I'm throwing money at my screen but nothing is happening? How do I buy this thing??
-
Save 20% on ALL SFF Network merch, until Dec 31st! Use code SFF2024 at checkout. Click here!
Completed DeskNano x300 - a 1.2L replacement for the DeskMini a300/x300
- Thread starter msystems
- Start date
First look at the custom switch kits. There are over 200 possible customizations of color and style. In the background is Rev 4.0 (production) which will be shown soon.
7/14/22 - Rev 4.0 (Production)
Rev 4.0 "Unibody" Core - Printed with ZERO support
Release Version
I am pleased to introduce the release version and final new features, although development will still continue on if I have time.
The complicated and flimsy modular part system of prior revisions, while excellent for rapid prototyping, has been simplified. DeskNano now consists of a single "Core", reducing four parts to only one. While installation is slightly trickier, this yields excellent dimensional consistency, rigidness, cleaner corners, and is optimized for production. With a quality printer, it can even be produced without any supports when printed inverted, and extremely minimal cleaning time.
There are 6 "Cores" to choose from with different face patterns - of varying difficulty depending on printer capabilities - including the familiar triangular lattices of prior revisions, slotted cutouts, "zipper" pattern, and the above - cubes.
New Feature - 3dMesh Lids
In addition to the "Cutout" style, DeskNano now has a variety of mesh lids to choose from. Thank you @morj for the inspiration.
The pattern and density is not part of the model, but created by the slicer, allowing the user to decide what the mesh should look like.
Although somewhat fragile, they look and feel incredible.
3dMesh L9a+120mm Slim - .4mm Nozzle, single wide "Lines"
3dMesh L9a+92 - .4mm Nozzle double wide "ZigZag"
Similar to the "Cutout" style, 120mm fans can be "press fit". No screws are required.
Not pictured, but supported, are 120mm x 25mm fans.
New Feature - Vesa Mount
Vesa attachment is now supported by way of an adapter plate that mounts to the tray. Vesa mounting with the Storage Bay is also supported (it can be sandwiched).
Project files and storefront should be available very soon.
Last edited:
a300 and x300 only, for now.
B660 is a possibility, but at the moment, it is poorly optimized and I don't recommend it until it gets further bios and microcode updates. It has worse integrated gaming performance, worse multi-core, and a 75 watt hard limit (x300 has 120 watts++). I have the B660 and only use it as a test bed for the on/off switches.
Last edited:
Case is available in Blue Sapphire + White. Allow 2 weeks for shipping. Custom orders with different colors and patterns will be available soon.
For DIYers, I just finished putting up the files & writing up a short DIY guide to assist with building the case. A more detailed assembly guide will come later.
Let me know if questions arise on the build process. There is also a storefront for the hardware and switches if you need it.
Last edited:
Hey @msystems that looks great!
Just a little advice, the threaded inserts you are using are made for cold press inserting, which isn't ideal in printed parts as they put quite some stress on the plastic and the hollow structure may not give enough counter-force to hold them into place, over time. Heat inserting them with a soldering iron is a better idea, but there are better inserts for that. You're looking for knurled nuts without slots, like these:
Just a little advice, the threaded inserts you are using are made for cold press inserting, which isn't ideal in printed parts as they put quite some stress on the plastic and the hollow structure may not give enough counter-force to hold them into place, over time. Heat inserting them with a soldering iron is a better idea, but there are better inserts for that. You're looking for knurled nuts without slots, like these:
That's a great idea! I will do some testing this week on supporting those style of inserts as well and update the guide.
Hasn't been tested scientifically, but to compare all the options, the best absolute performance is likely still the "Turbine" configuration (which was L9a 92mm + 2x 40mm for memory) which could handle about 85 watts sustained TDP while also keeping the memory under control (< 40C always). However the drawbacks were the expense, the wire management, and the noise of the 40mm's going up to max rpm. There's also no way to properly control the 40mm's to respond to memory temp in the Bios. I will eventually make a new Lid to be able to use this "Turbine" config with the DeskNano.
-The Wraith Stealth is a little worse than an L9a, but an economical solution. If you aren't overclocking, the Wraith Stealth is fine honestly. Memory has no cooling.
-L9A with it's stock fan is the best for it's size, but barely cools the memory or ssd.
-My favorite is using a 120mm slim fan swapped on the L9A heatsink, since it cools everything at once (VRM, CPU, M2, Memory), at the expense of higher core temperatures during multi-core loads. During normal usage and gaming (1-2 loaded cores), the cores aren't loaded enough to overwhelm the 120mm. This might not be a good choice if you do heavy computation tasks, like blender, but it's superior for gaming.
Last edited:
Thanks for your detailed thoughts @msystems, as I am running an A300, I should be fine with a DNano setup.
I fitted a Wraith Stealth recently but haven't tested it yet. I think I will go for an 80 mm Noctua first, then a 120mm later on.
To be honest, it's not worth me experimenting at the moment as v.091is still running an Athlon 200GE with a single stick of 4GB DDR4 2133 bumped to 2666!
I fitted a Wraith Stealth recently but haven't tested it yet. I think I will go for an 80 mm Noctua first, then a 120mm later on.
To be honest, it's not worth me experimenting at the moment as v.091is still running an Athlon 200GE with a single stick of 4GB DDR4 2133 bumped to 2666!
Found some heat set inserts that seem to work perfectly with the existing design of the tray. Tested them into PLA and ABS versions, and they held very well.
Specs:
OD: 4.6mm
Length: 5mm
I'll be updating the guide, and making these an option for the diy hardware kit.
Specs:
OD: 4.6mm
Length: 5mm
Amazon.com: in-saiL 100pcs/lot M3x5.0mm, OD4.6mm Heat Staking Threaded Inserts Quik-Tek for Plastic Brass C3604 : Industrial & Scientific
Amazon.com: in-saiL 100pcs/lot M3x5.0mm, OD4.6mm Heat Staking Threaded Inserts Quik-Tek for Plastic Brass C3604 : Industrial & Scientific
www.amazon.com
I'll be updating the guide, and making these an option for the diy hardware kit.
That's a sweet looking mesh! What is your workflow to configure slicer only for selected area of the part to make it a mesh? I have been experimenting with that and were only able to "stack" two different parts with different settings together to make a complex "combined" mesh. But maybe you have a better workflow?
EDIT2: looks like it happens automatically with the right slicer config because of the way you have designed the walls to be "slanted".
EDIT: also in my experience the part of the mesh that is NOT touching the printer bed (so the "upper" one when printing) looks more nice/consistent. But that might be just a personal preference.
Last edited:
I can't control the mesh directly, but noticed its possible to use chamfers to constrain where the vent (mesh) begins. So I set the chamfer in the model then check the alignment in the slicer. Then put the wall thickness to be just enough, so the walls overlap and the chamfered parts appears solid (otherwise you'll get tiny holes in your chamfer).
If the chamfer is at a shallow angle it will require 3mm+ walls. But a sharper chamfer needs only 2mm walls or less.
I'm printing these inverted - because it looks a little better than using supports, or bridging the area where the fan goes. Its not a flat panel.
They still can be improved further by forcing the extruder not to travel across the print and ruin the pattern
Last edited:
Yeah, I ended up finding your STL and playing with it in CURA. No special tuning is required indeed.
I printed a small additional mesh lid for my build by combining 2 parts into one in slicer and applying different infill/wall settings to them, but it's more fiddly, so that's why I was curious about your workflow.
Huh, you are right about that. Mine also has this defect. Do you know how to avoid it?
Possibly by enabling Combing, it might follow the perimeter instead of cutting across. There is also a setting that forces Z hop to "avoid the print", but I don't know if that will do anything
Edit: try stuff and preview it in cura, and have it show the nozzle movements too not just the layer
Edit: try stuff and preview it in cura, and have it show the nozzle movements too not just the layer
Similar threads
