> If we use ATA Flash RAM for data storage (Expensive, BUT quite >possible for those with "simpler" Wearables!), battery life can be >greatly increased over using regular HDs; the same interface (basically) >can be used for either PCMCIA HDs or PCMCIA SunDisk style Flash - Generally you don't want to go with flashrom with an ATA interface... This makes it essentially the same thing as a hard drive, too complicated to interface too. Simple wearables can directly use the flash array of say intel or AMD's flash, just like accessing an SRAM chip. Intel now has the new Strataflash memories which are ultra high density and low cost. >I see >nothing wrong with reducing clock speed for power savings, either, >provided it's what the user wants & needs (and if Flash RAM was used, >maybe it wouldn't be wanted OR needed.) I own a bunch of old 386/486 >machines with "Turbo" switches, same concept of clock speed reduction >for the same reasons (I change my 486 laptop to slow mode once it's >booted, to conserve battery power when on the road; same for the PC110 >much of the time.) I can figure out & remember to turbo a wearable up >when I need to web surf, etc. <G> IMHO nothing's wrong with letting >those who want/need speed, have it, or those who want battery life, have >that <G> Basic power management... But older processors (like the Z80) don't have these powerdown modes. There are a number of ways modern processors conserve power, but most of the time it doesn't have to do with slowing down the clock rate. Occationally a processor will have a "shutdown" modem which stops the oscillator, but generally today's processors have varying degrees of idle or shutdown modes which don't often mess with the clock rate. >> <snipped> >> with how much time you're willing to spend developing. While the NEC >> VR chips are cheap, super high performance, and use very little power, >> the simplest VR chip is a 120 pin quad flat package! Most of the VR's >> suitible for wearables (the VR41xx series) are 180 or more pins. >> There's *no way* you can homebuild a circuit board in which this chip >> can be mounted. You must resort to expensive board fabrication > > Sorry, I "have to" say it: Wirewrapped might do it. SCARY thought ><G> but, as we say, "Never dare an engineer" <VBEvilG> Don't think it's >very practical, agreed, it's horrible to think of making 5 or 10 without >PC boards... My hands hurt, just thinking about wirewrapping one <G> Hehe well the issue relating to the high density chips is actually getting physical connections to the chip's pins. They are often as close as 0.05 inches apart, there's no room for error when making the printed circuit board! But, fab houses can get traces as thin as 0.008 inches without blinking. Plus you must take into account that there's very few ways of hand soldering these chips, except for the neat little "gob the solder on the pins and copper wick it off" trick. The big places use wave or reflow soldering techniques... So, if you could actually access the pins with some sort of adaptor (which, by the way, either don't exist or cost hundreds of dollars), wirewrap might work, but in that case breadboarding would be easier :). -Mel ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -http://www.egr.msu.edu/~tsaimelv/expander.htm
From Wear-Hard Mailing list Archive (WH)
Maintained by R. Paul McCarty
Archive created with babymail