SSD boots even with power switched off ?

[Sebastian42]

I was hoping someone could just make sense of the strange phenomenon I describe below.
I use four HDDs each on three PCs. I need to change between HDDs a lot, and I would rather avoid removing and replacing leads, so I bought 4-switch units with the form-factor of a strip such as is normally on the end of a PCI card. There is a LED indicator associated with each switch. After some weeks, the 4 switches lost their independence and so the unit became less useful. So I bought a number more of them but the experience was repeated with each 'set of 4 switches'. I have also bought one at twice the price of the originals - its form-factor resembles a floppy disc drive. So far it has behaved as required and desired, but it is early days yet.
The difference experienced in the three PCs :
PC1 - Four spinning-disk HDDs - the 'strip' switch works fine.
PC2 - Three SSDs and one spinning-disk HDD - The strip-switch LEDs light up 'randomly'; the more expensive switch has worked up to now.
PC3 - Three spinning-disk HDDs and one SSD - The strip-switch LEDs light up 'randomly'. Having figured that the mixture of SSDs and spining-disk HDDs on the same strip-switch spells trouble, I connected the strip-switch to only spinning-disk HDDs and so far it has behaved well.
To be able to isolate the SSD, I fitted a molex-power switch between the SSD and the PSU, with (obviously) a molex-to-SATA power adapter to the SSD. The SSD ignored the On/Off setting of the molex-power-switch, and so the PC booted from the SSD irrespective of the molex-power-switch setting. I have tested the switch unit of the molex-power-switch with a digital multimeter which showed low resistance when closed and high resistance when open, so the switch WORKS.
1. It seems that an SSD remains powered, when a spinning-disk HDD would be off.
2. Swtching with both both SSDs and spinning-disk HDDs connected, causes problems EVENTUALLY.
Any enlightenment would be welcome.

 

[Tim Locke]

Tell us more about these switches. At least tell us what brand and model number so we can look them up.

 

[Bighorn]

Ditto. Pictures of the setup would also help, there's a lot of hardware mentioned. Are the computers set up as experimental or as Servers? Does the switching hardware work with the drives as "hot-swappable"?

 

[Sebastian42]

I include photos of the 'strip', floppy-disc-form-factor and molex switches. The trip brand is 'Star Empery Model PT628; the floppy-disc-form-factor is PELOTE Technology Pilot model HD-PW4101, and the molex switch is unmarked, so I can't deduce a brand or model.

I have often heard the term 'server', but do not have a clear understanding of it; all three PCs are standalone towers in a home. I believe that 'hot-plugging' applies to the way I sometimes operate them. All three PC's support hot-plugging, although they are AT LEAST 6 years old.

The pictures below are shown in an order NOT-planned by me; top : the strip; middle : molex switch; bottom : the floppy-disc form factor.

 

[Bighorn]

A more-modern version of "the floppy-disc-form-factor" case insert would be the USB/Memory card reader/Headset/Microphone case insert as seen in Desktops of the last few years.

But the question remaining in my mind would be about the hot-swappable part. Back in my 80486 days I used a switch to change drives but always had to first shutdown, flip the switch then boot back up. I had 3 x 345MB drives in the computer and that was the only way to get the drive changes recognized.

 

[Clintlgm]

I really don't understand what your doing with disconnecting drives and all? USB 3.0 or eSATA external drives would make more sense and you would just switch them on or off.
As far as your SSD's staying hot with the switch on I don't think that is possible a power supply is required to run any ssd. So I would suggest a continuity test on that switch either power is leaking through that switch or your ssd is getting power from another source.

 

[Sebastian42]

Whatever your experience with 80486 was - the floppy-disc-form-factor unit has no problem with hot-switching the SSDs and even the spinning-disc HDD; and AT FIRST, neither did the 'strip-form -factor' unit (bit that did not last). I can live with having to get the more expensive 'floppy-disc-form-factor' switch for a mixture of SSDs and spinning-disc HDDs; but I am baffled by the SSD booting, when when the power switch to it is turned off.

 

[Sebastian42]

'I really don't understand what your doing with disconnecting drives and all?' I don't think you need to understand my complexities. I'm not looking for a work around, but for UNDERSTANDING the phenomena I described.

"eSATA external drives would make more sense" They may well, but I do not have them and do not see the need to invest in them. My system WORKS - I just want to understand the bizare phenomena I described.

'a power supply is required to run any ssd.' I totally agree, and that is why I have posted the observation that the SSD will boot even when the molex switch in its power supply is turned off.

'always had to first shutdown, flip the switch then boot back up.' That may well have been your experience, but my floppy-drive-form-factor switch works fine with hot-plugging, and so did the strip-switch AT FIRST - and even now, so long as only spinning-disc HDDs are connected.

'So I would suggest a continuity test on that switch' I reported that I had done that. When the switch was open resistance was high, and when closed resistance was low.

'or your ssd is getting power from another source.'
This is the most likely scenario, and the reason I did the post - to get to understand about the difference in the way that an SSD is powered from the way a spinning-disc HDD is.

 

[Bighorn]

"'So I would suggest a continuity test on that switch' I reported that I had done that. When the switch was open resistance was high, and when closed resistance was low."
.
My Navy training for Aviation Electrician showed that when a switch was open or in the Off position there should be no continuity when tested with a multi-meter or Ohmmeter. Showing high resistance or low resistance instead of no resistance/no needle movement on the meter suggests a bad switch. The Off position of a switch should have the same effect as if cutting the wire to it.

 

[Sebastian42]

I understand that and agree with it. Since my analogue multimeter has died, and I am left with just a digital, I can not report on a NEEDLE movement, but only on the digital display value.

 

[Bighorn]

I understand that and agree with it. Since my analogue multimeter has died, and I am left with just a digital, I can not report on a NEEDLE movement, but only on the digital display value.

That's the difference between back in '60 and the change of technology. I like the small inexpensive multi-meter from Ace Hardware as it is easier to see needle movement in bright sunlight.