Discussion in 'Project Logs' started by No X, 30 Apr 2021.
Thanks for this link, this is the best PSU calculator I ever seen ...
I like it to but would/could be better if detailed info was displayed on power use per component and peripheral equipment, IMHO.
Regarding to your worklog, and the tremendous amount of details you're caring about before even starting to build; I guess you have quite a scientifical approach of things ... your cookbook is a mine of informations about computer operating and it's impressive , me...I'm the opposite on the bell curve , I like unknown paths and put myself in jeopardy , that's perhaps the reason why I don't end all what I start on schedule ... but honestly, the only two PSU calculator I knew were the Bequiet and the CoolerMaster's one, Outervision is way more complete and would certainly fit to 98% of user's queries, but your expectations are raising the bar in terms of accuracy .
The reason I am looking at this project in such detail is that I feel in love with the concept of computers when I was 3 years old back in the early 1970's watching Star Trek TOS (If you didn't live before mobile phones this part might be hard to understand). I loved tech and its true when they say, be aware you might get what you want. I wanted to work with computers and I got what I wished for. The problem with that is it takes away some of the mystery over time, it becomes work not play. I have had the fortune of planning, shopping and installing any type of components that companies work with and I used to love unpacking and getting the equipment up and running. This has slowly changed for me in the last years.
Building this case with the Monitoring and control system is a chance to revive an old love and learn some new stuff about her. Hopefully taking some of you with me on the journey. Getting feed back when working on a project this detailed both positive and negative is a huge advantage and I hope to hear more from those of you that follow this thread. I will in advance apologize for not being that polite I am from Denmark and we don't use a lot of please, may I etc. because for us it's implied.
The Monitoring and Control System has to be a reflection of the real world hence the need to understand the real world mechanics and physics. This is also the reason I might seem obsessed about this part, its an unknown that I might fail so this is what I putt my time and energy into, so I won't fail.
When I feel that I understand , I will have a working math model that to within a few percent is a mirror of the real world (the cooling system), at-least this is my target.
I will have putt together the loops with sensors and have tested if it works (receiving data). Until it works moving forward on other parts of the project seems futile since the shape and functionality of the frame and case to a large degree depend on the SAS4MC and its data, I will explain why later but it is to some degree already obvious in the notes about the sensors and the design of the case fan ducts etc. I will use some time while creating the case to program the SAS4MC. This will likely be a trial and error part of the project, needing real life data and time to complete.
Building the case of glass fiber and carbon fiber fabric I am Only concerned about the precision of the parts, the details should be fixed before I begin.
Until I get the details locked in, to many parts of the design are still in flux. To be honest I like that its not just business as usual.
I am learning a new CAD program onshape its free online. I used to use Tinker-CAD and Auto-CAD but it has become to limited or expensive in its functions for my purpose.
Its like always a steep learning curve, but it progresses to my satisfaction, at the moment.
I have switched my focus to 3D printing since it might make working with Fiber and Epoxy a tat easier. I started with the Motherboard Tray aka the Spine and decided to try and make a negative image of the piece, then placing the fiber on top of the "form" and cut it to size continuously while adding more layers before starting a vacuum pump etc.
This is my 1st 3D sketch in years and I have only been testing this service for 2-3 hours per day during the last 3 days so be gentle.
I have not yet placed all holes, wire and tube routes. Their is a lot more to be done.
This is a follow up to my thoughts about precision being an issue, a 1st draft.
Why a negative? Primarily for precision but also to be able to lay the Glass , Aramid or Carbon fiber material so it looks good. I have checked some videos about car carbon fiber and clothes and a tailors work process to get inspiration of how to do it better.
I took a day off from learning, instead I tested dimensions etc - proof of concept (Pictures are below).
I did also wander into some new software that other makers mentioned its freeware - www.simscale.com for design validation of air flow simulation and more (FYI). I haven't tested it yet will do when I have a better understanding of the new CAD service.
1st the frame then the reservoir, the specs are OK at 1st glance.
A sheet metal mock up for fun!
Found my 1st error on the Reservoir dimensions when adding the pressure valve, its a bit to high and will conflict with the Shell, corrected the dimensions now within specs. To hot to work or study today have a hard time focusing on whats at hand.
I expect more of these errors and will enjoy correcting them before they become a problem when building.
The unwanted height difference can be seen on the first picture, a correction of the side and top of 6 mm and we are in the clear.
A short update I am still learning OnShape CAD, testing my skills and measurements. The following pictures are of the Cooling Rig meant to house the Radiators + Fans, Pumps, sensors and the control system for the liquid cooling.
I decided some time ago to create radiator specific exhaust chambers in the cooling rig, to avoid problems with the performance, when the fans on the Radiators are not running with the same RPM.
I haven't implemented the plinth for the pumps or the boards running the control system on the Cooling Rig, yet. *Oh yea still on my to do list are wiring channels for this part.
The mechanism to secure the Reservoir is now close to hit the mark.
Cooling Rig V 0.7
Radiator side of the cooling Rig:
The Vent chambers when I get around to create the cut outs will have felt on the edges up against the Shell
A detail the Reservoir mounting mechanism
It is always a challenge when your skill level doesn't compare to your imagination and goals!
I am trying to put some curvature into the front and the fan ducts. Curvature like a 1940 pinup is the target if I had the skill.
These are some of the initial pictures there is still a lot to learn and many mistakes to be made.
To see the individual parts of the front I have raised them 5 mm higher then the end result and the surplus material will be sanded away . My initial idea for this part is to 3D print it, when I bye a 3D printer with enough real estate, and then use the print as a negative for a form for the final piece in Fiber material. As per usual agreement more pictures below.
Having more options with this 3D CAD program I got ambitious and looked on my account at www.pinterest.com to find some of the inspirational pictures aka as options that I have saved, lets just say it comes in handy. My ambitious detail is an Air Tunnel in the water cooling side that I have been considering for some time but decided it was way out of my league and it still might be.
1st picture is the inspirational, then comes the Tunnel design that doesn't work yet, but it will. I am if anything persistent and patient. Then my progress pictures for the front.
I am also working on my control system, I did burn 1 board so its a good thing I am training on some cheap boards 1st.
Almost forgot, the reason the cut outs areas are not see through is the curved front surface AFAIK, will have to check up on that.
I did crack the method for the Tunnel took a few starts and resets. Pictures below .
Thinking about using a 3D printer which produces a softer material where it is easier to fix small issues is likely a good decision. The BAD thing about more and new options are well options.
Creality CR 10 S5 is still my printer of choice if any of you have alternatives with a printing area of 500 x 500 mm I would love to hear about it!
Today's update and time to go to bed
An Antenna, An idea for a mounting point and prof of concept the first series of pictures, The antenna is hidden beneath the Shell until it needs to be used. I am not convinced about this solution time will tell.
Next the Water Tube Channels, two top and two low, now embedded into the frame. A channel for the Spine though the Water Tube Channels is cut/created.
Since all this will be 3D printed and then forms made for the Fiber parts, the Water Tube Channels will likely be mounted into the form for the fiber as a silicone shape to make it easier to remove.
The second part to this is the QD3 Male Quick Disconnect No-Spill Coupling that has to be secured to the Frame & the Spine to make it possible to well quick disconnect. I believe zip ties will do the job of securing the QD3 to the frame. A spot for the QD3 is now also embedded into the frame for 1 point the form will likely change.
1st it's a panel to mount the Fans Internally it will be glued in place with epoxy and then, a thin silicone layer/part will be made to fit the plate and insulate the fan vibration and air flow (from slipping away).
Next a mechanism consisting of 2 buttons that are mounted on the inside of the Frame. They are on each side of the Cooling Rig and are to glide into the holes available on the Rig.
The Mechanism is the Button, The encasing for the Button, a Spring and a bolt to adjust the length the Button can travel.
Last but not least the Shell that will glide around the Frame and is locked in place by a similar mechanism.
Their is still a lot to be done End plates that define the route travel by the rig and stabilizes the enclosure in this area is just one of a group of elements.
The challenge with Fiber forms and molds is less of a problem when you only have to connect 2 parts/surfaces, in this case there are multiple areas that have to fit together so it takes a bit of planning especially when one does not do this on a daily basis.
Here comes the picture show:
The marked circles are the space the Spring is mounted into
This marked circle is where the bolt is inserted
A few changes to the present design after looking for errors:
1st a change to the water tunnel channels, the two on top and the two below, on the frame on the spine Since we are working with fiber fabric 90 deg angel is a no go unless your purpose is to break the fiber an weaken the design. Soft bends are implemented and I am still working on the QD3 quick connect couplings mounts. See the first 3 pictures.
The holes for the bolts securing the spine to the frame do also have to be moved so we have 1 hole between the water tube channels. 3rd picture.
I checked the amount of space available above and below the motherboard and the GPU's to be sure that the water channels and QD3 Mount have a minimum of 30 mm of space, we are in the clear.
Cheeking the dimensions to make sure that there is enough space inside the case after implementing the Fan ducts, so a model of a GPU was inserted as box actually 2 with the size of the a RTX 3090. Looking at it and the positions of the Fans it dawned on me that the air flow going under the GPU's ( or lack of) is as a minimum suspect so I stated wondering if this can be done better.
I have been looking at a series of solutions for this problem before and here is he 1st; Mounting the Fan inside at an angle, this is the 1st option that came to mind and I have changed the model for the 3rd Fan duct and thus fan. Looking at this option the 2nd fan should likely also be mounted at an angle. See he remaining pictures.
FYI I was looking at flexible filaments for Panel hinges and springs with 3D printing might be interesting for some:
The difference in the shapes:
The Air Flow:
Suspect air flow if imagining the air under the bottom GPU:
Likely a better air flow more to follow.
The impact on the tunnel seen from the front is not a factor:
A late night and I have added the work flow from my desk top, its 0400 here and time passes very fast when a good audio book (Military Scifi) is is combined with some ideas and a CAD program.
From the end, I have had time to reconsider the solution to a a wire clip after seeing the solution from Tilia v.2 by Age in this forum (Tilia - a minimalistic open case). This is one of the solutions I had in mind a 3D printed clip in 3 parts that is inserted into the Motherboard tray and secure with bolts.
I will add text to the individual part to be added.
3 Part Wire Clip, Well as the header implies this is a clip in 3 parts that
An Update to the Front Panel of the case still with some problems with my lacking skill level. Here its easier to see what direction I am going in.
The Lock for the Frame /Shell connection its mounted on the motherboard tray with buttons going through the frame and into the Shell.
1st idea to present form
3 button will be mounted in the above pictures like in this picture below.
The Frame has also got some additions
1st a base plate for the Cooling Rig (the green plate)
A Red Plate for the Arduino Due and other boards for the sensors.
An added mount on the Water Tube Channels (WTC) for the QD3 Male Quick Disconnect No-Spill Coupling has been added like before it will be mounted with Zip ties. This can also be seen on the picture above.
A Channel with a multi purpose feature in yellow/gold is added for stability also working as a wire channel and for sensors. The Base plate for the cooling Rig is mounted to it and the case back plate is too.
The Back Plate
Back Plate with visible LED cuts in the middle.
I started playing with some, I call them leafs, inserts for the hexagon air vents. The idea being for the leaf to open because of the air exiting the vents. This is for now just an idea since the work space is very limited but it could be beautiful in a geeky way.
I also made a Air Filter frame for the front, just forgot the picture those will be added in the next update.
Mentioned on the last update the Front Frame Filter - the 2nd and 3rd picture. The Filter mesh material will be glued to 1 side of the frame, Magnets will be embedded into the frames and a small piece of fabric like the 1st picture is used to pull the Filter Frame out for cleaning or to turn it around when traveling to minimize the chance of damage to the material.
The Filter Frame is made with an indentation for a belt in the center/bottom that is going back to the Main Frame/Skeleton where it is mounted and can be used to drag the frame out of the Shell.
Last but not least a 1st try on the Docking Bay Mounting, a successful try? not yet but its in the right direction lacing the right air flow, a bolt mount for the 2½" drive and the mount on the Axis going top to bottom. To be honest this example is also a problem for me trying to imagine how to make forms of it, it works as a 3D Print but I am far from sure it would work in daily use it not strong enough. I have on an earlier posting added some drawing of how they are supposed to work.
Latest model revision of the 2½" docking Bay V 0.4 has been improved to now be able to be 3D Printed and then made of Fiber materials and/or of Aluminium. I would love to have this part in Aluminium or at least the frame, lets se how the next generations do. I have tested he Air Scoop a paper mock up with smoke to see how close to the air flow it had to be and it looks good. The Air scoop/Curved back of the frame could and should be made of Fiber materials.
I am still not good enough at creating curved forms to get the model to the point that is aimed for.
The Air Scoop if now functional and has a fin in the front to split the air flow to both sides of the SSD.The Fin will also be used to pack the connections into.
The Top Rail now has a spring that should be able to hold the SSD with an active process and is now functional by my best estimate.
The springs have been added for the Tool less SSD Lock. The (finger) Tap is still to fat and needs to be slimmer to not interfere with the SSD when the drive is being inserted.
There are still some thing to be done, the wires and the axis, to mount the Docking Bays (DB) on are still a concept and not clearly defined. The same is true for the SSD connections aka - SATA (7+15) Pin Female to SATA7Pin +Molex4Pin Power Combo Cable ((SATA22Pin to Data+Power)
I have updated some other parts and will be back soon.
Hi adding some pictures I will follow up with how and why time is at a premium atm.
Reservoir still need to add rubber seals
Water Cooling Rig WCR side mounts and a 45 deg mount panel on top connecting to the Frame
WCR Arduino Mnt on Acrylic panels
Frame insert for WCR with RED lock buttons
Next Picture on Top right MNT Panel for WCR (the RED one).
Fillet Arduino for the RED Panel on the Top left with a round locking add on.
Capacitive Touch sensors actually Hex shaped metal plates with wires going to a board
Front Panel getting closer to final
3 Button Lock now with LED Mounts. The Lights are for the button on Top of the Shell.
Below: Docking Bay for SSD. I like this version but am sill not sure it can't be done with more appeal.
The top rail on the the Docking Bay is able to putt pressure on the SSD since we have different size SSD and his is a tool less Mount.
With SSD size part
Below - An insert for a locking bolt / screw on the curve/back of the bays.
The Docking Bays seen from the front mounted in the air flow from the front Fans.
The entire Docking Rig is able to pivot outside the frame with a toolless lock.
Nice with a Sunday and a few hours to work on projects of interest.
I revamped the front (panel) of the case to get a better viewing angle of the OLEDs and a better air flow. The OLEDs where moved from the center to the low right corner of the panel giving a less complex air flow and most importantly a better view of the OLEDs when the user is placed to the left of the case, as mentioned before if I had been left handed the case would have been mirrored.
To avoid a catch22 where a fault in the locking mechanism could make it hard to get into the case I have decided to mount the front and back panels with screws into the frame. The back panel is the emergency access.
On the front panel extrusions for Mounting have been added.
Sensor panels - Small square panels are added to the front to have the option of accessing and exchanging the sensors with the smallest amount of work possible. It also makes it feasible to have different shapes/angles inserted in the sensor panel area. An example could be laser distance sensors with an angle pointed from the top down to the table 20 cm from the case.
The front Filter frame is also inserted into the model without fabric atm.
Have a nice Sunday!
Almost forgot holes for magnets inserted in to the filter frame have also been added.
* The model is not 100% correct atm since the Frame has not been added before the Shell
Separate names with a comma.