With some staff changes here at TweakTown, the opportunity to review cases has been put back into our lap. Rather than delivering reviews as we have in the past, we are stepping up the game a fair bit, but not to the point where much of what we find becomes irrelevant to the average buyer. In this quest, it makes a ton of sense for us to sit down, see what we want to do this time, and hopefully explain it to our readers in a way that eliminates questions that may come up along the way.
In our earlier reviews of cases here, we looked at them in a few different lights. The first view of a chassis covers fit and finish, with aesthetic appeal tied into that, and this is an assessment of the product as soon as it is removed from the box. We then move onto external features that may improve access or airflow, and we move into the interior next. Much of the same applies here, but we look with a more discerning eye inside, as it tends to have a lot more going on to look at. The build is also a great time in our process to get to know the chassis in question, and once done with that; we move onto testing, which is where things are going to change a bit in the way we attack cases from here on out.
Shannon brought forth some thermal and audio results, where we used to go with more of a pass/fail system for both, but we like the idea and are taking the concept and putting our spin on it. While we could keep a mess of thermal probes in various areas, we opt to make it even simpler. We have taken the test system and tested it in an open-air environment, with the office kept at 22.5°C at all times. This gives us what we will call the best thermal conditions for the system, without any restrictions to any of the components used.
We will then compare those open-air results to what the cases offer with the same gear inside. Seems simple right? We will also take audio measurements, both with PWM in control, if it applies to the fans supplied in said chassis, and we will also report a maximum decibel level of the pre-installed fans. If no fans are shipped in a chassis, we obviously will not mention sound, but we will test the chassis as it is shipped thermally.
The Test System Components
Before we show off the hardware we are using, we would like to take this time to give a shout out to ASUS for being a huge support system for us over here at TweakTown. We also owe a special someone over at NVIDIA much thanks for supplying us with one of the many unicorns in the GPU category to stay up to date and have a system that is not only relevant but visually appealing.
- Motherboard: ASUS ROG DARK HERO (AMD X570) - Buy from Amazon
- CPU: AMD Ryzen 5 3700X - Buy from Amazon
- Memory: Team T-Force XTREEM ARGB 3600MHz 2X8GB
- Graphics Card: NVIDIA GeForce RTX 3070 FE - Buy from Amazon
- Storage: SanDisk Extreme II 120GB SSD
- Case: Hydra Bench Standard
- Power Supply: ASUS ROG Thor 850W - Buy from Amazon
- OS: Microsoft Windows 10 Home 64-bit - Buy from Amazon
- Software: AMD Ryzen Master, AIDA64 Engineer 6.32.5600, and CPU-z 1.94.0 x64
The major component of any chassis build is the motherboard, as it is the one part that everything else connects. Without it, a PC is literally just a bunch of components with no way to make anything happen. We had a couple of other choices as to which ASUS motherboard to settle with, but we opted to go with the ROG Crosshair VIII DARK HERO for a few reasons.
We love our HERO Wi-Fi boards, they work well for testing and is a system we are used to rifling through on the BIOS level. We wanted to dumb down the RGB to allow the chassis to shine and not depend on the system for all of the bling factor, simply by using a windowed panel. Lastly, we have other plans for this system as well, and it required a motherboard with more PCI-e slots, which this AM4 X570 motherboard provides.
We are using an AM4 motherboard, limiting us to AMD, and due to current environmental and biological conditions, we were even limited from 5-series availability. In the end, we went with the AMD Ryzen 7 3700X. We will be locking this CPU down in testing over allowing PBO to control it, and we will not be allowing AIOC either. We are going old school.
To go along with that shiny new 3700X, we picked through some of our samples and found the TEAM T-Force Xtreem ARGB kit sitting there, looking amazing, and designed with tight timings and a fair speed that suits this build perfectly. We did not want to go RGB-less, and this RAM will add a nice but contained glow of lighting without being obnoxious. With 3600MHz and 14-15-15-35 1T timings, we could not pass on this match.
Storage is nearly irrelevant, as we just need to house an OS to test thermals and noise levels, and even occasional software options for cases. We did start with an M.2 but ran into lane mitigation issues. We then dug deep into the parts box and found this old SanDisk SDSSDXP-120G 2.5" SSD that is up to the task and makes us use trays and HDD cages when assembling these bits into a chassis.
To fill the motherboard's midsection and test the alignment of the expansion slots and cover plates, we were lucky enough to get our hands-on an NVIDIA GeForce RTX 3070 in Founder Edition clothing. The call here was relevance to the market, visual appeal, and power requirements came into play as well. The RTX 3070 is also only as wide as the motherboard, so we will not run into chassis fitment issues.
We again decided to power the system with an ASUS PSU, a matching ROG Thor 850W that we use in the CPU Cooler Test System, but this time we came out of pocket to obtain it. Not only should it blend into any build in the PSU mounting locations of various cases, but we also have sleeved cables that will dress up images and make wire management systems work harder to contain what we have.
As to the cooler, this can vary at times, as not all cases are created equal. We opted to go with the Castle 240 from, DeepCool for many reasons. They sent it to us for testing, so it is here and ready to be used for one reason. It offers RGB, but again, not too much, it seems standard for what most users will opt for when cooling a CPU by size, and at this point in time, mid-tower and full-tower cases should allow the use of this by default. Should a chassis not support this cooler in some sense, we will still provide thermal and audio results, but to the same setup, not compared to this cooler's results.
We snapped an image of the system all assembled, going through testing and stabilizing our BIOS settings. It is a down and dirty build on the motherboard box, but it will allow us to see how well the setup does thermally, with ambient airflow around the system always kept at a constant. We noticed that the VRM is much warmer when run like this, but otherwise, CPU and GPU temperatures are normal, and what we will be referencing in our comparisons of temperature.
UEFI Settings and Open-Air Testing
If we allowed AMD to control the CPU functions, we would be consistently hitting a temperature wall, as AMD overclocks these CPUs much like a GPU boost feature. Knowing that thermals would be an issue, we had to stop off in the UEFI and adjust some things to get the system to act the way we want it to, delivering a repeatable thermal load that we can compare.
However, when we open the EUFI and slide over to the Extreme Tweaker tab, we do run into the option to employ DOCP, which we absolutely did, to get all of the juice out of those TEAM sticks. No point in wasting it running SPD speed and timings. It is after that where we set the CPU core ratio to 40.00.
Sliding down the same page, just to show what all is being set by DOCP or anything we may have adjusted along the way, we are still under the Extreme Tweaker heading, but you can now see that we are into the voltage settings. With what we are attempting to do, using the 3700X at 4.0 GHz all-core, we found we needed to set the CPU voltage manually. We ended up using 1.15V to this CPU, and that is with a touch of padded voltage to enable this system to run like this indefinitely.
We left the SOC voltage to do its thing, and we will see in a bit that it is at 1.08V in Windows. The last but to note is the DRAM Voltage, but that value is applied when we use DOCP on the TEAM Extreem ARGB.
You will have seen a line to the left in the previous image that says Precision Boost Overdrive. PBO relates to how the speed of the CPU reacts to various forms of load being used. With these options left enabled, our CPU would run to the thermal limit and just sit there, which we cannot use, as all cases would then have the same CPU temperature; HOT! To make the 3700X And the DARK HERO do what we want them to do, we disabled FBO Fmax Enhancer, PBO, but found that those two settings were enough to limit the ramping condition we were seeing, leaving us with what we were after, an old school, all-core overclock.
In windows, And happy with what we accomplished in UEFI settings, we allowed the office to come down to temperature overnight, and the next morning we booted up our system, let it relax for roughly thirty minutes with monitoring open in the background, so we can assess where we are in an open-air test environment. The major points to see in this image are the component settings such as speeds and voltages, CPU Package temperature average of 30.1°C, average GPU Diode temp of 35.2°C, VRM temperatures at 40°C, and in the large AIDA window at the right; we keep track of fan speeds for anything we can, so we can reference those speeds when quoting a decibel level.
Since we left the system to rest for thirty minutes before getting the average values of our components, we figured that if we loaded the system for the same amount of time, things saturate, and hotspots develop in cases by then. It seems fair, to not have us loading the system for hours, to achieve the same results. At this time, recalling back to the same components again, our CPU PKG temp is now averaging 56.9°C under load of AIDA64 System Stability Test. The GPU averages at 69.5°C, the VRMs are 47.7°C now, and you can see the fans have ramped up in speed to compensate. All of this is done with the room kept to 22.5°C, with half a degree (margin of error) either way of wiggle room.
We treat our chassis reviews much more like a CPU cooler review, but we feel that is a good thing. Test the product as it is shipped, develop a solid methodology and eliminate as many variables as possible, use sound meters and software that have proven to be valuable over the years, and go with what we know. As with any way one wishes to test something, there is no way to eliminate all variables, and that quest is tougher in cases.
There is always a curveball around the corner due to styles, layouts, dimensions, feature sets, choice of fans, etc. All we can do is try to be ready for them. You may have also noticed that we made no mention of secondary motherboards. Right now, we are testing only ATX motherboard supportive cases, as we currently do not have the gear to go smaller. If we eventually get to that point, we will test that chassis just like we are showing here, but with other components and open-air values to compare to.
When it comes to a chassis, we feel that much of what you like is visual, beyond anything else. If it is a turd, you will not want it in the office or on the desk in the first place. We will be making comments on body lines, use of materials and coatings, making sure it looks well-built and things match, but beyond that, we will not go too far. Our energy will be spent looking at things like filters, airflow openings, PSU covers, drive storage, wire management, usability; you get the idea. It seems, to us at least, that if we are just straight forward, with no magic sauce clouding anything we report in this segment, we leave the final decision to the potential buyer without any haze clouding the reality of what they are about to have shipped to their door.
With less opinion on matters and a drive for just the facts, we hope that our efforts to put forth a better, more informed and detailed review of cases, leaves our readers in a clear head about what they would buy next. The whole point of our job here is to lay it all out on the table, see what works, what doesn't, dive into listed features, and ensure you are getting what you are lead into buying. Along with thermal and audio results being delivered, we feel like we are doing the best we can for our readers, which is the only reason we are here in the first place. Without readers, what we do would be non-existent, and if we are going to go through the effort to deliver top-notch chassis reviews, we also need to be able to adapt to what readers want.