• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

ZFS Super capacitor alternative.

E

enuro12

Gawd
Joined
Jan 25, 2004
Messages
781
Would a capacitor on the 12v line running into the SSD provide the same function as a supercap allowing it to finish writing?

Building it would be ultra simple and provide a second or two of power without any trouble. Considering the write time is in the micro seconds, could that be a viable alternative? (Or double backup to my zil with super cap anyway)

It seems to me the only way it wouldn't work is if a supercap ssd gets a signal that power loss has happened and it's time to flush writes. I was thinking "we got power, but no data, lets finish these writes" would be the DIY supercap solution.


This isn't a discussion of if you should just buy a ssd with a super cap. The question is would it do the same thing or not.
 
Would a capacitor on the 12v line running into the SSD provide the same function as a supercap allowing it to finish writing?
Click to expand...

No. Not as simple as you say.

Considering the write time is in the micro seconds
Click to expand...

You would have to send a command to tell the ssd to flush. Also disconnect the ssd from the rest of the system so that the charge from the capacitor does not discharge to the rest of the system.
 
It wouldn't do the same thing -- at least, not simply. Since the firmware on the SSD doesn't know it has a capacitor, it isn't prepared to check for imminent power loss. At that point, an SSD that has a supercap refuses further writes, isolates in-process writes, and flushes confirmed writes that are in volatile onboard memory. Externally, you'd have to issue commands to do the same.
 
From the electrical perspective, two points to note:
a) A diode should be placed before the supercap, in order to keep the SSD powered longer than the rest of the motherboard.
b) Some SSDs are powered from the 5V rail, so the supercap + diode should be placed there.
 
As others above where hinting at, this does not solve the problem.
Yes, lots of ssd these days are only using the 5v line.

But making that like have power from a cap, will not solve the problem, the ssd sure, will stay powered up longer, maybe even a FULL second longer, but since it is powered up, it doesn't know it lost power, and will not flush it's cache, therefor you have the same problem, and it is not solved.

You could power the 12v and 5v line FOREVER to solve it though.
 
I got a 480gb intel 730 SSD with a capacitor for £99 in the black friday sales /gloating

On a more helpful note you could just get a UPS.
 
Also an external cap is going to cause the input voltage to slowly droop causing a brownout instead of just straight loss of power. Problem with a brownout is digital circuits work on the difference in voltage to determine 1 and 0, if the input voltage droops you can have a condition where the circuits can no longer tell the difference between the 1 and 0 possibly leading to corruption. Additionally any cap you put on the input to the SSD will not only power that SSD but anything attached to the same 12V rail. If you have a single rail PSU then basically that cap will be trying to power the 12V on the entire computer. You'd have to put a diode on the line before the cap but diodes have voltage drops associated with them so you'd then be feeding ~11.3V to the SSD instead of the expected 12V.

Generally speaking randomly throwing more/bigger capacitors into a computer is a bad idea.
 
MAuVE said:
From the electrical perspective, two points to note:
a) A diode should be placed before the supercap, in order to keep the SSD powered longer than the rest of the motherboard.
b) Some SSDs are powered from the 5V rail, so the supercap + diode should be placed there.
Click to expand...

A supercap needs a charging circuit, otherwise it would melt away or get destroyed otherwise by the absurdly high current influx when the power is switched on. A diode is not a good solution to block the current flow to the rest of the system. It will cause a voltage drop that can easily exceed the specs of the 5V rail. The disconnect should be done by a power FET controlled by a brown-out detection circuit.

This is generally not a problem, there are ICs available like the one dresherjm posted that already do everything by themselves. What they can't do is to notify the SSD of the power loss.
 
Last edited:
First off it's great to see most of the community understand both the problem and my reasoning. Thanks for all the input guys

I too came to the same conclusion that even though i can create a 'diy super cap' it's still not going to flush the writes. Given enough power, some models will simply flush because they have nothing to do. However that's not really good enough to call a solution.

@Nicholars, you'll likely find anyone looking for a ssd with a supercap is using a ups, but wants to protect from the just as inevitable power supply failure.
 
You must log in or register to reply here.
Back
Top