Here are the results from different Tophat configurations - AIR, let's discuss the results!
-
Save 15% on ALL SFF Network merch, until Dec 31st! Use code SFF2024 at checkout. Click here!
Prototype The world most flexible SFF case? LOUQE Ghost S1
- Thread starter PaChalski
- Start date
D
Deleted member 4779
Guest
Thank you for publishing the tests above @PaChalski!
The MS configuration has interesting spikes. I would have expected the general trend to match the other two configurations (like XS and M already do). Instead, the temperatures seem to rise against XS+M at some key points. I wonder what's going on there.
Would you happen to have specs on the fans by the way? If the MS configuration had slim fans in the bottom, they would probably have had lower CFM ratings relative to the 120mm in the topmost hat right? The reason I ask is because the thin profile of the small hat + the low profile rubber feet doesn't seem like it would give the fans much clearance to pull in air.
I wonder if either the slim fans in the M hat, or the 120x25mm fans in an L hat would have sufficient clearance to pull in air more effectively.
I'm ultimately interested in a positive pressure setup with 120x25mm fans in an M hat in the bottom; I suspect that when using exhaust fans at the top only, xpockets of hot air might get trapped at the bottom of the case below the motherboard and also under the GPU (if using an aftermarket open-air as opposed to blower).
It would be interesting to see results for intake-only setups too. Specifically:
1) S-Hat with 2x slim fans on the bottom
2) M-hat with 2x slim fans on the bottom
3) M-hat with 2x regular (25mm-thick) fans on the bottom
3) L-hat with 2x regular (25mm-thick) fans on the bottom
The MS configuration has interesting spikes. I would have expected the general trend to match the other two configurations (like XS and M already do). Instead, the temperatures seem to rise against XS+M at some key points. I wonder what's going on there.
Would you happen to have specs on the fans by the way? If the MS configuration had slim fans in the bottom, they would probably have had lower CFM ratings relative to the 120mm in the topmost hat right? The reason I ask is because the thin profile of the small hat + the low profile rubber feet doesn't seem like it would give the fans much clearance to pull in air.
I wonder if either the slim fans in the M hat, or the 120x25mm fans in an L hat would have sufficient clearance to pull in air more effectively.
I'm ultimately interested in a positive pressure setup with 120x25mm fans in an M hat in the bottom; I suspect that when using exhaust fans at the top only, xpockets of hot air might get trapped at the bottom of the case below the motherboard and also under the GPU (if using an aftermarket open-air as opposed to blower).
It would be interesting to see results for intake-only setups too. Specifically:
1) S-Hat with 2x slim fans on the bottom
2) M-hat with 2x slim fans on the bottom
3) M-hat with 2x regular (25mm-thick) fans on the bottom
3) L-hat with 2x regular (25mm-thick) fans on the bottom
This is looking very promising, best of luck guys. Any pics of the ash grey colorway?
Thank you for publishing the tests above @PaChalski!
The MS configuration has interesting spikes. I would have expected the general trend to match the other two configurations (like XS and M already do). Instead, the temperatures seem to rise against XS+M at some key points. I wonder what's going on there.
Would you happen to have specs on the fans by the way? If the MS configuration had slim fans in the bottom, they would probably have had lower CFM ratings relative to the 120mm in the topmost hat right? The reason I ask is because the thin profile of the small hat + the low profile rubber feet doesn't seem like it would give the fans much clearance to pull in air.
I wonder if either the slim fans in the M hat, or the 120x25mm fans in an L hat would have sufficient clearance to pull in air more effectively.
I'm ultimately interested in a positive pressure setup with 120x25mm fans in an M hat in the bottom; I suspect that when using exhaust fans at the top only, xpockets of hot air might get trapped at the bottom of the case below the motherboard and also under the GPU (if using an aftermarket open-air as opposed to blower).
It would be interesting to see results for intake-only setups too. Specifically:
1) S-Hat with 2x slim fans on the bottom
2) M-hat with 2x slim fans on the bottom
3) M-hat with 2x regular (25mm-thick) fans on the bottom
3) L-hat with 2x regular (25mm-thick) fans on the bottom
Yes, those spikes do look funny. Could it simply be an offset of the data series? We could try to align them better and see if it makes more sense?
Looking at the average numbers we see the slim bottom fans in MS run are helping, both in terms of temps and lowering GPU fan speeds, resulting in a more quiet system. All case fans were set to 800 RPM where they don't create any noise.
Fans used in the bottom is our only NF-A12x15 PWM (still waiting for two more to arrive that we found in Denmark) and the Shuriken fan that comes with the cooler - Not an ideal choice of fans =)
Top fans are Corsair, the ones that come with the H100 v2.
My thoughts are that AIO up top & dual 25mm thick fans down below (with an open air GPU) might be the way to go...
With dual 120mm x 25mm fans in the bottom set as intake & the AIO fans exhausting, this should be more of a positive pressure setup...?
But I would definitely swap the current feet for ones two to three times taller, for improved airflow from below...!
I believe you could skip the top fans in the AIO, go for a dual M-TopHat setup and put the fans in the bottom. Or even go for S-TopHat in the bottom and use slim fans for cooling.
There has been a enough discussions about the small feet that we will do some tests with an elevated case at different heights(ex: +5mm, +10mm, +15mm). If this proves to have radical positive impact we will evaluate if taller (optional) feet should be included in the kit or available as optional add-ons.
So, what might be interesting is to use two top hats, one for CPU AIO cooling, the other for GPU AIO cooling. Did not expect much performance increase but nice to show it is DOable!
Here are some pictures with the S24 AIO installed in the bottom. A very nice choice for the Ghost btw!
- Quiet
- No cable clutter
- Great thermals
- Swedish
- Quick easy installation
I think there are too many variables changing here to get a clear comparison between air cooling and liquid cooling with a 240mm radiator. The temperature and the noise changes for each of the tests.
The tests would be much more meaningful if you constrained both tests to either the same noise level or the same temperatures.
Also, without dBA measurements, we sort of have to just take your word for it about what the noiseless level means. It would probably be better to get a dBA reading 0.5m from the case or something else empirical.
That said, it looks like you could draw the conclusion that there isn't as much benefit to going to a 240mm radiator for cooling instead of air cooling.
The tests would be much more meaningful if you constrained both tests to either the same noise level or the same temperatures.
Also, without dBA measurements, we sort of have to just take your word for it about what the noiseless level means. It would probably be better to get a dBA reading 0.5m from the case or something else empirical.
That said, it looks like you could draw the conclusion that there isn't as much benefit to going to a 240mm radiator for cooling instead of air cooling.
I'm having trouble understanding the results. Agree with HooglyBoogly; I think it would be more intuitive to set a temperature threshold (say, 70C for the CPU, etc) and measure the noise while not constraining the fan speed. Then the tests would show the noise levels produced to stay under the temperature threshold, which imo is more similar to the average user's typical use case
Firestrike Ultra is more of a GPU test, which can warm up the cpu a bit under the physics and combined section. If you really want to push the cpu to the limit, then go prime95 and disable AVX if you're worried about avx instructions or AIDA64 for a bit less stressful test.
I believe that the aio on the top will have a overall better temp result, due to not having to pull air from the ground. However with open air gpus, the hot air from the gpu will go up and warm the cpu so in situations where both are pushed to the max (folding is an easy example), it should be easy to see how an open air card like the SC2 1080 ti can negatively affect the cpu temps.
I believe that the aio on the top will have a overall better temp result, due to not having to pull air from the ground. However with open air gpus, the hot air from the gpu will go up and warm the cpu so in situations where both are pushed to the max (folding is an easy example), it should be easy to see how an open air card like the SC2 1080 ti can negatively affect the cpu temps.
All very good points regarding the tests, methods and reports, when we have more time, during the campaign, we will do more tests and present them in a more structured way so that you will be able to decide what tophats fit your needs most. Tophats will be selected after the campaign is over. now it is a question of having one (two, three) or not.
Seems like for air cooling the l12 is the most well rounded in your air cooler tests, another good bet may be a c7 cooler with a 25mm noctua fan, at 58mm it should avoid the turbulence, and fair much better than the l9i. With the copper c7 (if that will ever come out) it might be an even better set up then the l12. Thoughts? If you knew the thermal conductivity and surface area of the heat sink could you get a rough idea of performance?
Are there any air coolers that would fit that would be suitable for a stock 7820x? A TDP of 140w is pretty high for what i've seen around, and my plan at the moment is to watercool and use a L top hat, but thought i'd see if anyone knew of any.
That sounds like a very strong competitor!!! Is it not better to test that exact combo for real an see how it stacks up though. With your help we will be able to hash out the 5-6 best Air cooler combos to test in the case for the real ("official") report that we will be able to publish on our website during the campaign.
Do you guys think I would be able to fit an MSI gtx 970 into the case? I've been looking around for the specs of the card and come up with 142mm which is just short of the 143mm you allow. Will it fit comfortably in the case?
Really looking forward to the campaign this Saturday. Will be tuning in from Denmark!
Really looking forward to the campaign this Saturday. Will be tuning in from Denmark!
I
If the GPU is no longer than 305 mm and if the power connectors don't sit at the very top of the card (standing on the PCB) it will fit perfectly!
If the GPU is no longer than 305 mm and if the power connectors don't sit at the very top of the card (standing on the PCB) it will fit perfectly!
Thanks for the reply. Will surely have to measure that when I get home!
EVGA needs to offer a 240mm radiator option with their Hybrid graphics cards, or at least with their Hybrid conversion kits...!
Looked at the specs and it looks fine to me, but double check the placement of the power sockets!
Similar threads
