Safe temp for SSD

[Nominal Animal]

Open-cell foam, the kind used in upholstery and tool cases, works surprisingly well as a fan holder and vibration insulator.  It's my preferred solution.

The one downside is that using hot wire foam cutter releases noxious fumes, and if using scalpels, the foam dulls the blades faster than anything else I've cut.  Very fiddly work, too.

(I've thought about suggesting the foam block solution for those wanting a really quiet cooling solution for their Rigol DHO800/900 series oscilloscopes, mounted outside using the VESA100 mounts.  You do need to drill new big holes in the fans for the VESA studs, and "isolate" those studs from the fan with foam cylinders, if you also want to use the VESA100 mounts to hold the oscilloscope in e.g. a monitor stand.  Choosing a 3-pin fan that is silent but has good pressure gradient capability at 8V, should make it an excellent solution.)

When fans produce noise without distinct peaks in the frequency spectrum, the "beat" issue cannot happen.  The issue is easy to hear: instead of a "constant" sound, you hear a "beat", when the distinct frequencies generated are off by less than a few Hertz. Its mostly a problem in push-pull configuration, between the pushing and the pulling fan across a heatsink or similar barrier, and especially when they are driven from the same PWM. Then, the small difference in airflow causes the pulling fan to rotate slightly slower, causing the "beat".

When fans are in parallel, the combined airflow tends to "sync" them, so that when driven from the same PWM, they tend to have exactly the same RPM.  Unless, of course, one of them is damaged/older/et cetera, or has an a small blockage to its airflow compared to the other(s), and for that reason runs just slightly slower.

Noctua nowadays sells pairs of 140mm fans (NF-A14x25 G2 PWM Sx2-PP, NF-A14x25r G2 PWM Sx2-PP), that are offset by about 25 RPM (≃ 0.4 Hz) at the same PWM duty cycle, exactly to avoid the "beat" when driven from the same PWM.  Although Noctua says it can happen when fans are in parallel, I personally have only seen (heard!) it in push-pull configuration through a heatsink.

I learned about the workaround when watching machining videos, back when Noctua's largest PC case fans were still 120mm.  CNC mills and lathes vary the RPM sinusoidally around the target to reduce chatter, which is a resonance phenomenon very similar to fan "beat".

[Wilson__]

I mention this because if you want a silent enclosure, and you have multiple fans, you'll want to keep their frequencies apart or slightly varying, to avoid "beat" and enclosure resonance.  For example, Noctua already provides pairs of fans for their CPU heatsinks where one is configured to be slightly slower than the other at the same PWM duty cycle, exactly because of this (although there are also other benefits in a push-heatsink-pull config like that).  Another option is to use different fans: on intake (with filters), ones with better pressure, and on exhaust, ones with more flow, with something in between for any attached to heatsinks.  If you use all same fan, expect suboptimal results!

Does Noctua fans come in a pair with different part number meaning they differ in speed n same PWM in.   OR, you put in the LNA (low noise adaptor) cable.  eem a resistor inside???

[Nominal Animal]

Does Noctua fans come in a pair

In pairs, with a designed ~ 25 RPM difference at the same PWM duty cycle.

with different part number meaning they differ in speed n same PWM in

I don't think you can buy half a pair, only a pair with abovementioned difference.  And they only sell two different pairs, both 140mm models.

OR, you put in the LNA (low noise adaptor) cable.

LNA/ULNA slows them way down.  A small difference in the supply voltage (say, using an 1Ω or smaller quarter-watt resistor) may not actually affect the midrange RPMs on PWM fans, depending on the fan.  Better use a controller that can drive the two fans with slightly different PWM duty cycle.  (For 3-pin voltage-controlled fans, a small-valued resistor should work fine.)

I'm not sure if you can buy them in stores/eBay/AliExpress, but making a "splitter" that reads the PWM duty cycle from the motherboard connector, passing it and the tachometer directly between the MB and the primary fan, while controlling the "slave fan" to a slightly different PWM duty cycle, is not difficult or expensive at all.  Even an ATtiny or Padauk might suffice, although using a beefier ARM or RISC-V with two separate high-resolution timers would make it easy, and you could control the actual RPMs of the slave fan or both fans instead of just the PWM duty cycle.  As to cost, you need at most three BSS138/NX138AK MOSFETs, a regulator to power the MCU from the 12V line, bypass capacitors for the MCU, and some resistors.  If you add a MOLEX 4-pin male-female connector to tap into the 5V line, you don't even need the regulator if the MCU can work on 5V.  Very cheap, if you don't mind heatshrink-sleeved construction with just 12 wires poking out to three 4-pin PWM connectors (one male, two female).

[Wilson__]

Open-cell foam, the kind used in upholstery and tool cases, works surprisingly well as a fan holder and vibration insulator.  It's my preferred solution.

I put 3M 1mm thick double-sided foam adhesive tape on 4 screws of fan and not to tighten screws too much.  Also on fan "tray" screws. Help to reduce noise a bit

[Berni]

Vibration isolation with still having a normal screw in doesn't really help much because the screw is still in rigid contract and transmits the vibrations across. It would also need some rubber washer between the screw head and fan frame.

If possible i like to use rubber isolation pegs like these for example:
https://www.amazon.nl/-/en/Xiatiaosann-Reduction-Silicone-Anti-Vibration-Computer/dp/B0C13FYZ6C

The downside is that they don't always fit in every situation since they are mostly just meant for mounting fans to the PC case, but a bit of modification with side cutters can help them still fit in some situations.

[Nominal Animal]

My quest for very low durometer silicone (in tubes, like regular gasket silicones) is in part for exactly that.  It needs to be soft enough for a pea-sized pebble to deform almost to a disk when pressed between two fingers.

The silicone can form the kinds of soft rubber pads that you see many PC case fans with nowadays, but you can also "plug" the fan mounting holes.  Then, one can use e.g. long (35-40mm) M3 screws with large washers (again with a thick coat of soft silicone rubber), so the only thing the fan contacts is the rubber.  Even then, using a foam pad between the mounting surface and the fan helps.  The silicone only ensures the vibration does not conduct through the screws.

If you drill the holes even larger, and for VESA100 mounts 3.5mm inboard (at exactly 100mm center-to-center distance), you can use foam cylinders and foam pads instead of silicone.

Note that for each screw, you'll want 3 washers and nuts: two surrounding the mounting plate, ensuring the screw is perpendicular to the mounting surface –– and here the silicone is much better than foam.  The third washer and nut is on the outside, and here foam suffices.

The simplest option is to get sheets of foam, one exactly 25mm thick where you cut the hole for the fan.  Gel-like cyanoacrylate glues have worked well for me, and you don't need much; center and corners has sufficed for me.  (I tried to cut the corners so that the fan would stay put even without glues, but didn't have the tools: scalpel deforms the foam, and hot-wire cutter needs thick enough wire to hold in a rigid L or U shape.)
Then, sandwich the fan in 5mm to 15mm thick sheets, depending on how much depth you can give to the fans.  These only have the round center hole in them for the air.  On the outer end you need a semi-rigid surface (rubber sheet is rigid enough) for the screw nuts to push against.

Additionally, against any rigid surface, you can use rubber-backed adhesive sheets with foam on top like SilverStone SST-SF01 and similar.  Open-cell foam has a lot of friction, and two sheets of foam will not generally slide at all.  The rubber is good for dampening vibrations on stamped sheet metal, and adhesive backing keeps it in place.  This way, the screws have almost no pressure at all, they only keep the assembly in place against gravity.  The second washer and nut can then be plain, crushing the foam between the rubber sheet and mounting surface.  The circle you cut out is still useful for the washers and nuts on the other side of the mounting surface.

If you don't mind the look, nylon locknuts and M3 threaded rod makes this quite easy to experiment with.  My local hardware store sells these, and they're cheap.  In the US, consider #4-40 threaded rod: it is thinner, but that's okay, as the weight of the fan and the foam is easily borne by even smaller.  I'd use M2 if that was more easily available, although it starts to bend a bit too easily at 40mm+ lengths.

For VESA mounting both the fan and the stand, I'd use thicker bolts, maybe M4 or M5, depending on the total device weight.  Another option is making a bent steel bracket with two VESA100 mounts side-by side (one on top of the other); the bends giving ~ 1-2mm sheet metal the required stiffness.  That way, the fan intake is not blocked by the stand mount.  (This is wrt. Rigol DHO800/900 series oscilloscopes and similar.)