Lead-free solder and tin whiskers: Older than you think

[999999999]

Re: Lead-free solder and tin whiskers: Older than you think

^ It's true, sleeve bearing fans do have better shock resistance due to a larger impact area than the contact surfaces of a few bearings that will deform if the force is great enough, and ball bearing fans last longer primarily due to use of grease rather than oil that dissipates over time. Relube a sleeve bearing fan before it gets dry and the lifespan is usually longer than other components in the system so the issue becomes irrelevant unless the fan is to be reused.

It's not thermal cycling that wears sleeve bearings as much as the mechanical abrasion of starting and the high temperatures reached thinning and drying out the oil. It's a bit of an apples:oranges situation as HDDs are made to be reliable as possible within a reasonable budget while (most) sleeve bearing fans are made as cheaply as possible... budget construction includes low QC on whether they are balanced and no balancing attempt made later, and they'll use a poor quality oil lubricant to save fractions of a cent in build cost, and no oil reservoir to save 2 cents. Industrial fans tend to have at least a felt oil reservoir, if not also an oil fill hole or even an oil pump if a large enough fan.

To be clear there are high quality sleeve bearing fans, and I prefer them for a desktop PC where you have to listen to it running but on my file server that I only check on a few times a year, dual ball bearing fans are the only type I use.

[Wester547]

Re: Lead-free solder and tin whiskers: Older than you think

^ It's true, sleeve bearing fans do have better shock resistance due to a larger impact area than the contact surfaces of a few bearings that will deform if the force is great enough, and ball bearing fans last longer primarily due to use of grease rather than oil that dissipates over time. Relube a sleeve bearing fan before it gets dry and the lifespan is usually longer than other components in the system so the issue becomes irrelevant unless the fan is to be reused.

I also heard that dual ball bearing fans are essentially sealed for life in regard to its lubricant (grease) whereas sleeve bearing fans are not with their oil (hence as you stated, the eventual leakage expedited by heat and power on hours, as well as usage). I didn't think that power cycles should affect fans much, though since fans are lead-free now adays and since I imagine their thermistor / control circuit is too, if some companies go cheap on those they could eventually stop working (if they get hot enough) by way of heat cycles, thus completely doing away with the functionality of the fan (I imagine that's how JMC / DATECH "dual" ball bearing fans stop working quickly, assuming the bearing is just that - a cheap and dead thermistor... or maybe they used oil instead of grease? or didn't seal the bearing?).

It's not thermal cycling that wears sleeve bearings as much as the mechanical abrasion of starting and the high temperatures reached thinning and drying out the oil. It's a bit of an apples:oranges situation as HDDs are made to be reliable as possible within a reasonable budget while (most) sleeve bearing fans are made as cheaply as possible... budget construction includes low QC on whether they are balanced and no balancing attempt made later, and they'll use a poor quality oil lubricant to save fractions of a cent in build cost, and no oil reservoir to save 2 cents. Industrial fans tend to have at least a felt oil reservoir, if not also an oil fill hole or even an oil pump if a large enough fan.

To be clear there are high quality sleeve bearing fans, and I prefer them for a desktop PC where you have to listen to it running but on my file server that I only check on a few times a year, dual ball bearing fans are the only type I use.

I know there are "special" and "improved" kinds of sleeve bearing fans as well (Sunon's magnetic "maglev", Delta's "SuperFlo", ADDA's "Hypro", "rifle" bearing, etc), though I still think they're essentially sleeve bearing. Not sure how good hybrids are (one ball, one sleeve) but I imagine nothing beats a dual ball bearing fan as far as longevity goes (fluid dynamic bearing fans may be comparable but they have to run at high speeds to operate, as stated, and are expensive, both of which is a pain). The only thing I abhor about sleeve bearing fans (since they can be relubed for longer life) is that you can't mount them anyway you please. I imagine PSU manufactures get away with doing that horizontally because they usually use 120m-140mm fans at that point which will still last longer than 80mm fans in the same condition. It's interesting that hard drives are built to be reliable - they seem to be the most unpredictable and unreliable components of a system, I think. I'd say it's easier to predict the quality of fans.

I believe that capacitors (since they fail most of all due to the dry out of electrolyte) are not subjected to thermal expansion either, even the lead-free ones, but just heat and power on hours (and faulty electrolyte formulas).

[999999999]

Re: Lead-free solder and tin whiskers: Older than you think

Hybrid fans tend to be the worst, you can't oil them much when the sleeve bearing starts to dry out because there's too little surface area and if any oil gets into the ball bearing it gets really noisy and wears faster, as well as the shorter length sleeve bearing wearing faster. Not sure why anyone ever made them as hybrids except as a marketing trick to be able to state ball bearing but cheap out on not having two ball bearings.

[Rulycat]

Re: Lead-free solder and tin whiskers: Older than you think

Are there lead-free solder formulas that don't have the solder whisker problem? I can't believe nobody has come up with a reasonable substitute, like when they removed lead from petrol. (though I know obviously it's not the same thing)

[Shocker]

Re: Lead-free solder and tin whiskers: Older than you think

ST360021A

Broke it.

Also have to add this: The drive in question is my only (if out-of-date) backup of my data files. On top of that I also run as an Administrator constantly, and don't have any antivirus utility at all.

Well that's just silly.

As I saw somewhere once.. "If your data doesn't exist in at least three different locations, it doesn't exist"

Well I just made a backup to the barely-used ST340014A. Does that make you feel better??? (I know the third copy, the thread subject, is out-of-date.)

Recent Seagate failures are largely due to A) firmware problems and B) Seagate designing their drives to fail. Yes they do. Western Digital do the same thing. That's why they put ST smooth chips on all their drives.

Speaking of which, I'm using two WD800JD-00LSA0s in my PC. Those have fluid-dynamic bearings like the Seagate drives. Sound alright??? Well just wait until you see the board...

They know what they're doing. They intentionally make the drives have in-built failure mechanisms by default so you'll buy more. Perpendicular recording technology is also an issue, and much more of a threat than lead-free solder to hard drives.

Well I agree that perpendicular recording is worse - which is why I value old hard drives so highly. In fact, if I had to, I'd confidently spend more on a 7200.7 than I would on a new 7K4000.

Hybrid fans tend to be the worst, you can't oil them much when the sleeve bearing starts to dry out because there's too little surface area and if any oil gets into the ball bearing it gets really noisy and wears faster, as well as the shorter length sleeve bearing wearing faster. Not sure why anyone ever made them as hybrids except as a marketing trick to be able to state ball bearing but cheap out on not having two ball bearings.

+10

In case anyone's wondering why I have no pics, it's because I don't have a working card reader .

Also, I think someone stole my cookies. not an acceptable substitute.

[Wester547]

Re: Lead-free solder and tin whiskers: Older than you think

Broke it.

How did that happen (if you don't mind me asking, even if I sound slow to ask ^^; )? D:

And if I may also ask, how many power cycles does your ST380011A have (if you remember)? Just curious as to how many power cycles it's survived with relation to its power on hours (though Seagate also, in their SMART info, categorize "start/stop" counts and "power cycle" counts as different).

I can't imagine HDD PCBs running hot enough to result in solder fractures either, though short-circuiting via whiskers is again a risk, as aforementioned.

[Shocker]

Re: Lead-free solder and tin whiskers: Older than you think

How did that happen (if you don't mind me asking, even if I sound slow to ask ^^; )? D:

Dropped about 1m, onto concrete, while running. Even made a dent in the motor. Caused by accidentally dragging the cord going into the PSU (that dangerous chain was surely going to break something sooner or later). It didn't seize, but my luck didn't get much better. I plugged it into my PC, ran KillDisk, and 11% through, 51 reallocated, 3 pending, and 3 uncorrectable sectors. I stopped there, figuring it would be a waste of time to continue. (I shudder to think of how many sectors would have been reallocated after finishing the test.)

Of course you wouldn't reasonably expect a hard drive that abused to continue running, and I'm surprised it still ran at all. Obviously something to try not to do in the future.

And if I may also ask, how many power cycles does your ST380011A have (if you remember)? Just curious as to how many power cycles it's survived with relation to its power on hours (though Seagate also, in their SMART info, categorize "start/stop" counts and "power cycle" counts as different).

I don't remember the exact number (and I'm not going to plug it in with the whiskers), but I think it was somewhere between 2000 and 3000.

[Wester547]

Re: Lead-free solder and tin whiskers: Older than you think

Any other drive would have probably seized completely. But those older Seagates are generally troopers. Since that one weigh 635 grams, that's no surprise. Aside from head parking (that's probably the one good thing about head parking - shock resistance, and maybe lower temperatures), I can't imagine how laptop drives (which weigh at least 6 times less than good desktop drives) have better shock resistance than desktop drives, at least not when powered off (if you take APM into account, which many desktop drives don't have). Not to mention, those 2.5" drives because of that have much less target space for the read/write heads and actuators/motors to work in, which perpendicular recording technology doesn't help at all, especially since it all weighs much lighter. I can't imagine whiskers helping those drives either.

[Shocker]

Not just the solder itself

The vias were also affected by RoHS, and are now made of something that tarnishes (which never happened before RoHS). The drive at the start of the thread is probably a worst-case in this regard (one of them is pretty much completely black). So I don't know if I'll be able to put the lead in them.

How screwed-up does it have to be??? The drive surviving all those years, but the board about to go kamikaze at any time now??? Unbelievable.

RoHS

uspoliticians

[mariushm]

Re: Lead-free solder and tin whiskers: Older than you think

Oh man.... reading this thread makes me feel like I want to facepalm myself.

So many assumptions and so many guesses...

Rulycat ... whiskers are caused by the TIN in the solder, so any solder with TIN will have whiskers. Any solder including the leaded kind has whisker problems... lead was just good at reducing the frequency and slowing down the whisker creation process.

Shoker, as far as I know there was no banned chemical due to RoHS involved in vias. VIAs are made with a relatively simple process, by electroplating copper, so there's no lead or other substances involved.
I don't get what you say by "tarnishes".. is it layer of oxide? Maybe the person smoked in the house? Anyway, why would it matter if it's "tarnished" - it's a damn via ... a copper line that goes through the PCB. Doesn't matter what gunk or whatever is over it.

Go through this walkthrough if you want and see how PCBs are made, there's a detailed presentation with short clips for each step... including RoHS compatible through holes and vias:

http://www.eurocircuits.com/index.ph...ctional-movies

Originally Posted by Wester547
They know what they're doing. They intentionally make the drives have in-built failure mechanisms by default so you'll buy more. Perpendicular recording technology is also an issue, and much more of a threat than lead-free solder to hard drives.

Well I agree that perpendicular recording is worse - which is why I value old hard drives so highly. In fact, if I had to, I'd confidently spend more on a 7200.7 than I would on a new 7K4000.

This is so STUPID on several levels.

Just because a technology has increased failure rate, it's bad that we use it?

The market demanded 1-2 TB drives, which were impossible to make without perpendicular recording. Maybe you don't WANT 1-2 TB drives, but maybe you're in the minority, maybe you're not their market.

Without perpendicular recording, you would have had 5-6 platters, slow rotation speed (4200-5400rpm), big access times due to inertia, hot drives. The failure rate would probably have been higher due to all these factors compared to just using perpendicular recording.

IF you care about your data, or you have concerns, there's NOTHING stopping you from using RAID 1, RAID 5, RAID 10, RAID whatever, to secure your data.

NO drive has infinite MTBF, all drives have a warranty, the manufacturer doesn't say they'll live forever, why do you expect that all drives would have the same reliability over time? Nobody guaranteed you anything.

By all means, stick to 40-80 GB if you're that grumpy old guy shouting "get off my lawn" while younger folks can very well pay the same price you would have paid for a 80 GB drive back then and buy two 1 TB drives and have them in RAID 1.
It's not Seagate's fault or any other manufacturer that you're too cheap to secure your data. They make drives with more disk space while cheaper than ever.

Oh man.. sorry about the rant, but seriously... one more thing. Just because YOU have a poor experience with some hard drives, doesn't mean that it applies to everything. Maybe something in your computer is causing these issues... maybe an imbalanced fan inducing vibrations, maybe it's power spikes that kill that chip, it can be lots of things.

If you want to see a very thorough test on hard drives, Google did one and published the results:

Failure Trends in a Large Disk Drive Population: https://cdn.badcaps-static.com/pdfs/...2e3239963e.pdf

They did find a lot of interesting stuff... read the pdf and learn something... but they didn't find anything out of the ordinary with any particular series of drives.

[Wester547]

Re: Lead-free solder and tin whiskers: Older than you think

^ I agree with most of your post, but the one reason why I separate my experiences from Google's study is because their study is of running drives (or servers) 24/7 whilst I always power cycle my system once a day (I'm sure heat/cold cycles augment thermal expansion and the eventual expansion of platters and such so it changes how long they last as far as I'm concerned). I wasn't saying return to longitudinal recording technology, as I know it reached its limit already (around 300GB?). But that they used perpendicular recording technology even with "newer" drives of longitudinal recording size is what irks me, though they probably did that so as to keep consistency.

Heat-assisted magnetic recording though should move storage forward and solve all the problems perpendicular recording technology has (longitudinal drives have a much greater chance of lasting longer, in my eyes - I agree with PCBONEZ and Shocker on that as with PRT the space to read/write is much smaller). I also disagree that 1TB and higher drives are fine. The higher you go, the more likely you are to have read/write errors because of data density so thick. I understand that some people need that much disk space but no me - I'd much rather have two 500GB HDDs, or even just one. 160GB+ is enough for me. No offense, but people have become spoiled now adays especially. "OMGORZ if it cant have 8 dual core and 16GB of RAM and 3TB of disk space and four graphisc cards and if it cant run the latest game well it sux0raz!11!" Not saying people talk like that at all on this forum, but on others... as for lead-free solder, as stated before, all they have to do is move onto something that takes flex better (which I'm sure can happen) and the whole "leaded vs. unleaded" solder quarrel will be over. That still leaves weaker packages like BGA, QFN, and CSP to cause potential issue, but the potential for that is well, much less, and I'm sure they'll eventually find a way out of that too (though I suppose the move to that was inevitable with the ever increasing amount of pins and contacts).

[Shocker]

Re: Lead-free solder and tin whiskers: Older than you think

Whiskers are caused by the TIN in the solder, so any solder with TIN will have whiskers. Any solder including the leaded kind has whisker problems... lead was just good at reducing the frequency and slowing down the whisker creation process.

What matters is that it's good enough. Leaded solder is. Lead-free? No thanks.

VIAs are made with a relatively simple process, by electroplating copper, so there's no lead or other substances involved.

I'm talking about the plating on top of the copper. ALL of my RoHS drives have tarnishing on there, while leaded stays shiny forever.

Maybe the person smoked in the house?

When you're near me, no smoking UNDER ANY CIRCUMSTANCES, and I had that ST380011A ever since 2004.

Just because a technology has increased failure rate, it's bad that we use it?

To be green, you need to be RELIABLE.

Without perpendicular recording, you would have had 5-6 platters, slow rotation speed (4200-5400rpm), big access times due to inertia, hot drives. The failure rate would probably have been higher due to all these factors compared to just using perpendicular recording.

Head crash??? Never personally experienced one that wasn't caused by shock. And I already mentioned heat.

And as a matter of fact that ST340014A I mentioned got bumped while running, but it got off with not a scratch.

Oh man.. sorry about the rant, but seriously... one more thing. Just because YOU have a poor experience with some hard drives, doesn't mean that it applies to everything. Maybe something in your computer is causing these issues... maybe an imbalanced fan inducing vibrations, maybe it's power spikes that kill that chip, it can be lots of things.

1. It's not ALL HDDs that I have bad experience with, just the ones with perpendicular recording.
2. I have mixed old and new HDDs in the same PC.
3. I'm not alone in hating perpendicular recording. Just search PCBONEZ.

If you want to see a very thorough test on hard drives, Google did one and published the results:

Failure Trends in a Large Disk Drive Population: https://cdn.badcaps-static.com/pdfs/...2e3239963e.pdf

I've seen that before.

I also read that the increase in failure rates of cold drives was caused by a mixture of 5400RPM BB (cooler, but unreliable) and 7200RPM FDB (hotter, but more reliable) drives.

but they didn't find anything out of the ordinary with any particular series of drives.

They did, just they didn't want to reveal which are the most failure-prone drives. And from the study:

When examining our population, we find that seek errors are widespread within drives of one manufacturer only, while others are more conservative in showing this kind of errors.

That's Seagate. All Seagate drives do that, and it's normal for them.

By the way, I got an ST340016A today, but it has a SMOOTH chip .

The largest drive using longitudinal recording is Seagate 7200.9 500GB (4x133GB platters, but no-one asks for a 533GB drive). The highest density using it is 160GB/platter (7200.9 160GB).

[mariushm]

Re: Lead-free solder and tin whiskers: Older than you think

To be green, you need to be RELIABLE.

No, you don't.

Glass bottles are very green but they're not very reliable... just let one drop from your hands and you'll see what I mean. Same with aluminum cans.

Green doesn't mean it has to last forever, it means leaving a small footprint after a product's life and taking measures to recycle the thing easily. It also means to harm the environment during the life of the device.

All hard drives can be recycled easily, just throw it in a shredder: http://www.youtube.com/watch?v=sQYPC...eature=related

The metals can then be easily separated from the fiber glass and epoxy and magnets, the metals and fiber glass are recycled ... no problem.

When done right, you do more for the environment by saving those few watts per drive compared to how much it costs to throw one away when it fails.

For one person, it makes no sense, but in quantity it makes sense. For example.. you have 150 computers in an office building like the old company I worked for as a programmer, and a policy which says never turn off the computers because:
* they're updated over night,
* it costs the company more in wages to wait 10-20 minutes per employee to start computer, wait for updates to apply, log on to network, start outlook, go for a coffee until pc is ready... and so on.

Those 2-3 watts x 150 computers x 24 hours x 31 days add up...

[Shocker]

Re: Lead-free solder and tin whiskers: Older than you think

What about those shipping costs to get the HDDs from the factory to us???

Also consider that failed equipment = lost productivity.

The 11-year-old Seagate Barracuda ATA IV is still my favourite HDD, and probably always will be.

If you don't need tons of space, perpendicular recording is a waste of time.

[mariushm]

Re: Lead-free solder and tin whiskers: Older than you think

Shipping is super cheap...

8-12k for a ~ 32 cubic meter (20') container ...15-20k for 40' (~ 65 cubic meter) container... and I believe the limit is up to about 30480 Kg, as you want to be able to carry it on the roads.

Deduct about 4000 Kg for the actual container weight and you're left with 26000 Kg of merchandise.

0.65 Kg per drive, 0.003 cubic meters per drive ... round up to 0.8 kg and 0.005 for packaging etc

65 / 0.005 = 13000 drives
13000 x 0.8= 10400 kg

Yeah, even if 10k drives are packed in a container, I don't think paying 60 cents to a dollar in shipping will be an issue.

Failed equipment... let me tell you something... drives don't fail so often when they're running 24/7.. my system only reboots for windows updates maybe once every 2 months.

I have four hard drives in my system... the oldest is a WD Black, 400 GB... keeping system all the time yet I had no issues with any drive yet.
See attachmd text files with the hard drives if you don't believe me.. look at power on time and power cycle count and you'll see I don't lie.






In an office building with 100+ computers, failures happen - the IT staff always had 2 or more full desktop systems and several monitors in the original boxes.

I didn't bother to ask them, but I don't think there were more than 1-2 drives failing each month, at 100-150 computers.

You just pop new drive in or take out the spare computer from the box and put it on the desk, then the IT staff installs a disk image through the network. In 20-40 minutes everything's fixed.

On the floor that I worked on, during the course of a year, with about 30-40 computers, there were no hard drive failures that I was aware of. I remember 2 or 3 monitors developing full lines of broken pixels. Hard drives were 250-400 GB in size, whatever was supplied with the systems (OEM HP systems with 19-20" lcds... this was several years ago)

[Shocker]

Re: Lead-free solder and tin whiskers: Older than you think

I want reliable equipment.
That's all.

EDIT:

It also means to harm the environment during the life of the device.

[mariushm]

Re: Lead-free solder and tin whiskers: Older than you think

I wrote that at 8-9 AM, without any sleep over night. It was meant to be "harm the environment as little as possible".

Well, you say you want reliable equipment but first you have to understand what reliability means. Reliability doesn't mean it will live forever, so you have to treat the device as if it would eventually die (being a mechanical device that's subjected to various stresses) and take measures to not be affected by that.

A car gearbox is reliable, but you don't think it will live forever, as there's mechanical stress there.
A car tire is reliable but you know it will wear out eventually, or you'll replace it after 2-3 years even if you don't drive your car because the rubber rots on the car.
I can go on with lots of things.

The reliability is influenced by a lot of factors:

* the humidity of the air in your country (gigabyte had high failure rates in india due to humidity causing shorts between layers of the pcbs they made so they started to use a different pcb material)
* the heat your drives and the oscillations of temperature (drives will fail more often if you power off and on your system often)
* the vibrations (a hdd can be as reliable as possible but I've seen one die when some kid got angry and punched his keyboard with the system on the same desk)
* the quality of the power you give it to

Simply put, hard drives of the same series can be more reliable in some countries than others and equipment manufacturers can't predict human behavior and take in consideration anything the owners will do.

It's so easy to say a manufacturer made some choices to limit the device' life when they could just as easily made them as a compromise between what the market wanted (bigger capacity, bigger speed) and other factors.

In addition, most consumers do not expect hard drives to last for more than 5 years or so or don't care, in fact most not tech savvy people throw away their computers and buy other computers in 3-4 years.

I've personally bought 200 GB (gifted to my sister now, still works), bought 400 GB, a year later bought 1 TB, another 6 months later bought 1 TB (wd green, sent this one to Holland to datacenter I have server at and it's in the server for a year and a half now running 24/7), bought 2 TB in pace, a year later bought 3 TB ... and that's the four drives i have in the system.
But both me and my sister treat the computers with respect.

And going back to the topic of this subject, RoHS is really the least responsible for hard disk issues.

[Shocker]

Re: Lead-free solder and tin whiskers: Older than you think

the quality of the power you give it to

Taken very seriously now. I'm never going to hook something I care about to a power supply with non-Japanese capacitors. My signature tells you about Teapo.

In addition, most consumers do not expect hard drives to last for more than 5 years or so or don't care, in fact most not tech savvy people throw away their computers and buy other computers in 3-4 years.

There's ya problem!!!!!

If you want to save the environment, that's where you need to take action!!!!!

And going back to the topic of this subject, RoHS is really the least responsible for hard disk issues.

Unless you have my luck. (or unluck, depending on POV)

[Shocker]

Re: Lead-free solder and tin whiskers: Older than you think

Alternative version:

Whiskers are caused by the TIN in the solder, so any solder with TIN will have whiskers. Any solder including the leaded kind has whisker problems... lead was just good at reducing the frequency and slowing down the whisker creation process.

Yeah, in the same way that:

• All caps, including those made by Panasonic and Rubycon, degrade over time. But bad caps are far worse.
• All CD-Rs eventually deteriorate to the point of unreadability. But something made by Taiyo Yuden will last a lot longer than one made by HongKongFlyApart® Co., LLC.
• My old drives using longitudinal recording may, eventually, lose data. But there has been a lot more data loss with perpendicular recording.
• Solid polymer caps won't last forever. But as long as they're made by a quality brand, they'll hold up better than electrolytics would under the same conditions.

So you can worry about bad caps. I have no problem with that. But worrying about good caps isn't really an efficient way to use your time.

I noticed something. HDDs using the SMOOTH chips (whether Seagate, Western Digital, or someone else) make a strange characteristic noise that sounds kinda like a brush motor .

[999999999]

Re: Lead-free solder and tin whiskers: Older than you think

In addition, most consumers do not expect hard drives to last for more than 5 years or so or don't care, in fact most not tech savvy people throw away their computers and buy other computers in 3-4 years.

Not true. Most non tech savvy people only replace their computer when a failure costs more to fix than the system is worth. The average computer running today is older than 4 years and that's in the US, not 3rd world countries! Consider that most people buy OEM systems with the latest version of Windows installed, but as of February this year, WinXP, an 11 year old OS, was still the dominant OS. Vista was released and sold through OEMs a little over 5 years ago.

"Most" non enthusiasts (outside of work oriented production) have no need for more performance to do office, email, and web surf. They just want to keep their computer running the cheapest way possible.

http://www.tomshardware.com/news/Win...-OS,14611.html