Drive Shutting down - Environmental Safeguards?

@ Bill - I suspect.

 

Can you elaborate on the conditions under which the drive will initiate safeguards to protect itself?

 

The experience that prompts this question is the following.

Out in the car (the noisy bumpy sporty one - which I have not been in with the 'air drive' before) and music stops playing.

Look at the phone and no wifi is showing.

Look at the drive and there are no lights on the drive.

Power switch is in the on position.

Cycle power switch (after some cursing) - drive comes back to life.

Connect up and continue - 3 minutes later - same thing happens.

Repeat this experience (five times) until I gave up and switched off drive.

The drive was self powered.

 

 

The does get HOT in that car, because of the thirsty lump guzzling fuel and the drivetrain heating the cabin - (only 85 were ever made - so its not mainstream) - and it is BUMPY because basically, its more a track car than a road car. The drive has been fine in my regular production car - which is normal suspension and has A/C.

 

I have the drive running on the desk now - self powered and it's not shutting off. (Never has done before) This leads me to assume the drive is initiating some sort of self protection. I wonder if you could explain what the triggers for that might be?

 

Thanks in advance Bill.

ok - so now I see the other thread on this, and I can repro the same issue in my normal car as well :-(( It cuts out, just the same as other people have reported.

 

I stuck the disk in the glove box which has an A/C vent to keep it cool - and it still goes off. So it's not the heat it's the bumps.

I did a test. Placed it on the desk at home and raised one end 1" and let go - drive cuts out. End of test. Conclusive, drive is highly sensitive to bumps and knocks.

 

 

@ Bill - is the air system software responding to the drives accelerometer protection mechanism, or do you have your own accelerometer in your hardware somewhere? In other words is the drive shutting off and then you are shutting off in response - or are you guys using your own accelerometer to determine whether to shut off the system?

 

I ask because "I guess" (unless this can be fixed in firmware) I am building a foam case, or fitting an SSD - the current situation is absolutely unworkable.

http://1drv.ms/1kaI1h4

http://1drv.ms/1kaI1h4

 

your video is normal for platter drives that get dropped on a hard surface,theres no fix for this

I do need to understand in clearer detail, precisely what is happening. There are specific questions in my posts, to Corsair product management that when answered will enable me to take a course of action to mitigate the behaviour.

 

So - "there's no fix" shuts down the conversation and I don't want that to happen until my questions have been answered - I do of course appreciate what you are saying and the information you have provided.

Q: Is my Voyager Air resistant to damage from falling or dropping?

A: The hard disk drive used in your Voyager Air has the same built in protections as a hard disk drive used in laptops. However, users should take proper care not to drop their Voyager Air or allow it to receive impacts, especially when the unit is powered on.

it appears its just protecting itself once an impact or strong enough bump is detected.

imagine an old record player, while a song is playing the player gets bumped. this causes the record to skip and the read data is not in the order it should be so it powers off.

 

thats the best i can do with 1 cup of coffee this morning.

Thanks for that feedback.

 

The operational shock load of a standard drive is:

 

of 2ms duration:

 

300g non operational

65g when reading

30g when writing

 

A laptop drive however is much higher:

 

2ms duration:

 

300g when reading or writing.

650g non - operational.

 

Data pulled from WD.com specifications.

 

Corsair Air uses:

http://storage.toshiba.com/storagesolutions/client/mq01abd-series

 

Environmental

 

Temp - Operating 5° to 55°C (41° to 131°F)

Temp - Non-Operating -40° to 65°C (-40° to 140°F)

Vibration - Operating 9.8 m/s² (1.0G) 5 to 500 Hz

Vibration - Non-Operating 49 m/s² (5.0G) 15 to 500Hz

Shock - Operating 3,920 m/s² (400G) 2ms

Shock - Non-Operating 8,820 m/s² (900G) 1ms

 

 

I can be MUCH rougher on my three laptops and they do not die in the way this unit is displaying. Other people are reporting the same experience.

If you take a look at my video I doubt whether the shock factor on the hard surface is more than 10g.

 

This makes me (and them) ask these questions - because it "seems" this device is way more sensitive than other devices with similar rotating disks.

 

This makes me ask whether Corsair have additional protective circuitry as per the post? The answer may be no, or it may be yes.

 

I am asking because the device as it stands is not fit for my purpose and I need to mitigate that, so that it does become fit for purpose.

 

I can do that in three possible ways.

 

1. Ask Corsair nicely if they would make an adjustment in their next firmware update.

2. I can replace the disk with a 1TB SSD unit. (and invalidate my warranty I know)

3. I can build a foam container to make sure the thing is not knocked or bumped - beyond its capacity under normal use.

 

However - its useless investing any effort, unless the source of the sensor detecting the excessive shock and it's feedback mechanism is correctly understood prior to corrective action.

 

I do hope this clarifies. . .

So - all you customers with this shutdown issue - try this.

 

Power on the drive and before 1 minute has elapsed "gently" throw the drive around. After two minutes have elapsed throw the drive around EXACTLY as you did before 1 minute has elapsed.

 

You will find the following.

 

Before 1 minutes has elapsed - the drive will NOT shutdown.

After 2 minutes it WILL shutdown.

 

Conclusion.

 

The safeguard is NOT coming from the hard disk protection mechanism.

If it was, it would react as soon as the disk was spooled up from power on (after like 5 seconds). However this is not the case.

 

Therefore this is a Corsair implementation issue and not a HD issue.

Watch this video to see for yourself.

 

Because the device takes 75 seconds to power on I am doing the test in reverse. First I am fully powered on (+ 5 minutes) and I am dropping and it is shutting off. Then I am powering on and dropping before the OS is initialised, but the disk is spinning fine and continues to spin just fine through out the boot phase.

 

http://1drv.ms/1kfLo6k

 

 

@ Corsair = can you help here ?

I hope you agree this protection is just far too sensitive to be practical for transportation purposes.

 

The lack of response so far from Corsair - makes me think this is hardware implementation and there is no possible adjustment short of surgery with a soldering iron.

The lack of response so far from Corsair - makes me think this is hardware implementation and there is no possible adjustment short of surgery with a soldering iron.

 

Please be patient and give Bill time to reply and don't jump to conclusions.

My apologies - your right I am getting too wound in - thank you.

I am very happy to wait.

I am working to get the specific details on when the accelerometer / protection circuit takes effect. I have been told that as a hardware function it is not adjustable / user-modifiable.

Thanks you Bill - I look forward to the specifics.

 

I had hoped that the hardware trigger had some interaction with the operating system to implement the shutoff. From that point, some intervention albeit via a re-write of parts of firmware, would be possible to adjust / mitigate the current behaviour. (From the latest video I posted it would seem that the shutoff behaviour is non operable (the drive does not shut off while bumping around) during the Linux boot phase 75 seconds from power on. Which leans more toward 'interaction with firmware', is possible.)

 

Given what we know at the moment, I don't believe an of my earlier ideas, would make an effective remedy.

 

Edit:

Sorry - slow brain here again. I guess the force detection part of your mechanism HAS to be in hardware. I guess it only signals when the trip point is reached and that is probably as you say not adjustable without a redesign and a new motherboard with the updated hardware part attached. So I suspect, ignoring the signal in firmware is the only possible option?

Here's what I got from our engineering folks:

 

We detect whether the drive is in a free fall condition in all 3 dimensions. The threshold is 0.2g. (The lower this threshold, the later detection is triggered.)

 

Still checking on whether it is possible to allow the user to suspend/ignore this (knowing that there is increased risk of fall damage).

Thank you Bill, for you support and efforts on this.

How far can the drive fall while operational before exceeding the manufacturers rated tolerance? (400g).

 

Assuming the distance traveled during the deceleration period is .001 meters or 1mm. (1/25") (This would not be valid for tile or concrete.) (Ignoring air friction).

 

.001 = 1/2 400*9.81 * t^2

 

Solving for t

 

V = a t

 

V = 400*9.81 * .001 = 3.924 m/s

 

If the dropped drive is falling at a speed of 3.924 m/s, neglecting air resistance, it must have been falling for

 

3.924 / 9.81 = 0.4 seconds.

 

and that means falling over a distance of

 

S = 1/2 (9.81 * 0.4^2) = 0.78m

 

Or 78 cm. (~ 2ft.) (Not quite enough for a desk to floor drop on a hard surface.)

 

This is just a theoretical calculation to provide some 'real world' meaning to the figures from Toshiba.

 

Having said that, the disk in the voyager air drive is connected to rubber cushions which are attached to the chassis. It's been really well thought through, designed and built. This would provide a longer deceleration distance in all cases, than calculated above. i.e. more protection and cushioning.

Hi Bill - can you elaborate on any progress in the discussions so far?

Nothing to report yet - a couple of key people are traveling.

Hi Bill - It's been a month since we opened this dialogue without any significant response, (I get your depending on others). Is there any chance that you could set my expectations around a resolution or maybe say when you think there will be communication on this topic from Corsair?

If there is a modification - and we don't even know if that's possible yet - it won't be available in a public release for several weeks. At this point it is best to assume that there will be no change.