Maximum permissible NVME SSD Height/Thickness in NUC11BTMi9 / 7 & NUC11DBBi9 / 7

[Win7ine]

Looking to find information about the maximum permissible NVME SSD height/thickness that can be utilised in NUC11BTMi9. There is only standard M.2 lengths information available but no mention of maximum height. Want to determine if I can mount a device with manufacturer fitted heatsink.

[n_scott_pearson]

You cannot use any heatsink. A thin heat spreader is all that you can use - or need. The NUCs provide conductive foam that allows the heat from the SSD to be dissipated through the NUC's chassis mass. While this foam works for SSDs with no heat spreader, IMHO, it can be more-effective if there is a heat spreader. Note that the more that the foam is compressed (by thicker heat spreaders and heatsinks), the worse the effective heat conduction.

...S

[Win7ine]

Wow, that is some statement @n_scott_pearson . What I can or cannot do is just what I am trying to determine hence asking for dimensions! You have obviously not seen the internals of a NUC11DBBi9 where the three available NVMe sockets have spring loaded

pressure pads attached to the (plastic) case cover above them with said heat conducting pads of say 1.5mm thickness and where the height and pressure of said pads is adjustable with a couple of screws both ends to adapt to different package heights of NVMe SSD's. 

The fact that the same SSD runs at 10C higher temp when idle in my NUC11BTMi9 compared to my NUC11TNKi7 is what makes me

want to use additional cooling for the directly to the CPU attached SSD, physically crammed between RAM and CPU.

 

 

[n_scott_pearson]

Hhmmm, I am not sure that you aren't reading something into this design that wasn't intended. AFAIK, the 'floating' design is to ensure that the thermal pads are not unduly compressed (which would reduce their thermal transfer efficiency). The higher temperatures may simply be the result of there not being as much metal mass into which heat can be dissipated. Still, I will contact the developers and ask (I agree there is no mention of this in any of the documentation). Unfortunately, with the U.S. Thanksgiving holiday, it may be Monday before a response is forthcoming...

...S

[Win7ine]

The design is such that in default position thin metal bracket in the plastic cover which has the heat conductive pad stuck to it is furthest away from the surface of the SSD. In fact, in this position, the pad barely and unevenly made contact with the 980Pro I use as the controller chip is slightly prominent and higher as the rest of the chips. Only when releasing the screws do the springs come into play and the entire assembly closes in onto the surface of the SSD. There is at least 5mm of adjustment in the screws/springs. But I am only guessing and would like to have numbers I can work with. Appreciate your help @n_scott_pearson 

[n_scott_pearson]

Surprisingly, I was able to contact the design team manager this morning. My original assertion was correct; there was absolutely no allowance for heatsinks to be used in any of the NUC models or NUC Compute Elements. A relatively thin heat spreader is the most that can be used.

Ok, that said, I am concerned with your assertion that the thermal pad is not making effective contact with your Samsung SSD. This would seem to indicate that the heatsink-fan unit is not flipping down properly against the board. Are you sure you had it all the way down and lock in? Can you reseat it and see if it is still not sitting properly?

...S

[Win7ine]

All is sitting and locked in properly. Thaks for checking @n_scott_pearson but the "relatively thin" heatsink size is what I am still trying to determine in accurate numbers. There is height adjustment in place for the heat conducting pads sitting on the brackets and what I need is the dimension between these and the surface of a M2 card when the cover is fitted and closed. In fact, I will remove these brackets+pads if total height allows for some heatsink incorporating NVMe SSD's to allow for an extra 2mm or so.

[n_scott_pearson]

Ok, I will ask.

...S

[n_scott_pearson]

Here's what they provided:

1. Thermal pad is 1mm thick, if removed you “gain” 1mm of clearance.
2. Internal heatsink is 2mm thick from flat to top of “ridges”, removal may gain you extra clearance…
1. However the heatsink post is countersunk into the post so really if you remove the heatsink the best you will gain is around 1.7mm of extra height.

Don't forget that you also have access to the STEP files, etc. here: Intel NUC Element Mechanical Drawings.

Hope this covers it,

...S

[Win7ine]

Thank you for the info @n_scott_pearson , I will take a peek at the STEP file after conversion to 3D PDF as I have no CAD Soft running the moment.

[Win7ine · Accepted solution]

Attached is the relevant excerpt showing clearly the spring-loaded adjustment capability of the brackets with the pads and available extra height if discarding it all. Only thing missing is an accurate number to work with still.

[n_scott_pearson]

I have gotten you all of the information that there is to be gotten. If you want to continue with this stated unsupported activity, you are on your own.

...S