Osprey :: Inverted Build in a Cooler Master Q500L

[mikeyw64]

Osprey is a dual boot (both W10) for both personal and WFH usage.

Specs are:
Asus Prime B550MA MATX mobo
Ryzen5 5600X
64GB Crucial Ballistix DDR4 3000 (4 x 16Gb)
Boot drive is a Crucial P2 M.2 500Gb
Secondary drive is a 2Tb 7200rpm Seagate Barracuda
Corsair Hydro H100X AIO
ESDIY-FAB Moonlight RGB Fans
Antec Sleeved cable set
Corsair SF600 8-+Gold fully modular PSU
EVGA GTX950SC 2GB (beggars can't be choosers !)
The monitors are a pair of 22" AOC G2260VWQ6 which I have owned for a few years plus an Elecrow 8" monitor


Elephant in the room, the choice of case.

I was aware before starting that the Q500L (along with it's sibling the Q300L) has a reputation for being a hotbox however I pushed ahead anyways as I liked the flexibility it offered especially when it cam to doing an inverted build. This was a requirement for me due to where she would sit once completed.

That said I can't complain about the temps I'm actually seeing with my config which is ::

Radiator in the roof with the supplied 2 Corsair Fans pulling air through it complemented by 2 EZYDIY fans pushing
Front Fan Intaking
Base Fans intaking (and blowing air directly over the VRMs, CPU area & RAM, possibly an advantage of the inverted build)
Rear Fan exhausting

With everything set at Stock & no OC's the CPU sits at between 50-52c above ambient (20.5 c at time of testing)  under full load (intel burn test + furmark for 25 minutes)  although the GPU does run a little warmer at 60 -62c above ambient . That's something I have a cunning plan for that doesn't involve it's 80mm fan screaming like a banshee under full load although that does admittedly drop 10-15c off the load temps



Although only built just over a month ago she has gone through a number of minor revisions::

Revision 1 the PSU was mounted front & bottom intaking from the case and exhausting aback into the case
Revision 2 the PSU was moved to front upper intaking through the front and exhausting through the top
Revision 3 Case lifted approx 60mm from the desk using a pair of case handles intended for the Q300P

 

The Desk View

 

 

Revision 1 Front & Rear




Revision 2 Front & Rear

Revision 3 Front


Sample Temps & Fan speeds during IBT & Furmark load test (ambient 20c)

[jaslion]

Your layout goes around one of the biggest points of carp design in the case and that is the psu in front.

 

If you ever need more airflow do what others have done and just circle drill in the front. The magnetic mesh filters cover it anyways so you don't really see the giant round hole in the case then.

[mikeyw64]

2 minutes ago, jaslion said:

Your layout goes around one of the biggest points of carp design in the case and that is the psu in front.

 

If you ever need more airflow do what others have done and just circle drill in the front. The magnetic mesh filters cover it anyways so you don't really see the giant round hole in the case then.

I hear what you're saying but with full load temps at 50-52c above ambient I personally see no value in cutting holes at the moment.

NB if I apply an OC to the CPU those temps go up to 58-63c above ambient. If I was running like that all the time I might consider judicious use of a dremel

 

[mikeyw64]

I've come to the conclusion that (at stock settings at least) 3 of my 8 fans are pretty much redundant.

Ambient temp was approx 21c

CPU (5600X) and GPU (GTX950) were both on stock settings

All testing was run using Furmark stress test running at 1920x1018 and 15 Standard iterations of Intel Burn Test

Test 1 was a warmup/presoak run
Test 2 was with all fans on default Asus & GTX curves
Test 3 was with all fans on my Daily Driver Curves (Base & rear fans cut out <-50c)
Test 4 was with the base & rear fans disabled

Radiator in the roof with the supplied 2 Corsair Fans pulling air through it complemented by 2 EZYDIY fans pushing
Front Fan Intaking
Base Fans intaking (and blowing air directly over the VRMs, CPU area & RAM)
Rear Fan exhausting

And the outcome is that pretty much across the board on the CPU I'm seeing ~50-52c above ambient with very little difference between the configs (although noted temps were starting to creep up on Test 4 so may have risen further on a longer run)

GPU is running closer to 60c above ambient on default lowering by about 8-10c with my daily driver curve (although GPU fan noise does start to become intrusive)

Motherboard temps only varied by ~1c

 

 

[mikeyw64]

Corsair H55 Radiator and pump test fitted to the rear position as part of my mad lad experiment to water cool my trusty old GTX950 .

Not too get it cooler(although that would be a bonus) but rather so I dont have to listen to the GPU fan under load lol

Going to need to come back out to fill it and connect it all up with new hoses etc


 

[thrasher_565]

6 hours ago, mikeyw64 said:

Corsair H55 Radiator and pump test fitted to the rear position as part of my mad lad experiment to water cool my trusty old GTX950 .

Not too get it cooler(although that would be a bonus) but rather so I dont have to listen to the GPU fan under load lol

Going to need to come back out to fill it and connect it all up with new hoses etc


 

this is a cringe build... gpu is getting no air flow as the ran fans are sucking it up. and not only that this case it REALY bad for air flow with out mods....

 

 

Edited May 18, 2021 by thrasher_565

[mikeyw64]

2 hours ago, thrasher_565 said:

this is a cringe build... gpu is getting no air flow as the ran fans are sucking it up. and not only that this case it REALY bad for air flow with out mods....

 

 

 

 

I can't disagree as you're correct. The GPU fan is likely  getting sub optimal cooling especially if I was pushing it all the time, but I'm not.

That said I have been considering moving the rad from the top to the front and relocating the PSU (fan side up) above the PSU and I may yet do that at some point. Whether that's before or after the mad lad experiment of wc the GTX I don't know

If I do push it using my non overclocked daily driver config  then  the temps (as based on a 1 hour test running Prime95 on Small FFTs plus Furmark running) the GPU averages 76.46c in an ambient of 22c and the fan running at an average 2356rpm ( GPU fan gets noisy at about 50% upwards ,max speed is around 3800rpm).

Normal day to day usage CPU hovers around the 40 mark most of the time with the odd spike to mid to high 50s whilst the GPU stays around mid 50's to low 60's

The CPU for the same period averaged 75.21c withe case fans all running at a shade over 1000rpm from a 1700rpm max

For reference with the GPU overclocked to the point I can get top of the leaderboard scores for my combination of GPU & CPU in both Fire Strike and Time Spy  then (admittedly with the GPU fan at 100%) the GPU temps averaged 64c on the Firestrike run and somehow I hit an average of 50c on the Timespy run. Not quite sure how I did that as both runs were with all fans at 100%!

[mikeyw64]

Typical low usage/idle snapshot  figures (ambient 19.8)
GPU averaging 53.4c
CPU averaging 39.57c



 

[thrasher_565]

18 minutes ago, mikeyw64 said:

I can't disagree as you're correct. The GPU fan is likely  getting sub optimal cooling especially if I was pushing it all the time, but I'm not.

That said I have been considering moving the rad from the top to the front and relocating the PSU (fan side up) above the PSU and I may yet do that at some point. Whether that's before or after the mad lad experiment of wc the GTX I don't know

If I do push it using my non overclocked daily driver config  then  the temps (as based on a 1 hour test running Prime95 on Small FFTs plus Furmark running) the GPU averages 76.46c in an ambient of 22c and the fan running at an average 2356rpm ( GPU fan gets noisy at about 50% upwards ,max speed is around 3800rpm).

Normal day to day usage CPU hovers around the 40 mark most of the time with the odd spike to mid to high 50s whilst the GPU stays around mid 50's to low 60's

The CPU for the same period averaged 75.21c withe case fans all running at a shade over 1000rpm from a 1700rpm max

For reference with the GPU overclocked to the point I can get top of the leaderboard scores for my combination of GPU & CPU in both Fire Strike and Time Spy  then (admittedly with the GPU fan at 100%) the GPU temps averaged 64c on the Firestrike run and somehow I hit an average of 50c on the Timespy run. Not quite sure how I did that as both runs were with all fans at 100%!

i dont get it thow why wouldn't you want good temps all around.

 

i guess your just min maxing w/e you have now? with a bit of tweaking and airflow your blow these numbers off the charts making all this testing pointless was my point but its your build you have to be happy with it im not.

[mikeyw64]

5 minutes ago, thrasher_565 said:

i dont get it thow why wouldn't you want good temps all around.

 

i guess your just min maxing w/e you have now? with a bit of tweaking and airflow your blow these numbers off the charts making all this testing pointless was my point but its your build you have to be happy with it im not.

No testing is pointless otherwise you wouldn't know what improvements/changes made what difference

She's doing what I need of her on a daily basis  without going up in smoke and the tweaking is just playtime

[mikeyw64]

7 hours ago, thrasher_565 said:

this is a cringe build... gpu is getting no air flow as the ran fans are sucking it up. and not only that this case it REALY bad for air flow with out mods....

 

so had a rejig of the case this morning

PSU has been re orientated
240 Rad has been moved to front (4x fans in push pull still exhausting as before) 
Top fan (which was at front) is intaking as it was previously (tied to CPU header same as radiator fans) and now now blowing towards the GPU, which has dropped the idle temps by Rear fan is still exhausting
Base fans are still intaking

And its made no real difference to the temps, if anything the CPU is now running about 1c hotter on an extended load test and the GPU is about the same. Ho Hum

Now the next logical step would be to switch the rad fans to intake rather than exhaust to see if that affects the CPU temps but that might be later (and probably wouldn't affect GPU temps).







 

[mikeyw64]

so I wandered away and set the Rad fans to also intake giving me a positive pressure overall.

This has dropped the average CPU temps slightly  during an extended Prime95/Furmark load test however the GPU temps have gone up by a couple of degrees.

so my overall conclusion is that , and this is for my specific use case, in day to day usage the actual config doesnt matter greatly.

Under a test load Rad on Top exhausting (apparently stealing the air from the GPU) has a lower delta between CPU & GPU temps  than Rad at Front intaking which has the largest delta between CPU & GPU temps with Rad at front exhausting sits somewhere in the middle


5 minute snapshot of Rad at Front intaking ambient 22c


 

[thrasher_565]

try taking off the side panel then test. then your know that the fans are starved for air.

Edited May 18, 2021 by thrasher_565

[mikeyw64]

ah maybe that's what I did when I got down to that average of 50c for Time Spy ;o)

[TVwazhere]

17 hours ago, thrasher_565 said:

this is a cringe build...

14 hours ago, thrasher_565 said:

i dont get it thow why wouldn't you want good temps all around.

 

i guess your just min maxing w/e you have now? with a bit of tweaking and airflow your blow these numbers off the charts making all this testing pointless was my point but its your build you have to be happy with it im not.

You're being pretty negative dude. You may be trying to help but comments like these aren't helpful and really dissuade people from wanting to take any advice from you. It doesn't matter if you're not happy with it, because its not your build. 

 

@mikeyw64 I know some people have suggested drilling out the front panel. Looks like you're still experimenting with the best layout so it might not be the best idea yet but once you find it, a couple of build logs (admittedly including my own) can be some good references if you're interested in something like that.

 

 

[thrasher_565]

4 minutes ago, TVwazhere said:

You're being pretty negative dude. You may be trying to help but comments like these aren't helpful and really dissuade people from wanting to take any advice from you. It doesn't matter if you're not happy with it, because its not your build. 

 

@mikeyw64 I know some people have suggested drilling out the front panel. Looks like you're still experimenting with the best layout so it might not be the best idea yet but once you find it, a couple of build logs (admittedly including my own) can be some good references if you're interested in something like that.

 

 

ya maybe your right i was just trying to help i guess

[TVwazhere]

2 minutes ago, thrasher_565 said:

ya maybe your right i was just trying to help i guess

You're without a doubt a very helpful member of the community. I've seen it a lot in the last couple years with good and thoughtful suggestions in the past, so I wanted to give you feedback because I know what you're capable of. 

 

God I sound like my high school teacher

[mikeyw64]

1 hour ago, TVwazhere said:

 

 

@mikeyw64 I know some people have suggested drilling out the front panel. Looks like you're still experimenting with the best layout so it might not be the best idea yet but once you find it, a couple of build logs (admittedly including my own) can be some good references if you're interested in something like that.

 

 

Cheers.

Deliberately avoiding going down the dremel rabbit hole 

At the end of the day If i was a hard core gamer that wanted to run OC all the time I wouldn't be in this case and I might possibly have something a little better than GTX 95 in there (if I could find something reasonably priced lol)

For what I need it for and for my desk aesthetic then at stock clocks this case suits me down to the ground and I don't think getting figures like the ones I got a little earlier of 48/49c above ambient (21.9c at time of testing) for the CPU (at sock clocks)  are too shabby under a test load (Prime 95 small FFTs + furmark) which in normal daily driving she will never see. NB that was with the OEM restrictive filters on top, front & bottom (bought a 3rd one from Coolermaster for a few quid)

GPU temps are a slightly different matter but then again I'm not running her flat out all the time and it will be interesting to see how my mad lad water cooling of the GTX 950 works out(old H55 pump/rad [test fitted]  into a chinese gpu water block [on my desk]  and a modified twin fan shroud [en route from china] for the card)

I 'll also freely admit that if , for example, I put her on Ryzen Master Auto OC settings and run the same test then things are a little toasty but again, I don't run her that way

5 minutesnippet mid way through test run




 

[mikeyw64]

One other advantage of relocating the PSU is I've been able to tidy up the backside a bit  

And yes I know that hub isn't straight, needs fresh sticky tape lol

 

[thrasher_565]

what is this? looks like a pump?

 

 

[mikeyw64]

13 minutes ago, thrasher_565 said:

what is this? looks like a pump?

 

 

 

That’s the test fit of the H55 integrated rad/pump that I’m going to be using to water cool the GTX959

[thrasher_565]

2 minutes ago, mikeyw64 said:

That’s the test fit of the H55 integrated rad/pump that I’m going to be using to water cool the GTX959

oh i just noticed the fan mounted to a rad i get it now.

 

i was thinking that i never seen a pump like that before...

[mikeyw64]

Not sure why as it makes sense. Well to me at least lol

[ivanrcks95]

Yeah as someone with a build in a Q300L, trying to work around the cases flaws just aint worth it. Dremel all the way. I ended up making 2x140mm holes in the front only. Two Arctic P14s up front, and a Arctic P12 out the back, thats it, Deepcool Gammaxx 400EX air cooler, and a de-shroud mod on the GPU with two arctic P12s. Both the CPU and GPU are OCed and the fans are controlled through speedfan. My fans dont go above 50% fan speed, and under load even with the OC, my GPU stays below 70c, CPU below 65c, this is just the recorded temps, not above ambient. Super quiet under load, nothing gets hot, and its all OCed, so yeah, i tried like you to make the case work in anyway i could, but at the end it wasnt worth it, and cutting holes in it solved all issues.

 

Oh CPU is a R5 2600 and GPU is a RTX 2070 Gaming OC

[mikeyw64]

9 minutes ago, ivanrcks95 said:

Yeah as someone with a build in a Q300L, trying to work around the cases flaws just aint worth it. Dremel all the way. I ended up making 2x140mm holes in the front only. Two Arctic P14s up front, and a Arctic P12 out the back, thats it, Deepcool Gammaxx 400EX air cooler, and a de-shroud mod on the GPU with two arctic P12s. Both the CPU and GPU are OCed and the fans are controlled through speedfan. My fans dont go above 50% fan speed, and under load even with the OC, my GPU stays below 70c, CPU below 65c, this is just the recorded temps, not above ambient. Super quiet under load, nothing gets hot, and its all OCed, so yeah, i tried like you to make the case work in anyway i could, but at the end it wasnt worth it, and cutting holes in it solved all issues.

 

Oh CPU is a R5 2600 and GPU is a RTX 2070 Gaming OC

 I may well do down the hole cutting route in the future but for the moment .........