Hard disk drive motors/actuators

[DerbyBorn]

Because 45rpm is so logical!

It is between 33 and 78

[Richard]

16 2/3 rpm actually. LP at half speed.

[tabbypurr]

ISTR hearing that one particle in 6 of house dust is extraterrestrial. It's well in the category of we just don't know - all one can say for sure is we have zero evidence of any alien lifeform being here.

NT

[Dave Plowman (News)]

;-)

I did once find the explanation why the original tape speed was 30 ips - but never why 78 rpm was chosen for records.

[tabbypurr]

IIRC it was originally 80rpm, with early recordings being inconsistent on speed. I suspect 78 was just somewhere among what we've already got.

NT

[whisky-dave]

The klingons used it for target practice :)

Why wouldn't they, if they were an inteligent species wanting to learn.

But might have the inteligence to work it out.

, we'd be more likely to think it was roughly made than to think each line might contain data of an image.

We didnlt think that about Eygpt, we didnl;t say oh lok pretty meaningless carving of animlas that doen't exists. We try to work out what they mean, currently trying to work out the 12,900 year old structures and drawiong of Gobekli Tepe.

just not interested enough to do anything with it.

or use it for target practice if not interested in other cultures.

oing. Decoding an unknown disc with lines on is a long way down the priorit y list.

Some spend quite a bit of time on ancient scriptures in fact some spend the ir whole life on them studying them working out what they mean some will ev en run theier life accoirding to what they say. It doesn't have to be a pri ority on some sort of list you get someone interested to make it their work . Same thing happened with the rosetta stone, was it a proiroty of your's it wasn't of mine.

onsider finding & invading earth, a planet they most likely never knew exis ted.

Maybe they like sport, they might dress up in funny clothes and go on hunts with 16 legged spiders that look like donald trump hair who knows, what st range things sentient beings might get up to.

sense of the images & sounds. 1000rpm might be beyond any speed they've use d, etc etc.

Mighty be, but unlikely.

[The Natural Philosopher]

"Earliest speeds of rotation varied widely, but by 1910 most records were recorded at about 78 to 80 rpm. In 1925, 78.26 rpm was chosen as a standard for motorized phonographs, because it was suitable for most existing records, and was easily achieved using a standard 3600-rpm motor and 46-tooth gear (78.26 = 3600/46)"

Of course that was America, where 60hz was the norm, here at 50hz you get 3000rpm motors

And of course ratios using exact ratios of gears are deprecated because of uneven gear wear, so we can throw that one out of the box.

It may be that there was some sort of motor/gear assembly that produced about 78 rpm, and that this got to be 'the standard' but there is no real reason for any turntable speed to be a standard.

Any more than 4' 8 1/2" is the ideal railway gauge.

[whisky-dave]

You picking on Trump and Farage ;-)

They are reasonabley like 'normal' humans so the aliens haven't done too ba d, not quite a s well as they did with dressing up lizards to pretend to be the Royal family though.

Well the latest rumour/idea is that some types of octopus, cattlefish and o ther things like them are from another planet as they aren't the same as an ything else on this planet in that they can at will change 50% of their own RNA (not DNA) which nothing else on the planet can do.

So we could already have aliens on earth that were here before us.

[harry]

It can be shown statistically that there is no complex life elsewhere in the universe.

[tabbypurr]

quite the contrary - though we've never found any evidence of it existing.

NT

[Richard]

Show me.

[dennis]

No! it can be shown statistically that we will probably never make contact with it, but there is almost 100% certainty that there is other intelligent life in the universe. Of course this will upset you as you will realise you are insignificant.

[Roger Hayter]

Any such proof must prove that we don't exist either. In fact, I don't think any statistical approach can give more than a probability as an answer.

In simple probability terms our existence has no influence on the likelihood of the existence of other examples. I think on any Bayesian analysis our existence must make the existence of other examples more likely.

[Johnny B Good]

The platter spindle motor is typically a 4 pole dc brushless motor with the drive electronics living on the attached circuit board containing the rest of drive electronics and microprocessor controller unit. Unless you're prepared to fabricate your own motor drive electronics board, there's no simple way to connect it to a 5 or 12 DC supply to make it run.

However, it ought to be possible to utilise a low voltage 50 or 60 Hz supply with a capacitor to run it as a low voltage ac motor at a fixed speed (1500/1800 or 3000/3600 rpm?). You could use a low voltage output mains transformer to provide the required 6 or 7.5 v ac voltage which could be an old fashioned (bulky and heavy) 10VA 6 or 7.5 volts AC output wallwart as your source of low voltage AC.

In recent years, spindle speeds have typically been 5400, 5900 and 7200 rpm in desktop HDDs. High performance varieties use speeds as high as

15000 rpm in order to both reduce seek latency and boost data transfer rates.

The voice coil actuators typically utilise neodymium magnets which are extremely powerful permanent magnets (size for size compared to the older traditional steel alloy based permanent magnets). The motivation for most people who take a fancy to stripping an old or busted HDD is in the harvesting of these very powerful voice coil magnets for other DIY projects or simply just for fun.

If you're thinking of employing the voice coil actuator for another purpose, you need to consider how it responded to drive current from the head seek control unit whose job was not only to make the head assembly rapidly traverse the thousands of concentric data tracks to reach a specified target track but also to lock onto said target track using servo data embedded within the data streaming from the track picked up by the read/write head.

The servo controller would be continuously making tiny adjustments/ corrections all the time the drive was spinning regardless of whichever track had been targeted. The head was electronically locked onto each track rather than relying on some physical detenting system as used by floppy disk drives and the very first HDDs. This meant that the head assembly was free to move right across the whole usable radius of the platter(s) without the need to jump from track to track as was the case with the stepper motor systems used by floppy disks drives and early HDDs.

This sped up seek speed considerably compared to the very first drives which might specify an average seek time of 75ms and, therefore, an entirely predictable full stroke seek time of 150ms (real figures for an ancient Seagate half height 5 1/4 inch 20MB 2 platter (4 heads) 615 track MFM ST225 HDD).

Later voice coil HDDs were able to quote track seek times of 17ms average and 21ms worse case (not the 34ms equivilent of old) because the steady plodding through the tracks one track per stepper motor pulse of old had been replaced by a motion that owed its basic characteristic ballistics to that of rocket science.

The head assembly will move according to an accelerating force determined by how strong a driving current is applied to the voice coil. Use double the current and you get double the acceleration. Whatever drive current is actually applied, the larger the number of tracks that have to be traversed the greater the average speed of the head assembly.

Indeed, since the whole head assembly behaves like a rocket on almost frictionless rails, the controller has to estimate the halfway 'turnover point' and apply 'reverse thrust' (reverse polarity of the driving current) for an equal time interval to allow the head assembly to slow sufficiently for the embedded servo data to become readable so the servo controller can identify the exact track it has chanced upon to allow it to home in on the target track using a fine tuning seek algorithm to complete the seek command.

The point of describing the ballistic behaviour of the voice coil actuator is to alert you to this behaviour which may be surprisingly different to what you may have imagined assuming you had imagined that so many "volts" or "milliamps" would correspond to so many degrees of angular movement from a reference or resting position of the head assembly arms.

Unless you're stripping a very early model of IDE drive which afair, did use a relatively feeble restoring spring to park the heads in the landing zone close to the spindle upon shutdown[1], you'll discover that the head assembly is entirely free to move anywhere within the mechanical limits of its sweeping range. It's only the efforts of the servo controller that 'locks' the head assembly into any one track location whilst it is powered up and properly initialised by the HDD controller.

If you apply energising current directly to the voice coil (perhaps using a 1.5v AA cell with maybe a 100 ohms or so resistor) expect the actuator to accelerate to the end stop with a nasty klunk and stay put until you reverse the polarity of your test current whereupon it will accelerate to the opposite end stop and stay put.

If you care to experiment with various resistor voltage combinations or make use of a variable resistor (rheostat), you might be able to slow the acceleration sufficiently to directly observe the unabated acceleration by the application of a fixed current, perhaps just a few milliamps' worth.

If you have a modern digital camera you may be able to record a movie at 'trick frame rates' such as 120 and/or 240fps to more closely observe the event in slow motion playback. Even if you don't have any such 'trick frame rates' available, it may still be worthwhile recording a movie or two since you can reduce the playback frame rate in multimedia players such as VLC to provide a similar if rather jerky slow motion playback effect which may suffice to observe the continuous acceleration due to the application of a fixed level of drive current.

Incidentally, if you really are interested in experimenting or using the bits of a redundant hard disk drive, there are several 'Howto' videos on youtube showing how to disassemble hard disk drives which should help you avoid needlessly damaging any of the more fragile bits of interest.

Good luck and have fun. :-)

[NOTES] [1] This neatly provided an auto-park feature even in the event of an unplanned power outage. The earlier stepper motor based HDDs required that a head parking request be carried out by the seperate controller adapter card it was connected to just prior to shutting the computer down so as to place the head assembly over a safe(r) part of the platter surfaces that was clear of the storage area and could allow the heads to land safely onto the surfaces of the platter(s). This safe 'landing zone' was either inboard of or actually the innermost data track where the effective linear speed was the slowest and least damaging as the disk platter assembly slowed to a complete standstill causing the heads to lose lift and come back into contact with the surfaces of the platter(s).

This all harked back to the days of MSDOS 3.30 well before the advent of windows 95 which could automatically issue such a command when issued a shutdown request by the user, so it was necessary for the user to type the "park" command at the dos prompt just prior to hitting the off switch.

The "park" command wasn't one of DOS's built in commands but an aftermarket utility program which had to be 'installed' (just a matter of copying a file to a subdirectory (folder) that was included in the path statement, in this case, a file called "park.com").

This was considered to be an advised command only if you thought there was any chance of the machine being relocated elsewhere before being booted back up again. If the machine wasn't being moved between sessions, the risk of damage to the platters and/or heads was vanishingly small and few users ever bothered parking the heads which, in practice, proved to be perfectly ok.

The later IDE disk drives were able to auto-park their heads without any such manual intervention, relying, afaicr, upon a relatively weak restoring spring to drag the head assembly towards the spindle where it would remain safely latched until it was unlatched by a solenoid when the drive was next powered up.

The only problem with relying on a spring to achieve an automated parking action on loss of power was the need for the servo controller to maintain a steady bias current to cancel the effect of the parking spring which compromised the performance to a small degree.

Modern drives have eliminated the need for such a restoring spring powered auto-parking mechanism by cleverly using the spindle motor as a generator, which extracts the kinetic energy stored by the mass of the spinning platter assembly to provide power to the servo controller during any sudden loss of power event which allows the heads to be reliably unloaded off the outer edge of the platters and parked onto a plastic ramp provided for both this purpose and that of providing an additional energy consumption reduction strategy during normal use (and, in the case of the Western Digital laptop versions where this strategy seems to have originated, a reduction in the risk of fatal head crash events due to mishandling of the laptop during normal use).

[FMurtz]

Does not even convert to metric evenly

[Mike Clarke]

[snip]

Of course not, there was no existing standard. Someone just picked a horse drawn cart at random, measured the distance between the wheels and said "that's how far apart we'll make the rails'. From then onward all wagons that were to run on the rails had to have the wheels that far apart. There was no previous standard for the distance between the wheels of horse drawn carts, if they'd picked a different cart in the shed the track width would have been different.

[newshound]

Thanks for a great description. Isn't it an astonishing bit of mechanical technology!

In the days of VHS recorders, I recall reading that the head spindles were the most accurate bits of mass produced mechanical technology ever made, having a runout of something like 10 nanometers.

Wikipedia says that hard drive heads now fly at 3 nanometers.

[D.M. Procida]

There go my plans...

[detailed explanation snipped]

Thanks, that was very informative and useful, and much appreciated. I'll have to reconsider my options.

Daniele

[Johnny B Good]

I did see that reference on a page describing disk platters after reading the following wikipedia article

All the indications are that the areal density of the nand flash chips used in SSDs has already surpassed that of the platters used in HDDs and is set to keep improving by 100% per year over the next few years.

It does look as though all that exotic engineering is destined to go the way of the Dodo in another 5 to 10 years. Let's hope the manufacturers archive every aspect of modern disk manufacturing technology over the last 3 or 4 decades and add it to all the other technology history that's hopefully being stored safe from the worst possible survivable disaster scenario we can conceive.

Such technology archives should be created and maintained even if it's simply to provide inspiration in other fields of use. It would be tragic to think that a medical breakthrough was needlessly delayed on account of the need to *re-invent* an existing technology that had outlived its original purpose.

Knowledge is Mankind's most precious resource. It's a resource that needs to be protected at all costs. Since there's no way of determining which knowledge ought be kept and which can safely be discarded, 'cherry picking' what goes into our precious archive and what gets left out is totally out of the question - we keep it all, no matter how trivial it may seem even if it means, "We're gonna need a bigger archive." :-)

[Graham.]

I think Network Rail still measures distances in chains.