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

Hi I am almost back, Real Life called.

Oh well I started by checking on new hardware and ran into new ideas implemented on hardware at Computex 2023 and 2024.

IF you have seen design ideas like these below give me a like.

Havn HS420 - Youtube - Der Bauer

Montech. - Computex 2024 - Gear seekers - pull tap

Kins95 Radiator placement.

Antec Performance 1M - https://www.anandtech.com/show/21443/antecs-miniitx-chassis-can-house-a-geforce-rtx-4090

Noctua quiet air flow mod - Fans air layering with focus on sound volume - there is room for improvement.

Fractals ERA2 & MOOD - A unibody its beautiful.

Pogo pins in PC cases showed in 2023 there was 1 more but I cant remember them right now.

Inspiration and ideas are contagious its just small things seen on Computex 2023 and 2024 - Well I approve good ideas should multiply.

I will add some new ideas I noticed on Computex as seen on YouTube later as Picasso said Good Artists Copy; Great Artists Steal.

 

"Longevity? A back-connected motherboard is a wrench in the gearbox!

I thought my plan for ensuring the longevity of this case was nearly bulletproof, but it's not. The expectation was that by allowing for the insertion of a new motherboard tray, any challenge posed by a new motherboard design could be addressed, thus securing longevity through modularity. However, it's not that simple."

"The current design faces a significant issue that is nearly a dilemma. Typically, a problem offers multiple solutions, whereas a dilemma restricts us to two options, both of which are undesirable."

"Looking at the challenge shown in the pictures below, the problems with inserting a new type of motherboard are evident. The new cutout on the tray for a back-connected motherboard interferes with the airflow (recirculation and inter change between the 2 chambers), power supply unit (PSU) and access to wiring and cable routing, with cable routing being a significant challenge on its own."

looking at the options:

1-1: Shifting the PSU
2-0: changing the dimensions of the case to:
2-1: get room for a secondary motherboard tray,
2-2: a larger Water Cooling chamber for the PSU
2-3: or a wider Spine for cable routing inserted, see 2-0.

The pictures only show part of 1-1 and 2-2 essentially the problem no solutions.

Here's a refined and coherent presentation of my current status and the issues i'm facing with the motherboard tray design:

Present Status: Rear View



Revised Inspiration for Cutout Location



Cutouts Seen from the Water Cooling Rig Side (Backside of Motherboard Tray)



The blocking of two likely power insert areas is obvious. There will also be problems with the cables from the PSU interfering with access to the motherboard.

Attempt to Spin the PSU to the Side ![PSU Spun to the Side



First, I tried to spin the PSU to the side. This didn't fix the problem with access and introduced the next challenge with airflow.

Creating access to cables, and te next problem is obvious.





Creating access to cables introduces the next obvious problem. Interference with Fans on the Water Cooling Rig and Air flow

I have several options in mind, but all are quite challenging. I'll be back with more solutions.

After looking at the 2024 edition of Computex and companies like Fractal and others looking as if they have been inspired somewhere close to home, I decided to step up my game and work on the MK 02 that is more scifi inspired. Meaning work on this design is atm a secondary priority that will be updated when I relax. I learned a lot from this exercise and hopefully will learn more

 

This is part of the SAS4MC aka the Monitoring and control system.

To make this work I use the sensors previously described. The temperature sensors before and after the radiator and the flow meter that gives us the water speed an amount kg/s. This is a data based approach from 1 loop. This is a way around the lacking data on radiators ability to disperse energy during specific conditions.

On a personal note the companies that make quality radiators should test each radiator individually and use the data for marketing.

This approach can be used for both energy added and dissipated. Its not a beautiful scientific method / approach but it works and meets my wish of showing/knowing what is going on in the loops. It combines the effect of fan and liquid speed through the system on the temperature in system. the sensors precision are part of why I use this approach.

To calculate the heat transfer in a PC radiator loop, we can use the basic principles of thermodynamics. The heat transfer rate (Q) in this case is related to the change in temperature of the water as it flows through the radiator, the flow rate of the water, and the specific heat capacity of water.

Assumptions:

The system is steady-state.
The water is in-compressible.
No phase change occurs in the water (it remains liquid).

Equation:

The heat transfer rate Q can be calculated using the following equation:

Q=m˙⋅cp⋅ΔT

Where:

Q is the heat transfer rate (in watts, W).
m˙ is the mass flow rate of the water (in kg/s).
cp is the specific heat capacity of water (approximately 4184 J/(kg·°C) at room temperature and sea level). 4184 J/(kg·°C)
ΔT is the temperature difference between the inlet and outlet of the radiator (in °C or K).

Steps to Calculate Heat Transfer:

1. Determine the Mass Flow Rate (m˙\dot{m}m˙): This is the mass of water flowing through the radiator per second. It's usually given in kg/s. 0,1 kg/s

2. Measure the Inlet and Outlet Temperatures:

1. Tin is the temperature of the water before it enters the radiator. Tin 40 ∘C
Tout is the temperature of the water after it exits the radiator. Tout 35 ∘C

3. Calculate the Temperature Difference (ΔT):

ΔT=Tin−Tout ΔT 5 ∘C

This gives the temperature drop of the water as it passes through the radiator.

4. Plug the Values into the Heat Transfer Equation: Once you have}m˙, cp, and ΔT, you can calculate Q.

Example Calculation:

Let's assume:

The mass flow rate m˙= 0.1kg/s.
The inlet temperature Tin= 40∘
The outlet temperature Tout=35∘C
The specific heat capacity of water cp=4184 J/(kg*°C)

Then:


2. Using the heat transfer equation: Q= 0.1 kg/s * 4184  J/(kg·°C)* 5°C 2092 W

So, the heat transfer rate is 2092 W

This means that the radiator is dissipating approximately 2092 watts of heat from the water in this example.

This approach can be used for both energy added and dissipated. Its not the approach I would have preferred but it works and meets my wish of showing/knowing what is going on in the loops. It combines the effect of fan and liquid speed through the system on the temperature in system. It also makes the Water Cooling Rig able to stand alone and be used to compare with parts of other systems.

i will add a few more articles to show the way I approach the data in the SAS4MC. If you got questions post a reply.

The specific heat capacity (cp) of a liquid refers to the amount of heat energy required to raise the temperature of a unit mass of the liquid by one degree Celsius (or one Kelvin). The specific heat capacity depends on the type of liquid used in the system. Will be used in the Cooling Loops.

Water Water is commonly used in cooling systems due to its high specific heat capacity, which allows it to absorb a significant amount of heat before its temperature rises substantially. 4184 J/kg\cdotpK

Water-Glycol Mixture a 50/50 water-ethylene glycol mixture The specific heat capacity of the mixture is lower than pure water, which means it will absorb less heat per kilogram than water alone. However, glycol mixtures are preferred in certain situations for their lower freezing points and corrosion protection. 3300 J/kg\cdotpK

Ethylene Glycol (pure) Pure ethylene glycol has a lower specific heat capacity compared to water, meaning it heats up more quickly for the same amount of absorbed heat. 2300 J/kg\cdotpK

Propylene Glycol Propylene glycol is similar to ethylene glycol but is less toxic and often used in applications where safety is a concern. 2400 J/kg\cdotpK

 

Next I will add some info on Initial conditions: Ambient since the entire System relies on environmental conditions, making it a dynamic system dependent on external factors such as air temperature, moisture, and pressure. This will be a detailed breakdown of my thoughts, math and physics as far as I understand it.

To be posted soon, i need to make it understandable first.

 

! A short comment on why I think that Air density is important. My initial idea was to be able to control the environment inside the case knowing how the ambient conditions are, how the parts of the cooling system work and are controlled (Fans, Radiators and Pumps). After considering the many variables I have come to the conclusion that it needs to be calculated and manipulated on the Go, the system reacts on input, so not predictive but reactive. This is my thought on this for now.
If I ever get to a point where I think Machine Learning "AI" is the solution I might add this but for now this is it.

The Stand Alone System for Monitoring and Control (SAS4MC) involves the Hardware Chamber and the Water Cooling System (WCR) that relies on environmental conditions, making it a dynamic system dependent on external factors such as air temperature, moisture, and pressure. This is a detailed breakdown of my thoughts, math and physics as far as I understand it.

The system incorporates sensors BME280 in all chambers including the 3 tech chambers on the latest case iteration.

BME280 is using the I2C communication protocol and delivers temperature, humidity, pressure, and estimates altitude. These 5 sensors are part of the base delivering data for calibration, ΔT, ΔRH%, ΔPa. The difference between the Ambient and single parts of the case is the key to get things on point.

Initial conditions, Ambient:

Altitude (from sensors): Above sea level manually or from sensor = 0 meter
Pressure = 101,325 Pa (standard atmospheric pressure at sea level) Manually added = 101,325 Pa
Temperature (From sensors): = 25 °C
Relative Humidity (RH%) (From sensors): (RH%) = 60%

Basics - Constants:

0 °C in Kelvin = 273,15 K
liters per cubic meter = 1000 liters (l) or kg at sea level.

Specific gas constant for dry air (Rd): (Rd): = 287,05 J/(kg·K)
Specific gas constant for water vapor (Rv): = 461,5 J/(kg·K)

H2O Temperature from Sensors in Reservoir, L1 and L2. equal to air temperature after longer inactivity 25 °C - More on this later!

Air Density (Ideal Gas Law): ρ=(P/(R⋅T))
p air density (kg/m³)
P air pressure (Pa)
R specific gas constant for air (J/kg*K)
T absolute temperature (K)

This Table shows common maximum saturation vapor pressure values at certain temperatures (in °C) at sea level, these values are used to calculate the needed values:


Temperature (°C) Saturation Vapor Pressure (kPa)

10°C 1,227 kPa

20°C 2,338 kPa

25°C 3,169 kPa

30°C 4,246 kPa

35°C 5,622 kPa

40°C 7,375 kPa

50°C 12,336 kPa

The air density directly affects the cooling performance of a radiator. When air density increases, the radiator becomes more efficient at transferring heat from the liquid to the air, enhancing cooling performance. Conversely, reduced air density decreases the heat transfer efficiency, reducing the cooling performance, even if the liquid flow rate remains constant.

When working with weather data think of a layer cake. The situation at the bottom defines the weather above, lets just say its all connected.

Saturation vapor pressure at 25°C (from a reference table): from Table manually inserted need to program 3,169 kPa

Step 1: Convert Temperature to Kelvin T(K)=25+273.15=298.15K

Step 2: Calculate the Partial Pressure of Water Vapor(Pv)
Pv=Relative Humidity×Saturation Vapor Pressur Pv=0.60×3,169Pa=1,9014 KPa

Step 3: Calculate the Partial Pressure of Dry Air (Pd)

The total pressure is the sum of the partial pressures of dry air and water vapor. Therefore, the partial pressure of dry air is: Pd=P−Pv Pd=101,325 Pa−1,901.4 Pa=99,423.6 Pa 99,4236 kPa

Step 4: Calculate Air Density (ρ)

Now, using the formula that accounts for both dry air and water vapor: ρ=((Pd/(Rd⋅T))+(Pv/(Rv⋅T)))

1. Dry air contribution: ρd=(99,423.67 Pa/(287.05 J/(kg\cdotpK)×298.15 K))*1000≈ 1,16170837 kg/m³

2. Water vapor contribution: ρv=1,901.4 Pa/(461.5 J/(kg\cdotpK)×298.15 K)*1000≈ 0,01382 kg/m³

Finally, add the two contributions: ρ= ρd+ρv = 1.163 kg/m3+0.0138 kg/m3 ≈ 1,17553 kg/m3

*This part was for fun and to see if I could. The Air Density impacts you fans and radiators performance in both the hardware and the Water cooling part. To make this data usable and easy accessible I am still working on a part that chooses / combines the best cooling option when looking at Fan - Water Pump / Air and water speed. The article above of watt dissipated is a good start to reach the goal.

**Depending on the parts CPU water cooler, GPU water cooler or Radiator. The lack of data on heat transfer at different water speeds/amounts kg/s makes it a jungle of the lucky and why do I think that? Because the weakest link sets the pace and performance. Its not by default a bad thing to be slow if you deliver or dissipate more energy from the liquid in the same time frame. The key is the ΔT (Difference in temperature) in the loop vs the ambient temperature. More time in a radiator for a given amount of liquid might make more energy dissipate, the only disadvantage might be a 5 degree C rise in CPU or GPU temperature but that might be within the norm, this part depends on your hardware specs and attitude towards noise that could see a significant reduction.

-- This next part was to satisfy my curiosity and to see what impact a change in altitude might have on a system (air density 2).

Altitude (Air Pressure) impact on the above.
Step 1: Calculate Temperature at Altitude

The temperature at a given altitude can be calculated using the standard lapse rate for the troposphere (up to 11,000 meters), which is approximately: Lapse rate= −6.5∘C/km =−0.0065K/m
Starting with the sea level temperature of See above°C ditto here K), the temperature at altitude h (in meters) can be calculated as: T(h)=T0−L⋅h (Here comes the Layer Cake)

Where:

T0= (sea level standard temperature) °K 273,15 °K
L=0.0065 K/m (temperature lapse rate) 0,0065 K/m
h is the 3000 meter

Temperature at altitude -At -20°C, the saturation vapor pressure over ice (since water is typically frozen at this temperature) is approximately: T(h)=T0−L⋅h 253,65 °K

Step 2: Calculate Pressure at Altitude

The pressure at a given altitude can be calculated using the barometric formula: P(h)=P0⋅(T0T(h))^(R⋅Lg⋅M)

Where:
P0=101325Pa (sea level standard pressure) P0 = 101325 Pa
g=9.80665m/s2 (acceleration due to gravity) g = 9,8066500 m/s2
Molar mass of Earth's air M= 0,0289644 kg/mol
Universal gas constant R = 8,3144598 J/(mol\c*K)
n = (R⋅Lg⋅M) 5,2557877

Pressure at a given altitude of 3000 meters P(h)=P0⋅(T0T(h))^(R⋅Lg⋅M) 68,653 Pa

*** We are now at 3000 meter

now re-running the 1st part

Air density considering humidity = 1,17553 kg/m³ at 0 meter vs at 3000 meter 0,95914 kg/m³

The efficiency of the fans/radiator because of the lower Air density is now only 81,5924 % or lets put it like this your fans need to run and be 18,4 % more effective same goes for your radiator. This is where the measuring of Watt comes in and combines both in a more efficient form.

Next part should be about Newton’s Law of Cooling used on the Reservoir. This is slightly more complex, why you might ask! Because of the impact slow cooling liquid has on the entire system when doing a Statistical Analysis or Calibration of the sensors thus when booting. Knowing what is in the system and how and why it works as it does gives me the option/ability to force a result to the systems advantage.

At present I am looking at Linear Regression: Using Excel's LINEST function to fit a linear trend to the temperature data. Exponential Trendline: On a graph of the temperature data, we can add an exponential trendline and display the equation on the chart. This can help model the cooling behavior. Forecasting: Use Excel’s FORECAST.ETS function to predict future values based on your temperature data series. This is a might add not a necessity.

I like to keep my options open. So next Newton’s Law of Cooling .

 

I did briefly comment on the Reservoirs role above, just posting the calculations will add comments/explanations later.

Newton’s Law of Cooling: To model how the reservoir loses heat to its surroundings (typically through radiation and convection to the ambient air):

dT/dt=−k(T−Tambient)

Where:

T is the temperature of the reservoir.

Tambient is the ambient temperature.

k is a constant that depends on the heat transfer characteristics of the system.

Suppose a reservoir initially at 40°C is cooling down to ambient air at 25°C in 21 h 20 minutes Reservoir running temperature 40 °C

Ambient air at 25 °C

Time to dissipate heat from 40 to 25 °C t 1280 min

Cooling constant -k: -0,0030706 min⁻¹
Contestant e is equal to: 2,71828183

After 10 minutes, the temperature drops to approximately 39,546 °C.

Ambient Temperature = 25 °C

Initial Temperature Difference = 15 °C

Exponent Value (-k * t) = -0,9825791


=EXP(-0,9825791) = 0,37434437


= 25 °C + ( 15 °C * 0,37434437 ) (Calculates final temperature) 30,6151655

After (t) minutes, the temperature would drop to approximately t °C.
10 - 39,546
20 - 39,106
40 - 38,266
80 - 36,733
160 - 34,178
320 - 30,615 - 5 h 20 min
640 - 27,102 - 10 h 40 min
1280 - 25,295 - 21 h 20 min

Assumptions and Parameters for calculations with Reservoir of acrylic:

Thermal conductivity of Acrylic ( kacrylick): 0.185 W/m*K (average).
Link https://en.wikipedia.org/wiki/List_of_thermal_conductivities

Thickness of the Acrylic Wall ( d): 0,01 meters (10 mm).
Surface Area of the Reservoir (A): 0.0864 square meters. 120*120*6 mm = 0,0864 m2
Specific Heat Capacity of Water (cp): 4184 J/kg*K.
Mass of Water (m): 1 kg (1 liter). (m) 1,0368 kg (liter at sea level).
Ambient Temperature: 25°C.
Initial Reservoir Temperature: 40°C.
Temperature Difference: 15°C - (40°C - 25°C).

1. Heat Transfer Rate Q:

The heat transfer rate is given by:

Q= (kacrylic*A*(Treservoir−Tambient))/d = ((0,185*0,0864*15)/0,01) = 23,976 Watt

2. Cooling Constant k:

Now, using the formula for the cooling constant k:

k= Q/(m⋅cp⋅(Treservoir−Tambient) k= (23,976/(1,0368*4184*15)) = 0,00036847 min−1

Final Result:

The new cooling constant k with a wall thickness of 10 mm (0.01 meters) is approximately 0,00036847 min⁻¹.

Update: Let me start by stating that the simple solution to the extreme long time for the water in the Reservoir to cool down to ambient temperature (20+ hours) can be mitigated by the cooling loop, that will keep running for a bit of time after the system is shut down. I could calculate this but the I have a temperature sensor in the Reservoir to make this easy

Why then do this calculation? My reason is the rest of system and sensors that will get incorrect data for an extended period, now I know what to look for and for how long.

 

The Next calculations should be for the radiator and Fans but the calculations are to imprecise. I will have to make long term observations and calculate what the real values for the fan and radiator efficiency is, this is the part that has a lot of variables and is to easy to ignore. Why easy to ignore because there are other ways (see the 1st part of the calculations) to get the options into the SAS4MC.

This data below is only some of the data displayed on 3 OLEDs of 7-15 OLED.

Data Displayed:

Reservoir:

1-1. MPR12 – Water Level: 80 %
1-2. H2O Temperature: 22.21 *C
1-3. H2O Quality MPR12 or TDS Meter w Sensor: 60 ppm – Data over time
1-4. LED/RGB RGB Color

Loop 1:

1. Reservoir *
2-2. Pump PWM: 1200 rpm - 2.1 lpm - 20% Load
2-3. CPU: 49 C - 59 % Load
2-4. H2O Temperature: 24 C diff 1.8 C
-

3-5. Radiator fan: 1400 rpm
3-6. H2O Temperature: 22 C Diff 2 C
3-7. Flow Meter: 2.1 lpm - INS-FM14 Coolant Flow Meter
1. Reservoir *

This data changes with a constant interval or if something NOT NOMINAL happens.

This part i logically showing the placement of sensors on the Reservoir and in Loop 1. The data from the sensors is used for the calculations to be sure that all is working as expected. If its not Nominal OLEDs, RGB and sound is used to get attention. This is before or during a hard shut down. This is a minor part of the SAS4MC there are more sensors in the case then on the Water Cooling rig (Reservoir and cooling Loops). A lot of work has gone into making sure that the data shown on the OLED is intuitive, still the work is not finished, more options keep popping-up.

 

A methodical approach to system design is the target. Breaking the system down into modules makes it much easier to isolate and analyze the key elements affecting performance. This is about Circulation in the System (Flow), frition and Pressure (Head Loss or Gain) and should be added to the previous introduction with "This is part of the SAS4MC aka the Monitoring and control system." from the 28 Aug 2024.

FYI: All the data below has been copied from my Spreadsheet (Excel) you might be able to copy paste most of it, if you are interested.

Last time I looked at this data was around 3 decades ago. Then I had to go to the Library to get access to this information now its all online, SO EASY!

I used some engineering sites/ tools and did test the data I produce on ChatGPT, then I iterate the process 2-5 times depending on the results. What i gained from ChatGPT was mainly a way to present the data.

*I would appreciate if any of you that might work or has worked with this would check my results, I am only 75+ % sure that most is on point. It looks reliable so here it is.

Why did I start this? I was interested in the maximum speed of liquid in the loops depending on tube size and other components in the loops, this naturally connects to energy that can be moved, efficiency and thus performance.

Then I also was annoyed with the reviews, regarding Radiators, CPU and GPU blocks I got online, from the makers of these parts and from online reviews from YouTube and decided to try and create a method to evaluate Radiators, CPU and GPU blocks. This is the 2nd part. I will try to create a complete setup (read check list) for this later and i hope to get our input to get there!


The 1st half today is taking offset in my water cooling system. The 2nd half is about the method that could be used to evaluate components and systems in general for reviews, I will post the 2nd part after this one !

The System - For water cooling specifically the tubes, resistance, Head Loss and more:

To calculate the exact pressure loss of liquid, we need either the flow rate or pressure.

Flow rate vs the environment Is depending on the pressure generated by the pump and is easy accessible for this purpose by measuring flow speed.

Calculating the velocity of water in tubes with inner diameters of 10 mm 13 mm and 20 mm (20 mm being for comparability). Given the flow rate of:

Pumps - Typical operational flow rate: 3,8 - 5,7 Liters/min

Pump - (Pressure Head): 3,7 meter

Pump - Maximum flow rate: 1500 Liters/kg Per Hour

Pump - Operational flow rate per hour: m: 240 Liters/kg Per Hour at sea level.

Pump - operational flow rate per minute: 4 Liters(kg)/ Minute

Seconds per hour - 60 minutes x 60 seconds: 3600 s

Flow rate per second (1500/3600) Q: 0,066666667 kg/s

Assumptions:

The density of water is approximately assuming room temperature and sea level.
ρ is the density of water (kilogram per cubic meter) kg/m3: 1000 kg/m3. This means that 1 liter = 1 kg

π≈ π 3,1415927

Calculations are made for tubes with an inner diameter (D) of 10, 13 and 20 mm:

1. Convert to meters: (xx mm/1000) since there are 1000 mm per meter

D10 = (10mm/1000) because 1 meter equals 1000 mm - 10 = 0,01 meter
D13 = (13mm/1000) because 1 meter equals 1000 mm - 13 = 0,013 meter
D20 = (20mm/1000) because 1 meter equals 1000 mm - 20 = 0,02 meter

The cross-sectional area (A) for a tube with inner diameter (D) is: A = (π*( D^2))/4

2. Calculate the cross-sectional area:

A10 = (3,14159265358979*(0,01^2))/4 = 0,0000785 m2
A13 = (3,14159265358979*(0,013^2))/4 = 0,0001327 m2
A20= (3,14159265358979*(0,02^2))/4 = 0,0003142 m2

To find the velocity (v), we need to first convert the mass flow rate (m) to volumetric flow rate (Q) using the formula:

Q = volumetric flow rate Q=m/ρ = 0,0000667 kg/s

Then, we calculate the velocity using the equation: v=Q/A

The velocity of water is approximately m/s in a tube with xx mm inner diameter

v = A10 Q/A = 0,8488 m/s
v = A13 Q/A = 0,5023 m/s
v = A20 Q/A = 0,2122 m/s

1. Flow Regime: The flow transitions from laminar to turbulent has a Reynolds number of approximately 2,300. In turbulent flow, the velocity can be much higher. However, very high velocities may lead to issues such as cavitation or pipe damage.
2. Reynolds Number: The Reynolds number (Re) is a dimensionless quantity that helps predict the flow regime and is given by:

Where:

ρ is the fluid density {kg/m^3) = 1000 kg/m3
μ is the dynamic viscosity of water (Pa·s). At 20 C = 0,001002 Pa·s

The DYNAMIC viscosity of water (μ) is temperature dependent so look it up.

Check the Link for more info and options. https://www.engineeringtoolbox.com/water-dynamic-kinematic-viscosity-d_596.html

Reynolds number is used to determine if the flow is laminar or turbulent:

Flow is considered laminar when Re < 2300

Flow is considered turbulent if Re > 4000

Reynolds number D10 Re = ρvD\μ = 8471,32099 Re
Reynolds number D13 Re = ρvD\μ = 6516,400761 Re
Reynolds number D20 Re = ρvD\μ = 4235,660495 Re

To calculate the maximum velocity where the flow transitions from laminar to turbulent, we will assume that the flow is at the verge of transitioning to turbulent flow (Reynolds number of about 2,300). Let's solve for the velocity (v) under this condition.

Reynolds number 2300 = ρvD\μ

The maximum possible velocity before laminar flow changes to a transition state before turbulent flow it's approximately:

Max Speed for Laminar Flow for Tube D10 vmax=((2300⋅0,001002)/(1000*0,01)) = 0,230460 m/s
Max Speed for Laminar Flow for Tube D13 vmax=((2300⋅0,001002)/(1000*0,013)) = 0,177277 m/s
Max Speed for Laminar Flow for Tube D20 vmax=((2300⋅0,001002)/(1000*0,02)) = 0,115230 m/s

Water volume per second before the transitioning into turbulent flow for different size tubes.

for Tube D10 (0,230460*0,000078540*1000) = 0,018100286 kg/s = liter per second
for Tube D13 = 0,023530372 kg/s
for Tube D20 = 0,036200572 kg/s

These values represent the upper limit for laminar flow. In turbulent flow, velocities can be much higher depending on the pressure and other factors, but this comes with increased friction losses and potential for other issues like cavitation.

Turbulent flow occurs when the fluid moves in a chaotic and irregular manner, often characterized by swirling eddies and vortices. In pipes or tubes, this typically happens when the Reynolds number exceeds 2,300, and it's common in situations with high velocities, large diameters or low viscosity fluids.

We already used the viscosity above, these are the approximate viscosities for liquids at typical in PC cooling with operating temperatures:

1. Distilled Water - Viscosity at 25°C (77°F): Water is commonly used due to its high thermal conductivity and low viscosity, which allows efficient flow through cooling systems. 0,89 mPa·s (or cP)

2. Ethylene Glycol-based Coolants - Viscosity at 25°C (77°F): Ethylene glycol is mixed with water in different ratios (e.g., 50/50 or 70/30) to create antifreeze and corrosion-resistant coolants. It has higher viscosity than water, which can slightly reduce flow rates. 16-20 mPa·s (or cP)

3. Propylene Glycol-based Coolants - Viscosity at 25°C (77°F)* Propylene glycol is another common component in coolants and is often used in environmentally safer or food-safe applications. It is more viscous than ethylene glycol. 40-60 mPa·s (or cP)

4. Specialized PC Coolants - Viscosity at 25°C (77°F): These coolants are premixed with additives to prevent corrosion, scaling, and microbial growth. They are designed to have a balanced viscosity to maintain effective flow while ensuring protection for the cooling components. 2-10 mPa·s (or cP)

Impact of Temperature on Viscosity - Viscosity decreases as temperature increases: As the liquid heats up during operation, the viscosity will drop, improving flow rates through the cooling loop. Conversely, lower temperatures increase viscosity, potentially reducing flow rates.


Calculations for tube resistance at a given flow speed, size, length and liquid temperature:

Tube length: 1.5 meters (150 cm) - for 1 loop, its an estimate.

Something I never thought I would do! Calculations for fluid dynamics. This is important for understanding so worth the work, and yea it does at times make me feel stupid until I understand the connections.

For fluid dynamics calculations at 20°C, it's important to use these specific viscosity values at a given temperature to get accurate results. Higher viscosity leads to increased resistance to flow, which can impact the performance of your PC cooling loop (e.g., reducing flow rate or increasing pump load).

Calculations step by step for determining the maximum speed of water in the tubes, considering turbulent flow and a tube length of 150 cm.

Given Data:

1. Flow rate (mass): continued from above - m˙ = 0,000066666667 kg/s

2. Tube length: 1.5 meters (150 cm) = 1,5 Meter

3. Tube inner diameters(ID): continued from above 10 mm (0.01 m) and 13 mm (0.013 m) 10, 13 and 20

4. μ is the dynamic viscosity of distilled water at 20°C: continued from above (0.001002 Pa·s) = 0,001002

5. Density of water: continued from above - ρ = 1000 kg/m³

Steps:

1. Convert mass flow rate to volumetric flow rate (Q): data from above Q=ρ/m˙ = 0,066666667 cubic meters per second (m³/s).
2. Calculate the velocity (v) using the equation: data from above v=A/Q

A= data from above πD2/4 is the cross-sectional area of the tube m2

D= data from above is the inner diameter of the tube.

3. Calculate the Reynolds number (Re):- data from above Re=ρvD/μ

v is the velocity of the water - data from above = m/s
μ is the dynamic viscosity of water - data from above = 0,001002 Pa·s

4. Determine if the flow is turbulent:

Flow is considered turbulent if Re>4000.
My Re is minimum double that so yes turbulent

5. Calculate the pressure drop using the Darcy-Weisbach equation for turbulent flow: ΔP=f*((L/D)*(ρ*(v^2))/2)

Calculate Friction Factor (f) Using Blasius Equation f=0.3164*Re^−0.25

f for D10 f=0,3164*(Re^(-0,25)) = 0,032979851
f for D13 f=0,3164*(Re^(-0,25)) ) = 0,035215554
f for D20 f=0,3164*(Re^(-0,25)) = 0,039219874

Calculate the pressure drop using the Darcy-Weisbach equation for turbulent flow: ΔP=f*((L/D)*(ρ*(v^2))/2)

ΔP = pressure drop due to friction (Pa)
f = Darcy friction factor (dimensionless, depends on the flow regime and tube roughness)
L = length of the tube (meters)
D = diameter of the tube (meters)
ρ = density of water (~1000 kg/m³)
v = velocity of the fluid (m/s)

The pressure drop on 3 different tubes with ID of 10, 13 and 20 mm using the Darcy-Weisbach equation for turbulent flow on a tube of 1500 mm length.

ΔP=for D10 ΔP=f*((L/D)*(ρ*(v^2))/2) = 1782,164026 Pa
ΔP=for D13 ΔP=f*((L/D)*(ρ*(v^2))/2) = 512,5270004 Pa
ΔP=for D20 ΔP=f*((L/D)*(ρ*(v^2))/2) = 66,23006686 Pa

Conversion kPa - meters of head (Pressure drop) 6,9 kPa = 0,704 meters of head (Pressure drop)
kPa to Pa = kPa * 1000 (k = kilo = 1000) = 6900 Pa

ΔP=for D10 ΔP_D10/6900 = 0,258284641 meters of head (Pressure drop)
ΔP=for D13 ΔP_D13/6900 = 0,074279275 meters of head (Pressure drop)
ΔP=for D20 ΔP_D20/6900 = 0,00959856 meters of head (Pressure drop)

Worth noticing the pressure drop is 3 times larger with a tube of 10 mm Inner diameter(ID) compared to a tube with a13 mm ID. Then again a 20 mm ID has a pressure drop that is 8 times smaller than a 13 mm ID. The 20 mm is fictional for now!

The head loss due to bends in a pipe is commonly calculated using the equivalent length method, where each bend is treated as an additional length of straight pipe that adds frictional resistance. The head loss due to bends can also be determined using K-values (loss coefficients), which depend on the type of bend (e.g., 45° or 90°) and the geometry of the bend.

BENDS 45° and 90° ** Yes I had a bit of time so wanted to make sure i didn't tell you something that wasn't true regarding bends impact on flow speed and Head Loss.

Formulas:

1. Head Loss Due to Bend:

ΔHbend=K*((v^2)/2*g) K*((v^2)/2*g)

Where:

K = loss coefficient for the bend (dimensionless)
v = velocity of fluid (m/s)
g = acceleration due to gravity (~9.81 m/s²) locally here 9,82 m/s²

2. Typical K-values:

45° bend: K=0.2
90° bend: K=0.5

These values are approximations and can vary slightly depending on the specific geometry of the fittings used.

For the calculations the data from above is used for the velocity (v) of the fluid in the tube.

Calculate Head Loss for Each Bend - Head Loss for 45° and 90° Bends:

Using the formula ΔHbend= K*((v^2)/2*g)

For 45° bends (K = 0.2):

10 mm diameter: ΔH45°=0,2*((0,8488^2)/(2*9,82)) = 0,00733713 meters of head
13 mm diameter: ΔH45°=0,2*((0,5023^2)/(2*9,82)) = 0,00256893 meters of head
20 mm diameter: ΔH45°=0,2*((0,2122^2)/(2*9,82)) = 0,00045857 meters of head

For 90° bends (K = 0.5):

10 mm diameter: ΔH90°=0,5*((0,8488^2)/(2*9,82)) = 0,01834283 meters of head
13 mm diameter: ΔH90°=0,2*((0,5023^2)/(2*9,82)) = 0,00642233 meters of head
20 mm diameter: ΔH90°=0,2*((0,2122^2)/(2*9,82)) = 0,00114643 meters of head

Summary of Head Losses per Bend:

For 45° bends:
10 mm diameter: 0,0073 meters of head (Pressure drop)
13 mm diameter: 0,0026 meters of head (Pressure drop)
20 mm diameter: 0,0005 meters of head (Pressure drop)

For 90° bends:
10 mm diameter: 0,0183 meters of head (Pressure drop)
13 mm diameter: 0,0064 meters of head (Pressure drop)
20 mm diameter: 0,0011 meters of head (Pressure drop)
'
Difference between H45° - H90° bend i % and meters of head (Pressure drop)
D10 ΔH45° - H90° in % and meters of head (Pressure drop) 150% = 0,0110 Δmeters of head (Pressure drop)
D13 ΔH45° - H90° in % and meters of head (Pressure drop) 150% = 0,0039 meters of head (Pressure drop)
D20 ΔH45° - H90° in % and meters of head (Pressure drop) 150% = 0,0007 meters of head (Pressure drop)
Typical K-values of 0,2 and 0,5 are the common denominator for the 150% difference.

***These head losses are per bend, so if you have multiple bends in your loop, you would sum them to get the total head loss due to bends.

In a water-cooled PC system, various factors like pump pressure, maximum lift (head pressure), and the number of bends in the tubing all contribute to the overall performance and efficiency of the cooling loop. Here's an explanation of how each of these elements interacts with each other:

1. Pump Pressure and Flow Rate

Pressure (Head Pressure or Lift): The head pressure of a pump refers to the maximum height that the pump can lift water vertically. In a closed-loop system like a PC water-cooling loop, the head pressure indicates how effectively the pump can push water through the entire system, including restrictions like blocks, radiators, and tubing bends.
Flow Rate: Flow rate refers to the amount of liquid the pump can push through the system, typically measured in liters per hour (L/h). A balance between flow rate and pressure is crucial to achieving optimal cooling performance.

Relationship:

Higher head pressure generally allows the pump to overcome more resistance in the loop (e.g., from tight bends or restrictive water blocks).

Flow rate can drop if the head pressure isn't sufficient to overcome the system's resistance. Resistance will go up with higher speeds in a turbulent system.

2. Maximum Lift (Head Pressure)

This refers to the maximum height a pump can push water in a vertical direction. In the context of a water-cooled PC, while there's usually not a significant vertical lift (unlike in large-scale industrial systems), the maximum lift still correlates to how well the pump can maintain flow against restrictions.

If the pump has a high head pressure, it can better handle long or complex loops with multiple water blocks and radiators, where resistance is higher.

3. Bends in Tubes

Impact of Bends: Each bend in the tubing introduces resistance to the flow of water. The tighter the bend, the more the water flow is restricted. This resistance can cause a drop in flow rate and require more pressure from the pump to maintain adequate flow.
Smooth vs. Sharp Bends: Smooth bends cause less resistance compared to sharp or kinked bends. Using fittings like 45° or 90° angles instead of bending soft tubing can help reduce the impact on flow.

Cumulative Effect:

The more bends in your loop, the more pressure is required to push water through efficiently. If the pump does not have sufficient head pressure, flow rates can drop, leading to decreased cooling performance.

4. Optimizing the Loop

Pump Selection: Choose a pump with adequate head pressure for your loop's complexity. For example, if you're running multiple water blocks and radiators, you might need a pump like the D5 or DDC, which offer higher head pressures.

Tubing Layout: Minimize the number of sharp bends in your tubing to reduce flow resistance. Where possible, use angled fittings to avoid kinking the tubing.

Flow Rate and Cooling Performance: There is an optimal flow rate for cooling performance. Too high or too low flow can result in inefficient heat transfer. Typically, flow rates between 3.8-5.7 L/min same as kg/min (1-1.5 GPM) are ideal for most water-cooled PCs.

Conclusion

Pressure (Head): Determines the pump's ability to overcome resistance.

Lift (Max Head): Important in handling the vertical and restrictive resistance of your loop.

Bends: Increase resistance; minimizing or optimizing them is key to maintaining good flow rates.

Balancing these factors—ensuring that your pump can handle the resistance from bends and restrictions while providing an adequate flow rate—is crucial for efficient cooling performance in your water-cooled PC.

Let’s walk through an example calculation to determine how pressure, maximum lift, and bends in the tubing affect a water-cooled PC system.

Scenario:

You have a water-cooled PC loop that includes a CPU block, a GPU block, a radiator, and tubing with several bends.

You are using a typical pump (e.g., a D5 pump) and want to calculate if it can handle the resistance introduced by the components and the bends in the tubing.

Assumptions:

Pump Specifications:

Maximum head pressure: 3.7 meters (3700 mm) = 3,7 meter

Water Blocks:
CPU Block pressure drop: 0,5 PSI (3,45 kPa) = 3,45 kPa
GPU Block pressure drop: 0,4 PSI (2,76 kPa) = 2,76 kPa
Radiator pressure drop: 0,2 PSI (1,38 kPa) = 1,38 kPa

Tubing Bends: Number of bends in the loop, with each bend adding resistance.

Estimate each 90° bend adds 0.1 PSI (0.69 kPa) of resistance. = 0,69 kPa

Total Pressure Drop Calculation:

We will sum the pressure drops from all components and bends to see if the pump’s pressure (head) can overcome the total resistance.

Step 1: Convert all pressure drops to the same unit

1 feet = 2,9870 kPa
2,31 feet = 1 PSI = 0,704 meters of head (Pressure drop)
1 PSI = 6,9 kPa
6,9 kPa = 0,704 meters of head (Pressure drop)
2,31 feet = 1 PSI

Convert each pressure drop to meters of head:

CPU block: (3,45 kPa *0,704 meters of head )/6,9 kPa 0,3520 meters of head (Pressure drop)
GPU block: (2,76 kPa *0,704 meters of head )/6,9 kPa 0,2816 meters of head (Pressure drop)
Radiator: (1,38 kPa *0,704 meters of head )/6,9 kPa 0,1408 meters of head (Pressure drop)
Each 45° bend: 0,0026 meters of head (Pressure drop)
Each 90° bend: 0,0064 meters of head (Pressure drop)

Step 2: Calculate total pressure drop (resistance) in the loops:

Now, add up all the pressure drops in meters of head:

Loop 1:
CPU Block: 1 = 0,3520 meters of head (Pressure drop)
Radiator: 1 = 0,1408 meters of head (Pressure drop)
Curved Tubes? - Diameter? - Still need to work on this!
Tube ΔP=for D13 - 150 cm = 0,0743 meters of head (Pressure drop)
Straight fittings ?
45° Bends x 9 = 0,0231 meters of head (Pressure drop)
90° Bends x 5: = 0,0321 meters of head (Pressure drop)
Total pressure drop L1 = 0,6223 meters of head (Pressure drop)

L1 pump’s maximum head pressure is 3.7 meters. 3,7000 meters of head (Pressure generation)


Loop :2
GPU Block: 0.282 meters of head 1 = 0,2816 meters of head (Pressure drop)
Radiator: 0.141 meters of head 1 = 0,1408 meters of head (Pressure drop)
Curved Tubes ? ?
Tube ΔP=for D13 - 150 cm = 0,0743 meters of head (Pressure drop)
Straight fittings ?
45° bends: 14 = 0,0360 meters of head (Pressure drop)
90° Bends 2 = 0,0128 meters of head (Pressure drop)
Total pressure drop L2 = 0,5455 meters of head (Pressure drop)

L2 pump’s maximum head pressure is 3.7 meters. 3,7000 meters of head (Pressure generation)

Step 3: Compare with the pump’s head pressure

Both pump’s have a maximum head pressure is 3.7 meters, with less than 1 meters of head (drop) per Loop this system is in the clear.

Step 4: Estimate Flow Rate

The pump operates at a flow rate of (3.8-5.7 L/min) at lower pressure drops.
The higher the pressure drop in meters of head, has a direct relation to the flow rate and will decrease it from the maximum, but it should still maintain a reasonable flow rate within the optimal range 3.8-5.7 L/min (around 1.0-1.5 GPM).

Tube Roughness

In this calculation, we used a typical roughness value for smooth plastic or copper tubing of ϵ=0.0015 mm\epsilon = 0.0015 . For rougher materials, such as stainless steel, the roughness could be higher (e.g., around 0.045 mm).

This was part 1 check also part 2 coming up next

P-S. I will try to add some comments in blue in the coming days how to interpret the data.

I hope to hear from you if you got anything on your mind. This data is for you to be able to make a "poll of expectations" towards a water cooling system. The next part is for reviews of you know what.

Still needs a superficial work over to look good. I will check up on this when possible.

My take on these numbers is that splitting up the cooling system into 2 parts is sensible even though it has a monetary cost. A point that these numbers do not con way is the amount of heat potentially generated over a long period from especially the GPU (possible x 2) can force the system/pump to speed up to a point where the turbulence effectively stops the flow, not only from getting faster, but actually slows it down.
In a GPU block we need turbulence, but with larger Diameter tubes leading the water to and from the GPU the resistance/pressure can be kept down for longer. I will add the Reynolds numbers for faster flow to show what I mean. The main problem with larger tubes is the lack of fittings and tubes for the purpose at present. The hardware hasn't followed the development in power usage.

So looking at CPU an GPU temperatures is for me about finding 4 different levels.
1. Where the hardware fails!
2. Where the hardware throttles down and, we know we are wasting money.
3. The golden-spot where hardware works optimal and we use the minimum amount of power.
4. The lowest temperature the cooling system can produce for overclocking (fun).

For this to be an option we need better data on hardware before we buy components or a larger wallets $$$ to test more components.

 

When ChatGPT doesn't understand me

How to calculate the pressure drop in relation to a change in velocity in the water cooling system, we can use the Bernoulli principle and the Darcy-Weisbach equation.

With the flow meter measuring the difference in velocity of the liquid when new part (like a radiator, CPU water block, or GPU water block) are added or removed. We can relate these new parts to the pressure drop/ change in velocity of the water flow.

The test setup: The test setup would be a 20-25 liter bucket acting as the reservoir and another bucket at the end to collect the liquid. Between the buckets are: tube - a pump - tube - Quick Disconnect No-Spill Coupling couplings - tube - flow meter - tube -. The Quick Disconnect will be used to add either the Reservoir, CPU or GPU blocks.
Velocity (v) of the liquid in the system is measured before the component is added (v1) then with it (v2). The 2 data points for the speed (v) are used in the equations with the previous data points.

The velocity and mass are directly related.

Bernoulli's Principle (Simplified for Horizontal Flow):

In the case of horizontal flow where height changes are negligible, the simplified form of Bernoulli’s equation is:

(P1+(2ρv1^2/2))=(P2+(2ρv2^2/2)) adjusted for excel

Where:

P1 and P2 are the pressures at two points in the system (Pa).
ρ is the density of the fluid (kg/m³) — for water, it's approximately 1000 kg/m³.
v1 and v2 are the velocities of the fluid at the two points (m/s).

Pressure Difference (ΔP) Between Two Points:

Rearranging Bernoulli’s equation to find the pressure difference (ΔP) between two points due to a change in velocity:

ΔP=P1−P2=ρ/2(v2^2−v1^2)

Where:

ΔP is the pressure difference in Pascals (Pa).
v1 is the initial velocity (before the component, e.g., radiator).
v2 is the final velocity (after the component, e.g., after the flow meter).

Step-by-Step Example:

Let’s say the flow meter measures:

Initial velocity v1 = 0.8 m/s
Final velocity v2 = 0.6 m/s

The fluid is water, so:

Density of water ρ=1000 kg/m3

Step 1: Apply the Pressure Difference Formula

ΔP=(1000/2)*(0.6^2−0.8^2) also adjusted to excel

Step 2: Calculate the Velocity Squared Differences

0,6^2 = 0,36 and 0.8^2 = 0,64

Now, subtract:
0.36−0.64 = −0.28

Step 3: Complete the Calculation
ΔP=(1000/2)*(−0.28)=500*(−0.28)=−140 Pa

The pressure change needs to be recalculated in pressure drop/gain in meters.

Final Result:

The pressure drop due to the velocity change in this system is 140 Pascals. The negative sign indicates that the pressure has decreased after the velocity drop (which is expected when adding a component that restricts flow, like a radiator).
General Formula to Find Pressure from Velocity:

If you only know the velocity change and want to find the pressure difference, the formula is:

ΔP=(ρ/2)*(v2^2−v1^2)


Using the velocity difference measured from the flow meter.
ρ is the density of the fluid (for water, use 1000 kg/m³).
v1 is the initial velocity before the component.
v2 is the final velocity after the component.

This approach allows you to calculate the pressure drop based on the velocity change measured by the flow meter.

Creating a review I would add the options on temperature, flow speed and Watt from the reviews.

This should be looked at as a theoretical approach that need to be ...........

This didn't go any better

 

Buenos Dias.

I watched an episode of JaysTwoCents named Front mount or Top mount -- of a AIO radiator. I did enjoy and appreciate that episode with some data on the environment and temperatures. I would love to see a 2nd part without the filters in the case and with fans in push vs push config and visa Versa. Also a setup with temperature info of still ambient and temperature of the airflow inside the case in real time.

This made me stay up for another hour. I got hit by the idea of the scientific method and I know next to nothing about it. I had ChatGPT give me some info, short and fast this wasn't satisfying.

The New Component check list:

So i thought about a general setup for testing component flow and resistance. I will add more on temperature even though my 1st post in a long time has the essential.

Objective

The goal is to measure the performance of each component in a liquid cooling loop, focusing on the pump’s ability to circulate liquid and the pressure drops across components. You'll test the system to ensure the pump is strong enough to handle the cooling demands at different flow rates, with a focus on how liquid speed and pressure drop vary across the system components.

Steps for Testing

1. Initial Setup of Loop

Components: Reservoir, Pump, Tubing (same length as Loop 1), Quick-connect fittings.

Purpose: To test the pump performance and establish a baseline for the system.

Procedure:

Start with a simple setup: 25 L reservoir, pump, and another 25 L bucket to measure flow.

Run the pump to circulate 20 L of liquid through the loop, measuring how long it takes. At different speeds think PWM.

Use this to calculate flow rate: Flow Rate=20 L / Time (in seconds) = l/s

Create a performance table at different pump speeds (10%, 20%, etc.) by controlling the pump with a PWM signal.

Performance Measurement

After the initial test, we introduce individual components one by one. Add one Remove one.

Components: CPU water block, radiator, sensors, bends, and tubes.

Each component placed on a scale and weighed, then it's filled with liquid to determine its liquid volume capacity (again placed on a scale and weighed with the liquid).

Quick-connect fittings are used to easily add/remove components from the test loop.

Pump Speed Intervals

The pump speed is controlled in 10% increments, and the flow rate is measured at each speed setting.

We record flow rates at different speeds to create a performance table for the entire loop and each individual component.

Pressure Drop Calculation

As each component is added, the system’s performance in terms of pressure drop is recorded.

Pressure drop (meters of head) is calculated for each component at various pump speeds.

This is important to evaluate how much resistance each component adds to the loop.

A curve of performance vs. pressure drop is generated for analysis.

5. Doubling the Tube Length

To verify the tubing's effect on the system, we double the length of the tube and compare the performance results (flow rate, pressure drop, etc.).

This helps evaluate how tubing length impacts overall system performance, particularly the pump's ability to maintain adequate flow.

Results Analysis

With the collected data, we should have performance curves that reflect the relationship between flow rate, pump speed, and pressure drop for each component.

This will allow us to determine if the pump is strong enough to handle the entire loop and if each component is performing as expected. Thee Pump ID are in some models also 13 mm.

If the pump is too weak, we might see excessive pressure drops or flow rates that are too low.

Potential Challenges

Pump Limitations: If the pump isn’t powerful enough, it might not be able to maintain the desired flow rate at higher loads or with more components.

Pressure Drops: Some components may introduce more pressure drop than expected, which could cause performance issues at lower pump speeds.

Tubing Effects: Doubling the tube length may result in significant performance differences, depending on the pump’s strength.

I would look for components that are the weak link, that offset the performance of the entire loop. Without being a positive influence on the cooling.

That's my 2 Cents for now. I do hope some of you got some input that I have missed or an alternative angle.

Nox out!

 

I have been working on expanding the check list for Testing components (aka the last post). I hit a wall with some of the material trying to understand equations on real life physics regarding resistance in new components. The snag is understanding its internal shape and the resistance created by the speed of liquid (read changing/faster flowing liquid) on the equations versus The resistance created by the surface. I will let it rest a few days.

At a greater scale I am working with the Arduino part of the projects, or more broadly the embedded systems, organizing components and logic into layers and grids that will help me designing these "complex" systems. Each part by itself is not complex, take a look at an Arduino with a BME280 sensor that can deliver air pressure, temperature, humidity... and post a lot of data at a low cost and complexity.


At this point in time, updating the system with new sensors this includes the bme280 sensor, has made it possible to significantly cut down on the total amount of physical sensors. This again has made it an option to get a few new sensors into the mix, its all about wire(s) connections and power consumption. New and better sensors with more abilities have been the main benefit of treading water.

The design of the case is continuously adapted.

For this project "SAS4MC" with layers including Information, Sensing, Effectors, and Command, I am defining a structured framework for a control system. I will break it down later, conceptually for each layer.

I did mention this before that my initial thought was to build a theoretical model for how all of this works, and to be honest even though I am failing to achieve this, it helps me find the right approach. The right approach seems, in most if not all situations, to be measure and get data and respond in the last layer Command. The understanding part is what i posted above.

Let's start from the end.


- The Command Layer:

The command layer handles decision-making and system control. It takes inputs from the sensing and information layers and determines how to control the effectors.

Decision Algorithms: Logic based on sensor data to determine appropriate actions. This could involve simple threshold logic, PID controllers, or even machine learning algorithms.
Control Systems: Commands are issued to the effectors based on the current state and the desired outcome.
User Input: This layer might also process manual commands from a human operator or inputs from higher-level systems.


- Effectors Layer:


This layer involves the actions or effects that the system performs based on the information and sensing layers. It involves:

Actuators: Devices like motors, LEDs, displays, buzzers, etc., which produce physical output based on control commands.
Effect Execution: Converting the decisions made by the command layer into physical changes in the environment.

An Example in Arduino:

Controlling motors or servos with digitalWrite() or analogWrite() to control their speed or position.

Activating relays to switch high-power devices on/off.


- Sensing Layer:

This layer involves the system's perception of the environment through sensors, which feed the Information Layer.

Sensors: Devices like temperature sensors, light sensors, motion detectors, etc., collect real-time data from the environment.
Signal Processing: Basic processing of raw data from sensors to convert it into a usable form, such as analog-to-digital conversion (ADC).
Data Preprocessing: Filtering, noise reduction, or threshold detection before the data is sent to the Information Layer.

Example in Arduino:

Sensors connected to the Arduino via analog or digital pins.
Reading sensor data using functions like analogRead() or digitalRead().
Applying smoothing algorithms to sensor data to reduce noise.


- Information Layer:

This is the layer where all data is collected and processed. It encompasses:

Data storage and handling: Logs and records of sensor data, environment states, and command feedback.
Communication systems: The way data is transmitted between different layers or subsystems, such as between sensors, effectors, and control units.
External information sources: Could be inputs from the cloud, APIs, or databases, such as environmental data or historical records that influence system decisions.

Example in Arduino:

Data from sensors (temperature, humidity, etc.) is stored in variables or external memory modules like EEPROM or SD card modules.
Serial communication between Arduino and external devices (PC, cloud server, etc.) to log data for further analysis.

This is a fast overview of what is going on atm.

NoX Out

 

Good evening all

I was considering some parts of my calculations and how much mathematical uncertainty is caused by the expansion of water between 0 °C (where water freezes) and up to 100 °C when it changes form to steam.

The density of water (H₂O) at liquid state:

The density of water is approximately: 1 g/cm³ or 1000 kg/m³.

This value decreases slightly with increasing temperature and increases as water cools, reaching its maximum density at 4°C, which is about 0.99997 g/cm³ or 999.97 kg/m³

The empirical equation known as the density-temperature relationship for water:

ρ(T)=999.8425+(6.793952×(10^-2)*C247)−((9.095290×(10^−3))*(C247^2))+1.001685×10^−4*T^3−1.120083×10^−6*T^4+6.536332×10^−9*T^5

T: 22

ρ(22°C)=999.8425+1.49466944−4.39916036+1.06688381−0.26241943+0.03346948

ρ(22°C)=997.77294295 kg/m3

Excel friendly version below:

Cell in excel used for the Temperature is C247 change it in the equation if you want to test it for your self.

ρ(T)=(999,8425)+(6,793952*(10^(-2))*C247)-((9,09529*10^(-3))*(C247^2))+(1,001685*10^(-4))*(C247^3)-((1,120083*(10^(-6)))*(C247^4))+(6,536332*(10^(-9))*(C247^5))

Well i was looking at water density-temperature on the entire interval, but truth be told I expect we all know that the water temperature interval that will be in use is 20 - 50 °C.

The good news is that the variation is less than 1% data is below.

Blue color is at 4 °C when water is as heavy as possible at sea level. Yellow is room temperature at 20 °C and looking at the % expansion up to 40 °C is 0,6% being the cause of a minor cooling loss. Then again it goes both ways in the Loops.

 

Water Quality!

it started with an idea, that it might be possible to use the MPR12 12-Key Capacitive Touch Sensor to measure the water level in the Reservoir. Then I expanded the idea to also use the same sensor (MPR12) and probes to measure water quality. Theoretically it possible.

With data from the following hardware components:

Arduino Due
MPR12 conductivity sensor
SD card module (to log data for easy export to Excel)
Real-Time Clock (RTC) module (for timestamping the data)

Starting with H2O Quality measured by an MPR12 by 2 points :

The parameters being measured are conductivity related to pollution.
The frequency of the measurements is hourly.
The format of the data will be added as a CSV file and the data arranged as timestamp, sensor1, sensor2, ..., sensor12.
Specific analysis we are looking to conduct is regarding trends, anomalies, comparisons measured between the two points

Data Overview:
Conductivity (in µS/cm) is typically used as a measure of how well water can conduct electricity, often related to the concentration of ions (salts, minerals, etc.) dissolved in the water.
Two Points: The data will contain two columns or sets of data, each representing measurements from two locations.
Hourly Data: Given the hourly frequency, you'll have enough data to observe daily and seasonal variations.

Data Transfer:
RTC Module: Used to timestamp each conductivity reading, which is helpful for trend analysis later.
SD Card: Logs data in a CSV format that Excel can open easily. The file will contain three columns: timestamp, conductivity1, conductivity2.
Conductivity Calculation: The analogRead() function captures the sensor values, which are then mapped to appropriate conductivity ranges (this will depend on your calibration).

.. and the Idea expands. What if we use 12 points in the same reservoir for measurement instead of just 2? with some added complexity and options.

Since the raw data delivered from the sensors is a value between 0-1023 we need to convert them to conductivity values in µS/cm based on calibration or a map.

Formula for Conductivity Conversion

Using the map() function or a linear equation for this conversion. Let’s assume that the sensor reads a minimum value of 0 when the conductivity is 0 µS/cm and a maximum of 1023 when the conductivity is 10,000 µS/cm.

The linear conversion would be:

Conductivity=(Analog Reading /1023)*10000 µS/cm

Example:

If the sensor outputs a value of 512: Conductivity=(512/1023)*10000=5004.89 µS/cm

We’ll need to calibrate our sensor, and the exact multiplier (e.g., 10,000) may change based on the range of the MPR12 sensor.

Update: What I describe as the sensors is not the board but the contact points in the liquid.



Source for this picture and project: https://www.adam-meyer.com/arduino/MPR121

Each sensor is made of brass and with a wire inside and then some layers of nickle. Before they are inserted into the reservoir with some insulating material around them.



Average Conductivity for all sensors. Moving Average (Time-based). Anomaly Detection Using Z-Score. Difference Between Sensors. Correlation Between Sensors is a b@st4rd with Formula for Pearson Correlation Coefficient. More on that later.

Still needs some math for the above but I will add that later.

Next question, how to calibrate for water quality?

Nox Out

 

I had an hours so here comes some math on water quality.

Average Conductivity Across 12 Sensors:

Average Conductivity=(C1+C2+...+C12)/12

Where:

C1,C2,...,C12C are the conductivity readings from the 12 sensors.

Example: If the 12 sensor readings in µS/cm are: C1=4500,C2=4600,C3=4700,...,C12=4900C

The average conductivity:

Average=(4500+4600+4700+...+4900)/12=4750 µS/cm


Moving Average(MA) (Time-based):

To smooth out the data and see trends over time, you can calculate a moving average. A simple moving average over 3 hours would be:

MAt=(Ct+Ct−1+Ct−2)/3

Where:

Ct is the conductivity at time t,
Ct−1,Ct−2 are conductivity values at previous times.

This helps remove noise and shows longer-term trends.

This is also used on data from other analog sources


Anomaly Detection Using Z-Score:

Anomalies can be detected using the Z-score. The Z-score tells you how far a particular data point is from the mean in terms of standard deviations. It’s defined as:

Z=(C−μ)/σ

Where:
Δ=Delta again Delta=Difference
C is the conductivity at a particular point,
μ is the mean of all conductivity readings, μ = mu
σ is the standard deviation of the readings. σ= sigma

If ∣Z∣>2, we can flag the data point as an anomaly (as it’s more than 2 standard deviations from the mean).

Example:

Mean μ=4750 µS/cm
Standard Deviation σ=100
A reading at sensor 3 is C3=5050 µS/cm

The Z-score for this reading:

Z=(5050−4750)/100=3.0

Since ∣Z∣>2, this indicates an anomaly at sensor 3.

Looking at the Reservoir i wonder if the water will layer pollutants when not circulating?


Difference Between Sensors:

You may want to compute the difference in conductivity between sensors. For example, if sensors are spread across the reservoir, you can measure how much conductivity varies from one point to another.

Formula for Conductivity Difference:

For two sensors Ci and Cj:

ΔC=Ci−Cj

Example:

C1=4500 µS/cm
C2=4700 µS/cm

The difference:

ΔC=4500−4700=−200 µS/cm

A negative value indicates that C2 has higher conductivity than C1.

This yea a reminder to calibrate my d4mn sensors first or again. This is essentially an alarm re calibrate.


Correlation Between Sensors:

To understand if there’s a relationship between the conductivity at different points, you can calculate the correlation coefficient.

I will transform these equation to excel usually makes them easier to understand. I can't find the right symbols here.

Some math like this Correlation Between Sensors might not be viable on an Arduino and will need post processing for a Graphical read visualization on another system. I would like to implement this data on the SAS4MC but I am not sure that I am able, time will tell.

Formula for Pearson Correlation Coefficient:

r=(∑(Ci−μi)(Cj−μj)) / √ ( r=∑(Ci−μi)2∑(Cj−μj)2)

Where:

Ci and Cj are the conductivity readings at sensor i and j,
μi and μj are the means of those readings.

The correlation r ranges between -1 and 1:

r=1: Perfect positive correlation (as one increases, so does the other).
r=−1: Perfect negative correlation (as one increases, the other decreases).
r=0: No correlation.

Example:

If you have two sensors with these readings over 5 time steps:

C1=[4500,4550,4600,4650,4700]
C2=[4700,4750,4800,4850,4900]

You would compute the correlation to see if their conductivity patterns are related.

I will create this in excel where I can see if the numbers and equations work as predicted "I Hope" This should make it easier to transform the math it into working code on the Arduino.

If any of you got input I would love to hear it, especially if its one of my likely many mistakes.

Excel and calibration to come,when i got time.

NoX Out

 

These 2 sensors are made for water quality information. They are also the most accessible option to calibrate and test the setup, and to be honest to see if its a viable idea using one or both of these sensors. its gonna be a party.

None of theses sensors fit into my current setup and ideally they should be mounted in the Reservoir.I got some ideas to make them fit.

https://randomnerdtutorials.com/arduino-tds-water-quality-sensor/



TS-300B Turbidity Sensor Module Water Quality Detection for Arduino


- The method the MPR12 (my rain child) is using is registering the conductivity of the surrounding material, not to the next pin.
- The TDS is measuring the conductivity between 2 pins with a know distance.
- The TS-300B Turbidity Sensor Module Water Quality Detection measure contaminants Parts Per Million (ppm) passing in between to small towers.

Are they comparable on a specific liquid, I don't know yet.

It might be a good idea to use more than one of these sensors. This would be on top of close to 50 other sensors. Not counting AMS1117 Multi DC-DC 12V-3,3V/5V/12V Step-Dow, MT3608 DC-DC Justerbar Step-up, Arduino® Due -2, Pumps, fans and more.

 

A minor update.

I fixed the design challenge with filters in the Water Cooling System, pictures will follow within 24 h.

I have been gone for some time, actually a month. This is partly because of My choice to putt an entire glass of coffee into the keyboard of the laptop, used for this project, then being approximately 1 sec to slow to disconnect the power cable and yea the usual short circuit sound followed, must be getting OLD.

When opportunity comes knocking: I have wanted to change from Windows to Linux on laptops and workstations for some time. So I pulled out an 18 year old laptop with a Intel® Core™2 Duo T6670 produced in 2006. Upgraded it to 8 GB RAM and using a 500 GB old style hard drive. Imagine, I can Run the CAD program on it! That was the 1st of the Linux PC still working on adapting my expectation and workflow but getting there-.

This reminds me in this world where the new, fast and expensive hardware is always on page one I would like to promote Wolfgang's Channel on YouTube .

Update:

Before the pumps there is a need for a filter to pick up any debris/parts coming off of the remaining parts in the cooling loop. Its been a challenge finding the right solution to fit this build, but here it is.

A mesh in stainless steel is formed on a '3D printed Curve, then a border is printed. Its inserted into a 1-2 mm frame cut into the Reservoir Body.

The water is entering the Reservoir from the left and leaves for the pumps on the right holes.








See you all soon here as usual.

Nox out.

 

Today’s Challenge: PC Air Filters, Air Permeability and Design.

When it comes to PC cases, cooling is a known a challenge. Many issues arise from overlooking airflow basics among these are air filters. Filters are the first line of defense against dust, but they significantly impact airflow, especially if the wrong material is used or is poorly designed.

Why Filter Fabrics Are a Problem

Many manufacturers of filter fabrics don’t provide critical specifications such as:
• Air permeability
• Thickness
• Breaking strength
• Weight
• Tensile elongation
• Working temperature
This lack of data makes choosing the right fabric difficult. Even when specifications are available, evaluating whether a fabric fits your needs can still be tricky. Let’s explore this using an example fabric: Polyester Woven Filter Cloth.
https://sffiltech2022.en.made-in-ch...-200-Micron-Polyester-Woven-Filter-Cloth.html

It has an air permeability of 32 L/m²/s (or 1920 L/m²/min). How does this work with a 120 mm fan running at 2000 RPM?

Matching Filters to Fans
1. Fan Airflow
A typical 120 mm fan at 2000 RPM generates 1700–2830 L/min (or 60–100 CFM).

2. Filter Air Permeability
The fabric’s air permeability is 32 L/m²/s, equivalent to 1920 L/m²/min.

Looking at the numbers its obvious that 1 Fan at 2000 rpm need more than 1 m² of filter fabric and lets be honest that is not viable (in a flat form). So we have to look at alternatives.

3. Surface Area Calculations
The filter's ability to support airflow depends on its surface area:
• Option A: Flat square filter up against the Fan (120 mm × 120 mm)
Area = 0.0144 m²
Max airflow = 1920×0.0144=27.65 L/min

• Option B: Circular filter up against the Fan (diameter 117 mm)
Area = 0.01076 m²
Max airflow = 1920×0.01076=20.65 L/min

• Option C: Large rectangle filter Not up against the Fan (400 mm × 210 mm)
Area = 0.084 m²
Max airflow = 1920×0.084=161.28 L/min

The fabric and design chosen makes the difference between hero or zero.

Summary Table:
Filter Shape Area (m²) Max Airflow (L/min)
Square (120 × 120 mm) 0.0144 27.65
Circle (Diameter = 117 mm) 0.01076 20.65
Rectangle (400 × 210 mm) 0.084 161.28

Key Takeaways
• The airflow capacity of this filter design is far below the fan’s output ability. For example, a 120 mm fan needs 1700–2830 L/min, while these options only provide 10% of that demanded Air.
• For adequate performance, either the filter must have much higher permeability or a larger surface area.

Real-World Design Example
In my prototype, I use a large rectangular filter (option C) for the hardware chamber to support 3 fans on one side of the case. Even so, the filter fabric only meets 10% of 1 fan’s requirement, meaning airflow must be improved 30× = (3000%) to meet the specification for 3 fans.
To avoid degrading the air flow, I split the case front panel into two zones with an acrylic divider, Why:
• Hardware side: 3 fans
• Water cooling side: 7 fans
This separation should ensure a better chance for a balanced airflow, as the fans on either side of the case run at different speeds.

MORE BELOW THE PICTURES.








Practical Issues with Bad Design

For a fan to function well, it needs to be able to create low pressure between the filter and fan. For example, Fractal’s "Mood" case has beautiful aesthetics but airflow challenges, as the fans inside need tight seals against the filters to work efficiently this can be fixed with a collar from the GPU to the filter or other components inside the case.

Solving the filter Air flow should be adapted on a case by case basis. :-D




Air Filter Types from ChatGPT:

Air filters come in various types, each designed for specific applications and levels of filtration. Here’s an overview of the main types:

1. Fiberglass Air Filters
• Material: Spun glass fibers.
• Use: Basic filtration in residential HVAC systems.
• Efficiency: Low; captures large particles like dust and lint.
• Pros: Inexpensive, disposable.
• Cons: Not effective for smaller particles like allergens or pollutants.

2. Pleated Air Filters
• Material: Polyester or cotton pleats.
• Use: Residential, commercial, and industrial.
• Efficiency: Medium to high, depending on the Minimum Efficiency Reporting Value (MERV).
• Pros: Captures small particles like pollen, mold, and pet dander.
• Cons: More expensive than fiberglass; may slightly restrict airflow.

3. HEPA Filters (High-Efficiency Particulate Air)
• Material: Dense mesh of fibers (fiberglass or synthetic).
• Use: Hospitals, clean rooms, high-quality home filtration.
• Efficiency: Extremely high; removes 99.97% of particles ≥0.3 microns.
• Pros: Excellent for allergies and asthma.
• Cons: Expensive; may require special equipment for airflow.

4. Electrostatic Air Filters
• Material: Self-charging fibers.
• Use: Residential and commercial.
• Efficiency: Medium to high.
• Pros: Washable and reusable; captures particles through static electricity.
• Cons: Initial cost is higher; cleaning required to maintain efficiency.

5. Activated Carbon Filters
• Material: Carbon-coated fibers or granules.
• Use: Odor and gas filtration (e.g., VOCs, smoke).
• Efficiency: Depends on carbon content; not designed for particles.
• Pros: Removes smells and chemical fumes.
• Cons: Does not capture particulate matter.

6. UV Light Filters
• Material: Ultraviolet light combined with other filtration systems.
• Use: Kills microorganisms like bacteria and viruses.
• Efficiency: High for microorganisms; not effective for particles.
• Pros: Adds a sterilizing layer to air purification.
• Cons: Requires electricity and periodic bulb replacement.

7. Washable/Reusable Filters
• Material: Metal mesh or electrostatic materials.
• Use: Long-term residential or commercial use.
• Efficiency: Low to medium.
• Pros: Eco-friendly and cost-effective over time.
• Cons: Lower filtration efficiency; requires frequent cleaning.

8. Media Filters
• Material: Layered pleated material (large surface area).
• Use: High-efficiency residential or commercial HVAC systems.
• Efficiency: High; comparable to HEPA but with better airflow.
• Pros: Long lifespan, excellent for high-particle environments.
• Cons: Bulky and more expensive.

9. MERV-Rated Filters
Filters are also categorized by their MERV rating (1-20):
• 1-4: Basic filters for large particles (dust, pollen).
• 5-8: Improved efficiency; captures smaller allergens.
• 9-12: High efficiency; captures fine particles.
• 13-16: Very high efficiency; suitable for hospitals and clean spaces.
• 17-20: Ultra-high efficiency (HEPA-level); for specialized industrial use.


Filters with 1000 l/m2/s:
Filters with a capacity of 1000 liters per square meter per second (L/m²/s) typically refer to high-flow filters designed for applications requiring substantial air or liquid throughput while balancing filtration efficiency. Such filters are often used in industrial, commercial, and specialized settings. Here's an overview of the potential filter types that can meet this specification:

1. Panel Filters
• Usage: HVAC systems in commercial and industrial applications.
• Flow Characteristics: Designed for high airflow rates like 1000 L/m²/s.
• Materials: Fiberglass or synthetic fibers, often with a pleated structure.
• Applications: Pre-filters in HVAC systems, protecting more sensitive downstream filters.

2. Pleated Filters
• Usage: Industrial air handling units.
• Flow Characteristics: Medium to high airflow, depending on pleat density and MERV rating.
• Efficiency: Can balance high throughput with moderate filtration (MERV 8–13).
• Applications: Factories, cleanrooms, and environments requiring particle filtration.

3. HEPA Filters
• Adaptations: Some HEPA filters designed for high-flow systems may reach these flow rates.
• Flow Characteristics: Typically require strong fans to overcome pressure drop.
• Applications: High-performance clean air systems in hospitals and laboratories.

4. Bag Filters
• Usage: Industrial dust collection systems.
• Flow Characteristics: Excellent for high flow rates.
• Efficiency: Varied, depending on bag material (e.g., polyester, polypropylene).
• Applications: Ventilation systems, dust collection, and large-scale air purification.

5. Electrostatic Filters
• Usage: Designed for high-flow environments.
• Flow Characteristics: Effective at maintaining flow while capturing fine particles.
• Applications: Environments where low-pressure drop and high airflow are critical.

6. Media Filters
• Usage: Used in advanced HVAC systems and specialized industrial applications.
• Flow Characteristics: Customized to allow high flow while maintaining filtration efficiency.
• Efficiency: Comparable to HEPA filters in some configurations.
• Applications: Data centers, pharmaceutical manufacturing, and other high-throughput systems.

Considerations for High Flow Filters
• Pressure Drop: Filters with such high flow rates often require careful selection to minimize pressure drop while maintaining efficiency.
• Efficiency Ratings: Choose a filter with the right balance of MERV rating or similar standards to suit your specific application.
• Durability: High-flow systems often demand robust materials to handle airflow stresses.

 

Status update.

A straightforward concept: The designed tech chamber in the front right side of the case, making it less accessible than the other components and could in a worst-case scenario, such as an electrical fire. be hard to access, so as an extra layer of safety I added a Fire bottle discharge port to the frame. The idea was copied from jet plane engines. Since I am putting the SAS4MC system together without support I thought it might be a prudent.

The Fire bottle discharge port is kept in place with a small spring mounted in the hinge. This is not yet added to the 3D drawing.

 

Food for Thought? Maybe.

I am still lingering on Filter Fabrics. Why you should ask! I can assure you its not anyway close to my main interest and as a concept its way to close to cleaning. So why? Well its because its close to environmental challenges. This made Me look into dust and lets just say NOT ENJOYABLE. But long-term airflow security has essentially the same challenges as in a Water Cooling Loop, it can get messy,

It all comes down to what environment your equipment is located in and how often you will clean it. Hard floors with few people aka as a low dust environment with next to No animal hairs and a high-flow mesh might be the way to go, the same goes for a server room. A teenage room, a 'garage or a workshop will likely all have different demands, then again maybe NOT. Placement of a PC, is the enclosure on a clean desktop or on the floor next to pets, and this is only considering dust and hair, not data security.

Creating calculation for a mesh since Air (gasses) are compress-able at sea level and around 20 degree C might be added later. The challenge is as always getting real life measurements. This might be an opportunity for a YouTube channel like major hardware LINK that measure fan designs and airflow. I can honestly say that I do appreciate the channels change in method when measuring the airflow. not to long ago.

So if ANYONE knows of a site with filter fabrics Airflow L/m2/sec or min please let me now! Thanks in advance.

* For this build I have considered and created an option that will allow the surplus airflow (over pressure) in the hardware chamber to be routed into the Water Cooling Chamber (WCC) . I still have to make an estimates/calculations of the impact of the 2 extremes. One being the lack of airflow in the WWC to the radiators versus hotter Air being blended into the WCC and the impact on the system versus cooling. The simple part is not routing air flow that is needed in the hardware chamber. The design part on this subject was fun with lots of options.

Don't you ever wonder why store bought cases are not marked with a maximum amount of watt that can be used inside!

I am considering a high flow mesh for my build but its still to be decided.

Nox Out

 

Merry Christmas and a Happy New Year!

I played/Experimented with creating an optimized Motherboard Tray. To be honest I am curious to see If This idea will spread. I started by widening the Tray to 25 mm. That again forced me to adapt the air intake in the Water Cooling Chamber. The wider Tray 10mm to 25 mm had me focus on slimming the Tray, a process many of us know is approaching after Christmas. Well for the Tray it has made me rework the production process and materials. I like options so reworking the Tray to fit wires inside seems like a bonus modding idea, this of cause brings up some problems with heat dissipation from the embedded wired. So here are some status pictures.

Slightly modified Raw Motherboard Tray:


'
Adding the Parts that Need to be there:



Bringing us here:



Checking that the Air intake Tunnels are as needed:



* After checking the Likely Route of the Wires Its more than 90% likely that I will redo the Air Intake Area to be able to fix some stuff on the Motherboard Tray. On the following Picture the rear of the Tray can be seen to have lost some weight for a LED area lighting the rear IO of the Motherboard.



To minimize the amount of storage area for Dust some cut outs for acrylic or glass fiber plates are made. they will be glued into the frame (atm). this part of the drawing is not done.



The wires will be led in trough cut outs in the frame, so might be a good idea not to use glue else it might be hard to fix wiring problems.



With Motherboard



This is the 1st part of the motherboard optimization:



Adding Components on the Motherboard to add Wire Routes:

Found a picture in the motherboard Manual and I can honestly say I like the X870E motherboards, they have the PCIE slots I have been waiting for.



So after adding the Motherboard layout picture to the CAD Program I used it to quick prototype the layout



This would bring us to wire routes but those are at present not good enough, so I will rework the parts that need it. See you all on the other side of 2024.

Nox Out