JonC

Could you elaborate on the windows limitation (KB or API documentation)? I'd be interested in looking into a workaround.

I know this is an old thread, but it appears still relevant. It's now been over two years and this still doesn't seem to have been addressed. I personally would like an option to pick the profile under lock or continue to have the active profile continue running until the machine is unlocked. I think a configuration option would be the best to make the most people happy. I fully understand the hardware lighting options, but they are limited and I don't get the logic for disconnecting the USB devices when the system is locked. It just causes a lot of flashing when I lock / unlock my profile even though I have one lighting profile I use when locking so that the main profile and hardware profile are in sync as the RAM doesn't default to hardware settings. It's very inconsistent due to the USB disconnect of the devices and the varying behavior across the various products. The hardest part for me to get is this works pretty much how you want it to when the machine boots up and no one is logged in. Why that can't continue into the lock screen and hasn't been able to be added as a setting in over two years is hard to comprehend. I really like the Corsair fans and lighting effects, but honestly this lack of ability to fully control lighting when the machine is locked which is a significant amount of time, will cause me to move to a different vendors solution for my next build. Hopefully adding some configuration options in this area can get on the dev plan.

@elgreco – Glad that was what you were needing to see. I think putting all 3 LLs on top with it sitting on the floor sounds like a good plan. My LLs on top from the pic are all blowing into the case. That will be the coldest air for the radiator, also if you have them exhausting the LED ring will be against the radiator and the X support from the fan will be visible. It still looks nice that way, but a little less LED effect. I have tried mine both ways, but had much better cooling with my radiators blowing into the case. Look forward to seeing some pics of your build when you get it together.

@elgreco - Here's the top down view Let me know if you one to see anything else. I have it apart today putting in some QL fans. So most things are accessible with a camera.

usna92, I started off with cardboard and blue tape to roughly seal up the areas of interest. I initially even started with the just the front side that was easily accessible to see if I got improvements. Mentioning improvements, I recommend setting up the thermocouples from the Co-Pro on the inlet and outlet side of the radiator (at a minimum). Setup a program that can generate a near constant thermal load (I use Prime 95 non-AVX Small FFT, just make sure your temps stay under 75 – 80C with the load with 75 being my target for a 24x7 build). Run this for 20 – 60 minutes or until you hit steady state. I pull the thermo couple graphs out from the Co-Pro on iCue and put them directly on the dashboard. Once you graph is close to a flat line or your min and max are the same you should be at steady state. From a cold loop this can take an hour+ on a hot loop it’s quicker but you need to be at steady state / have your loop heat saturated at a constant load to have a repeatable baseline / reliable comparison. I also like to watch water temp on the Co-Pro as that is really what we are aiming to lower. As for the specifics on the blocking after I saw improvements with cardboard and blue tape, I made some mockups on foam board considering where I was going to drill holes in the fan tray, overlap needed etc. I dry fitted those a couple of times with blue tape putting the trays in an out of the case. Your front one can be a little wider than the rear as you will need a bit more clearance for the for the wires to go between the weather striping and the case. There were some small variations (1- 3mm) between the front and top fitment so if you want them tight you will need to make 4 templates (front and back for each fan tray). Take your time with this as it’s much easier to recut foam board than acrylic. Make sure to mark the foam with the line on how much you want to overlap the tray. This is important for mounting consistently and marking drill locations later. You will also need to offset for the bulb weather stripping. I cut my foam board full sized and then offset it for the weatherstripping. Remember the bulb will collapse about half but the hard vinyl will not compress. I think I used about 0.25 inch offset, but I cannot remember the exact number and it will vary from brand to brand. Once everything fits the way you want you will need to start fabricating the baffles. I used a 12x24 inch 0.118 thick piece of cast acrylic translucent black. I made my seals / edge dressing from Trim-Lok Trim Seal with Top Bulb for 1/8” edge. Hardware wise I used m3x8mm screws with countersunk heads and m3x0.5 nylon lock nuts both in black. You will also need a 3mm drill bit with counter bore for the acrylic. Once I had the materials, I covered my trays and acrylic in blue tape and traced off my first two acrylic pieces. I clamped mine between two pieces of wood using the top piece as a fence for a jig saw (not the best tool for the job, but what I had available). With acrylic keep your tool speed fairly slow, too fast and it will melt making a messy cut and sticking to your blade / bit. Once I cut the first two pieces I marked and cut 3 and 4. With those cut I then transferred my overlap lines from my foam to the acrylic sheets. I also transfer them to the metal fan shrouds. At this point I lay out my drill locations on the metal trays. I used 4 per side just outside of the fan mount holes. See photo below showing the countersunk screwheads installed. Once the trays are drilled you will need to debur (larger drill bit running backwards or a box knife will do if you don’t have a deburring tool). Next line up the appropriate piece of acrylic with the fan tray and tape / clamp in place. This alignment needs to be fairly precise just like it was with the foamboard templates or you could have a gap or be too wide and things not fit inside the case. Once the acrylic is drilled there’s no fixing a mistake and you would have to cut another piece. So take your time I took a break before starting this part. Once the acrylic is aligned and taped up I went to the drill press and drilled the holes passing the bit through the hole in the metal tray and then drilling into the acrylic as this guaranteed my 4 holes would match up. Just be careful not to slide the acrylic relative to the tray as you reposition. I kept a scrap block of wood under the acrylic to drill into and minimize chipping when the drill breaks through. After drilling all 4 pieces it’s time to take the countersink bit and cut the cone for the screw head to recess into. Do a little test the screw and repeat until the head sits flush. Once all the countersinking is done you may need to debur the acrylic. It depends on your tool sharpness and cutting speed. But do any cleanup you want at this point before pulling the tape and protective plastic. Once they are cleaned up remove the plastic film, cut and apply the bulb weather stripping. I won’t lie this took some time and some grip strength to get on straight. Note if you have a bad edge (cut) on your acrylic you can plan for it to be outside before drilling and then hide it under the weatherstripping. At this point it is mount them up and putting in the 4 screws per baffle. You want these snug but not too tight so as not to crack the acrylic. Dry fit the trays at this point to make sure there’s clearance for wires, the door closes fully, etc. If it’s too big you can probably remove the weather stripping and trim. I got lucky and all of mine fit first try. At that point mount up the radiators and fans. Then clean off the fingerprints and make them shine. reinstall the loop being careful to not leave finger prints and rerun your baseline test and see what kind of improvements you get. I think it also makes the front look nice as the only light you can see is from the fans and their reflection. Good luck with your mods.

Elgreco, that's a good point and I definitely have positive pressure in my case. I like a small amount as it keeps dust out of the case which I always prefer, but think is critical on a glass sided build like the 1000D. Note from what I have read negative pressure provides better cooling efficiency, but a lot of what I have read is generalizations and I suspect there are variations based on air cooling vs water cooling and how over / undersized the loop is. The case itself has a good bit of open area to equalize pressure differences. The back has a lot of honeycomb sections and there is a gap around 3 sides of each glass door (2 on mine as I weather stripped the top to further minimize hot air intaking into the radiator). So, the positive pressure does generate additional airflow out of these vents. I did some testing removing the mini ITX cover and temps dropped somewhat indicating that there is less than ideal exhaust flow. I didn't write down the numbers, but it was less than 20% the drop of limiting the mixing of hot and cold air streams at the radiators. I put the ITX cover back on as it didn't suit the aesthetics of the build to have a large open hole. One thing I was able to change to reduce the pressure difference is replace the PC slot covers with high flow versions which did realize a small, but measurable internal temp drop. I think in my situation less airflow with cooler intake temps yields the best result as probably 90% (guestimate, based on heavily overclocked 12 core CPU) of my total heat is coming from the CPU and going to a loop with 5 x 140 radiator with is significantly oversized to work well at low CFM / slow fan speeds. What I saw with thermocouples in intake and exhaust on both radiators is that my inlet to outlet air temps was about 6 delta T (dT), but over sustained load both air temps would rise as much as 8 degrees due to hot case air getting pulled in the inlet. The delta T on the air temps stays the roughly the same, but the inlet continues to rise, which causes a 1 to 1 increase in water temp and roughly a 1 to 1 at the CPU. With my radiators I get an outlet water temp that is about 4 – 4.5 degrees higher than intake air so the 8 degree rise in inlet temps was causing my water temps to be 12 – 13 degrees over room temp vs 4 – 4.5. So, the extra airflow was more than offset (3 or 4 times over) by the rising air inlet temps. I expect results would vary based on loop design (radiators area to heat in watts and radiator design to airflow / fan static pressure). I also tried the top radiator with reverse airflow, I will spare you the details, but it was not as efficient. Cold radiator intake air seems to be the most important variable in my setup. I’m testing a couple of loop changes currently to try to reduce the dT between the CPU and the water as that is my current bottleneck (non-linear change between water temp and CPU temp). Depending on those results I may come back to the airflow issue as there is still some room for improvement (maybe 1 degree of water temp vs the 6 I got on the baffles). I’m not sure if I will do that, but my next step would be to fashion an exhaust fan into the 3.5 inch drive bay (as I only use one 3.5” drive for backups). That would require cutting a hole in both the bottom of the case and the removable foot and then mounting the fan (I haven’t investigated what would / could fit yet / would require ducting). I’m running two Co-Pros so hooking it to an extra channel and setting it to turn on / ramp when case temps exceed a certain temp would be possible. I really like the flexibility of multiple fan channels and temp channels which can all be named in iCUE. it really makes it nice to be able to watch things under various loads and see if changes like opening the door or removing a panel or sticking in a cardboard and painters tape baffle make a difference, before fabricating changes As an aside (I may just be missing something in iCUE), it would be really nice to be able to define a dT to use to drive a fan curve so that it would be possible to start extra fans when case temp is more than X degrees warmer than ambient, etc. I currently drive my radiator fans by water temp, but I would prefer to have that as dT between water and ambient temps. Thanks again for you input on case pressure, I enjoy this type of discussion and look forward to any additional feedback you may have.

Zotty, first off I wanted to say thanks for the informative thread. I read through it a couple of time before starting my 1000D build. The diagrams for wiring up the Co-Pros and fans saved me some iterations of assembly and disassembly. I also thought I would post some a few things about my experience on getting low delta T on the water loop. I am using LL 140s on the Radiators with the 3x140 plates front and top. Based on the Co-Pro temp probes I was getting a good bit of hot case air pulling back to the fan intakes. I was getting about 9 - 11 C Delta T Air to Water in that default config. I modded the trays and added some acrylic and weather stripping to better separate intake and exhaust air. Here's a pic of the modded fan tray. I also removed the top of the case and added spaces adding 5 mm to the metal plate and 15 additional to the glass plate on top to allow the top Radiator to have better airflow. This got my delta T down to 4 - 5 degrees C and got roughly a 6 degree C drop on the processor temps. Probably not relevant for everyone as I'm pushing 12 cores at 5 GHZ and except for stress testing the LL140s stay around 700 - 800 with the fan speed controlled by the Co-Pro and Hydro-X Water Temp Sensor. Pretty pleased with the performance to noise ratio. Here's a pic of the top tray slid about 1/2 into the case that also shows the increased clearance on the top vents Here's another pic of the running system in a near complete state. I did a similar mod with LEDs on the outside of the bottom case to what you did with a similar style metal track. I don't have immediate plans to put in the ITX board so I made a U Shaped LED strip to go on top of the Power Supply Cover which shows in the above pics. It wasn't too bad to cut the strips between LEDs, expose the traces and solder in some 90s in the wiring and dress it up with hot glue and heat shrink. I am a few lights short of a full two strips, but I just turned off those LEDs in iCUE and it seems to adapt without issues for the various effects. If anyone else wants to do that I do recommend putting a continuity tester in the connected to make sure none of the traces got bridged when soldering. Thanks again for the informative thread. This one and your iCUE FAQ saved me tons of headaches for things that were not in the Corsair docs. Next week I am putting in some QL fans on the inside of the radiators (push pull setup) to get a bit more LED effect inside the case since both of my radiators exhaust into the case to intake the coolest air.