Um you do realize that every fan in your computer is impedance protected, don't you? If you stop the fan with your finger the current does not rise significantly at all, and the very design of most of these fans does nothing to self-cool.
Moral of this story - mechanical engineer thinks about the bearings, electrical engineer thinks about the motor and stator, software engineer thinks about RAID-5.
And systems engineers (like me) think about everything.:-) Admittedly I don't design computers (embedded industrial control systems is where I hang my hat) but I stand by my statements. Overspeeding motors of any kind often causes bearing damage which leads to premature failure of the motor. Contamination of the bearing lubricant also causes premature failure of the motor. There is a happy middle ground and it involves not overspeeding the motor when you hit it with the air gun.:-)
I'd presume that the switching regulator IC would have some sort of thermal protection in it, even if it just failed when it got too hot.
Most linear regulators do this (the entire 78xx/79xx family is great for this exact reason) but chips like the venerable LM723 (I think that's the right number) and practically the entire line of Unitrode parts has no built in thermal protection whatsoever. You can build that in to the feedback loop or the power sequencing part of the supply, of course, but that costs money.:-)
Silicon won't necessarily save you, because the tempco could be negative (bipolar transistors, for example, have a negative tempco) which means that as they get hotter, they conduct better. Conducting better means they will draw more current, and the resulting positive feedback loop quickly leads to thermal runaway. MOSFETs don't have this problem but the actual switches aren't the only thing to go wrong in switchers. IIRC the opto used in the feedback loop often employs a bipolar phototransistor which leads to supply brownouts when it gets hot. (i.e. the Current Transfer Ratio goes from 1:300 to 1:500 so when the LED is at half intensity the supply thinks that it's putting out a higher voltage than it really is and overcompensates.)
There are very good reasons why electronics designers are often heard mumbling and grumbling along the lines of "there's no faster fuse than silicon," and "if you use a $0.05 fuse and a $35.00 transistor, the transistor will blow apart to protect the fuse."
Sure it will cause some extra wear. You might even be able to measure it with the most sensitive equipment available. But is there really any reason to believe that doing this will cause the fan to fail any sooner. Not counting some statistical likelyhood to fail three seconds sooner.
Not true. Those fans are designed to spin at ~4500rpm. Just like a car engine, rev too high and you can spin the bearings. And despite sounding like what they're designed to do, spinning bearings and/or scaring the raceways they run in is a bad thing. It causes premature failure that is measured by years, not seconds.
You know, just because you know the ins and outs of VB and Perl doesn't make you a mechanical expert as well.
For some reason your uid rung a bell... you're not from K/W, ON, are you? If so, why is it I seem to meet more people _here_ from K/W than anywhere else on the 'net? 8-)
All computer power supplies have temperature sensors.
You may be an MCSE, but you're no electronics guru. I've got four different manufacturer's AT and ATX power supplies here, open, and none have thermal sensors.
Now there's a really expensive one in my server which does, but it uses it to regulate how fast to spin the fan. I see fuses, MOVs, NTCs and the usual array of electronic devices in these power supplies. That's it.
Air compressors are so much fun for cleaning computers. Try aiming the output directly at a fan. Spins faster than it was ever designed:). Yeah, I know it's not smart... but it's FUN!
Doing that will likely destroy the bearings, so you'll be replacing those fans soon, too. Please, if you're going to use compressed air to clean out the computer, stop the fans from moving so that you don't destroy them.
Destroy itself? Hopefully it wouldn't destroy itself in a firey ball of flame...
that is one of the things that UL certified equipment is tested for, numbnuts.
Basically they put cotton "fleece" around all the openings and cause a catastrophic fault. If any of the cotton burns, you fail.
With our power electronics equipment, the fault consisted of shorting out the load side of our equipment while we were connected to a bus capable of delivering 100kA. We passed just fine, but it was the fuses which afforded us that protection. Remove the fuses and the results are... well... spectacular.
What sucks is they dropped the commercial VPN client totally, the freeware version is still around (or was a couple weeks ago) but it only supports machine to machine, no machine to network connectivity, that was only in the commercial version.
That doesn't suck at all, unless you're using Win95/98. Win2k has built in IPSec and it works quite well with FreeS/WAN (I am using it every day). vpn.ebootis.de (funny name, great documentation) shows you how to patch FreeS/WAN to use X/509 certs, and how to generate the certs, and how to make win2k and FreeS/WAN play nice together. PGPNet for Win2k was a little bit of a goofy thing.
Geeesh, be thankful he actually told you number 1. Next time, I think he should probably stick with number 2 and just tell you when the fix is out - at least then you can't whinge about it.
I agree. 1 is better than 3, but 2 is what he has been doing for a long long time. This was a stunt to try and make everyone upgrade to 3.4 with the nifty privsep code that isn't fully working on all platforms yet. I suppose he's running out of alpha/beta testers and needed more or something.
...or does anyone else think that the way that this particular exploit was handled was, to say the least, irregular...
I agree totally. What is that dude Theo smoking? We've had remote root exploits before and they were handled properly. This one was like a forced push for untested brand new privsep code. No thanks. I just added "ChallengeRepsonseAuthentication no" in my config instead.
To me it seemed like a PR stunt. Why it was handled this way I don't know, but I'm not going to jump when he demands everyone upgrade or else.
I'd go as far as to say I want to see a working exploit for this. It sounds fishy. An integer overflow (not a string overflow) causing remote root? I find that very hard to believe.
Results: I get an extra 5 degrees Celsius trimmed off of my CPU heat, and I get to run my HSFs in "quiet mode." Suddenly, my poor XPs can run a bit faster, and life is good.
Not to mention the benefit of funnelling the water which is dripping from the sides of the case down into a cup.:-)
Seriously though I hope you are either in a dry climate or you insulate the case from the outside world lest your poor computer drown.
I'm not trying to start a flame war here, but I am curious as to why slackware packages are crappy. Perhaps you wouldn't mind explaining...
I was being flip, but I belive that the slackware packages are not compiled --with-objprelink which *really* helps reduce startup time. The other thing I had mentioned was using KDE CVS code, which I run on my notebook. We're talking 400MHz Crusoe compred to 1G P3m, but I am fairly certain that the KDE CVS code has numerous speed increases (including a prelinked Qt library too).
I'm a diehard Slack fan but for things like GUIs and web servers I compile my own instead of relying on a package. I'd like to see if GCC can optimize for Crusoe as well... -mcrusoe?:-)
SonicBLUE offers both Linux and Windows. And so does FICA.
We bought a couple of win98 ones for $599 each, and I blew away the win98 partition on the one and installed Linux. The guy that works with me put win2k on the other.:-)
It is probably the crappy slackware packages, but KDE3 was dog-slow on these things. I am going to put the latest CVS compiled with prelinking and gcc3.1 to see if I can't get the startup times down and maybe better speed through the updated code in CVS.
Not bad little machines. 1024x768, 5G HDD, 128M RAM, 400MHz Crusoe processor. The graphics chipset is a Silicon Motion Lynx3M which is totally supported with XF410. The wireless network card is an OEM Orinico with a funky antenna arrangement: there are two jacks on the card "I" and "II"; I goes to a long coax cable to the other end of the unit, and II goes to a very short run of coax to a similar aluminum-plate antenna but which slides over the PCMCIA card itself, which leads me to believe it's not a dual antenna PCMCIA card. At any rate the only thing I can't get working is the gay scroll joystick, which isn't a big loss.
I've been half-heartedly porting the qtopia popup keyboard to qt/x11 so that I can get away from plugging the USB keyboard in.:-)
ext3 has failed me more than once on my Red Hat systems, and the performance plain sucks (from what I've seen).
What were you doing to break it? It's true that performance isn't quite what ext2 is but for this notebook and my playing with ACPI, having a jfs that stores not only metadata but also the actual data in its transaction log is very nice. (at least I belive it does store everything, not just metadata...)
I didn't create this tech but I am using it to replace shitty HTML+Jscript+prayers+sacrifices web-based interfaces. There are some other guys like XWT too but XWT is simple, straightfoward and fast. It essentially projects UIs -- do you forms locally (on the client machines) but all your business logic sits on a server where it belongs. Talkes XML-RPC and SOAP. Very cool. Way way way better than what I would call traditional "web services."
Oddly enough, it's usually the people that react the way you just did that really can't drive and talk on a cell phone at the same time (they think they can, and they're very adamant about it, but in reality, they can't).
The proof is in the pudding, of course. I just really resent any legislation which tries to pander to the lowest common denominator because of the overreacting soccer moms screaming "think of the children!"
And yes, I do feel that way about a lot of laws.:-)
drivers phoning up to say they'll be late - in standing traffic, I hope
Blow it out yer ear. Just because you can't chew bubble gum and walk at the same time doesn't mean the rest of us should be hampered. If you can't drive without being distracted you shouldn't be driving, period.
The problem with OFO's, is that they tend to require a lot of up front setup, and people just "assume" that it works cleanly with Oracle and what not.
Amen. The only way I've been able to able to back up a postgres database "live" is to pg_dumpall to a gzipped tar which went to tape. Since it uses normal database operations to pull the data out it would work correctly and not disturb transactions or require the server to go down. While I use afio for my other backup needs (inside a perl backup program called flexbackup), afio has no advantage over tar since it's one big honkin' file.
But 128kbs is indistinguishable from CD quality, dontchewknow?
I've got a few LAME-encoded 128kbps CBR MP3s that are indistinguishable from the CD. In particular, my Melissa Etheridge album actually startled me when I ran across it in the playlist. "Like the Way I Do" has an intro filled with highhats, cymballs and tambourines -- instruments which lossy encoders have enormous trouble with -- but it was so clear I had to check to see if it wasn't a 384kbps or.wav that made it into my playlist.
This was several years ago that I encoded it, I wonder what settings I'd used...
If my $.99 bought me the raw stereo PCM data to burn, MP3, ogg, or sample then I would consider this reasonable.
Totally agree. I would buy tons of singles if I could buy a $20 card (or account) and download singles from 15 different artists and get the raw CD audio I would be a very happy camper. As it is now I haven't bought a CD in several years, partially because I get far more bang for my buck with other purchases, and partly because I dont' like many popular artists enough to buy the entire album (and purchasing a physical single is silly, IMO). My MP3 collection is from all my CDs from before then, and the few used CDs I've picked up to fill in the gaps.
I don't understand the gigantic piracy attitude that others have posted here. "I don't care if they're free, I'll still share 'em!" These turkeys just don't get it. Oh well.
Tell me, what exactly do you download that you would download continuously and consume this much bandwidth?
Far be it from me to tell you how you use your connection, but when I got DSL I also watched my consumption. I'm well under 10kbps average on the daily, weekly and monthly graphs. Sure there are bursts in which I peg the connection in either direction but at least for me the best aspect of high speed is the low latency.
I wonder if there are low-latency, low bandwidth connections.:-)
Slashdot Mirror
stuck fan overcurrent
Um you do realize that every fan in your computer is impedance protected, don't you? If you stop the fan with your finger the current does not rise significantly at all, and the very design of most of these fans does nothing to self-cool.
Moral of this story - mechanical engineer thinks about the bearings, electrical engineer thinks about the motor and stator, software engineer thinks about RAID-5.
And systems engineers (like me) think about everything. :-) Admittedly I don't design computers (embedded industrial control systems is where I hang my hat) but I stand by my statements. Overspeeding motors of any kind often causes bearing damage which leads to premature failure of the motor. Contamination of the bearing lubricant also causes premature failure of the motor. There is a happy middle ground and it involves not overspeeding the motor when you hit it with the air gun. :-)
I'd presume that the switching regulator IC would have some sort of thermal protection in it, even if it just failed when it got too hot.
Most linear regulators do this (the entire 78xx/79xx family is great for this exact reason) but chips like the venerable LM723 (I think that's the right number) and practically the entire line of Unitrode parts has no built in thermal protection whatsoever. You can build that in to the feedback loop or the power sequencing part of the supply, of course, but that costs money. :-)
Silicon won't necessarily save you, because the tempco could be negative (bipolar transistors, for example, have a negative tempco) which means that as they get hotter, they conduct better. Conducting better means they will draw more current, and the resulting positive feedback loop quickly leads to thermal runaway. MOSFETs don't have this problem but the actual switches aren't the only thing to go wrong in switchers. IIRC the opto used in the feedback loop often employs a bipolar phototransistor which leads to supply brownouts when it gets hot. (i.e. the Current Transfer Ratio goes from 1:300 to 1:500 so when the LED is at half intensity the supply thinks that it's putting out a higher voltage than it really is and overcompensates.)There are very good reasons why electronics designers are often heard mumbling and grumbling along the lines of "there's no faster fuse than silicon," and "if you use a $0.05 fuse and a $35.00 transistor, the transistor will blow apart to protect the fuse."
Sure it will cause some extra wear. You might even be able to measure it with the most sensitive equipment available. But is there really any reason to believe that doing this will cause the fan to fail any sooner. Not counting some statistical likelyhood to fail three seconds sooner.
Not true. Those fans are designed to spin at ~4500rpm. Just like a car engine, rev too high and you can spin the bearings. And despite sounding like what they're designed to do, spinning bearings and/or scaring the raceways they run in is a bad thing. It causes premature failure that is measured by years, not seconds.
You know, just because you know the ins and outs of VB and Perl doesn't make you a mechanical expert as well.
Math nerds can calculate the percentage error.
Why the percentage error is exactly 3.14157429...%
For some reason your uid rung a bell... you're not from K/W, ON, are you? If so, why is it I seem to meet more people _here_ from K/W than anywhere else on the 'net? 8-)
Yes I am, actually. Email me.
Do you have over 1000 comments? Why Not?
Actually I have over 1200. What's your point? :-)
All computer power supplies have temperature sensors.
You may be an MCSE, but you're no electronics guru. I've got four different manufacturer's AT and ATX power supplies here, open, and none have thermal sensors.
Now there's a really expensive one in my server which does, but it uses it to regulate how fast to spin the fan. I see fuses, MOVs, NTCs and the usual array of electronic devices in these power supplies. That's it.
Air compressors are so much fun for cleaning computers. Try aiming the output directly at a fan. Spins faster than it was ever designed :). Yeah, I know it's not smart... but it's FUN!
Doing that will likely destroy the bearings, so you'll be replacing those fans soon, too. Please, if you're going to use compressed air to clean out the computer, stop the fans from moving so that you don't destroy them.
The hottest day I ever had to endure was 35 in Taejon, Korea.
It was 36 degrees in St. Catherines, ON yesterday. Yes, Canada. :-)
Destroy itself? Hopefully it wouldn't destroy itself in a firey ball of flame...
that is one of the things that UL certified equipment is tested for, numbnuts.
Basically they put cotton "fleece" around all the openings and cause a catastrophic fault. If any of the cotton burns, you fail.
With our power electronics equipment, the fault consisted of shorting out the load side of our equipment while we were connected to a bus capable of delivering 100kA. We passed just fine, but it was the fuses which afforded us that protection. Remove the fuses and the results are ... well... spectacular.
What sucks is they dropped the commercial VPN client totally, the freeware version is still around (or was a couple weeks ago) but it only supports machine to machine, no machine to network connectivity, that was only in the commercial version.
That doesn't suck at all, unless you're using Win95/98. Win2k has built in IPSec and it works quite well with FreeS/WAN (I am using it every day). vpn.ebootis.de (funny name, great documentation) shows you how to patch FreeS/WAN to use X/509 certs, and how to generate the certs, and how to make win2k and FreeS/WAN play nice together. PGPNet for Win2k was a little bit of a goofy thing.
a flock of pigs
I'm sorry, but the mental image is too much for me. Thanks for the laugh. :-)
Geeesh, be thankful he actually told you number 1. Next time, I think he should probably stick with number 2 and just tell you when the fix is out - at least then you can't whinge about it.
I agree. 1 is better than 3, but 2 is what he has been doing for a long long time. This was a stunt to try and make everyone upgrade to 3.4 with the nifty privsep code that isn't fully working on all platforms yet. I suppose he's running out of alpha/beta testers and needed more or something.
I agree totally. What is that dude Theo smoking? We've had remote root exploits before and they were handled properly. This one was like a forced push for untested brand new privsep code. No thanks. I just added "ChallengeRepsonseAuthentication no" in my config instead.
To me it seemed like a PR stunt. Why it was handled this way I don't know, but I'm not going to jump when he demands everyone upgrade or else.
I'd go as far as to say I want to see a working exploit for this. It sounds fishy. An integer overflow (not a string overflow) causing remote root? I find that very hard to believe.
Pay no attention to the man behind the curtain...
Results: I get an extra 5 degrees Celsius trimmed off of my CPU heat, and I get to run my HSFs in "quiet mode." Suddenly, my poor XPs can run a bit faster, and life is good.
Not to mention the benefit of funnelling the water which is dripping from the sides of the case down into a cup. :-)
Seriously though I hope you are either in a dry climate or you insulate the case from the outside world lest your poor computer drown.
I'm not trying to start a flame war here, but I am curious as to why slackware packages are crappy. Perhaps you wouldn't mind explaining...
I was being flip, but I belive that the slackware packages are not compiled --with-objprelink which *really* helps reduce startup time. The other thing I had mentioned was using KDE CVS code, which I run on my notebook. We're talking 400MHz Crusoe compred to 1G P3m, but I am fairly certain that the KDE CVS code has numerous speed increases (including a prelinked Qt library too).
I'm a diehard Slack fan but for things like GUIs and web servers I compile my own instead of relying on a package. I'd like to see if GCC can optimize for Crusoe as well... -mcrusoe? :-)
SonicBLUE offers both Linux and Windows. And so does FICA.
We bought a couple of win98 ones for $599 each, and I blew away the win98 partition on the one and installed Linux. The guy that works with me put win2k on the other. :-)
It is probably the crappy slackware packages, but KDE3 was dog-slow on these things. I am going to put the latest CVS compiled with prelinking and gcc3.1 to see if I can't get the startup times down and maybe better speed through the updated code in CVS.
Not bad little machines. 1024x768, 5G HDD, 128M RAM, 400MHz Crusoe processor. The graphics chipset is a Silicon Motion Lynx3M which is totally supported with XF410. The wireless network card is an OEM Orinico with a funky antenna arrangement: there are two jacks on the card "I" and "II"; I goes to a long coax cable to the other end of the unit, and II goes to a very short run of coax to a similar aluminum-plate antenna but which slides over the PCMCIA card itself, which leads me to believe it's not a dual antenna PCMCIA card. At any rate the only thing I can't get working is the gay scroll joystick, which isn't a big loss.
I've been half-heartedly porting the qtopia popup keyboard to qt/x11 so that I can get away from plugging the USB keyboard in. :-)
ext3 has failed me more than once on my Red Hat systems, and the performance plain sucks (from what I've seen).
What were you doing to break it? It's true that performance isn't quite what ext2 is but for this notebook and my playing with ACPI, having a jfs that stores not only metadata but also the actual data in its transaction log is very nice. (at least I belive it does store everything, not just metadata...)
i know it sounds like a trendy buzz-word but i think it's here to stay and some seriously cool stuff will start to happen soon (look at Google).
The web is dead.
I didn't create this tech but I am using it to replace shitty HTML+Jscript+prayers+sacrifices web-based interfaces. There are some other guys like XWT too but XWT is simple, straightfoward and fast. It essentially projects UIs -- do you forms locally (on the client machines) but all your business logic sits on a server where it belongs. Talkes XML-RPC and SOAP. Very cool. Way way way better than what I would call traditional "web services."
Oddly enough, it's usually the people that react the way you just did that really can't drive and talk on a cell phone at the same time (they think they can, and they're very adamant about it, but in reality, they can't).
The proof is in the pudding, of course. I just really resent any legislation which tries to pander to the lowest common denominator because of the overreacting soccer moms screaming "think of the children!"
And yes, I do feel that way about a lot of laws. :-)
drivers phoning up to say they'll be late - in standing traffic, I hope
Blow it out yer ear. Just because you can't chew bubble gum and walk at the same time doesn't mean the rest of us should be hampered. If you can't drive without being distracted you shouldn't be driving, period.
The problem with OFO's, is that they tend to require a lot of up front setup, and people just "assume" that it works cleanly with Oracle and what not.
Amen. The only way I've been able to able to back up a postgres database "live" is to pg_dumpall to a gzipped tar which went to tape. Since it uses normal database operations to pull the data out it would work correctly and not disturb transactions or require the server to go down. While I use afio for my other backup needs (inside a perl backup program called flexbackup), afio has no advantage over tar since it's one big honkin' file.
But 128kbs is indistinguishable from CD quality, dontchewknow?
I've got a few LAME-encoded 128kbps CBR MP3s that are indistinguishable from the CD. In particular, my Melissa Etheridge album actually startled me when I ran across it in the playlist. "Like the Way I Do" has an intro filled with highhats, cymballs and tambourines -- instruments which lossy encoders have enormous trouble with -- but it was so clear I had to check to see if it wasn't a 384kbps or .wav that made it into my playlist.
This was several years ago that I encoded it, I wonder what settings I'd used...
If my $.99 bought me the raw stereo PCM data to burn, MP3, ogg, or sample then I would consider this reasonable.
Totally agree. I would buy tons of singles if I could buy a $20 card (or account) and download singles from 15 different artists and get the raw CD audio I would be a very happy camper. As it is now I haven't bought a CD in several years, partially because I get far more bang for my buck with other purchases, and partly because I dont' like many popular artists enough to buy the entire album (and purchasing a physical single is silly, IMO). My MP3 collection is from all my CDs from before then, and the few used CDs I've picked up to fill in the gaps.
I don't understand the gigantic piracy attitude that others have posted here. "I don't care if they're free, I'll still share 'em!" These turkeys just don't get it. Oh well.
I d/l over a gig in a single day sometimes.
Tell me, what exactly do you download that you would download continuously and consume this much bandwidth?
Far be it from me to tell you how you use your connection, but when I got DSL I also watched my consumption. I'm well under 10kbps average on the daily, weekly and monthly graphs. Sure there are bursts in which I peg the connection in either direction but at least for me the best aspect of high speed is the low latency.
I wonder if there are low-latency, low bandwidth connections. :-)