Motherboard grounding questions

All hardware or software issues that do not have a specific forum go here.

Moderator: The Mod Squad

Re: Motherboard grounding questions

Postby LakaWaka » Thu Jul 28, 2016 7:40 am

Karlsweldt wrote:I am not the only one here at the Forum with many years of computer experience. Many others, too, have long years of experience. Some are professionals, with their own shops.


You're basically the only one who has helped me, besides Twisty, so thanks.


Most I/O shields have springy tabs at some holes, to make contact with shields around ports. Some also have a metal tab or two to contact the top of port shields. Yet some I/O shields are no more than a thin plate with just holes.. relying on hard contact with the case. In the event of no proper grounding of the shield, a short length of wire could be connected with small screws to the shield and case.


This I/O shield is basically a plate with holes lol.

Here's an image I found online to give you a better idea.

Image


What fully encompasses "proper grounding." My main computer has just a plate that I stuck on. Is it supposed to have some sort of wires...????

I would assume every I/O shield works as intended, but the issue in my case is not having it connected on all sides, thus the wire is an extra precaution in this case?

I can connect 3 sides, would that work, or would I need all 4? Should I connect the wire on the 4th side? I could also possibly use the metal that came with the case, and get someone to cut the metal that came with the case and have the i/o shield fit to size. I was also thinking if having the metal drilled for the ports I will be using, so I'm curious would this thick piece of metal work better if I can drill it?

So the ports themselves have to be grounded, or is it that they are sticking on the outside of the computer? I'm just not sure if it just completes the loop, or if it's just because it's sticking outside of the case? I'm assuming that PCI slots are also grounded then? They are all metal though.

All small DC fans have one rotation. They are 'brushless', being driven by a tiny servo motor. If a brush-type motor were used, it would create a lot of unwanted static and RF noise. Same with AC fans. Not intended for use inside computer cases!
The heat produced in a computer setup is from resistive loading on circuits and chips. The CPU and GPU are the major heat producers. The PSU is the highest heat source.
Cooling fans for a computer case should be installed so that there is more air forced into the case than drawn out. This is so the PSU is not starved for its cooling air needs.
The fewer count of fans, the quieter a system is. Too many fans, and it sounds like a bee hive in full-on alert!
Several small fans would require more current, and still not be as efficient as one larger fan.


Thanks for the info. So why do the fans need 2" clearance? How would heat get pulled back in?[[[

As to measuring currents in working circuits, most ammeter designs use a slightly resistive bit of metal that causes a minute drop in voltage. The shunt has connecting wire leads at its ends, and feeds a volt meter which is calibrated to read amperes. The other type of ammeter has an aluminum bar that can be placed over or wrapped with the wire, and picks up the current flow by parasitic means. Either type can be almost 100% accurate.
That Kill-a-Watt device is a nice tool for the price, and has good accuracy.


Thanks for the info. I'm just a bit confused what topic this is referring to? Does this have to do with the pico and the votage/current issue you were expressing before?

The kill a watt should hopefully do what's needed, and hopefully I don't draw too much power.

Some older computer designs had a shroud over the exhaust fan by the CPU, and pulled case air over its heat sink and out. But with higher speed CPU types and more transistors on the die, a lot of heat is produced.. too much for that design. Ducting to the CPU fan inlet does still provide the most efficient cooling air source, from outside the case.


Ah, so basically creating a duct to pull the air directly onto the CPU? If I use the fan that comes with it, I can pull air from up top, however it is too tall that way.


I had removed the fan, and the ram and tried adding the smaller fans. They definitely seem to fit, but the issue is there isn't much clearance. I was thinking since the CPU is right next to the left side of the unit, if I couldn't use a few smaller fans and pull in air from the bottom, also possibly the top.

I also have an issue with the fan cables going all over, so I tried putting them under the mobo. Is there an issue with the insulated cables touching the bottom of the board? There is enough space, I just am not sure if something bad could happen. Which is why I ask you :).

I could do it above, over some parts or the heatsink, but I'm scared of the heatsink getting too hot and possibly melting the insulation.... and I definitely do not want to breath that in....

I can do 3 fans on the side, that basically covers the Head sink, issue is the fans are placed a bit below the board and the PCIe slot covers a little of the area, but the heat sink can still get air.

I could also do some fans from the top, I can fit 4x 40mm fans on the heat sink. I could also do 4x 60mm fans, just in case I need more cooling.

According to Noctua's site the 92mm fan that comes with the HSF has

Airflow 50.5 m³/h

Static Pressure 1.64 mm H₂O

--------------------------------------

the 40mm fans have


Airflow 8.2 m³/h

Static Pressure 1.78mm H₂O

------------------------------------------

If I get the 60mm fans which I hope aren't that much wider than the 40s.

Airflow 29.2 m³/h

Static Pressure 2.18 mm H₂O

------------------------------------------


Now I'm not sure how exactly I add up these fans to get a total airflow/static pressure, but it seems the 92mm has much higher airflow than the 40's, even if I add up 7 of them it's not that much. however if I could do 4 60's then it might work (not sure if they will fit over the heat sink, will get some and test), while also having 3 direct intakes from the outside.


The issue comes with getting the top open.

I am really looking to just carve out the "creative" logo.

It is stamped in, so I was looking at "deburring tools" and found this one http://www.noga.com/nogaProducts.php?catID=scra

which I figured could work and allow me to precisely cut out what's needed. There are other tools that could work from this product line, but this blade looking one seems to be the best looking for the job.

The issue comes in that it's in the middle of the computer so if I did cut it open I would have to move the motherboard closer to the center, which means that it's farther from the right intake, but closer to the top. I also lose precious room, so I'm not sure how that will go down...

Some of the letters of the logo do allow the fans to get air inside, but not sure about the ones that don't have anything above them. I don't want them to burn out (not sure how much room they need, maybe that's the 2" business you were talking about, but in that case my case is only 2" tall :P)????


----------------------------------------

Motherboards do have temperature sensors, so too the CPU and GPU, and possibly the hard drive. Some high-end fans also have temperature sensors for speed control. They may not be as accurate as designed units, but do the intended job.
Altering the memory heat sinks will likely void their warranty. Do so at own risk.


Yeah, I just am not sure if all computers come with software that monitors temps. I know my new one does, but not sure about this older sandy setup. Hopefully yes, but I'm not sure if I should get my own temp sensor/thermometer to test the air myself???

Yeah... I guess this g.skill ram does have a lifetime warranty, but I don't know what else to do in this situation for the ram.....

You might want to consider a "mobile" class CPU. They are designed to go to an idle state when not processing data. Less waste heat, less overall current needs. For any CPU, it has a set maximum temperature limit. At this point, it would start to throttle down, or may even cause the system to halt until the temperature drops to a safe zone.

[/quote]

Is that what some laptops use?

Yeah, the cpu throttles, but I'm not sure how fast. Could the CPU damage itself without enough cooling, or is that the whole point of it throttling to avoid that damage? Just not sure if too little fans or even no fans would damage the CPU? Just not looking to ruin any parts.



So overall I could go for a smaller class CPU, if I can run regular Operating Systems on it, but I have this setup and figured it would be cool to get higher end gear into this case.


I think I have a good plan, if I can go with the many smaller fans (up to 7) blowing directly onto the cpu. The 3 side fans will have direct access to the outside air, and at least 2 of the top fans would have direct intake, but the issue comes with the other 2 fans and if the fans have enough room with only a few mm to get air, but the air will be fresh air brought in by the other fans so it could work.

We are then presented with a dust issue, but I can just clean that out.


The ram will unfortunately have to be a victim in this experiment, but I will do that only at the end when I'm sure it will work with the fans.


I just need to make sure that the wires are okay under the board so I can get to the fan splitter.

Finally, I would just need to power it up and set it up, hoping all works. Then I will try to get the Pico.

I had to re-read your comment on it's power intake

That Pico PSU may be ideal for mobile needs, but requires a large amount of 12 volt DC current.. more than 15 amps. And does that Pico PSU provide all the required voltages for a motherboard?


Which is what I'm assuming you were referring to when talking about measuring current in the system? I guess I would want to do that before getting one of these units. Where would I connect the ammeter on the mobo?

This is the power supply they provide for the Pico.


http://www.mini-box.com/12v-16A-AC-DC-Power-Adapter

which there's information about it in the description.


From the picture it gets power in to the 24-pin, and that outputs a SATA, a perif, and a 4-pin CPU plug, with the abilty to do another set of Sata/Perif cables..


Also, I figured I would add this manual of information which has efficiency ratings and other stuff that could be useful.

http://www.mini-itx.com/store/informati ... 160-XT.pdf

I'm not sure if I would go over any of the limits on any rail, but I'm really not doing much...

I guess I will have to test each rail then? How would I do that?


Thanks for the help, this is getting to be a big/annoying project :P.


But it's all worth it itn the end... right? (as long as it works :) ).[


Thanks a lot for all of the help much appreciated.
LakaWaka
Initiate
Initiate
 
Posts: 77
Joined: Sat May 14, 2016 11:41 am

Re: Motherboard grounding questions

Postby Karlsweldt » Thu Jul 28, 2016 4:12 pm

That image of the I/O shield is of a better design type. All those tiny tabs act as ground points, to ensure RF emissions are kept to a minimum. Typically, thin steel sheeting is what most shields are from.
If only three sides of the shield fit properly, then the motherboard is not mounted properly in the case. But if an 'adjustment' is needed to get the shield to correctly fill the opening, a tin snips might be useful.

That notation about ammeters was just FYI as to how they work, which design is best.

Some I/O ports are separately shielded, or they may have RF problems with other data bus frequencies. All exposed port shields are grounded via a main neutral bus in the motherboard "sandwich" to the ATX main power plug.

In fluid dynamics, there is a term called "vortex" which can be critical to desired movement of fluids or air. Over the intake of small fans, a slight vacuum is created, which spins naturally, influenced by the Earth's rotation. This pulls in more air, from a desired point. If not enough free space, then any air to the side of the fan is pulled in, which may be warmer than desired.
You can see this effect if some very cold vapor or chemical smoke is dripped over a small fan intake from more than 2 ft. distant! On large bodies of water, they are known as 'whirlpools' which can suck anything down.

That older design using ducts over a CPU heat sink was so air could be exhausted out of the case. Very few blew fresh air in over the CPU heat sink. But that is just as effective, if not more so.
Should be no issue with wires running under the board, as long as no sharp points could puncture them or they obstruct air flow. Insulation should be rated at 85°C, which is sufficient.

One of the best hobbyist tools is a Dremel Motor Tool. Thousands of bits, grinding and buffing wheels work with it. Some are AC powered, but there also are 'cordless' models. Ideal tool for smoothing rough cuts or even making cuts in thin metal.

Almost all motherboards since the Socket 370 do have thermal sensors for CPU and other needs. The BIOS monitors them separately. But an add-on kit can monitor them for desktop or front panel display.

The suggestion of using a "mobile" CPU type in that system would reduce unwanted heat, and not draw as much current as a standard CPU. Check if the board supports it. Yes, a standard intent with laptops. The idea of going with one of those "mobile" motherboard kits would be even better. The circuits still require a +3.3 volt, +5 volt and -12 volt source. But that is provided by on-board regulators. Just a raw but pure 12 volts DC may be all the power source needed. Might run out to 15 or 20 amps total. The info provided for the Pico PSU noted a 165 watt power limit, which at 12 volts DC would be almost 14 amps. No less than AWG 14 gauge wire should be used for the feed line! Could run directly off a suitable 12 volt battery, or with the AC adapter.
The only way to measure input current to a motherboard is with a clip-on ammeter to each rail lead, and adding the total for each together.. the +3.3, +5 and +12. That would be the sum total of all current. Each source would have its own sub-total sum.

The smaller ITX motherboards should have no problem running a current OS version, if 2 GB of system memory is present, and a minimal 1 GHz CPU. But some features may not be 100% compatible, perhaps your model.
F@H.. to solve mankind's maladies.. in our lifetimes!
Karlsweldt
Mobo-fu Master
Mobo-fu Master
 
Posts: 20687
Joined: Wed Nov 12, 2003 11:57 am
Location: 07438

Re: Motherboard grounding questions

Postby LakaWaka » Thu Jul 28, 2016 8:06 pm

Karlsweldt wrote:That image of the I/O shield is of a better design type. All those tiny tabs act as ground points, to ensure RF emissions are kept to a minimum. Typically, thin steel sheeting is what most shields are from.
If only three sides of the shield fit properly, then the motherboard is not mounted properly in the case. But if an 'adjustment' is needed to get the shield to correctly fill the opening, a tin snips might be useful.


Wow this crappy I/O shield is actually a good one? haha.. It's hard to get on though, so hpefully I can get it.

So I want those pins to be bent around the sides of the ports, or in front? It's hard to tell what is supposed to happen with these.

The shield will hopefully fit over the I/O ports, but the issue is that the case is wider than the shield, so there's the 4th side of the case it's not connecting to.

The only adjustment I can do in this case is add more metal, but not sure what should be done? It seems you might have misunderstood me, as it seems like you're talking about the i/o shield not fitting properly on the mobo, which it should, but just not fitting uflly in the case's 4th side. I guess a picture is in order, but 'I thought I took one already previously.
That notation about ammeters was just FYI as to how they work, which design is best.


Gotcha, thanks, was a bit confused.

Some I/O ports are separately shielded, or they may have RF problems with other data bus frequencies. All exposed port shields are grounded via a main neutral bus in the motherboard "sandwich" to the ATX main power plug.
How would I know if it's "exposed?" Just no shield in front of the port or...???

In fluid dynamics, there is a term called "vortex" which can be critical to desired movement of fluids or air. Over the intake of small fans, a slight vacuum is created, which spins naturally, influenced by the Earth's rotation. This pulls in more air, from a desired point. If not enough free space, then any air to the side of the fan is pulled in, which may be warmer than desired.
You can see this effect if some very cold vapor or chemical smoke is dripped over a small fan intake from more than 2 ft. distant! On large bodies of water, they are known as 'whirlpools' which can suck anything down.



Hmmmm.[..

I guess a picture will be needed, but essentially the top fan is pretty much up against the inside top of the case, so it should have air fully coming from the outside only, but you're the expert, and I'll have to get your final approval.

That older design using ducts over a CPU heat sink was so air could be exhausted out of the case. Very few blew fresh air in over the CPU heat sink. But that is just as effective, if not more so.
Should be no issue with wires running under the board, as long as no sharp points could puncture them or they obstruct air flow. Insulation should be rated at 85°C, which is sufficient.


Thanks, then I guess I'll run the wires under the board. Now the real issue is making sure I get enough air from the fans. Do you think these smaller fans would work, or is it just too risky?

The ducts sound interesting, but yeah it seems that those sorts of things are used in premades or older comps :).

I'm just trying to figure out if I'll be able to get enough air to the CPU to remove enough heat.


I hear that CPUs should be given "Static PRessure" fans instead of airflow. Is this true? I notice the static pressure is pretty good on these smaller fans, so it might work with many of them, but nos ure what your thoughts are?

One of the best hobbyist tools is a Dremel Motor Tool. Thousands of bits, grinding and buffing wheels work with it. Some are AC powered, but there also are 'cordless' models. Ideal tool for smoothing rough cuts or even making cuts in thin metal.


I've been meaning to get a dremel but..

1. No place to really use it.

2. Not really sure what I would need it for fully.

I was going to use one to cut parts of my main case, but I ended up being able to drill out rivets and other stuff to get whatg I neeeded.

It seems you need a face shield and overall sometimes the discs break and stuff, so I rather not mess in the danger zone if I don't have to....


Also, in this case, the letters are WAY too tiny for a dremel, which I figured this sharp little deburring tool should do the trick. It shaves metal, so I figure the tiny area between the pressed in logo and the side shoud NOT be that thick are all...



Almost all motherboards since the Socket 370 do have thermal sensors for CPU and other needs. The BIOS monitors them separately. But an add-on kit can monitor them for desktop or front panel display.


My first custom was a core 2 duo, and I don't think it had that, but not sure what socket that was?
The suggestion of using a "mobile" CPU type in that system would reduce unwanted heat, and not draw as much current as a standard CPU. Check if the board supports it. Yes, a standard intent with laptops. The idea of going with one of those "mobile" motherboard kits would be even better. The circuits still require a +3.3 volt, +5 volt and -12 volt source. But that is provided by on-board regulators. Just a raw but pure 12 volts DC may be all the power source needed. Might run out to 15 or 20 amps total. The info provided for the Pico PSU noted a 165 watt power limit, which at 12 volts DC would be almost 14 amps. No less than AWG 14 gauge wire should be used for the feed line! Could run directly off a suitable 12 volt battery, or with the AC adapter.
The only way to measure input current to a motherboard is with a clip-on ammeter to each rail lead, and adding the total for each together.. the +3.3, +5 and +12. That would be the sum total of all current. Each source would have its own sub-total sum.


Yeah, the mobile system honestly looks PERFECT for my needs. I dcould attach a couple fans also, and even use an mSATA port for a hard drive to eliminate the SSD I have!!!

Also, what exactly is the difference between the mini PCIe and the mSATA ports? I hear that they are pbasically the same, but have heard they are used for different things. Some say the wiring isn't exactly the same.

I gues I'm just curious what goes in what? I think I've seen wireless cards go in one of them as well, so that could also be good if I could use a wireless card inside the pcie and a hard drive inside the mSATA :).

It seems that they only need a 12V DC, BUT I do see a 4 pin CPU connector, so I'm wondering if that's supposed to be a secondary power source, and not something that is required1?!?! I don't see why they would require both, sounds extremely silly.


----------------------

The PICO goes to 160W, but max 200W PEAK apparently (not sure how long that peak is meant for).

The power adapter has 192W @ 16A, so that should work then?


----------------------------

The feed line looks beefy, it has a special 4 pin plug also. Hopefully would work...


AS for reading the currents, I guess a picture is needed, because I'm not really sure where it should be.

Id id find a power chart at the top of this forum, so if that is a good indication of what's needed, I might be able to figure out if the pico would work or not.

From what the Pico employee told me, it should work. I'm hoping that is correct though.

The smaller ITX motherboards should have no problem running a current OS version, if 2 GB of system memory is present, and a minimal 1 GHz CPU. But some features may not be 100% compatible, perhaps your model.


Yeah, I'm not too sure.


I really really really like the idea of the fanless DC power motherboards, but the issue is taht these CPUs are terribly slow.


I did a passmark score for the N3160 in the one I was looking for, and it's score was 1260. My i5 2400 was 5600 and my new i5 6600 was 7600 scored.

It's funny because the 4th gen i7s have 11k scores which doesn't make a lot of sense against my 6600, but maybe tahts' just the hyperthreading at work, possibly the fact the newer tech in the 6th gen hasn't fully been used in applications.


I just fee lthe N3160 is so trashy compared to my 2400, and it's a 5 year difference; howerver the N3160 has 89 less TDP than the 2400, so... yeah....

I was thinking of MAYBE getting another cpu that could be less heat producing, and try my fan idea, but I don't know.


Would you say I could test the cpu without fans, without damaging it, or should I always have a fan? I've heard comments that the cpu will fry if there's no heat sink/fan setup, but what if it's just a heat sink? From what I've been hearing, the heat sink is what really plays the biggest part, as the fans are basically all of the same.

So I wonder if.....

1. This heatsink will be enough, possibly with a bunch of fans.

2. Get a bigger heatsink, possibly getting rid of the 4 intake fans on top, but having the side fans to blow air...


Not really sure... What do you think?



Overall, I really do not want to spend 200$ more on another mobo setup, not including extras like a mSATA if I go that route.


I would like to really try what I have on hand, and see if I can make it work, but I'm scared as hell of frying my cpu...

I would have to buy a pico to complete the closed case, as there's no way this seasonic will actually work in this case (sadly spent a bunch of money on this seasonic that is essentially useless, so maybe I'll sell it... or just keep it ias a tester/backup).


Also, I tried booting the computer up once, without an SSD, and got some error to enable something, or something like that, but figured that I needed my SSD and such setup to get it to work correctly, but maybe I don't, since I should be able to get to BIOS, without an OS... right?


Hopefully this mb works LOL... (sajdfaslkjfasd).




So yeah... the next thing is, I guess, to try and make sure that I get enough airflow.... I think I need an ambient thermometer, no clue what to get though... Are you knowledgeable in brands of ambient thermometers?




This is kind of bumming me out, but I must trek!


I feeel that if I can get this sandy bridge to work, then I possibly could get a skylake to be much esier to work with, with much less TDP, even though my 6600 is 65W, which seems to be a lot?

I'm not really sure how big of a difference 1TDP is to the next, but 30 seems like a decent amount, and 89 difference is even crazier.


Ij ust see this 6tdp without any fans and just a heatsink, and feel that I wont be able tot get this 95w tdp sandy to do what I want....


Also, I noticed some welding like marks on the inside of the heatsink, after I removed the fan, so I'm not sure if that's damage.l.. or what.


I've heard there is liquid inside the heat sink pipes, just like in a liquid cooler, but it's encased in metal inside the heat sink.


When I got this unit it seemed to have been usedd, which I confirmed by seeing the used thermal paste syringe.


I didn't take off the fan to checkd inside, but it seems there are a couple of dings, which could be factory, but just adds to the whole notion of used...

Just want to make sure that, with all of this other crap, that the heat sink will be okay... LOL....l




I think my plans of action should be.


1. Get the computer set up and working.


2. Open the case and allow regular fan to blow air onto CPU, and record temp.

3. Then try removing fan and see how much the temp rises.

4. Make sure temps don't get to crazy, then add smaller fans and test their capabilities.


From there I guess I can see if the CPU will be able to handle it or not.

For what it's worth, my current machine sits around 30C, with noctua industrial fans, but the thing that I wonder is if I'll get more air from the higher RPM 40mm fans, and get much cooler air to the CPU, by being closer to the CPU?




It sucks there's so many mixed feelings on this unit, but I really want ot get something that works, and that I could build multiple of these without much issue.

Again, it seems the issues are going to be getting the proper heatsink/fan setup for the CPU, as well as getting proper height ram kits.


Any idea on shorter ram heatsinks? :P.


Going to look up some other heatsinks, but I don't think there are any that can actually compare to this one in size, so I feel I'm pretty much stuck right now.....



I guess I shall start my testing, but would like to hear your advice if using smaller fans instead of the cpu fan will work, and if a larger heat sink would work better, with the smaller fans as well?


I guess, in the end, it really doesn't matter what fans I use, as long as the CPU's temp is below something like 60C, right? That's all I should care about in the end... right?


What temps should I look for in this unit?


I'm also not sure what I will be using that would ramp the temp up.


It seems that some units will idle at a decently high temp regardless if they are really in use, or not, so I will have to check my idle temps.


This computer will NOT be used for any sort of gaming. I would assume the most I am going to do is web browsing. I looking to use it as a testing machine for applications, so I am not sure if that might be an issue as well...


I guess we will find out.........
LakaWaka
Initiate
Initiate
 
Posts: 77
Joined: Sat May 14, 2016 11:41 am

Re: Motherboard grounding questions

Postby LakaWaka » Fri Jul 29, 2016 1:25 am

So I'm going to be a bad boy and double post because my posts keep requiring moderator intervention, so I cannot edit my post until it's approved, which seems to be by you every time, and then you post your reply... so edits are useless :(...

How long until I am "Trusted" enough? Is there a way that you could manually shut it off, since you personally have been interacting with me this entire time?


--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

So anyways I decided to try to start up the new computer to do some testing and it wouldn't start....


The power LED for the mobo was on, so it was getting power, but wasn't sure if the seasonic PSU was working as intended, or not. I went back to the PSU in my main machine which I know works for both my main and this new one, and it still didn't power on...


I ended up realizing, after awhile, that I had placed the power connector in the wrong place.... There is 4 pins above the front panel connector. The weird part is that I also tried placing both the power and reset buttons, so I don'\t know why it iddn't work then. I think I would have noticed if I skpped over one of the rows, but I'm just happy it works.


I had to make sure all of the plugs were properly seated, since I read on another forum that someohne killed their graphics carfd becuase they didn't have their 8 pin plugged all the way and it seemed to short... Don't want that happening.

--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------


My Display didn't turn on, on my main, at first, but after a shutdown and start it appeared.

Not sure if something odd just happened, or if that's a sign of a bigger problem.




I wonder if my pci(e) cards are all seated properly, as it seems my case's rear panel isn't situated properly. I had to struggle with the case during modding, so I did bend it a little by accident, but it looks fine, so I would have to finely check everything, but it's been working fine for 7 months already. I just need to make sure the ports are fully connected.


Sometimes my monitor flickers, and sometimes pages on the internet turn white, until I scroll up, or refresh the page, so I'm not sure if that could be some sort of weird GPU issue, but that's rare when it happens, or why the page would be fine after I refresh it...

There is also some sound card issue, which is rare, but the drivers are 3rd party, so it's hard to tell.

Sometimes my monitor doesn't recognize the connection and will revert to this other computer's display.

So I wonder.....

--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------



Anyways back to the new computer at hand.

I went into BIOS and recorded my temps.


Essentially with the main 92mm fan I had a rise from about 29C to 34C in about 5 mins.

I then turned off the computer and replaced it with the 3-40mm fans.

It rose from 30-34 within 5 or so minutes. From there I changed to the side position which I thought I might use instead of on top, just in case I didn't have enough room to suck air in. It rose a couple of degrees over a few mins, and then I decided to remove the fans and it rose a degree to 38C, which is the max it went to, but I only allowed it to be fanless for a min or so.

Then I decided to put the fans back on top, and it was able to get down to 35C after about 5 mins, and stayed there, so the smaller fans do work, but how well... I'm not sure.


Just as a mention, these measurements were taken outside of a case, in the open air. There was no fan turn on inside my room, but there was cool air entering the room from an AC.

I am not sure how hot the case will make it, or how much the temp will rise outside of the BIOS, but it doesn't seem that the 40mm fans are that bad, and 3 of them compared to 4+ that I want to use, it should work.

The only thing is that the 40mm fans were maxxing out about 4500RPM, and the 92mm fan was around 1300RPM with a max speed of 2200RPM, so it wasn't close to max, and was able to get about the same as 3 small ones.


--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

However, after trying to place the mobo directly under the creative logo, I realized that my case... will actually close.... I don't know how I didn't try that another time, but it works....


The thing is... Is it possible for the fan to be directly touching the top of the case, and not have direct airflow? Some of the fan will be exposed to the air through the logo, but most wont. Will this damage the fan, or as long as there is some air, it's okay?

Is the damage from overheating, or an issue with the blades? I never really fully understood what the problem was from blocking a fan's intake. I always thought that it might have to do with blocking the fans blades and not letting them do their thing(spin), but I have accidentally covered my main case's front where I modded the fans, and the fans still worked, so I'm assuming the issue is overheating due to no airflow?

... If I have some airflow... Will it work then?

--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

I rather use the fan that came with the heatsink if the air flow will be enough. Then I will be able to get direct airflow from the top, but also will still be using the 3 fans on the side of the case, sadly not as close as originally planned, but should still work nicely. I can also have 2 sets of fans, one next to the outside, and some closer to the mobo so that the airflow is directed appropriately.

--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------


It's funny that I figured it the case actually does fit... After doing all the tests with the fan.

But I'm glad the computer still boots up and all that jazz as well...

--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

The only problem I see now is that bot sides of the i/o shield are NOT touching the case.

The thing is.... People test their computers outside of their cases to make sure it works, so it's not a huge deal for this RF noise it seems??

I guess for long extended periods of use we should shield the computers then?

This I/O shield business is annoying the crap out of me, especially now that I have to worry about 2 sides.... I really want to work this slab of metal.

Does the shield have to allow the ports to stick out, essentially creating a seal, or are we allowed to just cover up the back and have holes?


I guess this similarly would be asked about PCI slots that don't have covers... Should we cover all open slots? I have a few open for airflow, but if this is bad, then I will cover.

But overall you cannot cover 100% of the outside, some people have dust filter covers as well, even my GPU has an exhaust, so I'm curious how exactly the RF noise gets in, and affects the system, and if there is just something special about the I/O ports needing a shield compared to other parts of entry into the case?

That's not even mentioning open fronts with fan slots, or back fan slots, etc.???

--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------


So I guess at this point we can move forward with the original setup if I can pull air from a small portion of the holes that I create on the top. I could just cut a giant hole, but that's no fun, I want this to look nice :).

Although, this is the first case test, with a high TDP cpu, so maybe I should just try to just get it to work and not think about aesthetics, even though I really just want this to look as good as it performs.

--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

Also, I thought I lost my standoffs, but found them. I think I could screw some nuts onto the standoff to increase the size to match the heat sink feet's length. I will go get a drill bit for it and drill some holes when I'm sure I found the perfect spot. I don't care much for what the bottom would look like, so I could mess up and redrill holes if needed.

In reality, would you say it would be a smarter choice to drill a bunch of holes on the bottom of the case for airflow? i'm thinking of having some leds on the bottom, so I figured some airholes would be good. I'm just not sure if the leds would cover the holes fully, or what, but still an idea overall. The underside doesn't matter as much as the top :).

--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------


I guess that's it, so basically if I can proceed then I will drill the holes, get the standoffs setup, and then try to cut these letters out with this scraping tools which I'm assuming will work as I need it, even though it's made for a different use.

I haven't seen any tools that do what I want though.......

Then, I will check out the guide at the top of this forum, and see if the Pico will work for all of my needs, and hopefully it fits the case (I don't see how it would not.


So overall so far so good, hopefully it continues on this path.
LakaWaka
Initiate
Initiate
 
Posts: 77
Joined: Sat May 14, 2016 11:41 am

Re: Motherboard grounding questions

Postby Karlsweldt » Fri Jul 29, 2016 6:48 am

Try to omit a lot of free spaces in your postings. One line free between notations is sufficient. Saves page length and data storage needs. Your 'turning point' for postings should be close. Our Administrators had to increase the new member post counts to deter spammers.

Mistook your note about the I/O shield not fitting, sorry. If smaller than the opening, then no real issue. If it fits mostly but has open space, then a metallic tape could cover the opening, if in purist thought.
Those tabs by each port opening are only meant to contact the face of each port.. not wrap around them. Only if there is a large tab at the top of an opening, that was meant to pass above that specific port.
Likely only the keyboard/mouse ports and the LAN ports would have integral covers (shields) as their use is considered priority, and interference with data signals would not be good.

If a fan intake does not have sufficient free air space, the fan may spin faster than intended.. and not move the desired amount of air. Could cause premature failure issues. So-called 'static pressure' just means that there is more motive force on air or fluid than when stagnant, to create a flow.

Dremel bits do offer tiny carbide rasps that are about 1/8" diameter.
The Socket LGA775 was the first design for the Core2Duo. https://en.wikipedia.org/wiki/CPU_socket
Each slot or socket for device cards is intended only for that type, contact pinout and keying are different. Not interchangeable.
If there is a 4-pin socket near the CPU, that requires a separate lead set from the PSU, for +12 volts. Motherboard traces are not beefy enough to handle all the current needs of a CPU. Separate regulators are near that power socket, for the CPU needs.
"Peak" power refers to the absolute max provided or needed, but not sustained. Otherwise, the PSU could fail quickly. Normally, there is an overload sensor that shuts down the supply if the maximum level is noted for more than a second or two.
AC generators have a "peak" and "running" power rating, the "peak" allows motors to start with a brief input of higher current than normal. A "reserve" limit. But the "running" load cannot be exceeded for long, or a circuit breaker would trip.

As to "fanless" CPU types, that only infers that the heat sink does not have a mounted fan. The cooling fan is elsewhere.
The "fanless" heat sink class may have a thermal-siphon design, where a liquid with low boiling point vaporizes and moves upward to where that heat is dispersed, then the gas condenses to liquid and renews the cycle. Or the "fanless" design may have an extended aluminum plate that transfers heat to an area where it can be cooled by air or contact with another metal plate.
I would not run any modern CPU without proper cooling needs, as it could overheat in seconds.. and possibly self-destruct. A few seconds with just a heat sink on the CPU may be acceptable for testing, but not for extended run time.

When a computer starts up, a POST process (Power On Self Test) checks that all is in order as to devices and assignments. If a problem, then the POST would halt the process and display an error. The BIOS pages should reflect exactly how the system is put together. After the POST process, then a hand-over or bootstrap allows the operating system to take over. If no suitable OS is found, or a bootable device, then the error code.
As to a thermometer, one of those indoor/outdoor types with remote lead may work well. The remote probe is small enough to place in needed spaces. A dab of hot-melt glue could keep it in place. Thermal probe sensors can be bought easily, for many temperature ranges.
Also there is the IR/Laser thermal meter, non-contact. Micro Temp is one brand. http://www.microtempusa.com/
F@H.. to solve mankind's maladies.. in our lifetimes!
Karlsweldt
Mobo-fu Master
Mobo-fu Master
 
Posts: 20687
Joined: Wed Nov 12, 2003 11:57 am
Location: 07438

Re: Motherboard grounding questions

Postby LakaWaka » Fri Jul 29, 2016 4:50 pm

Mistook your note about the I/O shield not fitting, sorry. If smaller than the opening, then no real issue. If it fits mostly but has open space, then a metallic tape could cover the opening, if in purist thought.
Those tabs by each port opening are only meant to contact the face of each port.. not wrap around them. Only if there is a large tab at the top of an opening, that was meant to pass above that specific port.
Likely only the keyboard/mouse ports and the LAN ports would have integral covers (shields) as their use is considered priority, and interference with data signals would not be good.


It's no problem at all :).
There is some space on the sides (need to take a pic), but the shield is also not very stable, and wont really stay on..... Metallic tape, I'll look into that. I would like to use the metal that came with it.
There are some with a tab on top.
I'm not sure how I could secure this shield properly without it going crazy. Maybe I could cut a hole in the metal and shove the I/O shield in there, instead of trying ot makeshift an i/o shield from the metal slab I have already? It seems the tabs are specific, so maybe I do need the shield as is, or...????

If a fan intake does not have sufficient free air space, the fan may spin faster than intended.. and not move the desired amount of air. Could cause premature failure issues. So-called 'static pressure' just means that there is more motive force on air or fluid than when stagnant, to create a flow.


So would it be a bad idea to do what I want? IT seems taht the only issue is extra spinning and possibly failure issues.
On a lot of cases, including my main one, there is a fan cover that covers part of the fan, as well as the front of the case that covered most of the fan, besides a few slits, and 2 side sections with holes.
The thing is that the front of the case had probably 3-4" of space to allow the air to come in through those tiny areas.
The smaller fans would probably have 1/4" or so of room to collect air, but I really do not know how powerful they will be, so I figure the original fan is what I should go with.
What temps should I be looking for max for both the cpu and motherboard sensors?
Again, in the end, the only thing that matters is if there is proper cooling... right?
I'll test the fan out, hopefully it works as intended.
I'm also thinking of maybe incorporating some designs around the creative logo to allow for additional airflow into the case :)... I think that will help out both in looks (hopefully) as well as airflow.

Dremel bits do offer tiny carbide rasps that are about 1/8" diameter.
The Socket LGA775 was the first design for the Core2Duo. https://en.wikipedia.org/wiki/CPU_socket
Each slot or socket for device cards is intended only for that type, contact pinout and keying are different. Not interchangeable.
If there is a 4-pin socket near the CPU, that requires a separate lead set from the PSU, for +12 volts. Motherboard traces are not beefy enough to handle all the current needs of a CPU. Separate regulators are near that power socket, for the CPU needs.
"Peak" power refers to the absolute max provided or needed, but not sustained. Otherwise, the PSU could fail quickly. Normally, there is an overload sensor that shuts down the supply if the maximum level is noted for more than a second or two.
AC generators have a "peak" and "running" power rating, the "peak" allows motors to start with a brief input of higher current than normal. A "reserve" limit. But the "running" load cannot be exceeded for long, or a circuit breaker would trip.


The area I am working with is super tiny, so even 1/8" is too big for some of the parts, I figured this scraper tool should do the job nicely, but not sure what you think?
For some reason I thought LGA775 was referring to the CPU, and not the socket on the mobo. I've read there are different designations, but maybe I'm confused.
So basically you would still need external power to the CPU then... wtf? How do laptops do it then, with only the 1 DC input???
I'm not sure if there's a sensor, but I hear that there is no regulation on the 12V rail so it could catch on fire if you overload it..... I hear peak can only be sustained for a second or so before blowing, so... yeah....


I would not run any modern CPU without proper cooling needs, as it could overheat in seconds.. and possibly self-destruct. A few seconds with just a heat sink on the CPU may be acceptable for testing, but not for extended run time.


I thought that all modern CPUS are equipped with an auto-shutoff when reaching too hot of temps, as well as the severe throttling?
I had only left the heat sink exposed for about a minute but the temps didn't go crazy.. I assume that's just because I was in BIOS, and not running anything?
If the temps stay low, is it okay with just a heatsink, or does it still need fans to blow air for other components besides the CPU?
I don't think I damaged anything, hopefully not.. :(

Also there is the IR/Laser thermal meter, non-contact. Micro Temp is one brand. http://www.microtempusa.com/

I'll check into this, thanks. I think I'll be okay with the onboard thermometers, since now I think going the original route will work the best.


I also would really like to get into some liquid cooling. Not sure if you have experience with that, but I really have liked it for awhile... Would be fun to rig one of these with some liquid cooling :).
LakaWaka
Initiate
Initiate
 
Posts: 77
Joined: Sat May 14, 2016 11:41 am

Re: Motherboard grounding questions

Postby Karlsweldt » Sat Jul 30, 2016 5:56 am

The opening for the I/O shield is more or less standard on all cases. But there are some designs that use a non-standard opening, intended for low-profile motherboards, as in rack systems. Small "nibs" on the shield lock it into place in the opening, if both are of standard design.
As noted, that extra 4-pin socket is for the extra current needs of the CPU. The system won't work if that socket is not fed power from the PSU.

Fan covers or shields may be present, but are intended to 'collect' or draw air from more favorable areas.. away from the heat sink exhaust. Small fans, under 3", may be desired in some uses.. but for a regular PC case, one larger fan could be more efficient than 3 or 4 of the smaller fans.. and need less current.
The more wattage (waste heat) given off by a CPU, the more cooling air is needed. Every CPU has a specs chart, noting current and voltage needs, nominal operation temperature, and maximum temperature. For average use, a 65°C limit is acceptable for a CPU monitor alarm threshold.

Whatever tool does what is needed, then good. Some tools are special-use only, but may work for other needs. Dental picks find wide use in a lot of other needs!
Laptops are not designed to use a standard CPU type. Only the 'mobile' type. Otherwise, excessive heat issues, rapid battery drain. And maybe melt-down of some components. 99% of PSU models do have regulation on all rails, to maintain the desired voltage and current. The -12 volt rail is not as accurately regulated as others, as it is only in non-critical use. Servo motors that require 12 volts may not spin as desired, and create data access timing problems.
"Heat soak" is the worst enemy of electronic components, next to excess voltage. If excess heat is not dissipated before shutting down, sensitive transistors and such could suffer permanent damage. Same if insufficient cooling air flow while in operation. All modern CPU types do have integral monitors to throttle down if temperature extremes are neared, or shut down entirely.
Liquid cooling is the most efficient, moving unwanted heat from one area to another. But can be bulky. And if the water is not kept sanitized, can harbor bacteria and algae.
F@H.. to solve mankind's maladies.. in our lifetimes!
Karlsweldt
Mobo-fu Master
Mobo-fu Master
 
Posts: 20687
Joined: Wed Nov 12, 2003 11:57 am
Location: 07438

Re: Motherboard grounding questions

Postby LakaWaka » Sat Jul 30, 2016 5:39 pm

Karlsweldt wrote:The opening for the I/O shield is more or less standard on all cases. But there are some designs that use a non-standard opening, intended for low-profile motherboards, as in rack systems. Small "nibs" on the shield lock it into place in the opening, if both are of standard design.
As noted, that extra 4-pin socket is for the extra current needs of the CPU. The system won't work if that socket is not fed power from the PSU.

Fan covers or shields may be present, but are intended to 'collect' or draw air from more favorable areas.. away from the heat sink exhaust. Small fans, under 3", may be desired in some uses.. but for a regular PC case, one larger fan could be more efficient than 3 or 4 of the smaller fans.. and need less current.
The more wattage (waste heat) given off by a CPU, the more cooling air is needed. Every CPU has a specs chart, noting current and voltage needs, nominal operation temperature, and maximum temperature. For average use, a 65°C limit is acceptable for a CPU monitor alarm threshold.

Whatever tool does what is needed, then good. Some tools are special-use only, but may work for other needs. Dental picks find wide use in a lot of other needs!
Laptops are not designed to use a standard CPU type. Only the 'mobile' type. Otherwise, excessive heat issues, rapid battery drain. And maybe melt-down of some components. 99% of PSU models do have regulation on all rails, to maintain the desired voltage and current. The -12 volt rail is not as accurately regulated as others, as it is only in non-critical use. Servo motors that require 12 volts may not spin as desired, and create data access timing problems.
"Heat soak" is the worst enemy of electronic components, next to excess voltage. If excess heat is not dissipated before shutting down, sensitive transistors and such could suffer permanent damage. Same if insufficient cooling air flow while in operation. All modern CPU types do have integral monitors to throttle down if temperature extremes are neared, or shut down entirely.
Liquid cooling is the most efficient, moving unwanted heat from one area to another. But can be bulky. And if the water is not kept sanitized, can harbor bacteria and algae.



There are some tiny things on the shield, but I think the issue is there is no back to catch and hold the shield, so it's flopping all over.
I don't know why there would be a need for the extra 4pin cpu connector, if they are using a 12V DC power adapter. Mobile Motherboards used in Laptops don't have additional power, so why would these itx desktop ones need it?

Currently it's set @ 70*C, but I'll set it down to make usre. I'll check olut the specs chart.
Yeah, I cannot find a proper tool, but this one seems like it could work... we shall see.
Laptops do use mobile CPUS, but they are changeable. Some are integrated into laptops are well though.

It seemed taht the PICO doesn't have regulation on the +12V rail, but yes on the 5 and 3.3v.. Why not on the 12 Idk.. sounds like a scary idea.
Heat Soak only occurs when you shut the computer down, or if the air is trapped inside? Not sure if it's instant once the machine shuts down... or what...
Liquid cooling seems interesting, but not sure about cleaning and stuff. I would like t go the "no chemical" route and just do distilled water, but maybe I would add Coolant and stuff later.... Just not sure how messy it gets, and how "friendly" these chemicals are.


I'm kind of bumming out, because I'm not sure what I'm going to be doing. I feel that this CPU might just be too strong, and maybe I should build a mini ITX build in a normal case, possibly with some water cooling instead... But that becomes a whole other project an a lot more money.

So far I have a seasonic PSU that wsan't really cheap, hanging around, and then would have an extra noctua heatsink hanging artound, but could reuse that in the later projct, as wel as the seasonic.


So... Idk what I should do lol. I do want to finish this project, but feel it just deserves some parts that wont produce as much heat, to be put in the tiny enclosure. Plus, maybe some shoter ram next time, as my DDR4 Corsair ram in my main just barely is above the ramsticks, whereas these g.skill ripjawsx are tallllll.

What would you do in my case? Would you buy a case for the sandy and just get it going and not sitting around not in use, or would you try to accomplish this project in the mini case?
If I go for a new case, would you think to want to watercool as a fun project, or just leave it as is?
I really enjoy putting computers together, building, etc, but don't want to waste time on things that wont work or could damage components/waste money..
Thanks a lot.
LakaWaka
Initiate
Initiate
 
Posts: 77
Joined: Sat May 14, 2016 11:41 am

Re: Motherboard grounding questions

Postby Karlsweldt » Sun Jul 31, 2016 5:30 am

Look at that I/O shield lip where it mounts into the opening. May be some tiny "pimples" that act as latches, or the lip may be rolled in a manner as to snap into place. Either way, it is intended to be secure in the opening.
As to water cooling, plain water will become contaminated over time. An anti-bacterial agent that is compatible to the workings is advised. And adding a bit of safe anti-freeze is good, as it aids heat transfer. The best kind is that used in the RV industry, propylene glycol. Found in some food products, considered only "slightly" toxic. Would I water-cool a system? If it is a super-fast hard-working system in a warm place, definitely. But I have no need for "supercharged" data chompers.

The +3.3 volt and +5 volt sources are critical to design needs, and must be kept as true as practicable. Timing circuits and 'gates' (data sorters) would not be accurate if the voltages were not exact. The +12 volt rail should be regulated also, but its main use is servo motor power and fan power, plus secondary CPU power. Not a critical need.
A 'mobile' class CPU is never in full operation continuously. It throttles back or idles when not working. A regular CPU is always fully active, and requires a lot of extra power.. thus the extra socket for its power needs.

For my thoughts, I would use a case that was favorable to the design and needs of the motherboard and features. But a smaller case that is almost proper could be used as a test setup.
F@H.. to solve mankind's maladies.. in our lifetimes!
Karlsweldt
Mobo-fu Master
Mobo-fu Master
 
Posts: 20687
Joined: Wed Nov 12, 2003 11:57 am
Location: 07438

Re: Motherboard grounding questions

Postby LakaWaka » Fri Aug 05, 2016 5:02 pm

Karlsweldt wrote:Look at that I/O shield lip where it mounts into the opening. May be some tiny "pimples" that act as latches, or the lip may be rolled in a manner as to snap into place. Either way, it is intended to be secure in the opening.
As to water cooling, plain water will become contaminated over time. An anti-bacterial agent that is compatible to the workings is advised. And adding a bit of safe anti-freeze is good, as it aids heat transfer. The best kind is that used in the RV industry, propylene glycol. Found in some food products, considered only "slightly" toxic. Would I water-cool a system? If it is a super-fast hard-working system in a warm place, definitely. But I have no need for "supercharged" data chompers.

The +3.3 volt and +5 volt sources are critical to design needs, and must be kept as true as practicable. Timing circuits and 'gates' (data sorters) would not be accurate if the voltages were not exact. The +12 volt rail should be regulated also, but its main use is servo motor power and fan power, plus secondary CPU power. Not a critical need.
A 'mobile' class CPU is never in full operation continuously. It throttles back or idles when not working. A regular CPU is always fully active, and requires a lot of extra power.. thus the extra socket for its power needs.

For my thoughts, I would use a case that was favorable to the design and needs of the motherboard and features. But a smaller case that is almost proper could be used as a test setup.


Yeah, I think it's just that I don't have the proper enclosure for the shield.
So is it just algae that contaminates it (without light?), I've also heard lubricants and maybe stuff in the rad (which you should have cleaned out). So PG is "anti-freeze" (we are supposed to use coolant, right)??
I've heard people using PG in food grade struff, so that's interesting it's actually toxic. I wouldn't really care to use a coolant, but would not like crap in my loop...

If I clean my loop every now and then would taht be okay? Rather have something that keeps everything in check, because tampering with loop = possible issues. I'm not looking to overclock or do crazy stuff, I just redally like building computers and I've wanted to do a water cooled loop for a long, long time. Just not looking to introduce toxic stuff not only around myself, but aroudn my components just in case. I've heard of AIO leaks corroding parts... Not looking to do that.

I think that your suggestion would be the best one. Buy a case for now, get to using this computer, and then I can worry about other stuff later... PErsonally I really do want to do WCing in this build, but not sure what case is suitable for that. I wanted Mini-ITX, but I think I would like more room in my case. Thoughts?


I'm looking to possibly join another forum that specializes in OVerclocking. I've seen Overclocked.net, any comments about that site? Thanks a lot for all of your time.
LakaWaka
Initiate
Initiate
 
Posts: 77
Joined: Sat May 14, 2016 11:41 am

PreviousNext

Return to General

Who is online

Users browsing this forum: No registered users and 0 guests