• Save 15% on ALL SFF Network merch, until Dec 31st! Use code SFF2024 at checkout. Click here!

Stalled Winter One -- 15.6L SFF case, 3090 Support, 3-slot GPUs, dual 280mm radiators, CFD Optimized Design

Post-Newt

SFF Lingo Aficionado
Oct 12, 2020
105
247
Fouzan has said that pull configuration are better because you want the exhausted air to be at a high velocity to prevent recirculation. Having the fans in push causes the air pushed by the fan to be slowed down by the radiator fins resulting in slower exhaust air. (This is is my interpretation of his recs).
All four fans configured to push and exhaust will look too damn good to pass up.
 

mxj1

Cable-Tie Ninja
Sep 13, 2020
179
451
Has it been verified either way whether it is better to push the airflow through the radiator by placing the fans under the radiator, or by sandwiching the fans between the radiator and the top of the case. Looking at the recent beauty shots in post #1,369 it looks like Fouzan has gone for the latter? Please forgive me if this has been answered (to death) already.

Water cooling rule of thumb should be: slow moving fans (<800 rpm) pull air through the radiator; fast moving fans (>1000 rpm) push air through the radiator. Just so happens to be the best for the case too... But I'm curious how this will impact bottom radiator placement, and the fitting colliding with the power cord when place at the rear. Fingers crossed.

of course with fans that create very high static pressure, it doesn't matter. 🙂
 
  • Like
Reactions: Post-Newt

WinterCharm

Master of Cramming
Original poster
Jan 19, 2019
428
1,941
Update #10 -- An Unfortunate Delay, Some Good News

Hello everyone.

I have unfortunate news. Circumstances around our supply chain forced us to delay the shipping date on Winter One to the second week of March. I know this is a disappointing announcement. It stings to hear that there's another delay. I feel like I've let all of you down. As always, I'm going into detail about the circumstances around this delay, and what we’ve done to address the underlying problem.

The Delay:

What Went Wrong: Our packaging supplier has not delivered.
This was the supplier we had chosen to replace the first packaging supplier that cancelled on us (one of the causes for the last delay). They were supposed to send us our order this week. We have not heard anything from them. We *need* those foam inserts to protect the case parts in transit.

Addressing the Problem:
This week, we worked with 3 different suppliers to make foam inserts for Winter One. What we thought would be a simple part of our supply chain turned into a repeated source of delay. To ensure this problem does not crop up again, we will negotiate a long-term contract with one of the new packaging suppliers.

Why Go Through All This Trouble?
We want to reduce delays on all future batches. The creation and pruning of a supply chain is a painstaking process. Each problem we resolve now is less likely to recur in a future batch. While proper solutions with redundancy take more time to put in place, they cost less time in the future. Choosing a single packaging supplier after the first delay was an error in judgement. I will not be repeating this mistake.

The Good:

Final Corrections:
I spoke to our manufacturing partner on Monday, and went over detailed feedback from early production. We even had a video call where I showed them each issue that came up during the build on preproduction parts. We outlined what needed to happen in order to correct the kerning issue on the rear panel engraving, and further detailed acceptable surface finishes.

Full Production Start:
Yesterday, our manufacturing partner informed me that corrections had been applied to the production process. On Monday, Full Production of Batch 1 / Black will start!

Batch 2 / Silver

I wanted to take a moment to address those of you who opted for Silver directly. You have all been waiting patiently for your cases. Batch 2’s turnaround will be much swifter than Batch 1, due to measures taken:
  • I have included internal components from Batch 2 in the Batch 1 / Black Production Run. This allows us to speed up turnaround time for those of you in Batch 2 who placed an order through Kickstarter. (Remember, all interior parts for Winter One are matte black. This optimization is a direct result of that decision!
  • Additionally, with the 3 new foam orders we placed, we will have more than enough packaging and stock parts on hand to cover Batch 2 orders placed through Kickstarter.
  • The last step on Kickstarter Batch 2 orders is producing the Matte Silver exterior panels and parts. This process will begin when our production facilities have completed the Matte Black production run.

The Next Few Weeks

We are on the last few steps before shipping Batch 1. Over the next 5 weeks, I have a few things planned in addition to the updates on manufacturing, and packaging. We’ll be releasing a new build video, the completed user manual, new CFD data, and thermal / noise testing, as well as another round of Q&A.
 

mxj1

Cable-Tie Ninja
Sep 13, 2020
179
451
Sad Cry GIF by Team Coco
 

rfarmer

Spatial Philosopher
Jul 7, 2017
2,669
2,793
Sorry to hear about the problems with your packaging supplier but with your shipping estimate so close to Chinese New Year this was almost bound to happen. Hope everything goes smoothly in the future.
 

WinterCharm

Master of Cramming
Original poster
Jan 19, 2019
428
1,941
This feels awful. The reason I delayed this update is because I wanted to have answers. I didn't want to be even slightly vague when it came to being able to answer when shipping would happen. A lot of time and attention this week went into squashing this problem by pulling all the stops.

It was my fault for underestimating the pressure on packaging suppliers caused by the explosive growth of e-commerce. What I had initially anticipated to be a simple fix that required submitting another order turned into a mistake that has affected everyone in the worst way. It's a painful lesson to learn when everyone trusted in you to deliver a product on time. I will make sure this doesn't happen again.
 

dotnetapp

Chassis Packer
Sep 25, 2019
18
24
Update #10 -- An Unfortunate Delay, Some Good News

Hello everyone.

I have unfortunate news. Circumstances around our supply chain forced us to delay the shipping date on Winter One to the second week of March. I know this is a disappointing announcement. It stings to hear that there's another delay. I feel like I've let all of you down. As always, I'm going into detail about the circumstances around this delay, and what we’ve done to address the underlying problem.

The Delay:

What Went Wrong: Our packaging supplier has not delivered.
This was the supplier we had chosen to replace the first packaging supplier that cancelled on us (one of the causes for the last delay). They were supposed to send us our order this week. We have not heard anything from them. We *need* those foam inserts to protect the case parts in transit.

Addressing the Problem:
This week, we worked with 3 different suppliers to make foam inserts for Winter One. What we thought would be a simple part of our supply chain turned into a repeated source of delay. To ensure this problem does not crop up again, we will negotiate a long-term contract with one of the new packaging suppliers.

Why Go Through All This Trouble?
We want to reduce delays on all future batches. The creation and pruning of a supply chain is a painstaking process. Each problem we resolve now is less likely to recur in a future batch. While proper solutions with redundancy take more time to put in place, they cost less time in the future. Choosing a single packaging supplier after the first delay was an error in judgement. I will not be repeating this mistake.

The Good:

Final Corrections:
I spoke to our manufacturing partner on Monday, and went over detailed feedback from early production. We even had a video call where I showed them each issue that came up during the build on preproduction parts. We outlined what needed to happen in order to correct the kerning issue on the rear panel engraving, and further detailed acceptable surface finishes.

Full Production Start:
Yesterday, our manufacturing partner informed me that corrections had been applied to the production process. On Monday, Full Production of Batch 1 / Black will start!

Batch 2 / Silver

I wanted to take a moment to address those of you who opted for Silver directly. You have all been waiting patiently for your cases. Batch 2’s turnaround will be much swifter than Batch 1, due to measures taken:
  • I have included internal components from Batch 2 in the Batch 1 / Black Production Run. This allows us to speed up turnaround time for those of you in Batch 2 who placed an order through Kickstarter. (Remember, all interior parts for Winter One are matte black. This optimization is a direct result of that decision!
  • Additionally, with the 3 new foam orders we placed, we will have more than enough packaging and stock parts on hand to cover Batch 2 orders placed through Kickstarter.
  • The last step on Kickstarter Batch 2 orders is producing the Matte Silver exterior panels and parts. This process will begin when our production facilities have completed the Matte Black production run.

The Next Few Weeks

We are on the last few steps before shipping Batch 1. Over the next 5 weeks, I have a few things planned in addition to the updates on manufacturing, and packaging. We’ll be releasing a new build video, the completed user manual, new CFD data, and thermal / noise testing, as well as another round of Q&A.
its ok i mean its not like i could buy a gpu or anything anyway :p
 

carlover670

Average Stuffer
Oct 22, 2020
74
100
Full Production Start:
Yesterday, our manufacturing partner informed me that corrections had been applied to the production process. On Monday, Full Production of Batch 1 / Black will start!
Woah! Thats really exciting to hear things coming together aside from the packaging delay. Second week of march is over a month away. Do you think batch 2 will be done or will start production before batch 1 ships?
 
  • Like
Reactions: WinterCharm

G23 Mr Gimp

Caliper Novice
Bronze Supporter
Sep 14, 2020
32
90
This feels awful. The reason I delayed this update is because I wanted to have answers. I didn't want to be even slightly vague when it came to being able to answer when shipping would happen. A lot of time and attention this week went into squashing this problem by pulling all the stops.

It was my fault for underestimating the pressure on packaging suppliers caused by the explosive growth of e-commerce. What I had initially anticipated to be a simple fix that required submitting another order turned into a mistake that has affected everyone in the worst way. It's a painful lesson to learn when everyone trusted in you to deliver a product on time. I will make sure this doesn't happen again.

This is more of a new venture for you than it is for us as consumers. The pressure resting on your shoulders under the weight of your own expectations must be immense, let alone that of trying to appease those of us who have come along for the journey with you.

I know I can't speak for everyone, but the regular updates and sincerity of your own disappointment in each of your updates where something hasn't gone to plan shows that (in mind anyway) I definitely made the right choice to back this project. You're doing your very best in some very weird times, and spinning as many plates as you are must be exhausting.

I don't feel let down. What you have already achieved is something to be proud of.
 
Last edited:

WinterCharm

Master of Cramming
Original poster
Jan 19, 2019
428
1,941
Woah! Thats really exciting to hear things coming together aside from the packaging delay. Second week of march is over a month away. Do you think batch 2 will be done or will start production before batch 1 ships?

Batch 2 will have started production once Batch 1 is shipping out. At that time, we will also start taking Batch 2 orders on the website. Delivery on the Kickstarter portion of Batch 2 will happen more quickly, since partial production will have happened by then. The Website orders for Batch 2 will come at the tail end of that batch, as production finishes up.

With each batch, our supply chain should become more robust. The ultimate goal is a 2 week turnaround from order to shipping, with JIT manufacturing of all orders. That's what we are striving toward in the long-term.

Here are some results from a pump block vs. standard block test. This test helped me decide what to do in the W1 so maybe it will help others.

The information below combines two different systems of measurement and weird table formatting. Read at your own risk.
This is really well done, and a handy resource for everyone in this thread. Thank you <3

I managed to order a EVGA 3070 XC3 Ultra from Scan today (£575 including next day DPD). I almost bought the ftw3 ultra they had in stock as it was actually cheaper by £25. I was hoping to get a 3080 but I think I'll wait until the end of the year and see if stock is any better. It might be too much for me anyway (1440p).

Was I right to be cautious of a 2.75 slot card with the solid side panels?
For solid panels, a 2.75 slot card would be really pushing it, in terms of airflow available to the GPU.

Just while we all eagerly wait for the next update, I was wondering if somebody could help me with a couple of quick questions about case fans.

I have 2 spare slim 15mm noctua fans which i plan to use in the bottom of the case as exhaust. Is it worth swapping them for the 25mm ones?

I have an x53 kraken AIO. I was going to swap the included NZXT fans with 2 x 25mm noctua fans which I already bought. Is this worth it? I suspect it'll be quieter but I'm wondering if I could save money by leaving the radiator alone and using the aforementioned 25mm noctua fans in the bottom of the case instead.

Cheers,
Nik

IMO it would be worth swapping the NF A12x25 fans onto your radiator. They are absolutely some of the best performing fans on the market in most situations. If you have to, put the A12x25's on the radiator, and put the A12x15's in the bottom of the case, or go with A12x25's on the top / bottom. Your other option is to grab a set of Arctic P12s in a pack of 5. They're an *amazing* value in the PC Case Fan space, and you can have very quiet 25mm fans on the top / bottom exhaust in your case.

So in summary, for me the best option is to:

- pass on the slim 15mm Noctua fans (and the NF-A9 92mm fan) all these were bought for an nCase m1 build but I sold the case when I discovered this thread.

- use my 2 x NF-A12x25's as exhausts in the bottom of the case.

- purchase 2 x NF-F12 PWM's to attach to the 240mm radiator.

Has it been verified either way whether it is better to push the airflow through the radiator by placing the fans under the radiator, or by sandwiching the fans between the radiator and the top of the case. Looking at the recent beauty shots in post #1,369 it looks like Fouzan has gone for the latter? Please forgive me if this has been answered (to death) already.

As far as i can tell, I do not need Noctua NA-SFMA1. I had a quick go mounting my NF-A12x25's to the kraken radiator and the included 30mm screws seemed to hold it in place just fine. Or is there a benefit i'm missing?

Once again, thanks in advance.

Based on these 3 options, use your NF A12x25's in the bottom of the case. You could even get another set of A12x25's for the radiator, rather than the F12s.

It's better to put the fans in exhaust and pull air through the radiator -- especially at lower fan RPMs, as you get steadier airflow, and higher exhaust velocity from the top / bottom, which maintains the exterior vortices that prevent mixing / re circulation of warm air.

However, if you go with a push configuration, higher fan RPMs should help maintain pressure and exhaust velocity. It will be a bit louder, however.

Question, can the Noctua NA-SFMA1 (with NF-A12x25) be installed in the winter one?

Yes. It'll be a tight fit, but if your radiators are 29.5mm thick, instead of 30mm, it should work.
 

carlover670

Average Stuffer
Oct 22, 2020
74
100
BTW, since you have a two cases. I assume you only need one for testing (granted Idk if your going to use two different cases to do air and water cooled testing). Is LTT getting one right now? Not sure if the defect is on both cases. Really excited to see LTT build in one. I'm just imagining that lattice pattern under studio lighting looking like an absolute treat!
 

Goatee

King of Cable Management
Jun 22, 2018
739
1,513
Batch 2 / Silver
I wanted to take a moment to address those of you who opted for Silver directly. You have all been waiting patiently for your cases. Batch 2’s turnaround will be much swifter than Batch 1, due to measures taken:
  • I have included internal components from Batch 2 in the Batch 1 / Black Production Run. This allows us to speed up turnaround time for those of you in Batch 2 who placed an order through Kickstarter. (Remember, all interior parts for Winter One are matte black. This optimization is a direct result of that decision!
  • Additionally, with the 3 new foam orders we placed, we will have more than enough packaging and stock parts on hand to cover Batch 2 orders placed through Kickstarter.
  • The last step on Kickstarter Batch 2 orders is producing the Matte Silver exterior panels and parts. This process will begin when our production facilities have completed the Matte Black production run.

I am reading into this that the Black batch is now planned for shipping mid march (approx. 4 working weeks after production start - account for CNY). Does that mean we hope to get silver batch shipping end march-early April?
 
  • Like
Reactions: WinterCharm

WinterCharm

Master of Cramming
Original poster
Jan 19, 2019
428
1,941
BTW, since you have a two cases. I assume you only need one for testing (granted Idk if your going to use two different cases to do air and water cooled testing). Is LTT getting one right now? Not sure if the defect is on both cases. Really excited to see LTT build in one. I'm just imagining that lattice pattern under studio lighting looking like an absolute treat!

I want to make sure the unboxing LTT has on camera is the same as everyone else. So we'll be waiting for packaging before shipping a unit to them. I want to make sure they get exactly what everyone else will get, so their review is an evaluation of what to expect for anyone who purchases a Winter One.

Second, the packaging plays a huge part in protecting the case during shipping, and I want to make sure these cases arrive in great condition, to everyone who orders one. It would be a shame to do all the work on QC, finishing, and more, only to have them get scuffed and dinged during shipping.

I am reading into this that the Black batch is now planned for shipping mid march (approx. 4 working weeks after production start - account for CNY). Does that mean we hope to get silver batch shipping end march-early April?

That is correct!
 

jupiterking

SFF Lingo Aficionado
Oct 23, 2020
90
163
This is too bad to hear but I’m sure it’ll be worth the wait. I’ll use this delay as an excuse to replace literally the last old component on my desk.... my mouse. 😅

If you delay any more I might be tempted to do a temporary cardboard box build. Been so excited to try out the new CPU. 😂
 
  • Like
Reactions: Koxx5D and mxj1

biopunk

Airflow Optimizer
Sep 24, 2020
248
359
I feel like a high flow rate target (1 gpm) is going too deep into the world of diminishing returns. With that said, one could fit this next to an FE card. The conversation of 280 space in this case has come up before and I'm just glad any 280 could fit at all.
The thing is, with a DDC pump at 20-30%, two water blocks, a restrictive radiator, and multiple 90° bends the flow rate would be way below 1 GPM, in the range where it makes the biggest difference for thermal performance.
According to your own tests of EK Velocity, going from 0.13 to 0.4 GPM improves CPU temps by 3.9℃. For comparison, swapping GTS 280 U-flow for X-flow improved the flow rate in my tests by roughly 0.25 GPM. I'm aware that X-flow doesn't fit in W1, it's just an example of a less restrictive radiator that I tested.

Of course, using D5 is ideal for W1 because it can push the flow rate above 1 GPM where 0.2-0.3 GPM extra would make only a marginal difference.
 
Last edited:
  • Like
Reactions: Post-Newt

WinterCharm

Master of Cramming
Original poster
Jan 19, 2019
428
1,941
The thing is, with a DDC pump at 20-30%, two water blocks, a restrictive radiator, and multiple 90° bends the flow rate would be way below 1 GPM, in the range where it makes the biggest difference for thermal performance.
According to your own tests of EK Velocity, going from 0.13 to 0.4 GPM improves CPU temps by 3.9℃. For comparison, swapping GTS 280 U-flow for X-flow improved the flow rate in my tests by roughly 0.25 GPM. I'm aware that X-flow doesn't fit in W1, it's just an example of a less restrictive radiator that I tested.

Of course, using D5 is ideal for W1 because it can push the flow rate above 1 GPM where 0.2-0.3 GPM extra would make only a marginal difference in thermal performance.

If you look at what's happening to the system In real world terms, it's pretty fascinating. achieving a true flow rate of at *least* 0.5 GPM is a good bare minimum. Ideally, you want 0.7+ GPM. That's around where diminishing returns start for most loops.

From a purely mathematical perspective, it's true that your heat transfer is directly proportional to flow rate through the loop, but there are other forces at play, at both the CPU block itself, and on the Radiator Side

Qtotal = m*Cp*∆T In the sense that, say you have a heat load of 250W from the CPU. Specific heat Of Water (Cp) is constant. The only thing that will change is ∆T and m. If your m is too low, your coolant temps will get very high, and that will also reduce the heat transfer efficiency
Q=U*A*(Tcpu-Twater) where A is the CPU block fins area from the CPU through the water block.

On the other side of things, your "diminishing returns" from increasing the flow rate further, comes from decreasing heat transfer efficiency on the radiator side of things where Q=U*A*(Twater-Tair), and A is radiator area. If the ∆T between air and water is too low, then the you cannot move heat out of the water. This is why, for most systems, coolant temps will slowly rise until they stabilize. This is because Q=m_air*Cp*(∆T), where the speed of air going past the fins will affect how much heat can be extracted from the water. Your ideal fluid temperature will maximize the heat transfer from the cpu/gpu blocks >> water AND from water >> air. Every Loop, based on the components used, has an ideal flow rate that sets the coolant temp to a particular point at which there is MAXIMUM heat transfer on *both sides* of the water. Finding this point at max load for your loop is the most important thing, as you can use it to set a steady flow rate for your pump, and set your fan RPM appropriately, minimizing noise.

What helps is to understand your watercooling loop as a series of heat transfer steps, and to use your temperature readouts to figure out how to optimize your loop based on where the heat flow restriction is. If anyone is curious, I might go write a detailed post on this. (brings back a lot of memories from my fluids and thermo classes back in my engineering days).
 
Last edited:

Post-Newt

SFF Lingo Aficionado
Oct 12, 2020
105
247
If you look at what's happening to the system In real world terms, it's pretty fascinating. achieving a true flow rate of at *least* 0.5 GPM is a good bare minimum. Ideally, you want 0.7+ GPM. That's around where diminishing returns start for most loops.

From a purely mathematical perspective, it's true that your heat transfer is directly proportional to flow rate through the loop, but there are other forces at play, at both the CPU block itself, and on the Radiator Side

Qtotal = m*Cp*∆T In the sense that, say you have a heat load of 250W from the CPU. Specific heat Of Water (Cp) is constant. The only thing that will change is ∆T and m. If your m is too low, your coolant temps will get very high, and that will also reduce the heat transfer efficiency
Q=U*A*(Tcpu-Twater) where A is the CPU block fins area from the CPU through the water block.

On the other side of things, your "diminishing returns" from increasing the flow rate further, comes from decreasing heat transfer efficiency on the radiator side of things where Q=U*A*(Twater-Tair), and A is radiator area. If the ∆T between air and water is too low, then the you cannot move heat out of the water. This is why, for most systems, coolant temps will slowly rise until they stabilize. This is because Q=m_air*Cp*(∆T), where the speed of air going past the fins will affect how much heat can be extracted from the water. Your ideal fluid temperature will maximize the heat transfer from the cpu/gpu blocks >> water AND from water >> air. Every Loop, based on the components used, has an ideal flow rate that sets the coolant temp to a particular point at which there is MAXIMUM heat transfer on *both sides* of the water. Finding this point at max load for your loop is the most important thing, as you can use it to set a steady flow rate for your pump, and set your fan RPM appropriately, minimizing noise.

What helps is to understand your watercooling loop as a series of heat transfer steps, and to use your temperature readouts to figure out how to optimize your loop based on where the heat flow restriction is. If anyone is curious, I might go write a detailed post on this. (brings back a lot of memories from my fluids and thermo classes back in my engineering days).
ron burgundy anchorman GIF