DAN HSLP-48: A powerful sub 50mm heatsink

Hey guys,

if you have an ASRock Fatal1ty AB350 Gaming-ITX/ac and want to instal an EKL Alpenföhn Black Ridge, you need to modity the cooler. See the images.
But the EKL Alpenföhn Black Ridge is better than the NH-L9a-AM4.
 

Attachments

  • 20181117_164023.jpg
    20181117_164023.jpg
    189 KB · Views: 0
  • 20181117_185057.jpg
    20181117_185057.jpg
    146.9 KB · Views: 0
mirgus If you install the Black Risge with the A12x15 blowing out without I/O shield you can improve temps much more.
 
Seems like more people are starting to receive their coolers, not sure where you guys are located but hopefully I'll get mine soon. I'm really curious to see how it stacks up against the LP53.
I'm counting on you to make the choice for me ;-)
 
Hey guys, registered to share my experience with the BlackRidge. My configuration at first: DanCases A4-SFX, ASRock B350 Gaming itx/ac, AMD Ryzen R7 1800X UC@3,4GHz/UV@1,024V, GTX 960 4G, 16GB Corsair DDR4 3000MHz CL15 LP. The Cooling of the CPU did a Cryorig C7 with a cardboard fanduct.
First of all building quality and ease of installation: building quality so so, the cpu plate has little bobbles at the rims of the surface, not very well polished. The fan is not very good. On its first run I let it run on max rpm and there where serious bearing noises which wouldn't stop after slowing it down again. Really shitty fan noise wise I would say. Installation as was mentioned before was not possible until I used my Dremel on the AM4/3 clamps. The form of them and the support of sockets AM3 and downwards is imo not necessary and creates compatibility issues...who is building sff with am3 or even downwards sockets???
But then I got it to run, here the results:

Cryorig C7+FanDuct; Prime maxHeat 20min; Room 18,9 °C
2054 rpm
64°C
51db
88,39W Power Consumption R7
1.040V VCore

Alpenföhn BlackRidge; Prime maxHeat 20min; Room 17,3 °C
2626 rpm
75°C
50db
91,65W Power Consumption R7
1.040V VCore

Alpenföhn BlackRidge/flipped heatsinkfan; Prime maxHeat 20min; Room 18,5 °C
2402 rpm
63°C
44db
88,23W Power Consumption R7
1.040V VCore

Remarks: the db was measured with a Samsung S7 from 10cm distance with closed vented site panels. Even though the absolute values might not be correct, subjectively the general message of the measurements is correct. Due to the fact that the default position of the BlackRidge Fan counteracts the ventilation philosophy of the A4 the results are really bad. Once the fan is flipped, it's quieter and a little bit cooler than the C7 with a fanduct, so there is a little headroom there. But to be honest I expected a little more of the BlackRidge. I thought defaultclocks! of the 1800X would be possible. Now I'm thinking about switching the fan to the Noctua... switching cooler to C7 Cu or switching Case to Ghost S1 with quiet a better CPU Cooler ;)
 
i thibk an test with the noctua 92x14 will show the real performance of the black ridge and with the 120x15 i think oc will be possible
 
Hey guys, registered to share my experience with the BlackRidge. My configuration at first: DanCases A4-SFX, ASRock B350 Gaming itx/ac, AMD Ryzen R7 1800X UC@3,4GHz/UV@1,024V, GTX 960 4G, 16GB Corsair DDR4 3000MHz CL15 LP. The Cooling of the CPU did a Cryorig C7 with a cardboard fanduct.
First of all building quality and ease of installation: building quality so so, the cpu plate has little bobbles at the rims of the surface, not very well polished. The fan is not very good. On its first run I let it run on max rpm and there where serious bearing noises which wouldn't stop after slowing it down again. Really shitty fan noise wise I would say. Installation as was mentioned before was not possible until I used my Dremel on the AM4/3 clamps. The form of them and the support of sockets AM3 and downwards is imo not necessary and creates compatibility issues...who is building sff with am3 or even downwards sockets???
But then I got it to run, here the results:

Cryorig C7+FanDuct; Prime maxHeat 20min; Room 18,9 °C
2054 rpm
64°C
51db
88,39W Power Consumption R7
1.040V VCore

Alpenföhn BlackRidge; Prime maxHeat 20min; Room 17,3 °C
2626 rpm
75°C
50db
91,65W Power Consumption R7
1.040V VCore

Alpenföhn BlackRidge/flipped heatsinkfan; Prime maxHeat 20min; Room 18,5 °C
2402 rpm
63°C
44db
88,23W Power Consumption R7
1.040V VCore

Remarks: the db was measured with a Samsung S7 from 10cm distance with closed vented site panels. Even though the absolute values might not be correct, subjectively the general message of the measurements is correct. Due to the fact that the default position of the BlackRidge Fan counteracts the ventilation philosophy of the A4 the results are really bad. Once the fan is flipped, it's quieter and a little bit cooler than the C7 with a fanduct, so there is a little headroom there. But to be honest I expected a little more of the BlackRidge. I thought defaultclocks! of the 1800X would be possible. Now I'm thinking about switching the fan to the Noctua... switching cooler to C7 Cu or switching Case to Ghost S1 with quiet a better CPU Cooler ;)
Great review. Ghost s1 is a great alternative for a less restrictive baller build. I had to go copper GPU heatsink 1080ti and copper CPU heatsink on my 4.5ghz 6700k to keep everything under control in my dan a⁴. Pretty hard to find parts, I got lucky.
 
Last edited:
i thibk an test with the noctua 92x14 will show the real performance of the black ridge and with the 120x15 i think oc will be possible
As shown in my results, the 92mm fans pretty much perform the same and a 120 fan is very noisy above 1100 u/min.

PantherB your results are pretty similiar to mine, it seems that the cooler isnt just perfomaning that much better CPU wise, but it helps pretty good with the temperatures in the case

u1nYpmI.png


Great review. Ghost s1 is a great alternative for a less restrictive baller build. I had to go copper GPU heatsink 1080ti and copper CPU heatsink on my 4.5ghz 6700k to keep everything under control in my dan a⁴. Pretty hard to find parts, I got lucky.

Copper isnt really that much better dissapating and the benefits are pretty much not measureable
 
Last edited:
73c with a12 fan set to intake and low rpm seems pretty incredible to me in that last chart....
 
On full rpm the A12x15 under the fin array is 10°C better as the L9i and 8°C better as the A12x15 with A9x14 @ full speed. I made this test on Ryzen. With the A12x15 attached on top of the fin array it is 12°C better as the L9i.
 
On full rpm the A12x15 under the fin array is 10°C better as the L9i and 8°C better as the A12x15 with A9x14 @ full speed. I made this test on Ryzen. With the A12x15 attached on top of the fin array it is 12°C better as the L9i.
Slightly confused, with the black ridge and a9x14 the temp is 8° better than the l9i? I assume this is in the case, with no fan ducton the l9i? Also is this with the Io shield in place?
 
My black ridge is out for delivery. Currently have family visiting but maybe later this weekend I can test it out versus the LP53 with nf-a9x14. I don't have any low profile 120mm fans. I also have a noctua NH-L12 (the old 66mm version) if anyone is interested in seeing it compared to a different small but substantial cooler? I realize most people are interested in this cooler for the Dan A4 so that noctua cooler might not be relevant
 
Fuck. Could you document your mods please?
Sure thing. As far as I can tell its only a coulpe of millimeters on the IO side of the fins that needs to be removed in order to make it fit. I actually thought it was in place until I placed the motherboard on a flat surface and spotted the offset - it's really that close.
 
I just tested compatibility with my AM4 boards inside Sentry. As it was said before, the combo bracket for AMD(AM4/AM3/AM2?) has collision with some vrm components on my asrock board while it doesn't happen on biostar unit. I've checked it online for other boards. Asrock/MSI are no go because of this, Asus/Gigabyte look fine.

Apart from that it is really bad that it goes over the PCI-E slot with the heat pipe endings and the last fin, because it means you can only use it in cases with flexible risers. I have no idea if it fits in anything with standard layout like M1 or cube cases.

I'll try to test this out in my case for comparison with other coolers I've tested so far (on R7 1700/R5 2600X/R7 2700X), but I'm not so keen on using for my daily driver PC unless it's really whole lot better, because it obstructs 2.5" drive slots ;/

Sentry.jpg

I'll probably mod the brackets on one side by mounting them in vise and bending the ends upwards. This way it won't be a combo bracket anymore.
 
Yeah, kind of what I expected. With a 120mm fan, the cooler is a game changer. With a 92mm fan, it doesn't perform any better than the other options we have, because you basically have a 92mm cooler + a tiny bit of extra passive cooling which really isn't enough to do anything in a case like the A4-SFX. So something like the C7 Cu + A9x14 would do noticeably better than the Black Ridge 92mm. The 120mm option is great if you don't want more than 16GB of slow DDR4 at a price premium, not sure that's really an acceptable compromise for many. I hope in the future we'll have more VLP options.
 
The cooler in Sentry 2.0 looks like this:

a0thvwcl.jpg


5BG2an3l.jpg


Note the way drive is laying freely between the rail and the PSU, no other way to do this ;/

OK, so I'm past the first CPU tests, the 2700X. The results are actually really interesting.

I have mounted the fan to pull the air from the outside through the fin stack because default orientation to push out air through it performed not that well inside Sentry, it was really similar if not worse than NH-L9i. I was not amused by the cooler at that point. It was most likely because it was pulling hot air from the GPU this way.

Noteworthy is that I'm comparing the temps to what I tested over few months now and it got cold recently so the ambient temp in the room is ~2 degrees lower than when I started testing few months ago, although luckily tests were done on overcast days so far.

I'm also using different board and different memory this time because of the cooler limitations. It's biostar X370GTN instead of asrock B350 itx ac and corsair LPX 3000 instead of g.skill ripjaws 3200, everything is set the same way though, the only difference may be the power section on the biostar board. I'll have to investigate that later though.

I'm testing it the same as all other tests that are already on our website in the gallery. 10 games/benchmarks ran with same settings on 1080p/max details/vsync-off in both horizontal and vertical orientation. Test are run in sequence without much time for the system to cool down. All coolers were run on aggressive fan curve for the fan to react quicker in those short tests, except for wraith stealth modded with titan fan because of the auto settings being really weird on this fan.

The footage for 2700X is already captured. I've noted the stable/sustained temps at the end of each test (so after heat up) while there's still the same load. I'm discarding spikes and max temps when the fan curve doesn't catch up to the momentary boost because it's not stable and repeatable. I'm not averaging temps for each test on their own either because that's not telling us anything really, but I'm averaging those noted sustained heat up temps from all those 10 games/benchmarks to have single numer.

Ryzen 7 2700X, temperature deltas for Black Ridge in vertical orientation:
vs Corsair H75: +12 °C
vs Noctua NH-L9i: -5 °C
vs Cryorig C7: +1 °C
vs Wraith Stealth w/ Titan Fan: -10 °C

in horizontal orientation:
vs Corsair H75: +5 °C
vs Noctua NH-L9i: -7 °C
vs Cryorig C7: 0 °C
vs Wraith Stealth w/ Titan Fan: -10 °C

Ryzen 5 2600X, temperature deltas for Black Ridge in vertical orientation:
vs Corsair H75: +10 °C
vs Noctua NH-L9i: -6 °C
vs Cryorig C7: +1 °C
vs Wraith Stealth w/ Titan Fan: -11 °C

in horizontal orientation:
vs Corsair H75: +6 °C
vs Noctua NH-L9i: -4 °C
vs Cryorig C7: 0 °C
vs Wraith Stealth w/ Titan Fan: -6 °C

In most games it sustains low 60's except for 3Dmark and Ashes where there is significantly higher CPU load.

It is also virtually silent/inaudible in idle/at low RPS while running at high 30s/low 40s, so quiet curve might be nice setting for it if you're okay with higher temps.

The size of the fin stack, especially where it takes space makes the build really hard. There's a really not much space to connect motherboard power/power switch/usb/sata. It also dangerously squeezes riser's ribbon just above the slot. Not that fun.

Considering the price point which is the same as NH-L9i I should recommend it, but fitting in just specific type of case/configuration (especially not fitting well with Sentry) and having issues with asrock boards, it leaves me with mixed feelings really. I'd really love it if the fin stack was centred above the socket, at least in the heat pipe direction, being above the ram doesn't bother me.

I'll be back (after capturing benchmarks for other CPUs) :)
 

Attachments

  • a0thvwcl.jpg
    a0thvwcl.jpg
    51.1 KB · Views: 1,393
  • 5BG2an3l.jpg
    5BG2an3l.jpg
    53.5 KB · Views: 1,243
Last edited:
According to the above test conditions, temperature deltas for Black Ridge in vertical orientation:
vs Cryorig C7: +1 °C

in horizontal orientation:
vs Cryorig C7: 0 °C

Makes it difficult to justify switching away from the C7 then. :cry:
Still waiting for my Black Ridge to be sent out from Overclockers UK.
 
Last edited:
Without a 120mm fan the Black Ridge can't be much better than the C7 or L9i because with a 92mm fan it nearly use the same fin size. Only a 120mm will unlock the full performance.

In my tests it actually shows that i's slightly better than L9i, but that's assuming the 1080p gaming scenario. There's obviously a physical limit how much air can 92 mm fan push, but you shouldn't expect an ultra sff pc to have 100W TDP CPU running all the time prime95-like loads.

On another topic - do you have a list of cases that should handle this cooler properly? Or is it strictly designed around A4-SFX ?
 
SaperPL: Keep in mind like we have a silicon lottery it is the same for heatsinks (there is a variation for heatsinks and soldering quality). So against your L9i the BlackRidge is much better but with a perfect L9i sample the difference is 2-3°C. This is why EKL used 6 heatpipes to compensate heatsink variation)

Compatibility question (I hope I can explain it without a drawing):

Classical Hardware Layout:
For standard ITX cases where the PCIe port will be facing to the bottom or the board is mounted horizontal so you have to mount the heatsink in that that the heatpipe bend will face to the IO (only possible on 115x) or PCIe port (possible on 115x and AM4). Then you can still use the PCIe port

A4 Layout:
For A4 Layout cases or Steam Machine like cases that has limited space around the motherboard you have to mount it in that way, that the heatpipe bend is facing to the bottom of the case. The fin array will be on AM4 boards above the PCIe port (because the AM4 socket is centered. These cases are using flexible riser so it isn't a problem.

richiegore: I never tested it.
 
Classical Hardware Layout:
For standard ITX cases where the PCIe port will be facing to the bottom or the board is mounted horizontal so you have to mount the heatsink in that that the heatpipe bend will face to the IO (only possible on 115x) or PCIe port (possible on 115x and AM4). Then you can still use the PCIe port

But that's only true if you have a recess above the IO backplate like this: https://images.anandtech.com/reviews/cases/roundups/super_towers2006/cm-z1.jpg OR you have holes matching exactly heat pipe layout to let their endings through the wall. In my unit the last fin in the stack is right above the edge of the board when mounted this way on LGA 1150 board.

A4 Layout:
For A4 Layout cases or Steam Machine like cases that has limited space around the motherboard you have to mount it in that way, that the heatpipe bend is facing to the bottom of the case. The fin array will be on AM4 boards above the PCIe port (because the AM4 socket is centered. These cases are using flexible riser so it isn't a problem.

Both in AM4 and LGA1150 boards the mounting point is roughly the same, at least for my boards - in both platforms I have the fins over the slot. If I were to mount it in opposite direction so the fin stack goes over the VRM side of the board, I hardly have ideas which cases would fit this.
 
Reading about the true potential only unleashed with a 120mm fan makes me wish for vlp ram. What's the cheapest way to get them?
 
I've made some rough mock-ups to see how many boards can handle this properly in something other than A4-SFX and it doesn't look good:

KJ3p6vUl.jpg


More board mockups here: https://imgur.com/a/Mt1eLzn

Note the fact that the cooler reference image is a bottom rendered shot that seems to be in perspective so the fin stack is further away meaning it's slightly smaller than it actually is (unless I greatly messed up matching the cold plate size to AM4 socket size), but it still shows some issues.

There is only one board vendor (gigabyte aorus) that has AM4 board that does not have its pci-e slot obstructed when fin stack is facing it.

On all of the intel boards black ridge seem to extend over the motherboard when the fin stacks are facing IO plate.

On all of the intel boards black ridge extends at least 2 cm over the motherboard when the fin stacks are facing away from PCI-E slot.

It seems like this cooler will work either when you don't need to use the pci-e slot, when you are using A4 or similar cases where the riser bends right around the pci-e slot, or when you are using a case that has a recess over the IO plate so mostly mass produced cases that don't have the PCI bracket going outside of the case.

This needs a thorough compatibility list.
 
Last edited:
Like SaperPL I'm having a difficult time figuring out how it will mount in my sentry base system. My z97 board has the cpu far from the pcie slot and even then with the end of the heatpipes towards the pcie slot its very close to not working. With the end of the pipes towards i/o its possible but only because my old z97 board doesn't have a huge vrm heatsink or i/o shroud, I haven't checked the clearance on the ram with the bend side of heatpipes though. I have a busy week with work so we'll see if I can get to it.
 
Hi everyone,

It looks like the new Alpenfön Black Ridge has some manufacturing problems. Some heatpipes are not soldered correctly with the base. Maybe you know I can’t give support by myself because it is not my product. My name is only printed on the carton because I helped EKL Alpenfön in the development process. However I forward the manufacturing problems to EKL and they ask me if I can post the following statement (I translate it from German to English):


“Dear customers, the first review samples were not affected by the problem, so we start producing the first batch with a rush order. This first batch is affected by the problem and is a result from quick manufacturing. We are very sorry about it. All already ready new batches and future batches will not have this problem. We informed all resellers. The resellers will contact you for a replacement process. We guarantee an easy replacement of the product. We feel very sorry that this happened due to quick manufacturing and we ask your forgiveness. - EKL”


Furthermore they confirmed that the current version with bad soldering will not perform as good as the flawless version. So everyone will get a new flawless version.

With best regards
Daniel Hansen
 
Back
Top