Noctua Has Competition: HAVN Performance Fans, BF360 Case, & Engineering Data
During our visit, the company showed off its new BF360 case, which follows up their HS 420.
Taking a closer look at the company’s new fan design, you can see some careful shaping to the fan’s hub, which is supposed to help with guiding air flow. We’ll have to see how it performs in our testing, but the idea is taking a thicker fan approach and thickening it even more. The company ended up using a 40mm-thick fan, which should help with pressure.
The company also uses 2x180mm fans mounted to the front of the case.
While we were there, HAVN prepared some CFD simulations for the flow and some simulations for mechanical stresses, on the panel for example, which is pretty interesting.
Taking a look at the BF360, the front panel pops out with a tug as it uses magnets to attach itself to the front of the case and has 3 plastic feet that help it snap in at the bottom.
The interior of the case’s front panel has a removable fan tray. Its design is somewhat typical but refined from what we’ve seen. HAVN has done a lot of iterations on it.
The top panel pulls off and its design is familiar to the HS 420 with its structural design. Once you remove the top panel, it exposes the case’s top fan/radiatory tray, which has rubber bumpers and a slide mechanism that lets you pull it out of the case, providing full access to the top of the chassis and should help with ease of installation.
A lot of the steel on the case is either .8mm thick and there’s some that’s 1.0mm, which is thick by today’s standards.
Taking a look at the backside of the case, we can see that it has marked cable management pathways, much like the HAVN HS 420 (read our review). This was kind of a nice thing for brand new system builders. Whereas the HS 420 used stickers, the BF360 incorporates it into the molding. Our understanding is that they may refine its design to try and get more light reflection so it’s easier to see. This is a nice attention to detail that adds to the ease of installation.
HAVN learned from the HS 420 on the 3 and ½-inch drive support and improved its implementation. There are 2 bays on the back side of the case for those drives and they can be removed. HAVN also changed how it handles the grommets with the case. The string-like material in between is the same but there’s now a little rubber tab that helps to prevent it from popping out by accident once it’s filled with cables. The point here is that HAVN is trying to learn from what they’ve done with the HS 420 and to refine their designs.
Taking a look at the BF360’s front panel, it’s supposed to look like stone, but is made of plastic as stone would be extremely heavy. Behind the front panel in the unit we saw are 2x180mm fans. HAVN has shoved a piece of steel at the top, which is supposed to prevent recirculation. This was a huge issue in old BitFenix cases, where due to the company’s design, it would recirculate hot air back into its fans. This design can improve performance in big ways.
HAVN tells us that the best setup for the BF360 is to have the top front fan be intake and the top back fan be exhaust, providing you’re using 180mm fans on the top. This matches our testing for other cases in the past. We’ve also found that if you’re populated the top slots, doing intake in front of an air cooler works better because otherwise air gets in the front and would get stolen by an exhaust fan out of the top.
The case we looked at had 4x180mm fans (2 in the front and 2 on top) and 1x140mm in the rear.
HAVN tried different types of power supply shroud designs. One had an angled scoop to bring in air from the front bottom 180mm fan.
One design had what we’re calling a “toilet-bowl” design, but that ultimately didn’t go through. The next design we looked at featured a more cylindrical/conical design. We asked one of the company’s thermal engineers why that design didn’t work better as it looks like it would project air towards the GPU, which would be a hot spot, and the rep told us that going with a wider design ended up performing thermally better. HAVN provided us some numbers for that, which we’ll take a look at below, but we’ll eventually do our own testing if we can get an early sample.
Taking a look at the first mock-up of the case, we can see some rails for the front fans. This design did not progress into the final. It uses a rail system and we’ve seen it in other cases. Corsair has implemented it and has a trademark for it which they call the “InfiniRail.”
Looking at the second mock-up, the company moved to vertical rails coupled with horizontal slats.
HAVN ended up ditching that design altogether and moved to a tray design, which is sort of moving to a tried-and-true approach.
The company incorporated another change where they’ve stamped and folded the tooling.
Then we took a look at a design that was closer to final, which has its full corner covered.
HAVN showed us 3 different prototypes for the front panel of the case. The first one we looked at featured a wood panel design, which we’ve seen on a lot of cases lately. Wood has certain manufacturing challenges to consider. The white panel you see above is made of aluminum, which is very expensive.
The last thing that HAVN is working on that they showed off are fans. There’s been a lot of fan development in the last few years. One of the big marketing phrases you hear a lot these days is LCP (liquid crystal polymer), which is something we talked a lot about at the last Computex. One of the benefits of LCP fans (which HAVN isn’t using here) is that you can get the blades way closer to the interior of the frame. The downside is that it’s incredibly expensive. We have an interview with Noctua’s Jakob Dellinger from 2 years ago that delves into how the company wanted to avoid using LCP for a long time but determined it was necessary to get the performance they wanted. HAVN isn’t using LCP but has designed a fan that has some “teeth” cut out on its blades, which is something we’ve seen before. The company has also shaped its fan hub to be able to scoop air in more. HAVN tells us that having the hub protrude out like a little mountain would offer the best performance but this presents clearance/compatibility issues, especially if you want to put a radiator up against it.
Taking a look at the 180mm fan, there’s a 2.0mm distance from the blade to the frame of the fan. We asked if HAVN tried to get it closer and were provided with a lot of data. The company showed us 3 revisions. 1 had a 1.8mm spacing, another had 2.0mm of spacing, and the third had 2.2mm. What they ended up seeing for pressure in mmH2O, is that a 2.0mm spacing ran at 2.19. The 1.8mm fan ran at 2.23, which is an improvement. The CFM (cubic feet per minute) for both was nearly indistinguishable at around 165. The 2.2mm spacing, however, had a big drop off, which is a design they didn’t go with. It had a 2.21 mmH2O pressure, which is more or less within error of the others, but the 159.9 is where that drop off occurs in the CFM. Comparing the 1.8 vs the 2.0mm spacings, they are basically hitting diminishing returns. Seeing this information is nice. We imagine that the benefits here weren’t worth the yields and costs because getting the blades closer to the inner frame could lead to a bad yield during manufacturing or might have the blades expand and hit the frame as it ages with time. This is what happened with a lot of Enermax fans in the past.
Now we’re going to go over HAVN’s first-part numbers, but we plan to do our own performance review of the case and fans once they’re available. First-party data should be looked at with some reasonable care.
Thermal comparisons should only be made against their own results here, not against ours, as the benches are different. The flat, typical shroud style was comparable to the rectangular type in result, with the rectangular type slightly better. We think this will see a larger impact in our testing, but we’d need the three types to know for sure. The important comparison is against the cylindrical type, where HAVN saw worse results by a measurable amount. Performance worsened by 1-2 degrees on the CPU and about 2 degrees on the GPU. That’s a large GPU temperature increase.
In their next test slide, HAVN tested various fan configurations with the flat shroud. The results were mostly as expected: CPU performance is hurt drastically by having only a bottom intake fan, to no surprise.
More interestingly, the ramp type shroud makes comparisons in “B” with a fan level with the top of it and “A” with a fan sunken to the floor of the ramp. The sunken approach resulted in marginally worse CPU thermals in A as compared to B, with GPU thermals mostly unchanged.
This slide shows CFD simulation and flow mapping for ramp angles at 120 and 135 degrees. HAVN has drawn a few highlights around areas of re-circulation or heat accumulation, particularly marked at “1” where the 120-degree ramp angle shows worse areas of heat build-up. HAVN tested this in 5-degree increments from 120 to 140, finding that 135 was a good balance when noise normalized.
In the above image, they show the fan spacing again, with a 20mm height increase benefiting performance.
HAVN didn’t shy away from competitive comparisons. The company says its BF 360 with its final front panel, shown above in light blue, had competitive performance against the H6. They also suggest significant reductions in flow resistance (which is a good thing) versus the Lancool 3, with the HS 420 also making some appearances here.
HAVN began studying its slot spacing and porosity of the ventilation slots in the panels next. The company found that a ratio of length divided by depth being ideally about equal to 6 was optimal for minimizing flow resistance, with the only downside being potential structural challenges. These charts show their flow performance with different slot sizes.
As shown in this set of images, the point is that HAVN is really mocking up a ton of different styles of panels, from the shrouds to the slot spacing, to try and determine the thermal performance and optimize for it. This is a major progression for them. It’s a lot of work to do all of this, which is cool to see. Of course, we have to test the product to see how it came together, but the CFM performance of type 4, as they called it, was the best, but we’re ultimately looking at tiny differences here. Still, all of those small differences across the case will add up in theory.
This image was cool: After all of this thermal testing, HAVN next did mechanical stress simulation for torsional forces against the panel, which allowed them to dial-in the thickness and makeup of the plastics.
That continued in this image, where the so-called “type 4” gets another highlight.
HAVN highlighted “Design C” for having a 0.8mm metal thickness with the chosen hole sizing.
By shaping the fan hub itself to guide flow, HAVN claims that it can better maintain pressure across the hub of the fan and along the inner wall of the blades. This is the area of worst performance in every fan, so optimizing here can also help minimize dead zones behind the hub. HAVN says that the molded flow guide in the hub benefited its noise significantly. The company claims its flow rate also improved significantly for the guided hub.
The image above is really cool. Referencing whale-fin evolution, HAVN used tooth-like leading edges on the fan to improve its thermal performance. We’ve seen this plenty of times in the past, but it never seems to stick around and we’re not sure why. In this simulation though, HAVN suggests that the air stream is more uniform along the wavy blade design.
Using a 30% glass fiber composite, HAVN saw performance with the new design slightly improve in CFM but largely improve in static pressure, going from 2.21 to 2.40 mmH2O.
There’s a ton more that HAVN did. The P/Q chart above shows the mock-up performance, including the stall region centrally. We’ll save all this discussion for the review, though.
HAVN had about 41 pages of this technical presentation, which is actually greatly appreciated. A lot of what remains will be covered in our review, so we’ll leave that for now.
The bearings are also interesting to talk about. These are called FDBs (Fluid Dynamic Bearings), depending on how you want to define that. We looked at one that was designed for the company’s 180mm fan. It was a 15mm-tall FDB. One of the things with fans, in terms of the support, is how the fan’s rod seats into the bearing. As we’re told, the longer the bearing is, the more it will help with the wobble of the blades.
Internally looking at the 15mm-tall bearing for the 180mm fan, you can see 3 grooved channels. In our bearings-factory tour, we learned that the grooves are part of the mechanism that allows the fluid to circulate to keep that pressure even across the bearing. That’s supposed to be what’s special here but we don’t test bearings individually, but the idea is that 3 channels are supposed to help with the stability of the fluids across the bearing.
For the 120mm fans, HAVN is going for roughly a 12mm tall bearing. The company claims that most are between 9-11mm tall for 120mm fans.
HAVN allowed us to look at the company’s thermal chamber. There are pros and cons to HAVN’s solution. The pro is that it’s supposed to help control the thermal environment as it circulates the air. The downside to the chamber is that we found it can influence the results, though it depends on the chamber. We found that for our testing specifically, having a larger, open-room environment works better. It is very environment-specific.
The company also had a flow-rate tester, which is similar to ours, that can do PQ charts, which the company used to test its new fans. It has a throttle control and a laser tachometer, which provides a reading of the speed of the fan. On one side of the flow-rate tester, there are tubes for the counter blower, which our system also has. Their flow-rate tester also has an air compressor, which is used for actuating some of the nozzles inside the system. We also saw that their flow-rate tester also came with an inexpensive desktop OEM computer (in this case, a ThinkCentre PC running an i5-12400).
HAVN also has an acoustic chamber, which has a pass-through on the outside coupled with a noise meter, which collects noise levels. One thing we liked seeing once we looked inside the chamber is that we saw a mechanical rigging for the fan, which allows it to be free flowing. They put their microphone in the corner, which we used to do when we had a chamber of a similar size. Putting the mic here allows you to maximize the distance of the chamber by going diagonal, providing maybe a half a meter of space. They also have a foam floor as well, which makes it closer to an anechoic chamber as opposed to something like our hemi-anechoic chamber.
HAVN is targeting a launch around September for its new products at which point we aim to run our own benchmarks to see how it all performs.
