN]:SCALe command) times the number of the horizontal scales (12 for MSO1000Z/DS1000Z). <<21:44
DocScrutinizer05so 10/5 samples per div, aka 0.2/s21:45
DocScrutinizer0510 | 521:46
DocScrutinizer05one sample every 5 or 10s21:47
DocScrutinizer05the discontinuities in the aggregated (stiched) waveform from delay for http://maemo.cloud-7.de/maemo5/usr/local/bin/rigol-waveform every 600s should be negligible21:48
DocScrutinizer05particularly since it only needs 12k download21:49
DocScrutinizer05(4*3k*8bit)21:49
DocScrutinizer05plus everything with nice display and ready-made "mux" (4 chan with matching probe cables)21:50
DocScrutinizer05you notioce how hard I try to get my frustration from a stupid battery-killing UPS, eh?21:51
DocScrutinizer05anyway such setup will be multipurpose and come n hande many times in the future21:54
DocScrutinizer05come in handy*21:54
DocScrutinizer05there are lots of batteries to charge and discharge, and lots of chargers to check21:54
DocScrutinizer05luckily I got a good power supply with Ilim to apply exactly 1.00A over the "shunt" wire I plan to exploit for current sensing21:56
DocScrutinizer05simply the battery connection cable21:56
DocScrutinizer05arrrg, my math sucks. Nevermind, the concept works22:04
DocScrutinizer0512*50s = 600s, 3000Sa/600s = 5Sa/s22:05
DocScrutinizer05just for fun: (12M[bytes mem] / 4[chan]) / 3000Sa/600s = 667 hours total theoretical sampling time, if the complete storage was usable22:09
DocScrutinizer05oops sorry, again. 667/4h22:10
DocScrutinizer05still nice, if it would work22:10
DocScrutinizer05still 167h22:10
DocScrutinizer05hmmmm, "waveform record"22:30
DocScrutinizer05meh, also not working to make any use of 24MB RAM for long term plots23:18
--- Sat Nov 12 201600:00

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