All this talk about helium leaking out doesn’t make sense to me. It’s hard to contain helium, yes, but this is because a helium balloon is a high pressure environment, creating the energy for the helium to escape the balloon, and causing the stretched plastic to return to it’s desired un-stretched state.
If the helium pressure is changing and causing drive failure, there’s two possibilities for why: The drives are pumped with helium, creating a high pressure environment, or the drives have everything not-helium sucked out, creating a low pressure environment.
A high pressure environment would cause more friction and higher temperatures than a neutral pressure environment, so if the helium is actually leaking out, I would think it would be a simple matter for seagate/WD to run diagnostics and update the firmware for the new parameters. However, pumping helium into the drive… Doesn’t remove other gasses, which makes me think this isn’t the issue. If it is, though, a helium-filled high pressure chamber would reinfuse the drive with helium also. Which is dumb.
A low pressure environment, which makes a lot more sense to me, means that the helium is leaking in. This makes sense because all you need to do is remove other gasses, and helium and hydrogen will move into the vacuum space to replace them. However, the pressure won’t be normalized if the helium is slow to move inside the drive, which it would be if helium balloons slowly leaking is any indicator. In this case, a vacuum chamber should, with a strong enough vacuum, pull the helium back out of the drive, creating the low pressure environment the drive needs. It might also be possible just to refactor the sensors and lower/or lower the expected drive target RPM to compensate.
I’m curious what’s actually going on, because nobody seems to be talking about it, besides “well, helium drives just seem to die, so the helium must be leaking out!”
I think you will find they are equal pressure… I actually owned some and what does happen is there is a little hole on the disk you are not allowed to cover, this hole will let a certain amount of the helium escape over time.
Why helium… I believe the answer is two fold:
Helium has a lower specific density than air thus it allows the platters to rotate with much less resistance.
Helium is a very good cooling agent.
In the drives I were using which were SAS3 ( think double SATA3 ) speeds the drives could of course get very warm the helium helped mitigate that issue.
“I’m curious what’s actually going on, because nobody seems to be talking about it, besides “well, helium drives just seem to die, so the helium must be leaking out!”
You are right that is exactly what is happening but if it was not i believe it could be possible to have mini bangs going on which would tank the drive anyway due to if I am correct helium expanding as it takes on heat.
I ran a bunch of them for many years on two 100TB raid 6 arrays, it was probably overkill but I got the drives unused cheap after the setup they were going to go into was sold due to a company liquidation.
The He8/10s I have doesn’t seem to have any breather holes you’d traditionally find on air filled drives. Or at least I can’t seem to find any. I believe they are designed to be hermetically sealed.
Sounds like these drives are purged of air and welded shut using lasers in an helium filled environment at atmospheric pressure.
I would be very surprised - those are formally Hitachi drives which and WD has taken over the HGST brand which is what I brought them under and they definitely had breather holes. I can’t state this with certainty but I believe helium also does the equivalent of BIO degrade and if I remember rightly it can in some cases pass through or be absorbed by - ( it has been a long time since I looked into it ) metals over time.
If they didn’t have breather holes and the helium expanded due to heating you would effectively have a 3.5inch IED.
For reference here is a picture of an HGST drive you can see the “breather hole” for want of a better name on the lower right top surface of the drive.
If the helium leaked out, though, the pressure inside the drive would drop, causing it to create a vacuum that would suck in helium through the same place/method it leaked out.
If it was absorbed into metals, well… In this case, I guess I could see the internal pressure drop as a small amount of helium is pulled closer to the metal surface(creating a metal-helide?) but in this case, I would think, like before, this could be overcome with recalibration, or infusing the metal with helium in the first place; something I would think they’d have done if it’s something that so easily happens over time.
Helium leakage is documented as a thing that can happen, that much I do know, but the question is really why. It’s everywhere, and if it can leak out at atmospheric pressure, it can leak in just as easily. More easily once the pressure inside the drive drops.
Helium is very light, once it leaks, it floats up and eventually escapes Earth’s atmosphere. If you have a leaky container, air will eventually take the place of Helium and you’ll have no Helium in short order.
This is correct, all helium filled drives are as hermetically sealed as possible.
Helium can’t leak back into the hdd because the atmosphere is less than 0.0006% helium, other undesirable gas molecules can however leak back into the drive (very very slowly).
Helium is so rare that it was literally observed/discovered on the Sun before we found any on earth.
There are two modes of helium depletion in defect-free hdds:
Helium diffusion
Helium diffuses across seemingly solid parts of the hdd such as the potted holes in the body of the hdd that carry signals from the pcb to the spindle/arm/head inside the drive while atmospheric gases very very slowly diffuse into the hdd. This process almost isn’t worth talking about because of how slow it is, a hdd could spend many decades without helium issue if this is the only depletion mode (there’s a FET joke in there somewhere), the spindle bearings are liable to go out first over these time spans.
Pressure-aided Helium dilution
As a helium filled hdd is run to temperature, helium within reaches equilibrium pressure with outside environment, the hdd is spun-down or powered off and grows cold causing a partial vacuum within the drive and atmospheric gases diffuse into the drive at an accelerated rate due to the pressure differential. Over many cycles appreciable losses in helium can occur. (this effect can also happen from cyclic altitude changes)
Helium is a unique gas, it is monoatomic where as pretty much every other gas save for the noble gases are atleast diatomic; that very roughly equates to a gas that is twice as small as the next “closest” gas which makes it very easy to leak or diffuse out any container.
While I’m on the subject of hdds, I don’t get why everyone is calling them spinning rust all of a sudden, the platters don’t contain iron anymore. Back in the day it would be accurate to call them that, but if we’re going to use throwback terms why not just call them the OG hdd slang, Winchesters? (which has an infinitely cooler origin story)
/Rant
If I’m not mistaken that value is derived from the amount of drag the platters experience rather than a bespoke sensor; more drag=less helium. The drag is calculated from a temperature corrected spindle motor current reading in firmware.
Ah, I see where I was mistaken then; I didn’t realize helium was so rare. I thought it would be like Hydrogen, where it just goes everywhere and can’t be kept out of things.
Still, that should mean that a lightly pressurized helium chamber aught to be able to restore the function of a helium-based drive that has failed due to helium leakage if you let the drive sit in it a while, but if the helium has been successfully replaced by other gasses, I guess not.
I wonder how much it would cost to build something like that.
Thanks everyone who relied, this became a much more interesting topic now.
What are the concerns of renewed from reputable sellers for these high density helium drives? Have you bought any? I have bought some about 6 months ago and none have failed (12 total). I am trying to mix and match my drives.
I would think it’s ‘appropriate’ pressure. Near sea-level where most human beings concentrate, the pressure inside the drive should be slightly lower than atmospheric pressure. At higher altitude where some human beings still gather and live, the internal pressure maybe slightly higher than atmospheric pressure
Are there a helium sensor to detect its concentration? How does it get a percentage of non-leakage?
Probably no need for an expensive sensor to analyse it chemically. Since it’s a very controlled environment inside the drive, some sort of temperature-derived sensor should be sufficient to monitor any leakage
Is there still a breathe-hole in helium drive?
I would think it most likely does. The outer seal is airtight. Internally you could think of two chambers inside. The ‘conventional’ HDD + the airtight outer seal. So the ‘conventional’ HDD portion still has the breathe hole…it’s breathing between the two chambers.