#garfield Logs
Sep 20 2014
#garfield Calendar
00:00 rue_shop3 170kb/s?
00:00 rue_shop3 would that be fast enough?
00:00 rue_shop3 tape drive here with a SG43660
00:01 dumbass iirc, the CBM drives were about that, but the serial cable was 360kbs,
00:01 dumbass i don't remember, i coldn't afford it when it came out, and when i cold afford the technology was better
00:01 rue_shop3 fflippy with a TEAC 7358-00
00:03 rue_shop3 maybe the head preamp of a vcr?
00:04 dumbass maybe, but again, likely very tight voltage input limits
00:05 dumbass magnetics dealt with uv *maximums*, and picoamps
00:06 dumbass you never said, what circuit willy ou put this comparator into?
00:06 dumbass 2Vp-p audio and a tri-wave?
00:07 dumbass or you wanna amplify the pwm output of an avr?
00:07 rue_shop3 comparitor for the switching pwoer supply
00:07 rue_shop3 tho, I feel like I already solved this....
00:07 rue_shop3 beh
00:07 rue_shop3 I'm having fun
00:07 dumbass ok
00:07 rue_shop3 LA7072
00:08 dumbass is it made any more?
00:08 rue_shop3 prolyl not
00:08 rue_shop3 its on a vcr board, in the video section
00:09 rue_shop3 digchip knows its sanyo, but thats it
00:10 dumbass noise canceler
00:10 rue_shop3 oh, LC7431?
00:10 rue_shop3 nothing digchip
00:11 rue_shop3 I think CRT cathode drivers are single ended io, fixed gain?
00:12 dumbass prolly
00:12 dumbass simple transistors
00:12 dumbass LC is sanyo too
00:12 dumbass VHS VTR Chroma Signal Recording / Playback Processor
00:13 rue_shop3 hmm
00:13 dumbass 4mhz vco
00:13 rue_shop3 TA7339?
00:13 rue_shop3 oo vco? which one?
00:14 dumbass LC7431
00:14 mbass happens to have the datasheet o
00:14 rue_shop3 have a sheett for it?
00:14 rue_shop3 !!
00:14 rue_shop3 can you send me that!
00:15 dumbass has a 5.3Mhz vco too, supports PAL, SEACM, and NTSC
00:15 dumbass SECAM
00:15 rue_shop3 :)
00:15 dumbass it's not made any more
00:15 mbass upload
00:16 dumbass want the brief for the LA70xx too?
00:16 rue_shop3 na, not into video processors
00:17 dumbass http://designerthinking,com/LC7431.pdf
00:17 dumbass oops
00:17 dumbass you can fix the url
00:20 rue_shop3 hmm I dont knwo if there is enough data there to use it
00:21 dumbass you don't have the vcr it was in? then you don't have the right parts to solder arouond it either
00:21 rue_shop3 I have the baord...
00:21 dumbass then power it up and use it?
00:21 dumbass or not,, it is a video processor chip
00:22 rue_shop3 I think the amp in this one is M51451AP, but I cant find a sheet for that
00:22 dumbass that sounds familiar
00:23 dumbass no data
00:25 dumbass VC, FM signal processor decoder, Mitsubishi
00:25 dumbass dip-30 ?
00:28 rue_shop3 about that
00:31 rue_shop3 oooh, not that I can read japanese, but LA7007 looks interesting
00:32 dumbass why?
00:33 rue_shop3 lots of amps in it
00:33 rue_shop3 4 of the amps have RF on them
00:35 rue_shop3 arg, I think I need sleep!
00:36 rue_shop3 if nothing else, I'v learned the symbol for a delay line!
00:36 mbass nods sl
00:37 dumbass delay lines made handy multiphase clock signals
00:37 dumbass but so did shift registers
00:38 dumbass dalay lines tended to be sensitive to the loads on each tap
00:38 rue_shop3 oh
00:38 dumbass shift regs needed very high clk signals
00:38 rue_shop3 varrying the delay?
00:38 dumbass 74LS ran out of steem at 30Mhz
00:38 dumbass yeas
00:39 rue_shop3 that could be handy tho
00:39 dumbass not really
00:39 dumbass if you had to put up with that, may as well pick caps an set a trimmer resistor to the exact delay you wanted
00:40 dumbass they were temperature sensitive too, like almost all chips
00:40 dumbass the ones not so sensitive to temperature had no buffers, and the load was critical
00:40 dumbass as you can guess, i didn't use them
00:41 dumbass fascinating concept, but left a lot of bad specs on the data sheets
00:45 rue_mohr2 oh drat, I forgot
00:45 rue_mohr2 I wanted to write that eprom chip today
00:45 mbass hopes she was hel
00:48 dumbass if you want variable delays, look at the 74121, 74122, 74123, 74221, 74423
00:50 rue_mohr2 I think I may order some 746
00:51 rue_mohr2 .5 off the top rail, 3.5 off the ground rail
00:51 rue_mohr2 ouch
00:52 rue_mohr2 that leaves 1V for my pwm signals
00:52 rue_mohr2 I got some fancy op-amps that were only good for 5V
00:52 mbass used a monostable as a motor controller once, used a transistor as a variable current source instead of a resistor to set the pulse w
00:54 dumbass if the back emf, with power off, was too low, the base drive changed and the next pulse was made wider to speed up the motor
00:54 rue_mohr2 oh, mcp6002
00:55 dumbass C64, one monostable, 3 transsitors, made a fantastic motor controller
00:55 rue_mohr2 how do you seperate the drive signal from the back emf?
00:55 rue_mohr2 alwasy wondered that
00:55 dumbass turn off drive, wait a certain time (dependant on the motor), then sample the motor terminal voltage
00:55 dumbass it's pretty accurate
00:56 rue_mohr2 I'v seen transistor circuit and not understood how they do it
00:56 dumbass i could make it stable at 60rmp up to whatever overspeed was, and from no load to some apeman mechanic trying to stall it
00:56 rue_mohr2 oh mcp6002 is rail to rail, but only 1Mhz
00:57 dumbass plan on using a -5 gnd +12, and reducing it as needed
00:57 rue_mohr2 I just dont want to have to stick in all the extra supplies
00:57 dumbass you'll need the +12/+15 to drive the big mosfets anyhow
00:57 rue_mohr2 :) why cant everythign just work on gnd,5V
00:58 dumbass but at the same time, you don't wanna be level shifting signals
00:58 rue_mohr2 its a P channel mosfet, which makes things a bit easier
00:58 rue_mohr2 tho none of my current fet drivers work at 250Khz
00:59 dumbass you'll need beefy bipolars, to-22 heatsunk, at least
00:59 e_mohr2 lioters on the 250Khz bar instead of jumping over the 30Kh
00:59 rue_mohr2 I have a .005R P channel
00:59 dumbass like i said, i was driving *AMPS* of gate current
00:59 rue_mohr2 oh yea, gate capacitance
01:00 dumbass rather made using the mosfets a stupid idea, when the bipolars may have needed 3amps at 1v, the fets needed 3amps at 15v , or 15x the power, *45 watts* of gate drive
01:01 rue_mohr2 I have a problem with the CE drop tho, I dont have much room
01:02 dumbass tip35/36 have very low CE Vdrop
01:02 rue_mohr2 yea, I think I have a tip36
01:02 rue_mohr2 been thru the pile a few times
01:03 dumbass you cannot get very low drop with the mosfets unless you have 12-15v gate drive, so you haveto accept the complications somewhere
01:03 rue_mohr2 but driving a P channel fet off the + rail isn't hard
01:03 rue_mohr2 mess with N channel
01:04 rue_mohr2 I cant use N channel cause the gate drive gets into the power supplies output
01:04 rue_mohr2 :) I'm being stubborn I know
01:04 rue_mohr2 I'll come down to reality at some point
01:05 rue_mohr2 I got another amp
01:05 rue_mohr2 before I got the LM393
01:06 rue_mohr2 LM6482
01:08 dumbass i dunno why, i don't trust cmos amps
01:09 dumbass The LMC6482’s rail-to-rail output swing is guaran-
01:09 dumbass teed for loads down to 600?.
01:10 dumbass yo cannot load it down
01:10 dumbass 1 V/us or slower
01:12 rue_mohr2 24V/us
01:12 mbass goes to put her paw
01:12 rue_mohr2 nothing on the op37
01:13 rue_mohr2 oh no, wait
01:13 dumbass you don't have the data sheet on the OP37?
01:14 rue_mohr2 op37 is 17V/us
01:14 dumbass yeas
01:14 rue_mohr2 ok, the AD is better than what I have then
01:14 dumbass or 5v @ 1/3 us
01:14 rue_mohr2 tho mine are good rail-rail @5V
01:15 dumbass you need rail to rail? i'd just use taller rails
01:15 rue_mohr2 :)
01:15 rue_mohr2 I'm trying to keep it between 0-5V
01:15 dumbass better current drive, nice clamping, etc
01:15 rue_mohr2 yea, prolly better linearity too
01:16 rue_mohr2 the 3842 has a charge pump in the app note
01:16 dumbass yeas, if you are comparing a tri to the audio, you wana enough tri at all time to be above the audio,, sames the signal going to DC, and lets you ignore the dirtiest part of the triwave (the peaks)
01:17 dumbass so if you clip the audio at +/-3v , you want the tri to be +/-4v , so you blown the 0-5v psu already
01:18 dumbass having a couple volts of signal, which you prolly will have anyhow, swamps a lot of noise
01:19 mbass goes to put her paw
01:19 dumbass prolly bbl
01:20 rue_mohr2 woof, rained today, the dialed up the choline in the water
01:20 dumbass huh?
01:20 rue_mohr2 less my audio is really the output of a regulator error amp
01:20 rue_mohr2 that thinks its driving a transistor
01:20 dumbass k
01:21 rue_mohr2 (bench power supply)
01:21 mbass
01:21 rue_mohr2 did you see the pics?
01:21 dumbass "choline" is a word, prolly none in the water tho
01:21 dumbass no
01:21 rue_mohr2 chlorine
01:22 mbass
01:22 rue_mohr2 http://hackaday.io/project/296-Bench-power-supply
01:23 rue_mohr2 I think I can do 250khz pwm thru those 6N137 optos, they are amazing
01:24 rue_mohr2 brainstorm, I should make the power board a seperate module and get it working with a PNP for now
01:25 rue_mohr2 then I can get the computer control and the regulation all happening
01:42 rue_mohr2 katsmeow-afk, thoughts?
01:42 rue_mohr2 the jewel power light was your suggestion iirc
01:42 rue_mohr2 I like it
01:43 rue_mohr2 lights up great with a 8mm led
02:35 katsmeow-afk i remember that project
02:42 rue_mohr2 what are libraries like up there?
11:11 katsmeow-afk rue, the one large bedroom of my house is larger than the adult section of the county library
11:12 tsmeow-afk ponders rue's robots conducting robot cla
11:12 rue_mohr2 ah
11:13 katsmeow-afk this is the south, learning is frowned upon
11:15 rue_mohr2 ok 1 hour till people arrive
11:17 tsmeow-afk imagines a robot "running" out as they arrive, waving it's arms and yelling "Danger! Danger! Dang
11:17 rue_mohr2 heh
11:17 rue_mohr2 "flailing robot!"
11:19 katsmeow-afk interrogating them for their "credentials and authority to be on this post"
11:19 rue_mohr2 http://www.next.gr/uploads/135-10046.png
11:19 rue_mohr2 ever made an osc like that?
11:20 katsmeow-afk not exactly, but a crowbar shunt on a psu can be easily made to oscillate if slightly off
11:20 rue_mohr2 yikes
11:20 katsmeow-afk the headphone impedance would make it osc, almost for sure
11:21 rue_mohr2 inductance required?
11:21 katsmeow-afk i've had them osc, turned out bad for the shunt and the linear regulator
11:21 rue_mohr2 :)
11:21 katsmeow-afk no, bad timing, or a cap in the wrong place
11:22 rue_mohr2 my thoughts on this are that R1 needs to be under the holding current of the SCR,
11:22 rue_mohr2 not sure how R2 is set
11:22 rue_mohr2 I suppose the holding current needs to be over the trigger current for it to work
11:22 rue_mohr2 drains C1 and restarts
11:22 katsmeow-afk the idea of the shunt is it's impossible for an overvolt to happen, but if the shunt current is too low, or the regulator takes too much time to shut off, well, poo happens
11:23 katsmeow-afk no, the timing in that one is the R1-C1
11:23 rue_mohr2 SCRs trigger by current tho..?
11:23 katsmeow-afk putting 9v*DC* on the hedphones isn't the best idea i have ever seen
11:23 rue_mohr2 KA current?
11:24 katsmeow-afk voltage
11:24 rue_mohr2 oh
11:24 rue_mohr2 zener off the top with a GK resistor might be better then, no?
11:25 katsmeow-afk scr is like a pnpn transistor, or 2 regular transistors, you can build a 3amp scr easy with 2 2n3055 or 2n2955 , they have tough bases
11:25 rue_mohr2 yea..
11:25 rue_mohr2 they would
11:25 katsmeow-afk no, you need the hhuge hysteresis of the scr for that to work, you can use a flipflop, or a opamp with some diodes
11:26 rue_mohr2 the idea was a 1 transistor sawtooth osc, without needing a jfet
11:26 rue_mohr2 that said
11:26 rue_mohr2 http://m.eet.com/media/1156683/296389-simple_sawtooth_generator_operates_at_high_frequency_figure_1.gif
11:26 rue_mohr2 that chip is a tiny sop
11:26 katsmeow-afk billions of 1-xsistsor oscs have been built
11:26 rue_mohr2 not by me :)
11:27 rue_mohr2 I did it once, by fluke
11:27 katsmeow-afk such chips are cheap as dirt on ebay too, if you look and bottom feed
11:28 katsmeow-afk note the D2 to add hysteresis, like i said, you can do it with an opamp instead of the '14 invertor
11:28 rue_mohr2 I rodered 100pcs for $9
11:28 katsmeow-afk ouch, i hope s&h was free
11:28 rue_mohr2 yup
11:28 rue_mohr2 its a sot23
11:28 tsmeow-afk
11:29 rue_mohr2 (lifetime supply)
11:29 katsmeow-afk the sorta thing you put under bigger parts and dare people to trace the circuit
11:29 rue_mohr2 :)
11:29 katsmeow-afk "wait.... you put the invertor inside this solder blob?!"
11:30 rue_mohr2 "look, I made this resistor oscillate, all on its own!"
11:30 katsmeow-afk putting such parts inside the pcb was a big deal once, made the pcb almost impossible to repair
11:30 rue_mohr2 I'm also waiting on 100pcs of '164 for $15
11:31 katsmeow-afk you ever see someone using pcb over .062, be suspicious
11:32 katsmeow-afk Tek has a scope that uses internal buried traces of a pcb as the inductor
11:32 katsmeow-afk but some people put their caps inside too
11:32 rue_mohr2 I seen that with some server power modules, they wind the traces and clamp the ferrite around the pcb
11:33 katsmeow-afk 100 pcs 74HC164 sop $11.39
11:33 rue_mohr2 :/
11:33 rue_mohr2 dip?
11:33 rue_mohr2 no, mine are dip
11:33 rue_mohr2 dip is a bit more $
11:34 katsmeow-afk yeas, ~$15
11:34 rue_mohr2 even with the amps I was looking at last night, dip is slightly more
11:36 katsmeow-afk 50 pcs $6.20 ; 100 for $12.xx
11:37 katsmeow-afk <student walks into rue's shop> <robot runs over> are you here to learn about me?
11:40 katsmeow-afk if you had a choice of 3 computers, 1) a 1Ghz cpu , 2) ten 100Mhz cpus 3) a hundred 100Mhz cpus ,,,, and memory, power use, were the same, which would you choose and why?
11:41 katsmeow-afk and cost is the same
11:42 katsmeow-afk oops, #3) a hundred 10Mhz cpus
11:42 katsmeow-afk ========
11:42 katsmeow-afk 1) a 1Ghz cpu
11:42 katsmeow-afk 2) ten 100Mhz cpus
11:42 katsmeow-afk 3) a hundred 10Mhz cpus
11:42 katsmeow-afk ========
11:49 katsmeow-afk [02:26] <kat> Tiggr knows some BM too :-)
11:49 katsmeow-afk [02:26] <eGG> lemme test her
11:49 katsmeow-afk [02:26] <eGG> kamu nak pergi mana, Tiggr?
11:49 katsmeow-afk [02:26] <kat> not a lot!!
11:49 katsmeow-afk [02:26] <eGG> LoL
11:49 katsmeow-afk [02:27] <eGG> tiggr pandai
11:49 katsmeow-afk [02:27] <[Tiggr]> i am a snart bot!!
11:49 katsmeow-afk [02:27] * [Tiggr] holds up a memory chip to show,, umm,, i forgot...
11:49 katsmeow-afk [02:27] * [Tiggr] hurrily plugs it back in. 10:10-5D
11:50 katsmeow-afk [02:27] <eGG> oh god
11:50 katsmeow-afk [02:27] <eGG> kewll
11:50 katsmeow-afk [02:27] <kat> snart - typo of smart
11:50 katsmeow-afk [02:27] <eGG> ah :D
11:50 katsmeow-afk [02:27] <kat> Tiggr cannot type, lol
11:50 katsmeow-afk [02:28] <eGG> yeah, its cute if she cant ;)
11:50 katsmeow-afk [02:30] <kat> i should goto bed
11:50 katsmeow-afk [02:31] <eGG> okies
11:50 katsmeow-afk [02:31] <eGG> take care kat
11:50 katsmeow-afk [02:31] <eGG> and good nite
11:50 katsmeow-afk [02:31] <kat> thanks :-)
11:50 katsmeow-afk [02:31] <eGG> dun fall down while in the dream :)
11:50 katsmeow-afk [02:31] <eGG> pleasure
11:50 katsmeow-afk [02:31] <kat> selamat petang :-)
11:50 katsmeow-afk [02:32] <eGG> oh thx, it is petang!
11:50 katsmeow-afk [02:32] *** Disconnected
11:51 katsmeow-afk [23:08] <AJ> tiggr bodoh atau pandai?
11:51 katsmeow-afk [23:08] <[TiggrBot]> i *am* snart! 4 =(°, ¿, °)=
11:51 katsmeow-afk [23:08] <robsz1> woohoo
11:51 katsmeow-afk [23:08] * [TiggrBot] holds up a memory chip to show,, umm,, i forgot...
11:51 katsmeow-afk [23:08] * [TiggrBot] hurrily plugs it back in. 3 87-5]
11:51 katsmeow-afk [23:08] <AJ> ahhha
11:53 katsmeow-afk logs from 7 years ago, back when i was smart and useful too
11:55 katsmeow-afk before humans got smart phone, when they still hated the idea of a smart computer
11:56 katsmeow-afk it's like Tiggr went from being hated to being obsolete, no middle ground of acceptance and usefullness
12:30 katsmeow-afk According to the study, the team not only created a waterproof set of tactile sensors, but were able to successfully build these sensors into a large robotic hand that can be attached to an undersea vehicle. The sensors are able to track changes in pressure as the vehicle descends into the deep ocean, allowing it to correct for what would otherwise interfere with tactile sensing.
12:30 katsmeow-afk tested the hand at a pressure equivalent to about four miles under water, and showed that the hand was still about 90 percent accurate at identifying a mug, a toy shark, and a chess piece.
12:31 katsmeow-afk http://www.natureworldnews.com/articles/9076/20140917/robo-hand-repairs-explores-deep-sea.htm
12:32 tsmeow-afk goes to find some breakfast before it's too late to eat lunch
13:40 katsmeow-afk http://cdn.rt.com/files/news/2e/11/80/00/cat.si.jpg
16:30 rue_mohr2 in reality it was an oil drum, a real shark, and the iffle tower :)
16:31 rue_mohr2 I'v never seen a lime tiger cat
16:31 rue_mohr2 before
16:32 rue_mohr2 it kinda has a "what the hell are you gonna do to me now" look on its face
17:21 katsmeow-afk replaced the leaking hose up under the turbocharger
17:22 katsmeow-afk let the engine warm up till the fan cycles on-off, no leaks, cept where i broke the overflow off the radiator neck
17:24 katsmeow-afk it warms up fast enough now i think i am not in a hurry to put the thermostat back in, just sometime this week,, after this weekend
17:31 katsmeow-afk i am thinking putting a small 120vac window fan in the center console, moving the existing console face towards me so it's all more reachable, and stick a 12x14 radiator in there too
17:32 katsmeow-afk i temp put such a fan there a while back, and finally had 100% of the windshield fog-free, as well as some of the side windows,, it was really windy
17:43 katsmeow-afk oh, i was wrong about the car's mileage, it's 194k
17:44 katsmeow-afk http://www.limestrong.com/images/cat_lime04.jpg
17:44 katsmeow-afk http://www.limestrong.com/images/cat_lime03.jpg
17:45 katsmeow-afk http://www.limestrong.com/pets.htm
20:01 katsmeow-afk well, the broken overflow is blowing onto the distributor cap
20:19 katsmeow-afk full-body spanks for astronauts, with heat-triggered memory wire for extra support
20:21 katsmeow-afk Lecturer Caught On Secrete Camera Doing It With A Female Student...
20:21 katsmeow-afk Take a look at this video and see how a lecturer was digging it with a female student who is suppose be his daughter, and as for the girls, please learn to focus on your studies and look up to God for His guidance.
20:28 katsmeow-afk how to use only first name on facebook: connect thru an indonesian proxy
20:28 katsmeow-afk Select Bahasa Indonesia langauge on facebook
20:29 katsmeow-afk Now click on your name i.e the edit link in front of it.
20:29 katsmeow-afk Bingo! you are done, now change language back to
20:29 katsmeow-afk english…
20:42 Tom_itx rue_house does ghost work on linux ok? 320GB -> 500GB drive?
20:43 tsmeow-afk wonders what Tom is up to nowa
20:43 Tom_itx little to nothing
20:44 Tom_itx or less
20:44 katsmeow-afk looking for something to do?
20:46 Tom_itx not really
20:47 katsmeow-afk soemthing little, praps, interesting in a way, that no one else is doing?
20:49 Tom_itx nope, pretty much given up doing anything nowdays
20:49 katsmeow-afk why?
20:51 tsmeow-afk thinks Tom has skills the world needs desperately, even if the world is too stupid to kno
20:56 tsmeow-afk will throw the idea out there anyhow..
20:58 katsmeow-afk how about a electromagnet, like a small relay coil, rigged to open a car tire valve, and what if the coil was shrunk down so small it operated on 6v, pulled 2amps, and would overheat if energised for 5 seconds,, so you put the whole coil in the hydraulic manifold, all the coils, where the hydraulics would cool them
20:59 katsmeow-afk i think the whole deal could be made so small, with little material, light weight, fast, high pressure or low, inexpensive once you license them out, and roboticists would adopt them like mad
21:00 katsmeow-afk i think they could be made so small, people would complain the hose barbs are too close together
21:07 katsmeow-afk "look, on ebay, 40 Tomvalves for $50!"
21:08 katsmeow-afk (Tom makes 10% on each sale without lifting a finger,per terms of the license)
21:12 katsmeow-afk ok, how cheap can you make a manifold hoding 4 conventional ball valves, or 4 springed ball bearing check valves?
21:12 rue_shop3 I need to make an eprom programmer that has lots of 4066 in it
21:13 katsmeow-afk "crosspoint switch"
21:13 rue_shop3 so many pins used for so many stupid thigns
21:13 katsmeow-afk or just instructions: red wire on this pin, blue wire on that pin, green wire on the other pin, now press "program"
21:16 katsmeow-afk http://www.ebay.com/itm/Pneumatic-Ball-Valve-Push-In-Fittings-Connectors-for-Air-Water-Hose-Tube-4-12mm-/351076457039 <<== you see, once you have made the first ones, the licensees will make them in plastics
21:17 katsmeow-afk by ganging them up, you reduce the number of couplings needed
21:18 rue_shop3 hmm
21:18 katsmeow-afk metal ones for a manifold mounting : http://www.ebay.com/itm/Air-Angle-Flow-Speed-Control-Valve-Connector-Tube-Hose-Pneumatic-Push-In-Fitting/400701146425
21:23 rue_shop3 ok kat, the 2764 is written
21:23 rue_shop3 have an operators manual for it? :)
21:23 katsmeow-afk for an eprom?
21:24 rue_shop3 for an eprom loaded with bitwise sinewave pwm
21:25 katsmeow-afk umm, no
21:30 rue_shop3 2764, how many address lines
21:31 katsmeow-afk 16,, 8 for dram
21:31 tsmeow-afk looks i
21:31 katsmeow-afk no
21:31 katsmeow-afk 2764 is an 8kx8
21:32 katsmeow-afk 13 addy lines
21:32 katsmeow-afk dip28
21:33 Tom_itx why does SD go read ony after copying only a few files in linux?
21:34 tsmeow-afk does not
21:36 rue_shop3 Tom_itx, because the card is faulty and there are errors
21:36 rue_shop3 linux resorts to going read only after the errors to protect the drive
21:37 rue_shop3 Tom_itx, delete and repartition the SD, reformat it, and try agian
21:37 rue_shop3 *trust me*
21:37 rue_shop3 if that dosn't work, ditch the drive
21:38 Tom_itx i did all that already
21:38 Tom_itx it's a new SD
21:38 katsmeow-afk already ditched the drive?
21:39 katsmeow-afk possibly some format of accessing it is wrong
21:40 rue_shop3 ok, dont trust me, apparently I dont know what I'm talking about
21:40 Tom_itx i recall you telling about that a while back
21:41 Tom_itx i set it aside and installed a spare hdd instead
21:43 theidiot4723923 ok I need a 17 bit counter
21:43 Tom_itx how do i copy a folder and all it's subs from terminal?
21:43 Tom_itx need root to copy the files or i'd use the gui
21:43 theidiot4723923 cp -rv src dest
21:43 Tom_itx onto another drive?
21:44 Tom_itx how do i address the 2nd drive?
21:44 theidiot4723923 the SD is prolly terribly corrupt and all the data is messed, if you copy it to a working drive, dont expect it to work properly when it gets there
21:44 theidiot4723923 mkdir /mnt2
21:45 theidiot4723923 mount /dev/whatever /mnt2
21:45 Tom_itx it's mounted already
21:45 Tom_itx how can i list the devices?
21:45 theidiot4723923 copy the data from where one drive is mounted to the other one
21:45 theidiot4723923 you have to mount all drives you want to copy files to and or from
21:45 katsmeow-afk 17 bit counter? 8kx8 = 64 bits = 16 addy lines , you can feed 13 to the 2764 and 3 to a 8:1 selector
21:49 Tom_itx they're both mounted
21:49 theidiot4723923 kat yes, I'm on that, I need a 17 bit counter
21:49 theidiot4723923 Tom_itx, so copy the data....
21:49 Tom_itx how can i list the devices?
21:59 Tom_itx what a fucking kludge
22:07 theidiot4723923 you can see the kernel find the devices in /var/log/syslog
22:07 theidiot4723923 sorry, more visitors
22:08 theidiot4723923 is it a device you can attach and detach?
22:08 theidiot4723923 I dont know what your doing, and I dont know if you have already given up
22:08 theidiot4723923 katsmeow-afk, is there a 16 bit counter in the 74 series?
22:09 theidiot4723923 it takes me 3 14 pin chips just to make the counter for this
22:09 Tom_itx i got it
22:09 Tom_itx had to call it some obscure name like 232l;k4hk2jh342k3jh42;;2l3kh4
22:10 Tom_itx half mile long but i got it
22:13 theidiot4723923 /mnt/something
22:13 e_shop stret
22:20 katsmeow-afk no, there's 8bits tho
22:20 katsmeow-afk 74269 = 8 bits
22:21 katsmeow-afk use two of them and a 7474 for the 17th bit
22:23 katsmeow-afk or 74579
22:25 katsmeow-afk got 19 pins on an avr you aren't using? 1) clock, 1) reset, 17) addy
22:28 katsmeow-afk most avr have 16bit counters, adding another bit with software or a 7474 shouldn't be too hard
22:33 rue_shop the avr cant count fast enough
22:35 katsmeow-afk if speed is an issue, the 579 is slower than the 269
22:36 katsmeow-afk http://www.ebay.com/sch/i.html?_from=R40&_trksid=p2050601.m570.l1313.TR0.TRC0.H0.X74+269&_nkw=74+269&_sacat=12576
22:38 Tom_itx rue_shop, can i install linux on one pc and transfer the drive to another one without too much trouble?
22:39 Tom_itx oh nevermind... the OS won't work on that pc anyway.
22:45 rue_shop Tom_itx, yes,
22:45 rue_shop linux dosn't care what hardware it boots on
22:46 rue_shop its not like windows where its installed specifically for the hardware available during the install
22:46 rue_shop kat
22:46 katsmeow-afk rue
22:46 rue_shop uh, whats my pwm clock rate supposed to be
22:46 rue_shop I'm at max on the freq counter and its not enough
22:46 katsmeow-afk the clk rate is what you designed the sample speed around
22:46 rue_shop 8 bit pwm, 256 samples per half wave
22:47 katsmeow-afk per half wave at what freq?
22:47 rue_shop 256 rom bits per pwm cycle
22:47 rue_shop 256 pwm cycles per half wave
22:47 katsmeow-afk per half wave at what freq?
22:48 katsmeow-afk if you are going to say 60hz, then you need ~31khz
22:49 katsmeow-afk at 8x that, ~248khz
22:49 katsmeow-afk depending on how you are suing the bits
22:50 katsmeow-afk if you say 60hz output, then if your clk is over 250khz you have done something wrong, because you're gonna run outa bits
22:52 rue_shop it looks like its in the order of 8Mhz
22:52 rue_shop so, output is 60Hz
22:53 rue_shop 1/60 is 16.67ms
22:53 tsmeow-afk bog
22:53 rue_shop so the half cycle is 8.33ms
22:53 rue_shop that 8.33ms is divided down into 256 pwm pulses
22:54 rue_shop which is .03ms
22:54 rue_shop that .03ms is divided into 256 pwm time slots
22:54 katsmeow-afk i am simply not seeing how you are using the bits to do pwm
22:54 rue_shop 256 bits per pwm cycle
22:54 katsmeow-afk pictures
22:54 rue_shop 256 pwm cycles per half wave
22:54 rue_shop k
22:54 katsmeow-afk are you making a 60hz signal now?
22:55 rue_shop I have an 8Mhz clock on it, and its close
22:55 katsmeow-afk oscope?
22:55 rue_shop yea
22:55 katsmeow-afk pics ;-)
22:55 rue_shop according to this calc, I need 7864320 hz
22:56 katsmeow-afk that's awesomely high to make 60hz
22:56 katsmeow-afk you have a 131,072 divide ratio
22:58 rue_shop its 8 bit pwm, with 256 pwm cycles per half wave...
22:59 katsmeow-afk for whatever reason, i think i understand, but i cannot picture it
23:00 katsmeow-afk i don't see a mistake in what you say, but it isn't clicking in my head
23:00 katsmeow-afk damnit, i hope i am not becoming normal
23:02 rue_shop I cant find any crystals close to 7.8Mhz
23:02 katsmeow-afk CB crystals /3
23:03 katsmeow-afk 14.xxxx secam crystals
23:03 rue_shop I'm thinking maybe 30M/4 is 7.5Mhz
23:03 katsmeow-afk got a 30mhz?
23:03 rue_shop maybe I need to set up a LC osc?
23:04 katsmeow-afk ok, make a crystal clk, then use it to phase lock the LC osc
23:05 rue_shop 33.33!
23:05 katsmeow-afk what do you have close to the desired 7.8?
23:05 rue_shop :/ too high
23:05 rue_shop I have 8
23:05 rue_shop 7.3
23:06 rue_shop 32Mhz
23:06 rue_shop heh
23:06 katsmeow-afk 8 / 32768 ?
23:08 katsmeow-afk 8 / 133,333 = wtf is wrong with my decimal points........
23:08 katsmeow-afk 8 / 133,333 = 60.000150000375000937502343755859 hz
23:09 katsmeow-afk 8 / 133,334 = 59.999700001499992500037499812501
23:09 katsmeow-afk i believe you can detune the xtal to cover that error
23:10 katsmeow-afk you can phase lock the 60 to a clock chip osc, and use that to tweak the divider from 133,333 to 133,334 and back once an hour
23:13 katsmeow-afk dividing the 8Mhz xtal by 133,333 will make it 1 sec fast every 6,666.65 seconds, or 1.85 hours
23:14 rue_shop you designed this ya know :)
23:14 tsmeow-afk tilts her head so rue can look into her
23:16 katsmeow-afk tweak the xtal, a few pf of trimmer cap shouod do it, 7,999,980 hz would be perfect to make time markers for a 60hz output
23:17 katsmeow-afk but with such a huge divider of 133,333 , you can toggle to ,334 every 2hrs or so, and 99.99999% of humans would be very impressed by it's accuracy
23:17 katsmeow-afk you can work out the math to get even more accurate
23:21 katsmeow-afk every 111 minutes, toggle tween 133,333-133,334-133,333-133,334- etc ,, but your freq reference will need to be linked to WWV or whatever time sync you have in Canadia,, you will be stupidly accurate
23:23 katsmeow-afk such accuracy will cost a small fortune, just use 133,333 and the 8Mhz xtal
23:23 katsmeow-afk you can pick off intervals in the 133,333 divider counters to toggle your pwm intervals, etc
23:25 rue_shop its just for a UPS
23:25 rue_shop :)
23:27 rue_shop 8Mhz is 61Hz
23:27 katsmeow-afk hmm, IRF830 is prolly not a good choice for a 100amp H-bridge switcher making 120vac off a 200vdc psu,, even tho it might be cheap
23:28 katsmeow-afk 8 / 133,333 = 60.000150000375000937502343755859 hz
23:28 katsmeow-afk do the math
23:29 rue_mohr2 I used the freq counter
23:29 rue_mohr2 maybe the crystal is otu
23:29 katsmeow-afk or the counter is, it's xtal controlled too
23:29 rue_mohr2 thanks, kat, this is really cool
23:30 katsmeow-afk PICS!!!
23:30 rue_mohr2 its really close, I'v tested it
23:30 rue_mohr2 comming
23:30 rue_mohr2 uploading now
23:30 rue_mohr2 the sine wave looks great
23:30 katsmeow-afk qty 100 of IRF830 = $10 , but even 100 of them in parallel is 0.015 ohms, or 1.5v drop at 100 amps
23:31 katsmeow-afk 100 of them spread out can use an awefully cheap plain heatsink tho
23:32 katsmeow-afk each one would dissipate only 1.5 watts
23:33 katsmeow-afk i guess that's not all that bad, throwing 300watts in the h-bridge to deliver 12,000 watts to the load ?
23:33 katsmeow-afk should i try to do better?
23:34 rue_mohr2 hmm
23:34 rue_mohr2 what about IGBT
23:34 katsmeow-afk possibly 2v drop per leg
23:34 rue_mohr2 100A*2V = 200W
23:34 katsmeow-afk i have a large pile of 800v 30amp igbt
23:34 rue_mohr2 **assuming 2V drop
23:34 katsmeow-afk 200w *per leg*
23:35 katsmeow-afk 2 legs on at all times in an H
23:36 katsmeow-afk and it will take less of them to make 100amps, but that will concentrate the heat, needing pricier heatsinks, instead of just a aluminum bar for the irf830's
23:36 rue_mohr2 http://ruemohr.org/~ircjunk/circuits/pwmsine/p1060734.jpg
23:37 rue_mohr2 there is no steering, across a split transformer, so its rectified
23:37 rue_mohr2 http://ruemohr.org/~ircjunk/circuits/pwmsine/p1060736.jpg
23:37 rue_mohr2 und my filter sucked
23:37 rue_mohr2 http://ruemohr.org/~ircjunk/circuits/pwmsine/p1060737.jpg
23:38 rue_mohr2 :)
23:38 rue_mohr2 your design rocks
23:38 tsmeow-afk blu
23:38 rue_mohr2 here I was gonna use an avr to pump values into a serial controlled pwm chip
23:39 rue_mohr2 the ONLY think I can think of it that it dosn't account for votlage drop, but from what I can tell, the only way most ups' do that is with relays that switch between different transformer taps
23:39 katsmeow-afk wait a sec,, is the scope pic off the rc filter, or off the transformer?
23:39 rue_mohr2 rc
23:40 rue_mohr2 there is no transformer yet
23:40 rue_mohr2 I have to work out if a modified sine inverter can be driven at pwm speeds
23:40 katsmeow-afk can i see the pwm pulses on one trace, and the sine on the other trace?
23:40 rue_mohr2 oh
23:40 rue_mohr2 ok
23:42 katsmeow-afk actually, you can boost/cut the Vdrop under load with pulse stretching,, 5% should be trivial to do, specially since you do have both the pwm and the sine there already
23:45 katsmeow-afk where your --|>-- buffers are driving the fets, insert a couple more buffers, cupled by R and maybe a C, then use anotehr resistor to pull that line up/down to stretch/shrink what the fet sees at the output of the final buffer/driver
23:45 katsmeow-afk prolly need to use cmos buffers, not ttl
23:46 rue_mohr2 not an easy photo
23:47 tsmeow-afk
23:48 katsmeow-afk i see where i was stumbling, your 8-bit pwm slots exist only in the 8kx8 chip, you are feeding a steady bit stream to the dac, it's not timeslotted there
23:48 rue_mohr2 http://ruemohr.org/~ircjunk/circuits/pwmsine/p1060739.jpg
23:48 rue_mohr2 the scope took it strange
23:49 rue_mohr2 by dac you mean the rc circuit?
23:49 rue_mohr2 its pre-dac'd
23:49 katsmeow-afk your dac to show the sine right now is the RC
23:49 rue_mohr2 yea, so its a bit stream thats pwm sine wave :)
23:50 rue_mohr2 I love that (it was your idea)
23:50 katsmeow-afk why is the pwm trace on the bottom showing droop at the top of the pulses?
23:50 rue_mohr2 I'm gradually starting to see the light of layered encoding
23:50 rue_mohr2 I dont know, there is something odd with it, might be the digital being mauled by the rc filter
23:51 rue_mohr2 I dont understand the fuzz
23:51 rue_mohr2 it should be... narrower...
23:51 rue_mohr2 (25Mhz scope)
23:51 katsmeow-afk try looking at any data out pin on the 2764
23:51 rue_mohr2 if I zoom way in, its good pwm
23:51 katsmeow-afk before the '151, should be cleaner and sharper and ttl levels, no droop
23:52 rue_mohr2 the exciting part is that it works
23:52 katsmeow-afk wanna hear what a perfectionist wold say?
23:52 rue_mohr2 I wish I didn't need 3 14 pin counter chips
23:52 rue_mohr2 do tell
23:53 rue_mohr2 ( I didn't even calc the filter, I just hocked it (there was a 470 laying on the bench and it took .22uF to see the wave)
23:54 katsmeow-afk the perfectionist wold say you need a latch tween the 2764 and the 151, slaved to the clock, to stabilise and hold the 2764 bits steady while the clock is changing the couters
23:54 rue_mohr2 its a ripple counter, too
23:54 katsmeow-afk arrgg
23:54 rue_mohr2 right now a 16 bit ripple counter running at 8Mhz
23:54 rue_mohr2 :)
23:55 rue_mohr2 I have some 74161
23:55 rue_mohr2 it would take 5 of them
23:56 katsmeow-afk what happens if you (power it down first), pull the bottom 5 addy lines off the 2764, and gnd them? leave the top 8 bits off the counters where they are
23:56 rue_mohr2 it would take less space to make a state machine out of a 16 bit eprom
23:57 rue_mohr2 it would drop the sample resolution
23:57 rue_mohr2 to 3 bits?
23:57 rue_mohr2 ...?
23:57 katsmeow-afk going to 8bits should give you 1/2 volt resolution, which is still pretty darned good
23:57 rue_mohr2 8 bit works...
23:58 katsmeow-afk no, there's 13 lines to the 2764, drop the bottom 5 lines to gnd, leave the top 8 (A5-A13?) where they are
23:58 katsmeow-afk 8 lines will reduce your counter to one chip
23:59 katsmeow-afk one 74xxx269
23:59 katsmeow-afk that's 1/2 volt steps vs the full 13bits, and if you reprogram to counter that, it will be as accuarte as it is now