Scratch Build – In Progress NUC NAS / HTPC / Desktop by Wolowitz

A big bezel front might fix your gauge woes IMO. -Might be clunkier than you want, though.
Mod the USB adapter?

 

Well I clearly said in my nomination file that the front electronic panel would be optional because it would never be possible to finish it until Septembre. Knowing this I think that my mod lost it best feature, so it was quickly put aside.
It's not an issue that I wasn't selected, I don't need to hurry and I can work on the project when I want.

And illuminate the plexi from the sides/bottom/top and not from behind which need more space?
It may works, but I fear that the luminosity won't be as bright.

The shortest connector available (that I found) is 19mm deep (that blue plastic piece):


The gap between the connector and the aluminum panel is 11+9mm.
I think I will drill a hole in the aluminum panel in front of the USB connector, I need that hole anyway to pass the SATA and power cable, but it's not an elegant solution.
It's not that important, 2 USB 3 on the rear panel (NUC based on Haswell CPU) is enough for me if I manage to place a USB 2 hub for slow devices in the HDD module.

 

Development boards
Tags: #Report_Electro, #Report_Firmware

Hello everybody

I changed my plans because I received last week the STM32F4 Discovery evaluation board, so that I could test for the first time the .NetMF environment and take some measure from I2C sensors I want to integrate in the front panel.




VCNL4000, luminosity & proximity sensor.
Unfortunately I can't get the values for the luminosity/proximity from the VCNL4000 sensor. The registers are all accessible, I can change the settings, but the sensor doesn't seem to make any measure. I think I will order another one, because I suspect this one to be broken.
Or maybe I will only take a luminosity sensor. The proximity sensing was initially a good idea, but the VCNL4000 sensor is limited to 200mm distance (theoretical value), in practice it seems more realitic to say max 100mm. In my case it would be even worse because I will need to use a small plexi light duct. For such a small distance I can replace it by a capacitive/proximity sensor.

NXP LM75A temperature sensor.
I reused an old Mcubed T-Balancer temp sensor I got for free several years ago. In the final design I will use Atmel AT30TSE758A sensors, they are not very different, except that they integrate none volatile memory to store an over temperature threshold value to drive the over temperature shut-down pin even if the I2C communication is broken.


Conclusion.
The firmware development will be very fast I think, it's a good news.
I will have to do performance tests, because I'm not sure if the STM32F4 (even if it is a very powerful controller) with the weight of .NetMF could manage all the LED controllers and color effects.



I'm also planing to remove every direct cable connection to the front panel by using spring loaded contacts.


It wasn't planned initially so it will be a little bit tricky to integrate.

I will need around 16 to 30 pins, depending on the extensibility I want. They will be placed on the bottom like this, I will only need to slide down the front panel, and that's it. But it won't be electrically hot-plug (I will avoird any risk of latch-up of the I/O).





EDIT:
I also got the luminosity/proximity sensor working. Apparently it doesn't like '1's on read-only bits, it prefers '0', I find that really strange.

 

Photo report 2: Base aluminum panel
Tags: #Report_Build

Plan​

​

Bad news, it went wrong, you will see why :-/

1. First I had to do some drawings
   
   
   
2. Some cuts with the Dremel grinder discs
   
   
   
3. Result after the folding process
   
   
   
4. The disappointing result
   The lateral vertical panels are too high, the base isn't wide enough there is 1mm missing. The consequence is that I can't fix the bottom aluminum profiles.
   (The feets aren't fixed. The vice tool is only present because there are no srews fixed )
   
   
   I'm disappointed, but honestly I was aware that the final folding process would be very difficult because there is nearly no headroom for those key elements. Remember that the aluminum panel is only 0.5mm thick, the profiles 1mm, the wood panel 1.5mm thick and the whole thing 140mm wide.
   It is very difficult to get the right folding without specialized tools.
   
   Since I needed 4 hours to build it (and I ruined it in the last 5 minutes...), I don't want to build 5 of them to get 1 really usable.
   So I will go back to my first idea. I will divide that component in 3 parts. 2 sides panels, and the central panel. The disadvantages are that it will add screws that will be visible, and it will also be visible that there is a cut between the bottom panel and side panel :/.
   

Meanwhile I also worked on the specification for the front panel electrical spring loaded contacts:


(The SPI is optional, I may replace it by the fan/temp controller hub or a seconde I2C. There are also some pins missing for an IR repeater)


Sorry for the disappointing results. :/

 

You have my attention!

 

It's hardly disappointing! It's staggering how much you have planned. If there's room, you could use a heavy epoxy on the 3 part version. No one is going to see the insides anyway.

 

Plans are worthless, but planning is everything.




I think I will give the one bloc version another try.

I looked closer today at the piece I built yesterday, it wasn't that bad finally, I identified how I could do it better. I also noticed that I used the wrong dimensions for the lateral vertical parts. I used the outer dimension instead of the inner dimension (40.0 vs 39.5mm), the difference is minimal but combined to other imperfections it results in a 1-1.5mm shift at the end. In fact I didn't know how the aluminum would react to the folding, would it stretch on the outer side? would it compress on the inner side?

I also will use 0.30mm or 0.40mm thick aluminum instead of 0.50mm, it will get sharper corners, but I will have to buy them over internet, it may last a week to get the delivery. So I think I will work on the wooden parts this weekend .

 

Agree with Cheaps here, It's a damn interesting project and although you didn't get the result you wanted, I'm glad you still posted.

 

PM me your shipping info and I'll send you my kit.

 

Photo report 3: Side wood panels and strips
Tags: #Report_Build

Plan​

​

The cuts were done with a simple cutter. It would have taken less time with a Dremel, but it would have required more rework I think and generated much more dust .

1. I also cutted the top and bottom wood panels, but I will cut the round corners only after having assembled the case structure, to be sure that everything will fit correctly.
   (The piece on the left was only a prototype )
   
   
   
2. They won't be assembled like on the photo. Actually, the base aluminum panel will be squeezed between the wood panel and the aluminum profiles we see here.
   
   
   
3. First assembly with all the feet.
   
   
   
4. Added the squared wood strips (the round ones for the bottom and upper corners aren't ready yet)
   
   
   
5. And another one.
   The feet aren't screwed yet.
   
   
   
6. 
   
   

----------------------------------------------------------------------------------------​

I ordered last week black aluminum mesh (Lian Li) on www.caseking.de, it should arrive this week.


I also got the luminosity/proximity sensor working. Apparently it doesn't like '1's on read-only bits, it prefers '0', I find that really strange.


And thank you for your encouragement!


@slipperyskip

Your kit???

 

Yes. I wish you would have responded earlier. All I can do now is recommend you get the kit.

 

You don't need it anymore???

EDIT:
Ok I read that you have abandoned the project on your thread.
Honestly your project were 1 of the 2 I really liked.

I hope everything will work out for you.


About the kit, well since we are only authorized to keep it if we finalize the project at time, I think I wouldn't be allowed to keep it anyway, except if they consider the mod without front electronic panel as finalized

 

There are bending allowance tables online if you haven't found one.
@Skip -You are too cool, man.

 

I have just added 3 photos (3, 4 and 5) to the last Photo report 3: Side wood panels and strips just 4 posts above.

@Cheapskate
Great, thanks!

 

so small, looks good!

 

Good design ~

Good design ~
Our company are interested in your design
I hope you can contact the NUC projects related to the development ...拇指：
Please reply to my E-MAIL
alek.chen @ gmail.com
Thank you ~拍手：