Post by ShrimpBrime on Jun 3, 2018 23:46:58 GMT -5
Imported Information and Pictures from another site.
Motherboard M5A78L-M LX PLUS 125w TDP max Cpu.
www.asus.com/Motherboards/M5A78LM_LX_PLUS/
TEC # 12715 Qmax of 136w - Temp difference (from cold side to hot) = Tmax 70c - 15v Max current. Max operating temp of 90c.
www.thermonamic.com/TEC1-12715-English.PDF
OK - now what I intend to do is example how a TEC can be used with newer technology and how we can get some freezing temps and see how that can effect the processors over all speed vs the voltage being used and perhaps some super high clocks past 5ghz. This will require shutting down some cores, but that's ok in my book. Been OCing this way for years already!
Any how, we are going to have a look at FX-8350 and FX-4300 processors. Both will be TEC cooled and the TEC will be cooled by liquid. I may cheat a little tiny bit and run tap to drain which is typically cooler than ambient air cooled water loops, so we'll have an advantage with temps. Not much but enough.
This is for example only. Experimentation purposes only. My opinions of any type of cooling and overclocking may not be the same views of others in this forum. Please do not attempt to use TECs on your own system. ClassicPlatforms/LowBuck PC are not responsible for the use of TECs on your system. This is a fire hazard. It's not a joke. Nor Admin or anyone here is responsible for your own actions if you so wish to chance of fire.
Thermal Electric Coolers can be Dangerous! Please keep in mind this rig set up is closely watched and fully and completely shut down after each time I use it. TECs can cause fires especially when under rated gauge wire is used. Most power supplies cannot handle a TEC. I have my power supply modified and is closely watched while the TEC is in use. I do not recommend people to use TECs with a standard PC power supply. Nor does this forum or any one who owns, runs and or maintains this forum condones use of TECs without knowledge of TECs. Always do research when trying to do something right. What you see here is for my own personal testing experience and do not expect people to think that what I do is ok.
Next few posts are reserved for data and pics.
This will be Low cost system straight out of the box. One high end Cpu to show motherboard durability and one low end Cpu for the "cost" effective factor. Plus some nifty newer TEC data. No body uses TECs, they really don't work super well, BUT if applied in the right manner, I believe they could have an edge over regular air cooling and liquid cooling systems.
NOTES
At around 1.7-1.9ghz I discovered UCbench to bring temps really close to un frozen. I do not like freezing and un freezing the cpu in this fashion, so again, my goal is to stay frozen 0c or less to be ideal this way less chance of condensation build up.
UCBench while loading the cpu up to four cores was running good temp between -4 and -6c. So from this I decided to DL 3DMark05 to sart. It should use up to 2 cores, I'll probably use a few different 3D benchamrks along the way to determine how they would run in a real world environment for benching.
I plan to use these benchmarks randomly as you would decide what to bench at this time ect. No real laid out plan here other than showing board and cpu durance under freezing conditions more often than just a DICE run on a particular day. In other words, you could use this probably more often than DICE and it'll be a little more cost effective.
Additional Comment: 2/28/2014
OK A couple Few Days in. Unlocked Cores at 1.1640v to Hexa (6) Core and found not the slightest bit of decrease in temperatures. A couple of theories, One is that even though 2 cores are off, they may still be generating heat or perhaps the cpu cache on the last 2 cores still generate heat. As if 1/2 the cpu needs to be shut off in order for the heat output to be dropped significantly enough to notice with our TEC cooling. More testing on a more stable cooling apparatus for better controlled temps would be required to better asses how this comes about. From running 4 similar tests, Temp results turned out to be the same within what I would consider a margin of error.
ADDITIONAL COMMENT: 3/2/2014
Another Boot up. It seems more and more voltage is needed to get it up there. But also in the same path the Cpu seems to require a lot less voltage at any clock speed I'd run on a regular water loop. Not sure it's going to stay frozen much past 5ghz. Anything but PiMod seems to get temps close to un-frozen. Which is OK, we will likely get some cpu-z screen shots or something as I try and push for a max clock speed.
The most interesting part and probably most fun for me is that as I add pictures and update, it's all under a freezing cold TEC. :smt023
ADDITIONAL COMMENT: 3/15/2014
I have posted the FX-4300 only a couple of times. From what I take notice, it acts very similar to the FX-8350 as a quad core at roughly the same voltages and clocks speeds (so far around stock clocks)
This processor was BNIB and posted bios the first time at -5c, and that is frozen. So I know the Cpu has a high voltage for this clock speed and lowered it down as far as 1.260v and have done some minor benching on it. Temps where reporting at -11c which I believe is right where it should be and loads running cinebench reached near -1c.
The water block Used to cool the TEC.
www.swiftech.com/apogeegtz.aspx
TEC # 12715 Qmax of 136w - Temp difference (from cold side to hot) = Tmax 70c - 15v Max current. Max operating temp of 90c.
www.thermonamic.com/TEC1-12715-English.PDF
Cpu FX-8350 Working Pstats I'll be using.
# of P-States 7
P-State FID 0x1A - VID 0x0A - IDD 13 (21.00x - 1.425 V)
P-State FID 0x19 - VID 0x0B - IDD 13 (20.50x - 1.412 V)
P-State FID 0x18 - VID 0x0E - IDD 12 (20.00x - 1.375 V)
P-State FID 0x12 - VID 0x16 - IDD 10 (17.00x - 1.275 V)
P-State FID 0xC - VID 0x1F - IDD 8 (14.00x - 1.162 V)
P-State FID 0x5 - VID 0x29 - IDD 5 (10.50x - 1.037 V)
P-State FID 0x10C - VID 0x33 - IDD 4 (7.00x - 0.912 V)
Motherboard M5A78L-M LX PLUS 125w TDP max Cpu.
43$ after rebates at Newegg
www.newegg.com/Product/Product.a ... 6813131795
PSU 1000w Antec CP series with Fan upgrade to serious CFM to keep the PSU cool while running the TEC.
store.antec.com/Product/power_su ... 100-8.aspx
+3.3V@28A, +5V@30A, +12V1@25A, +12V2@30A, +12V3@30A, +12V4@30A, -12V@0.6A, +5VSB@3.0A
Motherboard M5A78L-M LX PLUS 125w TDP max Cpu.
www.asus.com/Motherboards/M5A78LM_LX_PLUS/
TEC # 12715 Qmax of 136w - Temp difference (from cold side to hot) = Tmax 70c - 15v Max current. Max operating temp of 90c.
www.thermonamic.com/TEC1-12715-English.PDF
OK - now what I intend to do is example how a TEC can be used with newer technology and how we can get some freezing temps and see how that can effect the processors over all speed vs the voltage being used and perhaps some super high clocks past 5ghz. This will require shutting down some cores, but that's ok in my book. Been OCing this way for years already!
Any how, we are going to have a look at FX-8350 and FX-4300 processors. Both will be TEC cooled and the TEC will be cooled by liquid. I may cheat a little tiny bit and run tap to drain which is typically cooler than ambient air cooled water loops, so we'll have an advantage with temps. Not much but enough.
This is for example only. Experimentation purposes only. My opinions of any type of cooling and overclocking may not be the same views of others in this forum. Please do not attempt to use TECs on your own system. ClassicPlatforms/LowBuck PC are not responsible for the use of TECs on your system. This is a fire hazard. It's not a joke. Nor Admin or anyone here is responsible for your own actions if you so wish to chance of fire.
Thermal Electric Coolers can be Dangerous! Please keep in mind this rig set up is closely watched and fully and completely shut down after each time I use it. TECs can cause fires especially when under rated gauge wire is used. Most power supplies cannot handle a TEC. I have my power supply modified and is closely watched while the TEC is in use. I do not recommend people to use TECs with a standard PC power supply. Nor does this forum or any one who owns, runs and or maintains this forum condones use of TECs without knowledge of TECs. Always do research when trying to do something right. What you see here is for my own personal testing experience and do not expect people to think that what I do is ok.
Next few posts are reserved for data and pics.
This will be Low cost system straight out of the box. One high end Cpu to show motherboard durability and one low end Cpu for the "cost" effective factor. Plus some nifty newer TEC data. No body uses TECs, they really don't work super well, BUT if applied in the right manner, I believe they could have an edge over regular air cooling and liquid cooling systems.
NOTES
At around 1.7-1.9ghz I discovered UCbench to bring temps really close to un frozen. I do not like freezing and un freezing the cpu in this fashion, so again, my goal is to stay frozen 0c or less to be ideal this way less chance of condensation build up.
UCBench while loading the cpu up to four cores was running good temp between -4 and -6c. So from this I decided to DL 3DMark05 to sart. It should use up to 2 cores, I'll probably use a few different 3D benchamrks along the way to determine how they would run in a real world environment for benching.
I plan to use these benchmarks randomly as you would decide what to bench at this time ect. No real laid out plan here other than showing board and cpu durance under freezing conditions more often than just a DICE run on a particular day. In other words, you could use this probably more often than DICE and it'll be a little more cost effective.
Additional Comment: 2/28/2014
OK A couple Few Days in. Unlocked Cores at 1.1640v to Hexa (6) Core and found not the slightest bit of decrease in temperatures. A couple of theories, One is that even though 2 cores are off, they may still be generating heat or perhaps the cpu cache on the last 2 cores still generate heat. As if 1/2 the cpu needs to be shut off in order for the heat output to be dropped significantly enough to notice with our TEC cooling. More testing on a more stable cooling apparatus for better controlled temps would be required to better asses how this comes about. From running 4 similar tests, Temp results turned out to be the same within what I would consider a margin of error.
ADDITIONAL COMMENT: 3/2/2014
Another Boot up. It seems more and more voltage is needed to get it up there. But also in the same path the Cpu seems to require a lot less voltage at any clock speed I'd run on a regular water loop. Not sure it's going to stay frozen much past 5ghz. Anything but PiMod seems to get temps close to un-frozen. Which is OK, we will likely get some cpu-z screen shots or something as I try and push for a max clock speed.
The most interesting part and probably most fun for me is that as I add pictures and update, it's all under a freezing cold TEC. :smt023
ADDITIONAL COMMENT: 3/15/2014
I have posted the FX-4300 only a couple of times. From what I take notice, it acts very similar to the FX-8350 as a quad core at roughly the same voltages and clocks speeds (so far around stock clocks)
This processor was BNIB and posted bios the first time at -5c, and that is frozen. So I know the Cpu has a high voltage for this clock speed and lowered it down as far as 1.260v and have done some minor benching on it. Temps where reporting at -11c which I believe is right where it should be and loads running cinebench reached near -1c.
The water block Used to cool the TEC.
www.swiftech.com/apogeegtz.aspx
TEC # 12715 Qmax of 136w - Temp difference (from cold side to hot) = Tmax 70c - 15v Max current. Max operating temp of 90c.
www.thermonamic.com/TEC1-12715-English.PDF
Cpu FX-8350 Working Pstats I'll be using.
# of P-States 7
P-State FID 0x1A - VID 0x0A - IDD 13 (21.00x - 1.425 V)
P-State FID 0x19 - VID 0x0B - IDD 13 (20.50x - 1.412 V)
P-State FID 0x18 - VID 0x0E - IDD 12 (20.00x - 1.375 V)
P-State FID 0x12 - VID 0x16 - IDD 10 (17.00x - 1.275 V)
P-State FID 0xC - VID 0x1F - IDD 8 (14.00x - 1.162 V)
P-State FID 0x5 - VID 0x29 - IDD 5 (10.50x - 1.037 V)
P-State FID 0x10C - VID 0x33 - IDD 4 (7.00x - 0.912 V)
Motherboard M5A78L-M LX PLUS 125w TDP max Cpu.
43$ after rebates at Newegg
www.newegg.com/Product/Product.a ... 6813131795
PSU 1000w Antec CP series with Fan upgrade to serious CFM to keep the PSU cool while running the TEC.
store.antec.com/Product/power_su ... 100-8.aspx
+3.3V@28A, +5V@30A, +12V1@25A, +12V2@30A, +12V3@30A, +12V4@30A, -12V@0.6A, +5VSB@3.0A
You purchase shipping and consider it done. I got one from Storm on the way I'm going to re-lid for him actually. Not sure who de-lidded his processor, I thought all the Phenom chips where soldered, so it'll be interesting to see when it arrives. He's having issues with heat spikes and the plate is a tad crooked on the PCB. I'm going to lap the core a touch and the edges of the underside of the plate to compensate. Hopefully it's still a working chip. Heat and voltage kills them pretty quickly. I'll remove the plate and test naked to be sure of functionality before I try putting the plate back on.
