smoothing capacitors over HDD..

[Corrie]

I made an external HDD caddy for 5 drives, the data is connected through Sata>USB>USB hub. The power takes one molex connector and splits it through 5 DPST switches that control the 5v and 12v, HDDs have a constant twin ground from molex. Any of the switches power(12v) a central fan through a diode (to protect the HDD from being switched on with back current from the fan). And each HDD has a red LED+resistor (maybe 450 ohms i forget) on the 5v in parallel with its HDD.

The problem; when 1 or more drives are on, and i turn another drive on the drive/s that were on are re-started. I think its because the voltage dips when a new drive is turned on and the other drives shut down for a second so they need to re-start.

Possible solution; I could put a smoothing capacitor in parallel with each drive (12v and 5v?), in theory they would charge up when the drive was powered up and protect the drive from voltage dips caused by other drives being powered up.

Any suggestions welcome, not sure what size of capacitors to use, or whether i risk damaging the drives

 

[dknguyen]

I made an external HDD caddy for 5 drives, the data is connected through Sata>USB>USB hub. The power takes one molex connector and splits it through 5 DPST switches that control the 5v and 12v, HDDs have a constant twin ground from molex. Any of the switches power(12v) a central fan through a diode (to protect the HDD from being switched on with back current from the fan). And each HDD has a red LED+resistor (maybe 450 ohms i forget) on the 5v in parallel with its HDD.

The problem; when 1 or more drives are on, and i turn another drive on the drive/s that were on are re-started. I think its because the voltage dips when a new drive is turned on and the other drives shut down for a second so they need to re-start.

Possible solution; I could put a smoothing capacitor in parallel with each drive (12v and 5v?), in theory they would charge up when the drive was powered up and protect the drive from voltage dips caused by other drives being powered up.

Any suggestions welcome, not sure what size of capacitors to use, or whether i risk damaging the drives

Not sure how much you would need exactly, but it's not application that requires a specific value. But I would guess 470uF is the minimum. I'd probably try for 1000uF for first try. You want large enough to smooth things out but not so large it causes excessive inrush current (and voltage drop) to charge the capacitors when first plugging the device in.

You could stick a cap next to each drive. But this would mean everytime you turn on a drive there would be inrush current to charge up that drive's capacitor (which may be very similar or identical to the problem you're dealing with to start with). You could also try adding an additional common capacitor BEFORE the switch so that caps are charged prior to switching.

Just don't try too large a capacitor as your first value.

Also, stick a voltmeter across the power terminals and see if it quickly dips whenever you turn on a new drive. That would confirm things. It may be too fast for you to get an exact reading but you'll probably see some signifcant, momentary change.

Oh yeah...electrolytics capacitors.

 

[Corrie]

Good idea putting one before the switches, lot easier too.

Was planing a kit oscilloscope build, maybe i will get that done before i continue with this, as it would tell me whether the 12 and 5 volt are experiencing a dip in voltage and give me a better idea of how much of a dip (would be more accurate than the voltmeter).

Anyway thanks .

 

[tcmtech]

Something to consider is that its not always a supply side dip with say the 12 V rail pulling down to 9 - 10 volts and the 5 V rail going to 4 or something like that.

Quite often its the ground return being pulled up from either being as a hard drive spins up it will have current being drawn from both rails going back through the ground line and the 12 having the HDD motor load on it will draw enough current to easily cause the ground line to raise thusly making the 5 volt rail appear to drop and go far enough down to drop out the logic circuits which shuts everything off.

As for capacitors they will help a lot if they are big enough. I used to have a USB to 3.5" external HDD adaptor that managed to spin up most lower powered larger hard drives from a 1.5 amp or greater rated USB port or a dual 1 amp pair without the need for a external power adapter and it used a rather hefty capacitor banks in both ends of the boost converter circuits for the two rails to help get things spun up without excessive voltage drops on either.

Downside was that on lower powered single 1.5 or lower amp limited USB ports the inrush current to charge up the capacitors alone would cause them to reset or lock themselves out due to over current and the adapter would sometimes do nothing.

 

[Corrie]

Makes me want to put protection diodes on all the earths, if course it wouldn't work its not really reverse voltage but reference voltage you are talking about.. Weird that a capacitor would solve that problem too though, but i can see how it would. thanks