Dan's Data letters #87
Publication date: 23 Jan 2004.Last modified 03-Dec-2011.
Full throttle
I've been trying to find out at which temperature the so called "speed throttling" (thermal monitoring) feature kicks in on a P4 2.4C CPU. The response from Intel was somewhat daunting:
"Hello Matt,
Thank you for contacting Intel(R) Technical Support.
I understand that you would like to know at what temperature the speed throttling starts on your 2.4 GHz Intel(R) Pentium(R) 4 processor.
The speed throttling starts when the processor reaches Intel(R) Technical Support Center in your area. The maximum operating temperature for the processor. for more information on the maximum temperature for your processor use this website:
http://support.intel.com/support/processors/pentium4/sb/CS-007999.htm#Specifications Please do not hesitate to contact us again if you need further assistance."Yyyup... speed throttling kicks in... when it reaches... right, whatever.
The other interpretation of this message is that the guy/computer-generated-message means that speed throttling kicks in when it reaches the maximum temperature of 70 degrees Celsius, as determined by the aforementioned link, which seems highly unlikely, to say the least. Any ideas?
Matt
Answer:
You'll be happy to know that I can give you a more comprehensible washing the dog at
least every two weeks, during summer.
Sorry about that. My "Dada Lock" key was down. Fixed now.
The reason why you can't find a simple answer about the P4 throttling temperature is that, as far as I can tell (well, actually as far as Matt and I worked out over the course of a couple more e-mails, cunningly combined into one answer here), different CPUs of the same model may do it at different temperatures.
P4s (and P4-based Celerons as well, I think) can do two kinds of throttling, "automatic" and "on-demand". Automatic throttling chops the CPU to something between a 30% and 50% duty cycle, with a commensurate reduction in heat production, but you can't set what duty cycle you get; different processors just have a different hard-set duty cycle. What processor has what duty cycle is not revealed, and neither is the temperature at which it happens.
On-demand throttling can chop the CPU to anything between 12.5% and 87.5% duty cycle, in 12.5% steps. The temperature threshold for on-demand throttling is user-settable, as is the duty cycle; you should be able to set it in the BIOS setup program, which may also let you turn automatic throttling on and off.
There's a PDF datasheet available for download here that lays this out reasonably comprehensibly, and says that the automatic throttling temperature limit is inherent to the CPU. This seems a bit odd to me, because when the P4 throttling issue first came to light (and, as normal for Secret Things About Processors, was widely treated as at least the end of sunshine and happiness, if not the world itself), everybody noticed that the same processor would throttle at different temperatures on different boards. Maybe the different boards were just reporting different temperatures, mind you, and some of the people doing the testing were probably somewhat clueless. I haven't done any research on this area myself, since my own strategy is to just make an over-cooled computer and thus avoid the issue ever coming up.
Since the throttling temperature isn't disclosed in any way, though, I'm guessing that Intel just can't set it to quite the same value for any given chip. Since there's an analogue component involved (the thermal sensor inside the chip), this isn't tremendously surprising.
Purple monkey dishwasher.
(Shortly after I put this piece up, a reader pointed out this excellent X-Bit Labs piece on the subject. The plot thickens.)
Broken? WinME? Never!
I installed a CD burner in my computer. It works, and it will tell me there's a disc in there, but I can't open the disc; it says "Windows cannot access the specified device path or you may not have the appropriate permissions to access the Item", and "D:\ is not a valid Win32 application". I'm using Windows ME.
Please tell me how to get the damn thing to work!
Ell
Answer:
This could have something to do with incompatibility between drives or something as
simple as a dirty lens on the new drive, but it's probably WinME screwing up (WinME
is, arguably, Microsoft's most problem-prone Windows since Win95). Try using System
Restore and going back to a restore point before you installed the new drive.
[Ell e-mailed me back - System Restore did the trick.]
Magno-Doom 1
I just read your latest rare earth magnets article, and this question popped into my mind: What do such strong magnets really do to your body? What about all that hemoglobin with its iron? Would this be attracted to or affected by the magnetic fields? Could a strong magnetic field cause hemoglobin to stop circulating in the blood, or could the iron even be pulled out of the hemoglobin? Could it be that people who are using strong magnets therapeutically are actually doing themselves harm?
Benjamin
Answer:
Strong fields do surprisingly little to people, as far as anyone can see. Slide into
an MRI machine with a spanner in your
pocket and you'll be lucky if all it does is rip your pants apart, but in the absence
of actual lumps of ferrous metal on or in your person, even incredibly strong magnetic
fields have no perceptible effect on human biological systems. Well, as far as repeatable
science has been able to tell so far, anyway.
If your blood's iron weren't tied up in hemoglobin but just sitting around as a metal, sure, it'd be attracted to a magnet. But iron bound up in hemoglobin is like sodium and chlorine bound up in table salt; it behaves quite differently from its un-compounded form, and is no more magnetic than any number of other compounds. Practically everything is either attracted to or repelled from magnets (paramagnetism and diamagnetism), but the forces involved are so small as to produce little effect in everything but the ferrous metals, unless the magnetic field's a lot stronger than those usually encountered on on Earth.
One of the frequently quoted explanations for why magnetic therapy "works" is that it attracts blood to a painful area that needs it. It doesn't, of course, but your counter-suggestion about iron being yanked right out of the blood cells is just as well supported by science!
Here's another link on this subject.
Magno-Doom 2
My physics lecturer once said that a 1 Tesla magnetic field could kill you, though I think he said it in an off-the-cuff way, not an I've-killed-a-man way. Wouldn't moving through a powerful magnetic field bugger up the electrons in your heart/brain/nervous system, creating eddy currents like in the old cylinder-magnet-in-an-aluminium-broomstick trick?
Bennett
Answer:
Your physics lecturer was wrong. People sit peacefully enough in multi-Tesla fields
inside MRI machines.
And why didn't he try to kill a man, anyway? And he calls himself a scientist...
If your nervous system were a bunch of electric wires, then a magnetic field could indeed present a hazard to you - provided the field was strong enough, and changed fast enough, that the induced currents were large enough to compete with the signals that were meant to be there. Just walking past an ultra-strong magnet could possibly do this, but pulsing the field would do it more effectively.
However, the human nervous system is not a bunch of electric wires; it's electro-chemical. Signals are transmitted through the nerves themselves by minute electrical potential changes, but the nerves are connected to each other by chemical activity. A big magnetic field could induce a current in a whole human just as it could in any other similarly wet and salty conductor, but it wouldn't affect individual nerves particularly, unless you threaded some wire through your skin or held the terminals of an induction coil or something.
Cannibalise!
After reading your excellent piece on rare earth magnets I started ripping apart all the old hard drives at work - my boss came in and thought I'd gone crazy. I was sitting there with five dismembered hard drives in front of me and a big grin on my face. Then I showed him what I was doing and he joined in.
We became a couple of little kids trying to find the biggest items we could lift. We managed to lift up a desktop PC case with two magnets!
But we can't get the magnets off the metal brackets they're attached to. I've tried everything short of using a hammer and chisel and nearly impaled my fingers with screwdrivers on numerous occasions. How did you manage to get hard drive magnets off the brackets?
Also, I've always been told that if you bang a magnet really hard it'll lose some of its magnetism (that's why I haven't used a hammer yet). Is this true?
Rob
Answer:
If the magnets are glued on, I think it's generally best to just leave them there. Makes
it easier to grab the blighters. They're so fragile that there's no good way to bust
them off; maybe horrible solvents would do it, but I haven't tried.
Often, though, the magnets are just sitting between some stamped locator dots and held in place by their own magnetism. Then they're reasonably easy to get loose.
Bashing-to-demagnetise only applies to old-fashioned iron/nickel-iron/Alnico magnets. Their magnetism will also leak away naturally over time (even if you store them with "keepers").
There's more on demagnetising here.
Off-topic department
For the benefit of those Americans, myself included, who have difficulty understanding Australians, much less New Zealanders, do you have any idea what the New Zealand Army's motto "Ngati Tumatauenga" means?
Matthew
Answer:
"The Tribe Of War", apparently.
The whole Maoris-are-incredible-badasses concept, most frequently expressed today on the rugby field and by the bouncers at Australian nightclubs, has considerable grounding in fact. It's fair to say that the indigenous New Zealanders not only put up more of a fight against the well-armed British military than the various other native peoples the British Empire subjugated, but in fact were, man-for-man, markedly superior.
Toy time
In a gas powered RC car, the servos to turn the wheels and operate the throttle and brakes are powered by batteries, usually about eight AAs. What kind of battery do you think would last longer, alkalines or some type of rechargeable? The last thing you want is the batteries going dead while you have the throttle open and the car hitting a wall at full speed...
Jack
Answer:
Actually, "receiver packs" are usually four AAs, or four or five rechargeables (probably
smaller than AA size, unless you're talking about a
big-ass car). Four
unloaded AAs manage 6 volts; so do five 1.2 volt rechargeables, only they can hold their
voltage under harder load, which they're going to face when powering a fast, high torque
steering servo. The receiver and the throttle servo generally aren't very demanding
loads in comparison.
Old electric cars needed a separate receiver pack as well, because the speed controllers in those days had no "battery eliminator circuit" (BEC), to deliver 6V for the receiver from the 7.2V (probably) of the main pack. Even mechanical speed controllers have a battery eliminator lead these days, though.
In answer to your question: Every serious gas car racer uses a NiCd or NiMH receiver pack. Alkalines just can't handle the load from a hefty servo for long, unless you use ridiculously large cells, and a sagging receiver pack will change the steering speed of the car, which is a Bad Thing when you're approaching the end of the main straight at 80 kilometres per hour.
Another live one
What do you think of this?
Chris
Answer:
I don't need to think anything - Bob Park's done
my thinking for me!
He's been on Randy "Hydrino" Mills' case since 1991. Blacklight Power (Mills' company, previously known as HydroCatalysis) still haven't demonstrated a damn thing, of course.
References (relax, it's only one paragraph per URL):
http://bobpark.physics.umd.edu/WN91/wn042691.html
http://bobpark.physics.umd.edu/WN92/wn102392.html
http://bobpark.physics.umd.edu/WN94/wn042994.html
http://bobpark.physics.umd.edu/WN97/wn050997.html
http://bobpark.physics.umd.edu/WN99/wn010899.html
http://bobpark.physics.umd.edu/WN99/wn012299.html
http://bobpark.physics.umd.edu/WN00/wn021800.html
http://bobpark.physics.umd.edu/WN00/wn031000.html
http://bobpark.physics.umd.edu/WN00/wn081800.html
http://bobpark.physics.umd.edu/WN00/wn102700.html
http://bobpark.physics.umd.edu/WN02/wn062102.html
http://bobpark.physics.umd.edu/WN02/wn090602.html
http://bobpark.physics.umd.edu/WN02/wn120602.html
BlackLight have faded into the background over the last year, thanks to putting a cork in their grand pronouncements.
All of this fun-poking would be irrelevant if BlackLight had ever demonstrated that their power-from-nowhere "process" worked, of course. But in all these years, they haven't. There's some more detailed info here.
Freezing the unfreezeable
I live in southern Canada, it does get rather cold here. I dutifully buy the -40 windshield washer fluid, watering it down in the summer and putting it straight in in the winter. When I first start my car in the morning, there is often some snow and ice on the windshield. My first reaction of course is to use the windshield washers to get it off. This works, but within a few seconds the fluid freezes to the window, starting me down a spiral of needing to keep squirting the fluid every few seconds to combat the freezing and maintain visibility.
I may live in Canada, but it is not minus 40 here, so why does it keep freezing?
Paul
Answer:
It's fairly likely that "minus 40 fluid" actually is an accurate description of the
product's properties rather than a mere brand name or catchphrase. The stuff's completely
alien to me, as I live in coastal Sydney, Australia, where on the very coldest nights
in winter there may be a slight frost, but its basic composition seems simple
enough. It seems that low-temp wiper fluid just has some glycol and a lot of alcohol
in it, to haul its freezing point down and prevent it from splitting the washer fluid
bottle like a frozen water pipe.
The problem with this is, I think, that the fluid's part alcohol, part water (plus whatever secret herbs and spices a given manufacturer adds), and when you spray it onto the windscreen and it starts to evaporate in the airflow, the alcohol evaporates enthusiastically and chills the remaining water enough that it freezes.
You could, in theory, just dump straight cheap denatured alcohol into your windscreen washer reservoir and enjoy its freezing point, approaching minus 100 degrees C. (Minus 40, by the way, is where the Celsius and Fahrenheit scales cross over.) There wouldn't be much of a freezing problem then - only water the wipers hadn't wiped away could freeze. But that'd leave you with a plastic bottle of flammable liquid sitting close to various hot things in the engine bay. You'd also be setting yourself up for an interesting insurance claim whenever you pressed the washer button.
Adding a solid extra splash of denatured alcohol to the fluid might be an idea, though; a 50% alcohol solution is still, technically, flammable, but no more so than strong vodka.
There are lots of other low-freezing-point liquids that you could try too, but most of them are highly flammable and would eat your car's paint.
I added a gadget to an old car of mine that pre-heated the washer fluid before it hit the windscreen - it was just a little water block that strapped to a radiator hose. The idea was that hotter water would clean the windscreen better, but it might combat freezing problems as well. (Or, as a few readers have now pointed out, it might just make the fluid warm enough to crack the freezing windscreen, if and when the radiator hose got warm enough to impart any significant heat to the fluid at all.)
Unfortunately, the extra hose length in the gadget I bought reduced the washer power enough that the device never worked very well. But it's still a thought.
