ChristianChalliner

Yeah I could use speedfan but do I have to manually enable it every boot up? I'd prefer something that was set and forget which is why i'd like to do the adjustment in the bios if possible.

I wasn't sure what section suited this best (air cooling, water cooling or m'board discussion so i posted it here as it's a m'board setting). Long story short, I have a custom loop with a 7700k and a 1080ti, 1x240mm rad and 1x140mm rad. The system has a ddc pump controlled via pwm, 2x80 mm case fans, 2x120mm fans on one rad and 1x140mm fan on the other. All fans are pwm and are connected as follows; CPU_Header ddc pump both 120 mm rad fans CHA_Header both 80 mm fans 140 mm rad fan The q fan control has adjustable ramping which is great except for the problem that a cpu temp of 75 is locked to 100 fan speed. To keep the ddc pump and fans at 'quiet' noise levels (silence is the goal here not large clocks) I've used the manual profile to create a very low and flat curve up to about 60 degrees (60C=40% speed) and this works exactly as I want, in Superposition benchmark I get good gpu temps of <50 and cpu temps (can't remember exactly but they're fine) so clearly my cooling solution is adequate at this low noise and low speed. At 20% speed on everything at idle `I've got gpu temp of 25 and cpu temp of 34 so I'm happy with that. But, this is the 7700k with it's dumb temperature spikes, so if i go and run something like aida64 and a core momentarily pings to 70-75 from 60 my ddc and fans decide to go full blast 100% (thanks ASUS). Not only does this suddenly make my system go from silent to sounding like a tornado it's also utterly pointless. Blasting the fans and pump up to crazy speeds is useless in a watercooled build since there's no way the fluid temp can drop quickly enough to make a difference. I figure there must be some way to change this upper point in the cpu curve? But it cannot be done on the graph so maybe it can be done elsewhere in the bios? Alternatively, while the CPU header must always be locked to the CPU temp the CHA header can reference off of other temperature sensors such as my coolant temperature sender. Therefore, I figure I could run my pump pwm signal and 120 mm fans off of this header to keep noise down whilst running my larger lower speed and quieter 140mm fan off of the CPU header? (In this i'd likely ditch the 80mm fans as I think their impact is minimal anyhow. Anyhow, I'd like to here what others think on this and if anyone has found a solution to this problem before but please note the following in your reply: - Silence is my priority, I do not mind my cpu running a little hot to keep it quiet (high 70's-low 80's i can easily live with) - I have no space for a fan controller so this isn't a solution (my case is a Corsair 250D and my 5.25 bay houses my res/ddc pump combo) - My preference is a 'set and forget' type solution hence the desire to use q-fan in the bios rather than a software based solution such as speedfan. Thanks ?

Best way to minimize the risk in a custom loop in my opinion is to use soft tubing while minimizing the number of joints in your system so only using any fittings other than straight barbs/compression fittings where it's absolutely necessary. No stacks of fittings (90 angle + 25mm extender + barb as an example). Minimum number of joints = minimum points for failure to occur. likewise soft tubing is much less likely to disconnect itself from fittings like hard tubing can. That said, i've used push compression fittings on very small diameter hardline tubing for automotive application before and they've not leaked/come loose. But I'd still go with soft tubing if you're really concerned.

Power supply. If it's a non-modular power supply there's likely a molex connector on there somewhere, if not, you can get a SATA to Molex adapter. If it's modular you have a cable somewhere which will plug into an unused outlet on the power supply which will probably have a few SATA power connectors on it plus a Molex.

ASRock out of the two but if you can get the ASUS board as it's much better. Not sure how useful 2 x ethernet ports are for most people but at least you get some more usb's in the stack :-)

Look at top flow/low profile coolers, the 212 evo isn't particularly thick (i.e. Fin stack is thin and it's not a dual fan or stack solution) so a bunch of top flow or lp style coolers can match or at the very least come close to the 212's performance. if you don't like Noctua colouring I'd suggest looking at Bequiet! I see barely anyone ever talk about them but they make some really good stuff. For a case perhaps look at the Corsair 250D? I've been building in one recently and it's very compact but can fit full size psu's, gpu's etc with very little wasted space. The official site says a max cooler height of 92mm but i'm pretty sure there's more space than that. I can measure it tomorrow if you're interested?

He still has a point though to a degree. Any X299 board shouldn't really be considered 'low end' so shouldn't have these VRM issues.

If you've got lots of glue residue that could be part of the problem. When i delided my 7700k all went well until i went to glue the IHS on, i followed the instructions that came with the rockit88 kit but after about 30 minutes i noticed there was substantial fogging due to the glue vapours. If the glue fogged over the top of the IHS as mine did that could be stopping you getting a good application of your thermal compound. Something slightly abrasive works good (like the cloth that comes with the coolaboratory stuff) or you can use nail polish remover to dissolve the glue just be careful with this method as i don't know what if any impact it might have on the pcb if you go nuts with it.

It's difficult to say without knowing exactly what grade of brass is used. self corrosion is a thing but not for regular items, it occours in strange heavy metal combinations. Something sortof similar to self corrosion relating to brass is 'dezincification' (no I'm not joking). Basically the zinc is pulled from the alloy and only copper remains, generally they add something like arsenic or tin in very small quantities to prevent this from happening when they know the item will be used in contact with water, chlorine, etc. These alloys are usually marked DZR (dezincification resistent) or CR (Corrosion Resistant). In short, your fittings should be made with this in mind, therefore, it shouldn't be a concern unless you manage to find some super dodgy cheap non-brand fittings. I'm not 100% on the chemistry though so don't take it as the be all and end all :-)

Thanks, will go down that route then, completely forgot the PWM signals were 5V=100%. With regards to a fan controller, i simply don't have the space to fit one in the case, so i guess my best bet is to swap the radiator fans for non-pwm ones and just run them at a fixed speed for now. it's somewhat irritating that the board only has one PWM header.

Thanks for the reply but I take it you didn't look at the pump (No, I linked the wrong one like an idiot! Good spot!). It has a 12V supply direct from the PSU via molex connector which is the feed for the pump. It then has an auxillaey connector which has a PWM signal line and a rpm signal line. The pump couldn't draw power via the motherboard even if it wanted to.

Hi, I have an idea of how this will work but i wanted to run it by a few people who might have done something like this before. In short, I'm building a system using an ASUS Z270I (itx) motherboard, the board has the following available fan headers(and pin outputs): CPU Fan - PWM (ground, +12 V, rpm, PWM Signall) CHASSIS FAN - (ground, +12 V, rpm, +5V) AIO PUMP - (ground, +12 V, rpm, +5 V) I have two radiator fans (120 mm pwm), 3 chassis fans (2x80 mm, 1 x 140 mm) and the pump. My original plan was to run the 2 radiator fans off the cpu fan header, the pump off the AIO header (note it has a 12 V power feed direct from the PSU also) and the 3 remaining fans off the chassis fan header. However, i did not realise that the AIO header did not support PWM, thus, if i use this header for the pump i will only ever have a fixed pump speed 42% of maximum, that doesn't seem all idle but also not the end of the world? But... I also know that this board can likely control fans via dc voltage on the none PWM headers, so my thinking was what if i connected the pump PWM signal pin to the +12V pin instead perhaps I could have DC control of the pump speed via the AIO header instead? After all, PWM is just a digital way to mimic a voltage so i doubt the pumps going to notice the difference between a 7V PWM signal and a 7V DC signal? Here's the link to the pump i'm using https://www.ekwb.com/shop/ek-ddc-3-2-pwm-laing-ddc-3-2-pwm Thanks, Christian