Scratch Build – In Progress Logic - V 01 - The cook book!

Looks like you can print it.
But there is one more issue to tackle.
The resin needs to be UV cured to achieve its properties.
You probably have to cure it from the outside first, and then hang a powerfull UV diode through one of the ports and cure it from the inside too.
Otherwise you retain a smeary surface on the inside, where expoxy won't adhere to.
The last thing you want to end up with is epoxy and resin gunk in your cooler fins.

And one more experience from bad judgement
Drilling printed holes to final size often leads to scrap and frustration.
The flutes from the drill tend to bite into the resin and crack the part, especially big ones (say > Ø6mm)
Best thing is to print marginally under size and use a reamer instead of a drill.
For bigger threads (like G1/4) the core hole can be printed directly.
For small treads I always use brass inserts.

 

I see a lot of complexity coming my way, then again we are not pushing the envelope by being safe and content like most people. I will look into alternatives just in case and see if I can combine some different approaches.

Thank you @dan297 for the input and pointers.

 

Good evening, everybody.

I am in the process of acquiring the skills needed to build the Stand Alone System for Monitoring and Controlling (SAS4MC) the sensors, fans and more.

History - On this forum I posted the inspiration on the 1st of May 2021, and on May 9th 2021 I jotted down some introductory ideas which I hereby start to update.

*A sketch of my current ideas will be posted within 2-3 days, and with my (hopefully) increasing level of knowledge, I will update these again soon.

I have changed the plans for SAS4MC monitor (screen) layout, since the TFT I wanted seems to be, not in store permanently. So my idea right now is to use the Arduino Due with a Multiplexer (TCA9548A) and 8 OLEDs (1.3" to 1.54" with I2C) with a resolution as high as possible which in the near future hopefully will be 240x240. Alternatively - OLED Display 128x128 1.5“ I2C SPI.

This is only ½ of the TFT update --------To be updated-------------------
The 2nd part is the other (slave) Arduino DUE that will be connected to the 1st also by bus I presume. It will have minimum 3 OLEDs displaying the data from the hardware chamber while the 1st Arduino Due shows and handles the Water Cooling system. Why the split to two system? Its to secure the water cooling system can be removed with a minimum of effort ( read not having to remove wires from sensors and Fans ).

Why the Multiplexer? The multiplexer is used since the OLED most often have the same I2C address. If I only used two OLED the address can be changed to 1 of 2 addresses this is not the case for 7-8 OLEDs. The Multiplexer uses its own in this case 8 inputs/outputs to address the OLEDs individually one after one. The OLED are on a 1 to 1 connection/bus on the multiplexer. Link to video You Tube

Why I want to use 7-8 OLEDs, its simple to have enough screen surface to post the important system information in a coherent manner. I could in theory use 1 OLED by cycling the information on the screens it would not make sense when the information appears with a long interval, IMHO.

How does this work with the low data transfer speed of I2C? Because the data shown on the OLEDs is mainly text based.

Examples check the links, its worth it if you are interested in options. All Links are on you tube.

Why an Arduino Due? because memory is the key for multiple displays and speed! Link1 and Link 2.

Why only 2 addresses on an OLED . Link with explanation

Difference I2C and SPI connection. Link explanation.

*As always please correct me if I'm wrong or you have an alternative or input that upgrades the process!

The other news to the build is also a structural part of the SAS4MC, these are Capacitive Sensor Modules (CSM). Why in plural, because I imagine that two CSMs can be used for different purposes.
- One for changing information on the OLEDs, turning on/off light diodes, fans, motors etc.
- And the Second could be used for monitoring the water level in the reservoir and possibly in the tubes. Detailed drawing of my initial ideas will follow soon.

In theory CSM's might also be used to check water quality, since deteriorating water quality might also mean a higher conductivity in the liquid, I think. To be determined.

The sensor of choice ATM is the MPR121 capacitive touch sensor.

How it works - capacitive touch sensor. Link.

Capacitive touch sensor and liquid. Link.

Ups almost forgot should have been on top - POWER for the project

I will try using an Arduino simulator for wiring, power and programming I will post link when I start on it. Should mention when I find a simulator that supports all the boards!

Late ad-on SPI vs I2C on speed. I might use SPI for 1 OLED for diagrams and graphs check video Link for explanation


I'll be back, very soon.

Press the pictures for a full view.

(Updated 22 of August 2021) Arduino to Arduino serial communication Video

A teaser of components that will be used for the SAS4MC in the Hardware and Water Cooling side more to follow.

 

Parts of the project sorted by complexity. This complexity is naturally not the same for all of us , its mainly about knowledge on a given subject when tackling a project like this, and I expect this to change for me in the future.

The order starts with the highest complexity

Creating a working Circuit design.

- I'm currently watching a couple of You Tube channels to stay informed (learn and be update) and I'm 80% sure I've chosen the right components for SAS4MC. What I am missing is a full understanding of low current circuits to be able to select the rest, this will change over time.

Arduino to Arduino programming with breakout boards and sensors.

- Essentially the same as the above. Not the same problem since almost any mistake can be fixed. It's Not like wiring a circuit wrong and roasting all the components.

Choosing fabric for the front (should likely be higher up the latter) .

- Still searching for optimal solutions. I will in the coming days ask for input. The fabric is supposed to work like an air filter and still be aesthetic to look at.

Choosing colors and fonts.

- Same problem as with the fabric, I'm a man. The problem with being a man and this topic is that we know it should not matter, unless we are in the process of creating eye candy / hardware p * rn.

Creating the mold (not the case) for the Carbon fiber PC enclosure.

- I will elaborate within 2 weeks. I got a plan A, B and C and am prepared to pay to make it work.

Creation of an overall design, with as many pros, and as few cons as possible!

- 99% done there is still room for improvement on the minutia. Tube and wire routes being among those.

Creating the Case (not the mold)

- Totally relaxed about this point, when the mold is ready, the case is just fun and joy, even including the challenges.

Connecting the water cooling loops .

- Been there done that, confident it will work.

Gathering the PC components.

- Just waiting for the case to be ready and being excited for next generation of components to appear, these will be included in the project. I don't see any real trouble here except availability and pricing.

Last but not least costs. I have a spreadsheet where I follow the expenses to be able to change supplier and get a better offer on parts.

(Updated on the 22. of August 2021) I am at present estimating the cost around 850,- U$D on the water cooling loops without GPU and CPU water cooling blocks and without the reservoir.

The SAS4MC is around 660,- U$D again I expect this will become cheaper when I start shopping around.
Update end

A personal note. Summarizing project parts helps me keep track of the overall status, this is part of why I do the planning online and in this forum. I have learned a lot and the journey is not over yet.

 

Design Follow up to Reservoir 3D printing. Still in the Design planning phase so this is a preliminary sketch.

Imagine that you look at the case and in the right side the reservoir is mounted. The reservoir has a pivot function. When you tilt it to the right it is out of the case/frame and you can fill the reservoir, with limited danger of liquid flowing back into the case.

The Pressure Valve in the top left (top part of picture) will be used as a fill port. When tilted it is still on top (lower part of picture).

The red line, is the frame/case border of the enclosure. The grey line connecting the black ring is the normal state on top of the picture vs the tilted state below.

I made some initial yellow/orange lines on the reservoir, that border on the middle center line in a soft blue color that is in a 45 degree angel from corner to corner. These orange lines are conductive paint or embedded wires that leads into the reservoir, on the other end all the wires is a Capacitive touch sensor connected described 2 posts earlier. This should make it possible to measure the water level and make it visible on the Arduino OLEDs.

Capacitive touch sensor and liquid. Link.

In the bottom of the reservoir there will be 4 x 45 degree rotary compression fittings for soft tube, a temperature sensor and a light diode. Rotary fitting have been chosen to make it easy to tilt the reservoir on the side. I do hope the tubes will follow on in an orderly way ( I have looked at some options to ensure that it happens).

The dimensions of the reservoir will be 80x80x80 mm or 100 mm on each side not 100% sure yet.

You can if you want to torment yourself look at my drawing from may 12th there you can see the reservoir in the case and a tilt function indicated.

 

Hi

An overview of data available on each single OLED display, the Water loops and the Arduino's with general connections are shown. The sensors are mounted directly to the 2 Arduino's. A total of 12 displays are on the frame. 6 displays on the water cooling rig (The water loop on the right below) and another 6 displays on the hardware side. Changes are still likely as I get deeper into the subject. If you have any input or questions they are appreciated!

I will try to make it more structured later



Overview - Awareness is a NICE TO part for now

 

Front and side views of the enclosure. Marked are Micro switches, the 3 Button Lock, Fans, Radiators, OLED's, Motherboard, GPU, PSU, CPU, Docking Bay, Pump, Reservoir, grip holders.

The lowest mounted Fan on the left is only active when the Docking Bay is open. The system has a Micro switch that activates the fan in this situation.

 

Now with a semi permanent hard drive dock for 4 x 2½" hard drives in the left side of the enclosure. A spring and 1 button that can be pressed to swing the dock open. The drives are normally mounted with a spring for short term use. There will be an option for mounting the drives with screws if needed. This is 1 of 2 docks.

 

Hi

I have been pondering and redesigning the front of the case. Why you might ask, its not an easy question.

I was following the specifications and power consumption for the next generation of GPU's and CPU's and they are rising. I had in the earlier design taken precautions with the option for an Emergency Cooling System (ECS) . This ECS can have many variations when we are talking the cooling part, from an additional radiator, to active cooling etc. I came to the conclusion that I did not want Fans spinning at high rpm 24/7 or have the ECS attached all the time so I decided to implement larger radiators, 2 x 360 mm and still with the option for an ECS solution.

To keep the form factor of the enclosure I decided to remove the front docking bay (dimensions 420x160x60 mm) in the lower front. I still need some docking options for larger drives and have implemented 1 in the front right of the frame and a second in the hardware chamber.

In the mock up I have added hexagons openings in the Carbon fiber shell for exhaust. This is my 3rd mock up and I am still searching for better options when considering weight, hardware security, efficiency etc.

The frame is not depicted yet details will follow ASAP.

 

Looks like you are getting at it...

 

Hi

Left side of the frame seen from the left

The key on this side of the frame is the 3 Button Lock(3BL) that controls the shell position, in the shell there are only 2 holes for the 3 buttons. To get access to the power button, USB and SD card reader requires to press 2 buttons red and yellow on the top of the case and drag the frame 4 cm out of the shell as seen on the last picture.

The 2nd picture, shows of the inside dock with the 4 x 3½" bays that are mounted on a pole. I will likely use glass fiber to create the red rails of the bays to protect the drives from electrostatic charges. The screw holes for the hard drives are not depicted yet. In the button of the pole there will likely be a rubber or silicone ball for chock protection. In the top of the pole there is a spring.

I will be back

 

Yea its getting there. I started collecting ideas for an enclosure that is more of a show piece and doesn't need to meet the requirements for sturdiness. Both will take time I am not moving as fast ahead on the System for Monitoring and Control as I wish, priorities sigh.

I am trying to conjure up new methods for creating a good looking frame of fiber material and found this (see next link) but its a bit more demanding then the process I am working on now.

https://static.dezeen.com/uploads/2...ersity-of-Technology-students_dezeen_ss_3.jpg

 

Right side of the case.

The Emergency Cooling System (ECS) connections are seen on the left side of the top radiator, inserted in the loop after the liquid has been through the GPU radiator in the bottom. Why did I choose that place for the ECS because the GPU's are the units delivering the highest Thermal Design Power (TDP) output and I might have two GPU on the loop. It could be argued that it doesn't matter where the heat is ejected since both cooling loops share the reservoir, but it makes me feel better about the setup and it has the advantage that the connections share the hexagon holes for exhaust. This should make it easy to insert a part of the quick disconnect no spill couplings.

I don't know if there is a U shaped loop in the radiator but I presume so until I get other information . I have made a point of sending the water into the radiator from the bottom to flush the system of air bubbles as fast as possible.

The CPU also has quick disconnect and no spill couplings. On the picture in the bottom right purple lines are used to indicate the water cooling and control system that is detachable from the skeleton in 1 piece.

The parts that are still involved in the cooling system: Radiators, Fans, Pumps, sensors, Control system (screen + Arduino's), Reservoir, GPU and the Quick disconnect fittings to the CPU. This is part of why I chose to use flexible tubes. This is also the reason why the Control system is divided on 2 Arduino's - 1 Arduino in the hardware side for the fans PWM, temperature sensors, Sound level sensor, Moisture sensor and maybe connected to the main system (to be decided). The first Arduino is connected to the other Master Arduino by Serial connection, the Arduino on the cooling system does the rest.

I have added Y connectors on the ECS connector to ensure optimal directional pressure inside the connection. I will elaborate later.

** After thinking about the ECS connections they might have to be moved or as a minimum the on/off switch needs to be moved to a place accessible when the shell is on the frame, why should be obvious.

 

I am still working on some alternatives but a dump of the front seems appropriate.

 

The Design - Dual chamber - Air intake area.

Filter = Air intake area into the Cooling chamber, in the right side.

20 x 40 = 800 cm2
20 x 40 = 800 cm2´

Total 1600 cm2

Partially covered area of intake area:
8,5 x 15 = 127,5 cm2
10 x 15 = 150 cm2
8 x 15 = 120 cm2

Approx 400 cm2 are covered of the 1600 cm2

The 6 x 120mm on the radiators.
6 x 12 x 12 cm2 = 864 cm2

A 7th fan of 120 mm inside the coolant chamber (just in a worst case scenario) increasing air pressure in the hardware chamber if needed.

a 8th fan in the PSU also of 120 mm in size.

I have looked for alternative designs but haven't found any non custom options, meaning handmade radiators (alternative forms and sizes) and more, that would work for my Expectation, Purpose and Hardware.

A 1 to1 intake vs Fan area works since the intake area is only partially block. I did state 120 mm as a square and not a round form while making the estimate so there is some wiggle room.

I will have to check the filter material for efficiency.

BBL

 

The pictures are now on IMGUR i am testing if the resolution is higher. The new images are updated.

I am working on the coolant frame on the right side atm. OK might just be waiting for inspiration that will make the design sexy with the limited space available.

Left side


Front



Right



Top



The bottom frame is still a work in progress. Damn the whole frame is still a work in progress.

 

The liquid cooling part that is unmounted and mounted in 1 complete part.


Updated view top and bottom

 

I just noticed that you are building a cube...
Cool
That is my favourite type of case.
I have difficulties to handle the required foot print on my notoriously cramped desks, but still a sexy design

 

Yes I like it too the cube format.

I did consider a rail on top of the case so it would be able to be mounted underneath the desk, but it would be counter my purpose for the PC. The cube is just the right format for stability, airflow and inside volume in this case.

The foot print is 42 x 42 x 48 cm

What project are you working on?

 

Well, I just finished my router...
My wife asked me to build her a kind of laptop or tiny portable.
Not to carry it around, but to take no space on her desk. (Which is even more cramped than mine )
Very low demanding hardware, so no graphic card required, etc...
Top KPI for her is as small as possible and nice looking ...
I am still browsing for ideas...
Either something with a small monitor integrated in the case, or a sub 4 l case with one of these extra displays for mobiles, or something under the desktop. Don't know yet...