IRC log for #edev on 20071127

01:02.47*** join/#edev Belxjander (n=Belxjand@203.173.153.238)
01:02.59*** part/#edev Belxjander (n=Belxjand@203.173.153.238)
01:21.33*** join/#edev prpplague (n=dave123_@ppp-70-244-175-101.dsl.rcsntx.swbell.net)
01:22.16prpplagueho ho ho
01:22.20prpplaguemerry freakin monday
01:28.00MonMothaisn't it a bit late for that?
01:38.23*** join/#edev prpplague^2 (n=dave123_@ppp-70-244-175-101.dsl.rcsntx.swbell.net)
01:45.42agaffneyMonMotha: nah, it's still monday :P
01:56.42*** join/#edev prpplague (n=dave123_@ppp-70-244-175-101.dsl.rcsntx.swbell.net)
02:12.34*** join/#edev davidc__ (n=davidc__@S01060002b360aacd.vc.shawcable.net)
03:03.04MonMothaI guess...
03:03.12MonMothahas anybody ever used the "place your bid" thing at pcbnet.com?
03:03.36MonMothaI need some cheap one-off PCBs and I don't care how long it takes
03:08.24davidc__nope, never used em
03:08.29davidc__how cheap / what specs?
03:09.14MonMothaI should be able to make do with 8/8mil line/space 2 layers, but the boards are huge by necessity and I would really like gold fingers
03:09.33MonMothalike 10x8" huge
03:09.43MonMothaok, maybe more like 8x6, but still very large
03:09.56MonMothaalso have a couple of unusually large holes (like 500mil) in them
03:10.11davidc__pci somethingorother?
03:10.58MonMothaneo geo cart :)
03:11.28davidc__hah
03:11.29MonMothawhich means there are two, count them, two boards
03:11.52davidc__might try talking to a local fab
03:12.04davidc__they might be willing to do the job given unlimited leadtime
03:12.08MonMothaI'd like to be able to get 5-10 each of both boards for no more than $15 per board, which I know is really tough in low quantity
03:12.13MonMothaI don't know any "local" fabs
03:12.35davidc__ya, that'd be really hard
03:12.40davidc__especially if you want e-test
03:12.45MonMothaI don't care about e-test :)
03:12.51davidc__might be able to get someone to do it if you skip the silkscreen / soldermask
03:12.55MonMothawith 2 layers, I can perform a visual inspection
03:13.01MonMothaI need soldermask, but silk is optional
03:13.14MonMothathere are some 0.5mm QFPs that are going on it, and I can't solder those without a soldermask
03:13.28MonMothaand silk seems to come cheap if you have a soldermask
03:13.33davidc__ouch.. yeah, those need mask
03:14.33MonMothathe best I've been able to do so far is possibly cramming them both onto a single multi-design panel and running it on the Advance Circuits $33/ea deal
03:14.44MonMothathat's $33/board + $50 for the lot to do multi-design
03:15.10MonMothaI can hack them apart then and have $16.50/board plus the lot charge spread out over it all
03:15.23MonMothathat assumes I can get both of them down to 60 sq. in. and still have them hold tight in the casing
03:15.53MonMothathat also doesn't get me gold fingers
03:16.09davidc__MonMotha: well, good luck.. let me know if you find a place that'll take ya up on it
03:16.19davidc__the masks alone for that large a board won't be cheap
03:16.28davidc__you might be best trying to find someone in china to do it
03:16.53MonMothaI thought about submitting it to the gold phoenix proto deal
03:17.31MonMothathe gold fingers adds $$$ to that though
03:17.48MonMothaI'm hoping to be able to take advantage of the "virtually unlimited" lead time
03:18.42davidc__MonMotha: eh, email em directly
03:18.51davidc__they're probably open to negotiation
03:19.01davidc__as then they can just do the fingers whenever they next need to do gold
03:19.10MonMothaperhaps
03:19.44MonMothaI've not actually done business with them, but every time I've been in contact they've been eager to accommodate
03:20.17MonMothaI could also email my sales rep from advanced circuits.  I've probably done 5 proto orders through them just on personal projects, let alone business stuff
03:20.54MonMothathey also owe me after shipping my boards 7 days late (albeit over the holiday and with something weird happening that delayed them further)
03:21.24davidc__heh... I like using a local board house.. I just drive out and pickup the boards in person :P
03:23.48MonMothaI'm not aware of any around here
03:25.05davidc__MonMotha: yeah.. they can be hard to find
03:27.01MonMothayeah
03:27.05MonMothahappen to know any in Fort Wayne, IN?
03:28.54davidc__nope
03:28.59davidc__only one I know of is in Delta, BC
03:30.24MonMothaI'm not even aware of any in Indiana
03:49.41*** join/#edev linac (n=lin@125.77.71.73)
04:40.33*** join/#edev DrZimmerman (n=theo@2001:5c0:8fff:fffe:0:0:0:6af1)
04:45.53*** join/#edev dijenerate_ (n=dijenera@72.22.144.160)
04:52.34*** join/#edev joe_bleau (n=joe_blea@adsl-65-67-67-84.dsl.fyvlar.swbell.net)
05:33.09*** join/#edev redb1ue (i=star@ppp045.108-253-207.mtl.mt.videotron.ca)
07:11.10*** join/#edev howtoo (n=howtoo@195.159.62.66)
09:31.53*** join/#edev hw (n=hw@p578b3905.dip0.t-ipconnect.de)
09:56.19*** join/#edev lyakh (n=lyakh@dynamic-unidsl-85-197-26-24.westend.de)
10:59.04*** join/#edev cathyal (n=cathy@unaffiliated/cathyal)
10:59.10cathyal:D
12:20.14*** join/#edev key2 (n=Ritual@193.33.36.20)
12:47.57*** join/#edev Off_Namuh (i=GPS@gateway/tor/x-166fa22374897b2a)
13:32.45*** join/#edev Off_Namuh (i=GPS@gateway/tor/x-5907d9c5e050bf2f)
13:59.55*** join/#edev GPSFan (n=kenm@12.10.255.246)
13:59.55*** mode/#edev [+o GPSFan] by ChanServ
14:02.51*** join/#edev T0mW (n=Tom@24.238.86.158.res-cmts.sth.ptd.net)
14:31.13*** join/#edev prpplague (n=dave@mail.americanmicrosystems.com)
14:44.42*** join/#edev chouimat|work (n=mchouina@64.26.181.242)
14:52.57prpplaguehehe funny, just after we announced the hammer, gumstix and triton both put out press releases about new products
15:01.07*** join/#edev linac (n=lin@121.41.55.59)
15:01.35chouimat|workmorning
15:02.07chouimat|workprpplague: http://wulffmorgenthaler.com/strip.aspx?id=656c268e-ff70-43a3-9e52-b4923c42e1f2 <--- worksafe
15:06.39Crofton|homewhat new product?
15:09.26prpplagueCrofton|home: we went live with the hammer board last week
15:10.32prpplaguechouimat|work: indeed that is exactly how most americans see the world
15:10.44prpplaguechouimat|work: which is a shame since it used to not be that way
15:11.33prpplaguechouimat|work: my kids are treated as odd balls at school(they are sorta, but thats beside the point) since they know all about other places outside of the US
15:12.20prpplagueCrofton|home: linux devices has an article up on a SDR for the gumstix
15:13.43*** join/#edev chouimat (n=mchouina@64.26.181.242)
15:16.24Crofton|homeI saw that demo
15:16.43Crofton|homeThey are good guys
15:16.51Crofton|homeI told them we would do a smaller one next year
15:17.01Crofton|homebut the size reduction would come in the RF
15:17.10Crofton|homethey used an old box they had lying around
15:19.45prpplagueCrofton|home: interesting
15:20.01prpplagueCrofton|home: what kind of mips power do you need to do a sdr?
15:20.27Croftondepends entirely on what waveform you are porcessing
15:20.37CroftonI suspect hammer could do NB FM OK
15:20.49Croftonthe CRC demo did not use the wmmx stuff
15:20.58Croftonbut we all want to use that sort of thing
15:22.28prpplagueCrofton: NB FM? narrow band?
15:22.40Croftonyeah
15:22.43prpplagueahh
15:22.51Croftonbasically, depends on the sample rate
15:23.13Croftonthey converted RF to near base band so they could sample with the sound card
15:23.51prpplagueCrofton: ahh interesting
15:24.16prpplagueCrofton: so like using a standard ac97 codec ?
15:24.53Croftonnot sure, you want want linear response through the DA conversion
15:25.02CroftonI am not great at sound systems
15:28.21prpplagueahh
15:28.37prpplagueCrofton: would be nice to see if the hammer could do a small SDR
15:28.56Croftondo you have sound in cpapility?
15:29.14prpplagueCrofton: not directly, but it could be added
15:29.26prpplagueCrofton: more likely the sledge hammer would be a better choice for an sdr
15:29.27Croftonethernet?
15:29.31prpplagueCrofton: yea
15:29.51Croftonthe ethernet solutions have a much higher sampler ate though
15:29.53chouimatI hate myself ... got something working last night but I forgot how I did it
15:30.04Croftonand I have no RF to sound card front end
15:30.22Croftonthe useful one
15:30.57prpplagueCrofton: ethernet solution? interesting
15:31.08prpplagueCrofton: you got a url with a summary of that?
15:31.13Croftonusrp
15:31.18Croftonwell usrp2
15:31.36Croftonor usrp attached to somethign with usb high speed and passing data over ethernet
15:32.17prpplagueCrofton: has to be high speed?
15:32.29Croftonyes
15:32.31Crofton:(
15:32.46prpplaguedoh
15:32.46Croftonbbl
15:32.51prpplagueCrofton: okie dokie
15:50.31*** join/#edev CosmicPenguin (n=nobody@163.181.251.103)
15:50.31*** mode/#edev [+o CosmicPenguin] by ChanServ
16:03.16key2hey
16:04.19key2prpplague: do you think that with the s3c2410  I could read 640KByte/s ?
16:04.44prpplaguekey2: depends on the interface
16:04.58key2the s3C2410 would be connected to an FPGA
16:05.07key2the databus and the addr bus are 32bits wide
16:05.07prpplaguekey2: depends on the interface
16:05.14key2prpplague: what do u call the interface
16:05.15key2?
16:05.34prpplaguespi/i2c/usb/sdio/bitbang/smbus
16:05.53key2bitbang
16:06.00key2well it's 32 wires for data
16:06.04key2and 32 wires for addresses
16:06.20adyerhow much data per interrupt?
16:06.27*** join/#edev ico2 (n=ico2@modem-3032.chameleon.dialup.pol.co.uk)
16:06.32prpplaguekey2: if you are doing address/data you'd just use the standard address bus
16:06.42key2prpplague: yeah thats the idea
16:07.28key2prpplague: but from that bus, 640Kb of data doesnt seem to be too big ?
16:07.33key2per sec
16:07.39prpplagueno
16:08.11key2thx
16:11.22*** join/#edev Off_Namuh (i=GPS@gateway/tor/x-134bbb611433d53c)
16:24.00MothaWorkgod damn RoHS
16:24.06T0mWLOL
16:24.16MothaWorkas much as I applaud the effort, it makes things a royal PITA for not really much gain
16:24.42MothaWorkthe lead free stuff is silly
16:24.52MothaWorknow, restricting the other stuff like Cadmium I can understand
16:24.59T0mWI am saying the same thing about the RTCVDD supply of 1.8v to keep RTC alive during powerdown.  Just how are you supposed to come up with a 1.65..1.95 battery?
16:25.01MothaWorkbut there's a darn good reason solder has had lead in it all these years
16:25.31MothaWorkLDO off a 2.5-3V?
16:25.38MothaWorkwhich is silly of course
16:25.45adyerT0mW low current so a LDO shouldn't hurt much
16:25.51T0mWyeah, right, at 30% efficiency
16:26.13adyerre: RoHS it's done, it wont go away, so you just have to deal with it
16:26.23MothaWorkadyer: the current is usually in fact so low that Iq becomes a significant factor
16:26.25*** join/#edev bbradley (n=bbradley@82-32-35-42.cable.ubr05.azte.blueyonder.co.uk)
16:26.53MothaWorkadyer: somehow I anticipate the lead free stuff going away in about 10 years when electronics start spontaneously killing themselves due to tin whiskers
16:27.22MothaWorkthe lead in solder is such a minute contribution to the nasitness of your average electronic good to begin with
16:27.28MothaWorkand it serves a very important purpose
16:27.52MothaWorkyes, there are disposal concerns, but there are disposal concerns even without leaded solder
16:27.57adyerMotha I doubt it.  Somebody will figure that one out in a fewe years
16:28.08MothaWorkthey've had like 10 years so far and nobody's solved it yet
16:28.29MothaWorknow, packaging should get better so at least the thermal issues during reflow assembly will hopefully get worked out
16:28.43adyerlol, it hasn't been jammed down everyone's throat til now, that the only way it'll get fixed
16:28.44MothaWorkbut nobody seems even close to solving the tin whisker problem yet
16:29.35MothaWorks/everyone/europeans/
16:29.43MothaWorksee, in theory I don't have to deal with this
16:30.07adyerand Chinese and Californians, and ...
16:30.09MothaWorkhowever, leave it to the french to come up with a system of measurement and name it "Le Systeme Internacional"
16:30.19MothaWorkchina rohs is not quite a strict
16:30.29adyerstill stricter than nothing
16:30.43adyerwould you have them name it in esperanto? :-)
16:30.44MothaWorkand of course you can still buy cars in the US that are emissions certified in 49 states
16:31.27MothaWorkthe issue isn't one of intent, the issue is one of regulators being unaware of the state of technology
16:31.37MothaWorklead free processing of circuit boards DOES cause problems
16:31.43adyerlast I heard French was still the language of diplomacy, and a standards efoort is nothing if not diplomacy in action
16:31.55T0mWadyer: 5ua of osc current, LDO is not going to work.
16:32.07MothaWork5uA?  jeesh...that's low
16:32.11MothaWorkI don't know any LDOs with Iq that low
16:32.24adyershunt regulator?
16:32.45Crofton|homeno wonder diplomacy has a bad name ....
16:32.49T0mWno, sorry, 50ua
16:33.21T0mWI guess this calls for a supercap
16:34.13MothaWorkhum, 50uA...
16:34.28MothaWorkthat's at least to the point where the regulator Iq is 50% of the current draw :)
16:34.35T0mWyeah
16:34.51MothaWorkhum, I wonder if that product is released yet
16:35.02MothaWorkwhen I worked at ON, we had this uber low Iq regulator they were touting
16:35.12MothaWorkbut it was still in development, but that was >1yr ago
16:35.19T0mWthat is why I was wondering WTF Samsung engineers were thinking of by using a 1.8v clock circuit
16:35.52prpplagueT0mW: whats that?
16:36.07T0mWprpplague: the RTC for the s3c2410
16:36.17prpplagueT0mW: ahh
16:36.48MothaWorkhum, may not have been available in 1.8V
16:36.55T0mWMonMotha: well, maybe a 2.0v LIoN cell with a series diode?
16:37.19T0mWdunno what the drop would be at that low current though
16:37.27MothaWorkT0mW: finding one that will have the voltage drop you need (reliably across all samples) at that low currentw ould be tough
16:37.39T0mWheh, maybe a 0.2v zenor diode?
16:37.57prpplagueT0mW: what is it you are looking at?
16:38.01MothaWorkah, there we go
16:38.06MothaWork1uA Iq LDO
16:38.06T0mWprpplague: ?
16:38.16T0mWMonMotha: where!?
16:38.18MothaWorkT0mW: check out the ON Semi NCP583
16:38.22T0mWthanks!
16:38.27prpplagueT0mW: the rtc for the s3c2410 works fine with a standard ldo to 1.8v
16:38.38MothaWorkeven avaialble in a reasonable SC-82AB package
16:39.03MothaWorkas well as teeny tiny SOT-563 (good luck hand soldering that...)
16:39.07T0mWWOW "Ultra-low Iq..."
16:39.14MothaWorkyeah, they aren't kidding
16:39.25MothaWorkthe other specs aren't that great, but Iq is LOW
16:40.06MothaWork1/4V dropout too, so you're good down to 2V
16:40.16MothaWorkassuming you can handle 1.75V (it probably can)
16:40.26T0mWMonMotha: how did you find that?
16:41.13MothaWorkT0mW: I know how to use ON's parametrics :)
16:41.22MothaWorkthey sadly don't have an Iq parametric
16:41.44MothaWorkbut I knew they had something like that, so I just asked for LDOs, single output, 1.8V, and scrolled through looking at descriptions
16:41.48adyermaybe max1725/6
16:41.52T0mWMonMotha: heh
16:42.38MothaWorkhum, 2uA Iq, not bad
16:42.47MothaWorkthough I know some people report sourcing issues with Maxim parts in large qty
16:43.08MothaWorkON's part is also like half the cost
16:43.09adyeroh yeah, maxim sucks as a supplier, but they make interesting parts
16:43.18MothaWorkoh, agreed
16:43.24T0mWMonMotha: always, I stay away from maxim unless nobody else has an equivilant
16:43.26MothaWorkthey do have some very neat little purpose-specific parts
16:44.10T0mWI'm using some of their iso-rs232 stuff, MAX3250, so far I've been able to get them via DigiKey
16:44.54MothaWorkT0mW: TI doesn't clone those?
16:45.05MothaWorkI know TI clone's a good chunk of Maxim's RS-232 line, but maybe not the iso stuff
16:45.10MothaWork*clones
16:45.30T0mWmaybe they do, but Maxim has delivered so far.
16:46.04T0mWTI may have picked that item up and put into production in the past two years
16:49.44*** join/#edev Crofton (n=balister@hc6521721.dhcp.vt.edu)
16:55.45T0mWMonMotha: heh, the part legend for the PCB is almost the same size as the SC-82A package
16:56.16T0mWs/part/silk screen
17:42.23MothaWorkT0mW: Yeah.  Most of my boards I end up turning the legend text size way down
17:42.40*** join/#edev ico2 (n=ico2@modem-148.cougar.dialup.pol.co.uk)
17:42.41T0mW40mil or less
17:42.44MothaWorkwhen lining up 0603 passives, 75mil text is too tall
17:42.53MothaWorkyeah, I usually do 35mil high text
17:43.16MothaWorkI haven't had to go to 0402s yet.  I try to avoid anything smaller than 0603 simply because I can't reliably hand solder them
17:43.32MothaWorksame with SOT-563s.  Can't do em
17:45.49MothaWorkI like my package sampler from ON.  They apparently were unable to solder down the SOT-963s and SOT-723s
17:46.40*** join/#edev ralphi (n=ralphi@122.168.24.246)
17:47.07ralphihello every1!!!
17:48.30prpplagueralphi: greetings
17:49.51ralphihi prpplague
17:49.53MothaWorkT0mW: I'm still trying to figure out how I blew up this resistor yesterday
17:50.10MothaWorkit popped audibly due to factors other than heat
17:50.21*** join/#edev diJenerate (n=dijenera@72.22.144.160)
17:50.30ralphican any1 suggest a software to crack rar file format
17:51.11prpplagueralphi: uh NO
17:51.13MothaWorkin here? don't count on it...
17:51.43ralphifrom where can i get 1??
17:51.53prpplagueralphi: not here
17:52.21T0mWMonMotha: prolly cracked it an it arc'ed over?
17:52.49prpplagueralphi: try Fnet or dalnet irc, i'm sure they have tons of channels dealing with that
17:52.53*** join/#edev MothaWork (n=monmotha@host-69-95-169-16.cwon.choiceone.net)
17:53.16chouimat|workprpplague: I found 3 with google ;)
17:53.27prpplaguechouimat|work: interesting
17:53.42*** join/#edev MothaWork (n=monmotha@149.164.193.61)
17:53.44*** join/#edev MothaWork_ (n=monmotha@149.164.193.61)
17:53.59MothaWorkmy god our internet connection here sucks ass
17:54.01ralphiwell i m kindda new here ...so i m not geetin....how can i get into fnet channel
17:55.13ralphihello????
17:55.31ralphisomebody...reply
17:56.52prpplaguefnet is an irc server
17:57.05cathyalanyone played with symbian
17:57.38*** join/#edev chouimat (n=mchouina@64.26.181.250)
17:58.45prpplagueralphi: http://www.efnet.org/
18:08.35T0mWMonMotha: that's grounds for leaving for another cubicle at another company!
18:08.47prpplaguehehe
18:09.07T0mW"screw the raise, just get me bandwidth!"
18:09.14T0mWheh
18:29.45T0mWprpplague: ping
18:29.53prpplagueT0mW: pong
18:29.56T0mWprpplague: on the ENC28j60
18:30.01prpplagueyes?
18:30.17T0mWprpplague: would you say that most of the current consumption is done in the transformer?
18:30.44*** join/#edev davidc__ (n=davidc__@S01060002b360aacd.vc.shawcable.net)
18:30.46prpplagueT0mW: yea, if you turn off the PHY the current drops like a rock
18:30.49prpplaguedavidc__: greetings
18:30.54prpplaguedavidc__: are you alive?
18:31.22T0mWprpplague: I am thinking of crash-dump detection of the input supply.  When the power is removed, have the kernel flush its buffers out to the SD card filesystem.
18:31.27davidc__prpplague: yup
18:31.29davidc__prpplague: not for terribly long though
18:31.50prpplaguedavidc__: hey, i've been trying to get in touch with someone at SFE about the hammer
18:32.00T0mWprpplague: I am thinking of putting the ethernet transformer on a separate 3v supply isolated from the main s3c2410 supply. Sound reasonable?
18:32.00prpplaguedavidc__: don't seem to be getting any response
18:32.18prpplaguehmm
18:32.36davidc__prpplague: who'd you email?
18:32.47davidc__prpplague: the best contact would be spark@sparkfun.com
18:32.51prpplaguedavidc__: tried spark and pete
18:32.54T0mWprpplague: so, the enc28j60 would still be powered but the xformer would not.  I don't see a problem with that do you?
18:33.10prpplagueT0mW: sounds logical
18:33.17T0mWnod
18:33.18prpplagueT0mW: might want to actually try it first
18:33.37T0mWprpplague: WHAT!? AND MISS A SMOKE TEST?
18:33.38davidc__prpplague: hrm... well, the guy that decides new products is on vacation right now, should be back in a week or so.. might be why you haven't received a response from spark
18:33.46prpplagueT0mW: you could alwasy just shut down the PHY
18:33.49davidc__prpplague: and pete is not known for replying to emails ;)
18:33.57prpplaguedavidc__: ahh ok
18:34.06prpplaguedavidc__: well we went live with the hammer last week finally
18:34.13T0mWprpplague: PHY meaning the magnetic, right?
18:34.17davidc__prpplague: heh, cool :P Just as I went back home
18:34.22prpplaguedavidc__: you back in the land of the maple leaf?
18:34.26davidc__prpplague: yup
18:34.47prpplagueT0mW: no, the enc28j60 has a seperate section for the phy
18:34.58prpplagueT0mW: essentially the enc28j60 is three chips
18:35.08T0mWoh
18:35.10prpplagueT0mW: a spi controller, a MAC, and a PHY
18:35.17adyerPHY usually means the interface to the magnetics, line drivers, receivers, clock recovery, etc
18:35.34prpplagueT0mW: you can issue a command via the spi interface to turn of the phy
18:35.57T0mWum
18:36.03T0mWsure
18:37.41T0mWI want "instant off" as I don't know how much time I will have before the 3.3v regulator finally drops out (hard reset), it may take some time for the driver to get access to the enc28j60 registers to command it off.
18:38.16T0mWthat's why I was thinking of a separate power supply to the beast sucking at all the current
18:39.58prpplaguedavidc__: http://www.tincantools.com
18:40.15prpplaguedavidc__: think they are still interested in resale'n the hammer?
18:40.46davidc__prpplague: could be.. they were interested when I mentioned it to them
18:40.54davidc__its a bit of an odd beast though, gobs of power, not much IO :P
18:41.23prpplaguedavidc__: ??
18:41.52davidc__prpplague: well, not much IO in terms of only 36 something IO pins ya know :P
18:42.06prpplaguedavidc__: compared to what?
18:42.23davidc__prpplague: oh, a header board or something like that
18:42.36davidc__prpplague: but anyhow - do talk to them. They were interested in taking a look
18:42.59prpplaguedavidc__: ok, i'll try again, but i'm curious as about your statement
18:43.28prpplaguedavidc__: i mean with usb you can add just about anything you want including port expanders
18:43.55davidc__prpplague: oh yeah - no, the IO thing isn't really a concern for most people :P
18:44.19*** join/#edev chouimat (n=mchouina@64.26.181.250)
18:44.20prpplaguedavidc__: "gobs of power" ?
18:44.28davidc__eh, its a very powerful board :P
18:44.33prpplaguedavidc__: didn't think the high end of 120mA was that large
18:44.40davidc__no, not power consumption
18:44.44davidc__I meant number crunching :)
18:44.47prpplagueoh ok
18:45.02prpplaguedavidc__: gotcha thanks for the feedback
18:45.08davidc__the 120ma is actually really nice
18:45.14davidc__didn't know it was that low
18:45.16prpplaguedavidc__: got a J-O-B yet?
18:45.24prpplaguedavidc__: actually thats the high end
18:45.26davidc__prpplague: not at the moment
18:45.31prpplaguedavidc__: low end is about 50mA
18:45.50davidc__prpplague: doing some contracting work right now.. but after that, dunno... might look for work, might do a starup, dunno
18:45.57prpplaguedavidc__: depends on what items you have configured
18:46.08prpplaguedavidc__: hehe, fun fun
18:46.19prpplaguedavidc__: *cough* design some hammer board stuff?
18:47.07davidc__prpplague: hah, maybe
18:47.14davidc__prpplague: got a wishlist of things that need to be made?
18:47.27prpplaguedavidc__: a whole list of stuff
18:47.59prpplaguedavidc__: most of the items are things i'm not familiar with, such as robotics
18:48.05davidc__prpplague: link? :P... anyhow, for the moment
18:48.26davidc__er, for the moment, I can't really afford to buy a bunch of stuff to hack on it
18:48.29prpplaguedavidc__: yea, i'll get some stuff together and email you
18:48.41davidc__but maybe in the future sometime when I'm working / get a big consulting payment ;)
18:48.44*** mode/#edev [+o prpplague] by ChanServ
19:09.49*** join/#edev davidc__ (n=davidc__@S01060002b360aacd.vc.shawcable.net)
19:21.00prpplagueT0mW: i don;t see any reason you couldn't cut the 3.3v to the phy
19:21.38T0mWyeah
19:21.50prpplagueT0mW: but also there is no reason you couldn't trigger an interrupt when a power loss occures
19:22.02prpplagueT0mW: if it doesn't shut things down
19:22.07prpplagues/doesn't/does
19:22.23prpplagueT0mW: thats what we do on our handhelds
19:22.30T0mWprpplague: microchip support was pretty response when I had the prob with the mis-wiring on the transformer.  I will check with them as well.
19:22.48prpplagueT0mW: good deal, let me know, i'm curious as well
19:23.01T0mWprpplague: I was afraid that if I cut power to the PHY, then the SPI would lock awaiting PHY powerup.
19:23.35prpplagueT0mW: no, the PHY is seperate from what i can tell
19:23.48T0mWprpplague: this way, I figure to cut power to the transformer, where all the current is, then the PHY can still operate but have nothing to push against.
19:23.54prpplagueT0mW: when the enc28j60 powers up after a reset, the phy isn't even turned on
19:24.40T0mWIs see that Pin 15 is the TX power on the PHY, I'll ask if I should also drop power to that pin as well.
19:24.46prpplagueT0mW: yea, i'm betting that the phy is drawing alot of power too
19:25.00T0mWI don't card about an orderly shutdown of the ethernet, just let it dies.
19:25.04T0mWs/dies/die
19:25.17prpplagueyea
19:25.38T0mWless hassle, less time if I just allow it to die and stay busy flushing the SD card.
19:25.38prpplagueproblem is that if you are running linux you could have a que problem
19:26.26T0mWprpplague: I could also block the ethernet driver when the power-down interrupt trips
19:26.41T0mWif that were to be a prob
19:26.45prpplagueT0mW: yea
19:27.38prpplagueT0mW: hehe put the whole enc28j60 section on a seperate power supply controlled with a gpio
19:28.12prpplagueT0mW: if a power loss interrupt occures bring the interface down, and cut the power
19:29.54T0mWprpplague: already thought of that, but then I would need a power FET on the card
19:31.49T0mWprpplague: input to the main +3v3 regulator is isolated from the main supply via a 1.5Amp diode with a 220uf cap.  I need to see if that 220uf needs to be larger or not, but the idea is that when powerfails, I get an interrupt, the +3v3 regulator bleeds the 200uf cap down while the system saves the SD card buffers.
19:32.32prpplagueT0mW: oh so you won;t be running on a battery backup?
19:32.57prpplagueT0mW: strictly just a cap?
19:33.23T0mWprpplague: power is applied to the unit externally.  There is a battery backup, but what happens if some jerk just flips the power switch on the unit?
19:33.57T0mWprpplague: this is not the hammer
19:34.00prpplagueT0mW: oh you mean something like an apc?
19:34.11prpplagueT0mW: upc backup?
19:34.28T0mWprpplague: yes, the unit I'm designing for has its own UPS
19:34.36T0mWthe UPS is external
19:34.41prpplagueahh
19:35.16prpplagueT0mW: this design going to be running linux?
19:35.47T0mWthe power supply in the unit is a DC/DC converter which either takes the external battery (UPS) or rectified AC (transformer driven) as the DC input (24volt).
19:36.31*** join/#edev MothaWork_ (n=monmotha@149.164.193.61)
19:36.32T0mWprpplague: yeah, this is that SPI card rack unit I was working on a year ago, then got sidetracked with all his other crap (8051 stuff).
19:36.50prpplagueahh
19:36.56prpplagueT0mW: s3c2410 based?
19:37.02T0mWprpplague: for now
19:37.15prpplaguehmm
19:37.34T0mWprpplague: we will see how well it performs, then maybe switch to a PXA270 or better if need be.
19:37.44prpplaguehow long do you figure before you loss power after disconnection from the mains?
19:37.57T0mWprpplague: about a second
19:38.03prpplaguenot much time
19:38.42T0mWprpplague: that is why I need only the s3c2410 subsystem powered at that time.  I have an on-board 3v3 supply on the CPU card.
19:38.44prpplagueT0mW: i wouldn't worry about the enc28j60 driver inside linux
19:39.06prpplagueT0mW: on a disconnect from main, i'd just drop power to the whole enc28j60 section
19:39.28T0mWprpplague: i just don't want to end up with an error on the SD card filesystem. Yes, I can run e2fsck, but...
19:39.55prpplagueT0mW: yea
19:39.57T0mWprpplague: correct, that is what I'm thinking as well.  Power off the entire section.
19:40.35prpplagueT0mW: hehe could always use one of those 10F caps
19:41.24T0mWhave: +3v3 and +VCCIO (alternate 3v3 supply) when powerfails, the +VCCIO would dropout, then +3v3 still runs the SDcard and s3c2410
19:41.56prpplagueyea
19:41.57T0mWprpplague: if they are small enough, you know, that just might be cheaper too.
19:41.58prpplaguehttp://www.sparkfun.com/commerce/product_info.php?products_id=746
19:42.19T0mWI've seen 4Farad caps too.
19:43.32*** join/#edev ico2_ (n=ico2@modem-4074.chimpanzee.dialup.pol.co.uk)
19:44.03T0mWOOOOO!  360Farad is the largest value!
19:44.13prpplagueT0mW: hmm, wonder how long two of those in series would run at 200mA
19:44.15T0mW:P <pant>
19:44.27T0mWprpplague: hour or two?
19:44.28T0mWheh
19:44.46prpplagueseriously?
19:45.02T0mWprpplague: charge 'em up to a couple of hundred volts and toss them to your co-workers.
19:45.09T0mWLOL
19:45.16T0mWI need a break!
19:45.23prpplagueT0mW: oh i think we are talking two different ones
19:45.31prpplagueT0mW: i'm refering to the 10F 2.5v ones
19:45.38T0mWyeah
19:45.40T0mWwe are
19:45.48prpplagueT0mW: if i did two in series, that'd be 5F/5V
19:46.15T0mWprpplague: My dad used to tell me that he would do that when he was working at rebuilding magnetos.  Charge up some caps and toss them to people
19:46.27prpplaguehehe
19:46.32T0mWprpplague: T=RC
19:47.00T0mWprpplague: I'd have to look it up too
19:47.30prpplagueT is what?
19:48.27T0mWTime
19:48.52T0mWexponential discharge curve
19:49.12prpplagueCapacitance * Voltage = I current
19:49.18prpplaguehmm
19:49.36prpplague5F * 5V = 25A ?
19:50.25*** join/#edev Tokemi (n=kvirc@5352D6FD.cable.casema.nl)
19:50.49T0mWyou would have to look at it from a power perspective, voltage would be declining, but, yes, you are on the right track, amps is power (factorially).
19:51.02T0mWprpplague: let's ask MonMotha, he went to college!
19:51.08prpplaguehehe
19:51.12T0mWheh
19:51.14prpplagueT0mW: 25A is a bunch
19:51.22T0mWnice smoke
19:52.40T0mWouch, $8 for a 10Farad supercap
19:53.31T0mWhmmm, 1 Farad at $4
19:53.34prpplagueT0mW: just 5 at sparkfun
19:53.51prpplaguehttp://www.sparkfun.com/commerce/product_info.php?products_id=746
19:54.13T0mWprpplague: oh yeah, I can just see me putting "sparkfun" as the vendor into the partslist.  Get a grip
19:54.25T0mW:?
19:54.44T0mWprpplague: I know, JameCo
19:54.46prpplagueT0mW: hehe, just means they can be had cheap
19:55.07prpplagueT0mW: if sfe is sale'n them for $5 you know bought them for alot less
19:57.25cajakahttp://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=P13764-ND $4.04 usd
19:58.23T0mWah well, I need 6.3v rated
19:59.05T0mW+5v supply, Underwriters Labs won't allow anything to run in excess of 70% ratings
20:00.58prpplagueT0mW: not even in series?
20:01.06prpplagueT0mW: i.e. 3 in series?
20:02.37T0mWwell, a 220uf 10v electrolytic is only 10cents, if I put 5 of them on the board, then I would have a Farad.
20:02.57T0mWder, no
20:03.08prpplagueT0mW: hehe
20:03.20T0mWgot my millis and micros mixed up
20:03.46T0mW0.47Farad would be 470,000ufd, right?
20:03.59T0mWyeah
20:05.21prpplagueT0mW: hmm if i'm drawing 500mA at +5v, that would give me 6Farads
20:05.34prpplaguesorry .1F
20:06.26prpplaguenow, a Farad is measure in seconds, so if i wanted to know what i needed for an hour
20:06.35prpplaguei would assume i multiply by 360
20:06.44T0mWLOL
20:06.48prpplaguehehe
20:06.51T0mWprpplague: you are out of control!
20:06.54prpplaguei doubt that is correct
20:06.58prpplaguebut i'm curious
20:07.06prpplagueseriously though
20:07.22prpplaguewe were considering using a small lit-poly battery as a backup
20:07.25prpplaguethis might work
20:07.29T0mWyeah, I know, I need food first, mind is slowing down.
20:07.34prpplaguewe jsut need a few minutes
20:08.38cajakaI had a yellow led emitting enought light to read by for 40 mins using 5F and 5v for a school project if you want something to compare to for time
20:08.53T0mWI have to check to see if the internal supply will have a powerfail indication.  Best thing would be for me to get a early indication that the +5v is dying.  The earlier I know, the better to do an orderly shutdown.
20:09.13T0mWcajaka: heh
20:09.51prpplaguecajaka: hmm, +3.3v led
20:09.56prpplaguecajaka: i assume
20:10.01prpplaguecajaka: and about what 20mA?
20:10.28cajakaIt was something between 15-20ma
20:10.40prpplagueyea
20:11.13cajakaAfk for a sec
20:14.33T0mWMax. Current (25oC)(3) > 3.1 A (3) 1 sec discharge rate to 1/2 VR
20:14.50T0mW3Farad out really work out
20:15.16T0mWs/out/would
20:17.13prpplaguehmm
20:17.20Tokemiis there anyone here who experienced problems(recursive dependency LIBXML) with the latest buildroot?
20:17.32prpplaguenot me
20:18.44Tokemiokay
20:18.45cajakaOk couldn't find it. Actually I think it was a 2.7v led, I remember debating over weither to use the caps in parallel or series
20:41.54adyerlatest buildroot is failing fro me beuilding freeetype (T*&^(&(*& buildroot)
20:43.15*** join/#edev dijenerate (n=dijenera@72.22.144.160)
20:43.26*** join/#edev davidc__ (n=davidc__@S01060002b360aacd.vc.shawcable.net)
20:44.05adyerand buildroot dropbear fails during make clean
20:44.23adyerand buildroot tries to fetch libungif from the wrong place now
20:44.49Tokemihmm, is there a place to turn to other then IRC to get help with buildroot that you know of?
20:45.31CosmicPenguinbuildroot, like all buildsystems, has a tendency to fall into bitrot when it isn't closely maintained
20:45.38CosmicPenguinI'm sure nobody will say no to a patch
20:45.44adyermostly I just muddle around until I fix it.  sometimes there's useful stuff on the mailing list
20:46.28adyerit just pisses me off that every time i do an svn update it seems like more stuff is broken.
20:47.34Tokemiyeah i can imagine that
20:48.18adyerthere's a lot of commit activity, but it seems like fix one, break two
20:49.15adyerit's only saving grace is it's not too complicated to see what's going on.
20:50.27Tokemihmm, well i have no previous experience with buildroot, its just that i wanted to try to make a image for a Edimax router
20:51.08adyeredimax == evilmax
20:51.46Tokemihaha so ive heard
20:51.59*** join/#edev MothaWork_ (n=monmotha@host-69-95-169-16.cwon.choiceone.net)
20:52.24*** join/#edev MothaWork_ (n=monmotha@149.164.193.61)
20:52.54adyerblew up one of my dev systems with one of their PCI cards that shorted rails in a 3.3V PCI system
20:52.56Tokemibut did it get that name because of the soft- or hardware?
20:53.40Tokemiooh..that answers my question
20:57.53Tokemithanks for sharing youre experiences with buildroot though, ill just try to see if i can find anything on the mailinglist or try to fix it myself
21:20.52MothaWork_T0mW: what are you looking at mulit-farad caps for?
21:25.17adyerI think he wanted to provide a 'holdup' supply for saving some data when power went down
21:25.40T0mWMonMotha: just to keep a processor circuit alive after powerfail until I could flush the SD card buffers.
21:44.30*** join/#edev ico2_ (n=ico2@modem-1909.cheetah.dialup.pol.co.uk)
22:03.56*** join/#edev Crofton|work (n=balister@ossie.wireless.vt.edu)
22:05.13*** join/#edev linac (n=lin@121.41.37.236)
22:17.27*** join/#edev Off_Namuh (i=GPS@gateway/tor/x-1b592c909fe54f11)
22:21.34*** join/#edev Crofton|work_ (n=balister@ossie.wireless.vt.edu)
22:33.08MothaWorkT0mW: hum...
22:33.49MothaWorkI had thought about doing such a thing on a Z-80 based system using FeRAM as the main system memory just using a ISR on LVI to flush the internal processor state, but I'd never considered it on a full DRAM based system
22:34.20T0mWMonMotha: well, I am going to be running linux
22:35.53MothaWorkthat pretty much precludes a Z-80, certainly, but not FeRAM or MRAM
22:36.04adyerwe did a card for a server with sdram, fpga, PCI, supercaps and NAND flash like this...
22:50.18*** join/#edev Crofton|work_ (n=balister@ossie.wireless.vt.edu)
22:52.01*** join/#edev davidc__ (n=davidc__@S01060002b360aacd.vc.shawcable.net)
22:56.05prpplagueCrofton|work_: you alive?
22:57.40flyback<PROTECTED>
22:57.40flyback[17:54] <flyback> I smashed my car into the side of the garage
22:57.40flyback[17:54] <flyback> I am just sleeping all night the hell with this
22:57.46flybackpaxil withdrawl sucks ass
22:57.58flybackprpplague how are sales?
22:58.31prpplagueflyback: please refrain from using that kind of language here
22:58.59prpplagueflyback: need i remind you that this is a professional channel and irc from work is a privilage not a right
22:59.06prpplagueflyback: about what we expected
22:59.14flybackok
22:59.18prpplagueflyback: we don't expect sales to jump for atleast a few months
22:59.27flybackwell I think it will :P
23:01.15flybackI doubt I will be around to see it but oh well
23:01.19flybackgood luck
23:01.51Crofton|homeprpplague, yeah
23:02.02Crofton|homeI got tangles up with opensrt this afternoon
23:02.18prpplagueCrofton|home: np i had some tangles as well
23:02.23Crofton|homeheh
23:02.42prpplagueCrofton|home: i had a look at the hardware ursb
23:03.26Crofton|homefx2?
23:03.26prpplagueCrofton|home: looks like it would be pretty darn easy to replace the cypress chip with the hammer
23:03.26prpplagueCrofton|home: yea
23:03.27prpplagueCrofton|home: got a few minutes for me clarify how the system works?
23:03.38Crofton|homenot tonight :)
23:03.45Crofton|homeafriend is coming over
23:03.50prpplagueokie dokie
23:03.52prpplaguenp
23:04.06Crofton|homelikely not a short discussion
23:04.16Crofton|homedo you have a need for a SDR?
23:04.34jbevrenprpplague: ping
23:04.39prpplaguejbevren: pong
23:04.49jbevrenso Ive found a simpler connection for the sid heh
23:04.54jbevrenI was reading on ucapps.de
23:05.05jbevrenthey connect the sound chip via SPI using a pair of shift registers :)
23:05.10prpplaguenice
23:05.12jbevrenthings are simpler. hehe
23:05.14*** join/#edev Crofton (n=balister@66.207.66.26)
23:05.19jbevrenits a hackjob but imho its pretty elegant
23:05.32jbevrenyou shift 16 bits out. 8 are data, 5 are address, and one more to reset the chip
23:07.33flybackdamn
23:07.36flybackthat is slick
23:07.46flybackisn't that too slow though for sound?
23:08.45jbevrenthis is a synthesis chip, not an audio dac ;)
23:12.05prpplaguejbevren: so is this going to be interfacing to a real SID or a fpga version?
23:13.14jbevrenI have a small box with several real sids in it that are from c64 repairs and harvesting
23:13.20jbevrenso I'll be using a genuine part
23:13.42prpplagueproblem is that it won't be replicatable unless you have original SID chips
23:13.45jbevrento date, none of the fpga-based implementations are sufficient, but then I own a real c64 so I know what it should sound like
23:13.53prpplaguejbevren: ahh
23:14.10jbevrenI'm also aware of the differences between the two major revisions of the sid
23:14.20jbevrenand I know of the changes over the years on the first revision :)
23:14.51prpplagueahh
23:15.32jbevrenas an example, the 8580 sid requires a different vdd (9v vs 12v), and has a different filter structure thereby requiring different filter caps
23:15.41jbevrenits also a better-designed filter :)
23:15.58jbevrenthe 6581 filter has the interesting ability to overdrive itself, creating an internal fuzz effect
23:16.10flybackjbevren
23:16.23flybackwhat about that chick's c64 sic
23:16.26jbevrenthe 8580 by contrast cannot, which gives you more depth in filtering :)
23:16.27flybackshe's damn sexy too
23:16.32jbevrenJeri's?
23:16.37jbevrenit doesnt have filters.
23:16.40jbevren;)
23:16.41flybackah
23:16.49prpplaguejbevren: yea what about that, doesn't it have something SID compatible?
23:16.53flybackshe's like the female version of me
23:17.01jbevrenprpplague: it has the digital section
23:17.06jbevrenthe challenge with filters is that theyre analogue
23:17.18jbevrenflyback: I'll respectfully disagree ;)
23:17.25jbevrenIve seen her HDL sources
23:17.28jbevrenshe's ingenious
23:17.40prpplaguejbevren: so you couldn't use the dtv as a base?
23:17.46jbevrenI have no doubt youre smart tho
23:17.51jbevrenprpplague: the DTV is a SoC
23:18.16jbevrenit's also lacking in filters and has a few cpu bugs ;)
23:18.21jbevren(being Jeri's design)
23:18.43jbevrenhaving square, triangle, sawtooth, and noise waveforms is cool
23:18.46prpplaguejbevren: right, what i was thinking is you could flash some code on it to make it a slave device
23:18.50jbevrenbut if you dont have filters, you lose a lot of effect
23:19.02jbevrenprpplague: I could.  afaik the CIA's serial ports are somewhere
23:19.15jbevrenalthough the radioshack hummer would be better if I were to use that
23:19.27prpplaguethe shack hummer still available?
23:19.34jbevrenI'm not sure
23:19.44jbevrenthe DTV is using version 2 (euro) hardware now
23:19.48jbevrenmy DTVs are v1
23:19.51jbevrenthe hummer is also v2
23:19.52flybackactually
23:19.56flybackif I remeber from her video
23:20.10flybackshe basically at least put out vias for adding origional c64 shit
23:20.11jbevrenprpplague: the dtv v1 uses a mask rom
23:20.15jbevrenits in a CoB blob
23:20.25jbevrenhowever, the rom CS line is accessible, as is most of the bus
23:20.32jbevrenthe hummer uses a flash device
23:20.51prpplagueyea, the hummer doesn't appear available from the shack
23:21.13jbevrenprpplague: its fairly moot, as the SID chips are no longer in production, and any you get now will likely be secondhand
23:21.26jbevrenthey have a sound I enjoy though
23:21.35prpplaguejbevren: yea thats why i was wondering if something in production could be used
23:21.47jbevrenand enough others like it to support projects like the SIDstation (big money) and the PCI sid cards for PCs
23:22.16flybackprpplague
23:22.22flybackyou seen eureaka seven yet?
23:22.43prpplaguei guess not, because i have no idea what that is
23:22.51flybackanime series
23:22.56flybackwas on cartoon network also
23:22.59jbevrenheh
23:23.07flybackawesome ending I get chocked up watching it every time :P
23:23.14jbevrenI saw the last episode of that 18 months before the first aired on toon net
23:23.15jbevren:P
23:23.45flyback:P
23:27.15flybackwhen renton said he was going down to see if he could get nirvash going, I had a feeling something cool was about to happen
23:27.18flybackand it did
23:27.20flyback:p
23:29.20*** join/#edev CosmicPe1guin (n=nobody@163.181.251.103)
23:36.45jbevrenmeh why do people insist on using a microcontroller to replace a pair of shift registers?
23:38.36*** join/#edev ico2_ (n=ico2@modem-3226.chameleon.dialup.pol.co.uk)
23:43.30davidc__jbevren: microcontroller might be cheaper? :P
23:45.12jbevrentwo shift registers are less than a buck :)
23:45.28davidc__jbevren: cheap pic is 37c ;)
23:45.36jbevrenlink?! :)
23:45.42davidc__pic16f54
23:45.44davidc__qty 1k
23:45.47jbevrenok well
23:45.54jbevrenI dont have 3700 bucks
23:45.58jbevrennext? :D
23:46.04davidc__88c on digikey qty 1
23:46.38davidc__er, and it would be 370 bucks @ 37c ;)
23:46.50jbevren18 cents per 1k for the reciever :)
23:47.08davidc__reciever?
23:47.23davidc__18c for two shift registers?
23:47.24jbevrennod, you need a pair of ICs.  One to recieve teh scan rows, the other to transmit the columns
23:47.33jbevren18c for 1k SRs
23:47.53davidc__well, you're at 36c then for two shift regs
23:48.02davidc__which can be replaced by a reprogrammable 37c micro
23:48.24jbevrenhehe
23:48.29jbevrenpoint taken
23:48.44jbevrenI however lack the time to learn a new cpu (although I would enjoy the knowledge)
23:48.46davidc__[btw, I hate that pic16f54 chip with a passion]
23:48.57jbevrenisnt the F54 an 8pin IC?
23:48.59davidc__yeah, pics aren't a fun cpu
23:49.02jbevrenI need 16 lines to scan :)
23:49.10davidc__nah, its more
23:49.13adyerthe incremental bang-for-the-buk is pretty good with microcontrollers, too.  find something else for it to do :-)
23:49.18jbevrenah, the 14pin one (looking)
23:49.20jbevren12 IOs doh
23:49.36davidc__its an 18 pin soic
23:49.41jbevrendavidc__: besides, I'd use my SX if I were to use a pic :)
23:49.42davidc__but IIRC its 12 IO's
23:49.52davidc__jbevren: hah, I hate those with a passion too
23:50.00jbevrenwhat do you not hate?
23:50.04davidc__jbevren: [wrote an ethernet firewall on an SX52]
23:50.07davidc__jbevren: AVR's
23:50.09adyeraround $1.50 is where the choices really multiply
23:50.13davidc__jbevren: thinks with flat memory spaces
23:50.16davidc__*things
23:50.21jbevrenah, you dislike harvard
23:50.24jbevrenI dont care for it either :X
23:50.25davidc__actually, I hate pretty much everything from ubicom
23:50.28davidc__eh, harvard is fine
23:50.29adyerluminary 3S101?
23:50.39jbevrenLuminaries seem quite cool
23:50.43davidc__but the SX52 has 16 global registers, and everything else is banked
23:50.51davidc__which is a huge PITA
23:50.57adyerPIC32 (really a mips)
23:50.57jbevrendavidc__: I programmed on a c64 ;)
23:51.01jbevrenI'm used to banking and overlaying
23:51.12davidc__jbevren: heh, cool - but at a 16 byte granularity?
23:51.14jbevrendavidc__: try memory you cant even read or write? :D
23:51.24davidc__jbevren: just exec outta? eh, thats normal
23:51.25jbevrendavidc__: 256 byte granularity in one case
23:51.30jbevrendavidc__: nono
23:51.37jbevrenits liek a storage device lol
23:51.45jbevrenyou tell the controller I want this data put here in my memory
23:51.45davidc__oh, like eeprom :P
23:51.50jbevrenof sorts yeah
23:51.55davidc__jbevren: eh - the newer ubicom products [formerly scenix] are even worse
23:52.05jbevrenthen the controller sends the data to memory via DMA (which is fun)
23:52.15adyerreminds me of the old joke data sheet for write-only memory
23:52.19jbevrensx20 et.al are scenixes arent they?
23:52.23davidc__think of doing a function call for reading/writing sdram
23:52.31davidc__jbevren: yup, and they became ubicom
23:52.36jbevrennod
23:52.42davidc__jbevren: which then created the bastard children known as the IP2k/3k/4k
23:52.53davidc__or maybe it was 5k instead of 4k
23:54.13jbevrenI always thought the sx would make a good microcode engine ;)
23:54.23jbevrenthats how I'd most likely use it in any case
23:54.57davidc__heh... personally, I don't want to touch any processor from that company again if I can avoid it
23:55.14davidc__AVR's and arms are so cheap these days... I'd rather use a processor that doesn't make me want to stab myself
23:55.45jbevrenhehe
23:55.53jbevrenI've peeked at arm assembly
23:56.01jbevrenI dislike high level languages somewhat
23:56.06davidc__arm assembly is delightful
23:56.15davidc__its my favorite asm so far
23:56.16jbevrennod, its 6502ish, which makes me all warm and fuzzy :)
23:56.33jbevrenthe beq and bne (bz, bnz) opcodes et.al :)
23:56.34davidc__but C on the arm works quite well, and interfacing C->ASM is pretty easy
23:56.45jbevrenyep
23:56.46davidc__well, everything can have those condition codes
23:56.48davidc__not just b
23:56.52davidc__so you could do addeq
23:56.53jbevrenhm
23:56.54davidc__or addne
23:57.18jbevrenthats another source for fun asm trickery
23:57.38davidc__well, you can do some really tight code with it
23:57.49davidc__and the integrated barrel shifter ahead of everything makes for fast fixed point
23:57.55davidc__anyhow, afk for 45

Generated by irclog2html.pl Modified by Tim Riker to work with infobot.