A DIY PC built from wood and rope

[DDY]

Modern parts tend to sip power and produce very little heat unless you start overclocking,etc.

The part consuming the most power and producing the most heat will be the graphics card,followed by the CPU if its overclocked or something like an FX. However,most people don't really overclock so the CPU really isn't consuming that much or even stressing the totally not bling motherboard I have.

I had a SD37P2 Shuttle SFF system,running an overclocked high VID Q6600 on a hot running 975X chipset,two HDDs,card reader,optical drive and cards like an overclocked HD5850 or an overclocked 8800GTS 512MB. I even had a Pentium D805 in there(overclocked OFC) at one point.

The system worked for 5 years fine,even though it the exhaust was stupidly hot on the side and the back. There was only one fansink to cool the whole system and it blew out the side of the case:

https://www.bit-tech.net/reviews/tech/shuttle_sd37p2/2/

I did install a better chipset cooler and put better fans in and ran them at a lower constant speed,but that system probably had tons of deadspots and Shuttle tended to not always use the best caps either. After 5 years I thought it had finally gone kaput - it was actually a dodgy cable on the HDD.

Another mate made an ammo box PC,with a single slot 9800GT,a Zotac 9300 mini-ITX motherboard and it ran stupidly hot since there was virtually no airflow. The motherboard and 9800GT probably still work even today(they were a few years ago).

My current SFF system has one 92MM fansink for exhaust and a single inlet fan running at low RPM. I have had that system for years.There are no VRM heatsinks,the chipset has a tiny heatsink,etc - and I use a AIO water cooler,so its not like the stock cooler where you get some air directed to the VRMs.


Also what about all those tiny NUC type systems in their tiny cases?? There is virtually no airflow to the motherboard either,as is the case for most laptops - yet they seem fine.

I reckon that the wood rope build is substantially more air flow constrained than the Shuttle. The Shuttle has venting and significantly more forced airflow, compare the forced exhaust of the CPU fans in the Shuttle versus about 20mm of a really slow spinning 120mm (140mm?) fan with the most direct air flow path directly in and back out again of the panel cut out, of the wood rope build.

The ammo box PC would have a greater internal air volume than the wood rope rig and assuming that there are fans inside, there would be enough airflow generated by them to circulate and thermally mix internal air, thereby enabling the entire ammo box enclosure to exchange heat with the outside and remain within reasonable temperatures. The metal case of the ammo box would have magnitudes greater level of thermal transmission than wood.

Besides, two examples are hardly representative of the potential effects of poor thermal management on longevity and reliability of hardware.

Intel's NUCs are notorious for getting really hot, they are known to thermally throttle and become noisy under load as the tiny fan does its best at keeping the lot cool.

The same can be said with laptops when it comes to CPU and GPU cooling. Also, laptops obviously don't have the luxury of air volume but rather take advantage of the direct thermal conduction of lower-power components, such as the chipset or power supply circuitry, with the laptop body or even through the keyboard!

[CAT-THE-FIFTH]

I reckon that the wood rope build is substantially more air flow constrained than the Shuttle. The Shuttle has venting and significantly more forced airflow, compare the forced exhaust of the CPU fans in the Shuttle versus about 20mm of a really slow spinning 120mm (140mm?) fan with the most direct air flow path directly in and back out again of the panel cut out, of the wood rope build.

The ammo box PC would have a greater internal air volume than the wood rope rig and assuming that there are fans inside, there would be enough airflow generated by them to circulate and thermally mix internal air, thereby enabling the entire ammo box enclosure to exchange heat with the outside and remain within reasonable temperatures. The metal case of the ammo box would have magnitudes greater level of thermal transmission than wood.

Besides, two examples are hardly representative of the potential effects of poor thermal management on longevity and reliability of hardware.

Intel's NUCs are notorious for getting really hot, they are known to thermally throttle and become noisy under load as the tiny fan does its best at keeping the lot cool.

The same can be said with laptops when it comes to CPU and GPU cooling. Also, laptops obviously don't have the luxury of air volume but rather take advantage of the direct thermal conduction of lower-power components, such as the chipset or power supply circuitry, with the laptop body or even through the keyboard!

A high VID Q6600 overclocked on a non-optimal 975X which ran very hot - I really don't think you get the concept of heat(it would heat a small home! ) or the fact unlike nowadays most motherboards were not using solid caps,let alone long life non-solid types. My main rigs for well over 12 years have all been SFF systems. All Shuttle form factor or mini-ITX ones,with full desktop parts,all the GPUs overclocked if required.

But the issue is not the only case of SFF systems I have owned,or mates owned which were thermally constrained - SFF systems are by their very nature,so its a compromise.

The same goes with millions of NUC type devices,bog standard desktops,laptops,etc. If you go by what enthusiasts on forum keep saying all these systems would suddenly explode within a year. Then when you go somewhere,you will find these systems which have not been maintained for years,full of dust,running stock parts still chugging along.

Many last years - having owned loads of SFF systems in over a decade,I honestly people don't seem to appreciate how modern DIY PC components are overbuilt compared to stuff made even 10 to 12 years ago,or even how efficient the parts are.

Plus the ammo box PC ran hot,as it had no airflow - often hitting something like 80C+ and all the parts still work today. The Q9300 in it had a slightly uneven heatspreader and it ran hot,and I told my mate perhaps the rig wasn't a good idea,and he still went meh!!

I save the best one for last - helped a mate out with a rig yonks ago(ATX one). He managed to unlock an Athlon II X3 to a full Phenom II X4 with L3 cache,but he used the auto-unlock function which shoves a ton of voltage throught it. I told him to replace the heatsink,drop down the volatage manually and see the lowest he could get away with, since the AMD alu cooler of fail would probably melt and put in another system fan,and to do it as quickly as possible. He had said he was ordering the extra parts and was going to do so.

Fast forward 4 years,he apparently never bothered. The unlocked Phenom II X4 was under that crap cooler for 4 years with a load of voltage,and it was running ridiculously hot. Yep,there was much mocking that day! The motherboard and CPU still work though!

This is why the next time I made sure he ordered all the parts,and installed sufficient cooling for the replacement build!

The fact is the system in the OP will be fine,especially when it has a largish gap above the motherboard box for air to rise upwards.

By the time the build in the OP might have issues it will be long past its warranty date and guess what?? If you go SFF,don't expect good airflow as it comes with the territory. Not those "MASSIVE" ones but smaller ones. People need to stop worrying so much.

If you go on like that you will never build anything other than in a massive e-ATX case with a 1000 fans pointed everywhere - even modern PSUs are not designed to be the best for cooling,but more for aesthetics and noise. Plus no one would ever use an AIO water cooler,only horizontal air coolers as they provide chipset cooling. Yet I barely see anyone worry about that.

Its like all the people on forums who "need" mahoosive power supplies. Literally every SFF system I had,had PSUs under 500W. In fact I run a 450W SFX PSU now - my last one was 450W too. I am running a GTX1080FE,IB Core i7,a few drives and the PSU is not really that taxed. I knew that since I looked at what Corsair specced for their own mini-ITX systems. A 400W version of the same PSU with a Core i7 and a GTX1080,and at the wall I was lucky to see even 280W.

Its the same with all the ones who said you needed to delid IB to make it useable due to "temperatures" and getting neurotic if the CPU breached 60C,etc so on whilst most examples in the wild are running crappy stock coolers and working fine.

The fact is Prime95 might look cool,but in most cases,people are unlikely to push their systems anywhere near that,and even if they do it would only for brief periods I suspect.

If I listened to half the stuff said,I would never have even done a SFF system in 2005,I would never even have bought a Shuttle(since they apparently would go kaput in a year,etc) since I would got scared it might blow up. Meh.

[DDY]

I think you've misunderstood my last two posts. Lets get back on track.

You've accused me of not understanding the concepts of heat or the thermal nature of modern electronics, and that my only solution for thermal management is to build water cooled EATX rigs. And that I take my advice from people who warn that machines explode within a year due to overheating.

Honestly, I feel somewhat dismayed by the tone of your post and to be pragmatic I'll put it down to a misunderstanding, although I've noticed and appreciate that you've edited your post and toned it down a little. I don't think I've written anything unreasonable, or anything which indicates an unsound understanding of thermal concepts or suggests that I take my advice from people who warn that machines explode within a year, or that I only build water cooled EATX rigs.

The point that I wanted to put across is that I would concerned about the lack of cooling for motherboard components for the wood rope rig.

Why?
Lack of air flow, due to obstructions, lack of venting and ineffective arrangement of the CPU fan.

Why would I be concerned?
Possible effects on performance, longevity and reliability. While running components at the temperatures likely to occur within the wood rope rig would probably be within design spec. I don't think anyone would disagree that at such temperatures, the component longevity, reliability and performance will be adversely impacted, but the extent in terms of time, risk and impact is up for debate.

Is this a reasonable concern?
It's extremely difficult to quantify the risk, not least because our assessment is based on the information off a Youtube video, but generally because there are many, many, more factors unknown to us such as stochastic variability of component manufacture, environmental conditions and usage scenarios.

But there are things that we know at a short-term empirical level that are probable, for example;
- We know power throttling may occur when motherboard power supplies reach high temperatures, overlocked or not, and performance may be impacted as a result, e.g. when the CPU under-clocks as a mitigation measure.
- We know that instability may occur when motherboard power supplies, e.g. to the CPU or the RAM, reach high temperatures, again overclocked or not, but of course substantially exacerbated by the former.
- We know that thermal throttling of the SSD may occur, we know this happens in well ventilated machines, imagine what will happen in this scenario?

For the longer term, would there be issues? Probably not, given the empirical evidence and design of modern electronics. But I don't think anyone is in a position to unequivocally predict the outcome, especially given the unique nature of this build.

Considering this, I feel it is a reasonable concern, although you may continue to throw anecdotes at me if you like.

But really, I mean, would it kill to punch a few holes somewhere?

[CAT-THE-FIFTH]

The point that I wanted to put across is that I would concerned about the lack of cooling for motherboard components for the wood rope rig.

Why?
Lack of air flow, due to obstructions, lack of venting and ineffective arrangement of the CPU fan.

Why would I be concerned?
Possible effects on performance, longevity and reliability. While running components at the temperatures likely to occur within the wood rope rig would probably be within design spec. I don't think anyone would disagree that at such temperatures, the component longevity, reliability and performance will be adversely impacted, but the extent in terms of time, risk and impact is up for debate.

Is this a reasonable concern?
It's extremely difficult to quantify the risk, not least because our assessment is based on the information off a Youtube video, but generally because there are many, many, more factors unknown to us such as stochastic variability of component manufacture, environmental conditions and usage scenarios.

But there are things that we know at a short-term empirical level that are probable, for example;
- We know power throttling may occur when motherboard power supplies reach high temperatures, overlocked or not, and performance may be impacted as a result, e.g. when the CPU under-clocks as a mitigation measure.
- We know that instability may occur when motherboard power supplies, e.g. to the CPU or the RAM, reach high temperatures, again overclocked or not, but of course substantially exacerbated by the former.
- We know that thermal throttling of the SSD may occur, we know this happens in well ventilated machines, imagine what will happen in this scenario?

For the longer term, would there be issues? Probably not, given the empirical evidence and design of modern electronics. But I don't think anyone is in a position to unequivocally predict the outcome, especially given the unique nature of this build.

Because I don't think it is as bad as you seem to think it is - most cases should be made of something like aluminium,but instead its steel,plastic,composites,wood,etc which are all not as great with regards to thermal conductivity.

Many normal cases also are made with aesthetics and "noise-blocking" in mind instead and not for optimal airflow. Most modern consumer PSUs are not ideally designed for airflow but the top down fans are much better for noise.

Look at modern AIO water coolers - they are made for aesthetics,and for easier mounting(they weigh less),but ideally they are not the best,since a horizontal air cooler would draw air over the motherboard chipset and VRMs. Its why the Intel and AMD stock coolers are horizontal coolers.

All of this takes advantage of modern components for DIY builds being overbuilt to a degree,so can run hotter than required and SFF builds(which this is despite the stand) you are sacrificing cooling for size and form factor and design aesthetics.

Like with a few of the other SFF builds linked in this thread,form factor and appearance are the number one point of many of them. The chap who built the wood/rope steampunk coffee machine PC,probably didn't drill some extra ventilation holes as it affected whatever aesthetic he was looking for and I also suspect it is partly structural. Those heatsinks he used are mahoosive and weigh a ton - it also makes me wonder whether removing more metal from the sides would lead the central platform too weak to handle the weight as the wood sides are just aesthetic and its aluminium not steel. The whole PC is attached in place by 2 bolts on each side of the metal box.

If you saw my current SFF system,you would be getting right nervous. It has zero ventilation holes outside the bottom air inlet and the rear exhaust since I prefer clean looking solid cases - and they also transmit less noise. My current SFF system has all the drives mounted vertically on rubber gromits underneath the side of the vertical motherboard tray,and is enclosed by the side of the case. So no active airflow for the drives.The drives are perfectly fine.The other side of the case,I modded with thick foam(traps heat though),to soundproof it with a slight gap at the bottom and top for air but again not ideally thermally too,but again none of the parts are that high power either,and the GTX1080 has a blower fan too.

However,my view is modern parts have enough leeway to run a bit hotter for the expected lifespan of the build(say 5 years or thereabouts) and if you have a SFF build in anything other than those ridiculously large "mini-ITX" cases,with non low-power parts expect things to get more hotter than a better ventilated larger case. It goes with the territory.

[dandyp]

I've watched his stuff for a few years now, and a few of his builds are more function over form. He's clearly benefiting from sponsors these days, which obviously allows him to do these kinds of builds, that does mean that his channel doesn't quite have the same low-fi charm that it used to.