Overclock.net › Forums › Cooling › Specialized Cooling › Peltiers / TEC › Peltier chilling system
New Posts  All Forums:Forum Nav:

Peltier chilling system - Page 8

post #71 of 167
I've got 2 x 18w DDC2.0's with 1.8gpm flow rate in my current loop, I'm happy with that, but it would be nice to have more channels and get the C/W losses a bit better, but it's very very good as is.
post #72 of 167
Thread Starter 
Quote:
Originally Posted by Puck View Post

Yea, Foxrena is making these blocks in a "Standard" version with fewer wider channels but low restriction, and if you ask he will do a higher performance version with a lot more(much smaller) channels recommended for people with multiple pumps or monsters like the 24v D5 Strong.

I wonder how deep the channels are. Theoretically you should be able to get low restriction and high surface area (high performance) with a similar design that iv'e shown here. You just need the same flow through area as the inlet or ID of the tubing itself. I surpassed that with this design by like .02". Theoretically this design will flow slower in the channels, than the rate of water coming into the inlet. Enough theoretical though. I'll be able to test as early as mid-next week. Need to order some acrylic for the top and some #1-64 hardware yet.

I still don't know what ill be doing for a cold water-block yet. Instinct tells me it should be an entirely different design than the hot-side. For now i'm going to keep the current cold blocks, and test these hot blocks, knowing where i was as a baseline with the old hot blocks. Once i get a baseline for how the new hot-block improves the heat removal, ill try another one with the cold side. Perhaps this would also chill effectively too. Although i'm thinking we would want far more surface area in the cold side than the hot side to pick up and bit of that cold.
post #73 of 167
Thread Starter 
Also i have the MPC-655 and i stalled that thing out at full pump speed. It's my issue though. My cold-side water blocks only run .1875 ID tubing. Its way too small. I need to add parallel circuits to it.
post #74 of 167
Quote:
Originally Posted by mrkubanftw View Post

I wonder how deep the channels are. Theoretically you should be able to get low restriction and high surface area (high performance) with a similar design that iv'e shown here. You just need the same flow through area as the inlet or ID of the tubing itself. I surpassed that with this design by like .02". Theoretically this design will flow slower in the channels, than the rate of water coming into the inlet. Enough theoretical though. I'll be able to test as early as mid-next week. Need to order some acrylic for the top and some #1-64 hardware yet.

I still don't know what ill be doing for a cold water-block yet. Instinct tells me it should be an entirely different design than the hot-side. For now i'm going to keep the current cold blocks, and test these hot blocks, knowing where i was as a baseline with the old hot blocks. Once i get a baseline for how the new hot-block improves the heat removal, ill try another one with the cold side. Perhaps this would also chill effectively too. Although i'm thinking we would want far more surface area in the cold side than the hot side to pick up and bit of that cold.

Here ya go smile.gif
Quote:
Originally Posted by foxrena View Post



Done Cutting water channels for two blocks, these are 0.02'' wide by 0.07'' deep channels, 58 channels in total!!
The area of channels is so much larger than normal water blocks. It takes 50 minutes to cut one block. 0.02'' with channel is a higher restriction channel design. For low restriction and higher flow rate, I can cut wider channels with same width/depth ratio. I guess most people who use this block have very good pump setup to cope with the restriction limitation thumb.gif

For comparison my Arqtik block is super low restriction but only has like 20 channels.
Edited by Puck - 8/3/16 at 4:16pm
Behemoth
(25 items)
 
  
CPUMotherboardGraphicsGraphics
Intel i7 4790K "Relidded" @ 5ghz <50c Asus Sabertooth Z97 MK1 OC'd RX580 8GB OC'd RX580 8GB 
RAMHard DriveCoolingCooling
2X4GB Patriot Viper Extreme 1TB Velociraptor 10k RPM + Crucial 128GB SSD Two Arqtik TEC Chillers in Series Archimedes TEC block w/TN669 TEC 
CoolingCoolingCoolingCooling
Two Ported/tapped MCP655 Pumps 2x Black Ice GTX 480 Gen2 Rads DD dual drive bay Res with LCD water temp monit... GPU Loop: Modded Enzotech Sapphire CPU blocks, ... 
CoolingOSOSMonitor
36x Globe 120mm 80cfm server fans Win 10 Pro Mint 15(KDE) 4K 28" Samsung U590D 
KeyboardPowerCaseMouse
Logitech G710+ Mechanical CM Silent Power M 1000watt Main, OCZ Z850W Slave Custom Mountain Mods Extended-Extended Ascension Logitech G502 Proteus 
Mouse PadAudioAudioAudio
Homemade 16ga 304 Stainless pad, Neoprene backed Aune T1 DAC Various Tubes: Mostly switching from 60's GE "S... AKG K7XX Headphones 
  hide details  
Reply
Behemoth
(25 items)
 
  
CPUMotherboardGraphicsGraphics
Intel i7 4790K "Relidded" @ 5ghz <50c Asus Sabertooth Z97 MK1 OC'd RX580 8GB OC'd RX580 8GB 
RAMHard DriveCoolingCooling
2X4GB Patriot Viper Extreme 1TB Velociraptor 10k RPM + Crucial 128GB SSD Two Arqtik TEC Chillers in Series Archimedes TEC block w/TN669 TEC 
CoolingCoolingCoolingCooling
Two Ported/tapped MCP655 Pumps 2x Black Ice GTX 480 Gen2 Rads DD dual drive bay Res with LCD water temp monit... GPU Loop: Modded Enzotech Sapphire CPU blocks, ... 
CoolingOSOSMonitor
36x Globe 120mm 80cfm server fans Win 10 Pro Mint 15(KDE) 4K 28" Samsung U590D 
KeyboardPowerCaseMouse
Logitech G710+ Mechanical CM Silent Power M 1000watt Main, OCZ Z850W Slave Custom Mountain Mods Extended-Extended Ascension Logitech G502 Proteus 
Mouse PadAudioAudioAudio
Homemade 16ga 304 Stainless pad, Neoprene backed Aune T1 DAC Various Tubes: Mostly switching from 60's GE "S... AKG K7XX Headphones 
  hide details  
Reply
post #75 of 167
yeah I have the 58 channel one, even though it is a "higher restriction" it has less restriction than my Apogee XTL, I get an extra 0.1gpm switching from Apogee XTL to the Archimedes. I think the only way to improve on fox's design in terms of performance would be to cross cut the channels to make pins for added turbulence, that would seriously increase the restriction but it should also up the performance significantly, but it might mess with the even cooling of the whole TEC, so it would need testing. The flow area of this block is ~52mm2 vs the area of a fitting at 10mm diameter with ~79mm2, so it is restrictive in that sense, but this is good as the velocity is higher through the channels which aids the heat removal a tad.

As for the cold side, I'm pretty confident that what works well on the hot side, works well on the cold side, perhaps with the caveat that it takes longer to chill the coolant than it takes to remove the heat from the copper block on the hot side, so more surface area and slower flow rate could be beneficial. I've always wanted to test a highly parallel extremely low flow rate cold side vs the same number of blocks in series to see which one works better or how they work differently, I think the parallel version would give the best results because overall loop flow rate would be higher, which is good for the actual CPU block, but the coolant is moving through the cold blocks at a much slower rate allowing it to be chilled more with each pass, but then it's X blocks wide by 1 block deep, vs 1 block wide by X blocks deep so it might be a wash except for total flow rate being lower in the series version, but with chilled water the slower flow rate through the CPU block isn't so much of an issue either. I think someone like Thick8 would have a far better understanding of this and be able to explain it. Only other option is to test. I have access to a manual mill so I may get around to that one day. Maybe when I need to make a chiller.
post #76 of 167
Thread Starter 
Thanks for posting those details. His is pretty restrictive in that regard. Slot width on mine will be very similar with one main difference being i designed mine around a 42mm TEC. Now more experienced members here have expressed to me on a few occasions that running a large TEC is going to be more efficient than me running 8 little TECs. I'm going to take that advice here. This setup has been a fun learning experience but i'm going to abandon the work i have here and begin working towards a 62mm setup.

I calc that his block has a 0.081" inlet surface area, assuming his slot array is roughly 62mm. My slots are .005" wider but keep in mind i went .170" deep. I achieve the high flow rate there. And while he has more slots i still have more inlet surface area at 0.1317". But lets calculate his functional copper surface area within those slots first.

Each slot is .02" wide and .07" deep. Each wall then 2.440(62mm)*.07= 0.1708 in2. The floor of the slot is .02*2.440(62mm)= 0.0488 in2.

Each slot has 2 walls and a floor so we do that math 2(.1708) + .0488 = 0.3904 in2 of surface area in each pocket. Having 58 pockets the total functional surface area would be 22.6432 in2.


By comparison my much smaller blocks still had roughly 19 in2 of functional surface area with half the slots and still maintains the inlet flow area needed for high flow. It would have been very interesting to see how it would perform for 42mm but its performance would be hard to measure as thats not even apples to apples. So i'm begging my design work on the 62mm block. This isn't a huge set-back for me as i still need acrylic for the tops that wont be in until next week. So i'm ordering copper stock now with the acrylic. Hopefully by next weekend i have a few blocks to test around with!

The goal: high surface area, high flow rate. The inlet (G1/4) will determine my inlet surface area, and my inlet surface area will determine the slot depth.
post #77 of 167
Does anyone know the channel layout on the US2 V2 block? I suppose I could pull it out of the box and count them ... my eyes are not ready for that though.
post #78 of 167
I believe the US2 V2 block uses finer channels than fox's block. Also the area that the channels cover is far larger than the 62x62mm TEC. It's like 80x90mm area or something like that, fox's is pretty much 62x62mm of channels. US2 V2 is very good for massive heat load transfer. Flow rate is very high because his channel cross section area is much larger than the area of tubing/barbs inlet/outlet. This means low velocity but with such a huge area and such thin channels it would work very well. Fox has 0.5mm channels, I imagine the US2 is anywhere from 0.4mm down to 0.25mm. Looks like about ~100 slots and 100 fins on the US2 block, so probably 0.4mm slots, maybe thinner fins, not sure exactly
Edited by LiamG6 - 8/4/16 at 9:18pm
post #79 of 167
Thread Starter 
Quote:
Originally Posted by LiamG6 View Post

I believe the US2 V2 block uses finer channels than fox's block. Also the area that the channels cover is far larger than the 62x62mm TEC. It's like 80x90mm area or something like that, fox's is pretty much 62x62mm of channels. US2 V2 is very good for massive heat load transfer. Flow rate is very high because his channel cross section area is much larger than the area of tubing/barbs inlet/outlet. This means low velocity but with such a huge area and such thin channels it would work very well. Fox has 0.5mm channels, I imagine the US2 is anywhere from 0.4mm down to 0.25mm. Looks like about ~100 slots and 100 fins on the US2 block, so probably 0.4mm slots, maybe thinner fins, not sure exactly


More good info Liam. Fox did right on the .5mm channel as far as manufactuability goes. Assuming he's using a controx or similar slitting saw. Getting bellow the .02" mark on a 1.0 or 1.5" diameter starts to get tricky. Blow up a saw in one of those blocks and all the work on those blocks is useless.

I'm going to stick with his .02 design for now. My slot area is larger though. I ran some numbers last night. 61 slots at .02 wide and .17 deep came out to around 58" of surface area with around double the inlet cross section surface area. IE water will move in the block at 1/2 the speed as it does within the tubing.
post #80 of 167
Of course this is only in my situation and my mobo, but if I could change one thing on Fox's block, I would make the "step" on the cold plate taller. I had to run less insulation then I wanted because the caps around the socket were causing interference and a poor mount.
Behemoth
(25 items)
 
  
CPUMotherboardGraphicsGraphics
Intel i7 4790K "Relidded" @ 5ghz <50c Asus Sabertooth Z97 MK1 OC'd RX580 8GB OC'd RX580 8GB 
RAMHard DriveCoolingCooling
2X4GB Patriot Viper Extreme 1TB Velociraptor 10k RPM + Crucial 128GB SSD Two Arqtik TEC Chillers in Series Archimedes TEC block w/TN669 TEC 
CoolingCoolingCoolingCooling
Two Ported/tapped MCP655 Pumps 2x Black Ice GTX 480 Gen2 Rads DD dual drive bay Res with LCD water temp monit... GPU Loop: Modded Enzotech Sapphire CPU blocks, ... 
CoolingOSOSMonitor
36x Globe 120mm 80cfm server fans Win 10 Pro Mint 15(KDE) 4K 28" Samsung U590D 
KeyboardPowerCaseMouse
Logitech G710+ Mechanical CM Silent Power M 1000watt Main, OCZ Z850W Slave Custom Mountain Mods Extended-Extended Ascension Logitech G502 Proteus 
Mouse PadAudioAudioAudio
Homemade 16ga 304 Stainless pad, Neoprene backed Aune T1 DAC Various Tubes: Mostly switching from 60's GE "S... AKG K7XX Headphones 
  hide details  
Reply
Behemoth
(25 items)
 
  
CPUMotherboardGraphicsGraphics
Intel i7 4790K "Relidded" @ 5ghz <50c Asus Sabertooth Z97 MK1 OC'd RX580 8GB OC'd RX580 8GB 
RAMHard DriveCoolingCooling
2X4GB Patriot Viper Extreme 1TB Velociraptor 10k RPM + Crucial 128GB SSD Two Arqtik TEC Chillers in Series Archimedes TEC block w/TN669 TEC 
CoolingCoolingCoolingCooling
Two Ported/tapped MCP655 Pumps 2x Black Ice GTX 480 Gen2 Rads DD dual drive bay Res with LCD water temp monit... GPU Loop: Modded Enzotech Sapphire CPU blocks, ... 
CoolingOSOSMonitor
36x Globe 120mm 80cfm server fans Win 10 Pro Mint 15(KDE) 4K 28" Samsung U590D 
KeyboardPowerCaseMouse
Logitech G710+ Mechanical CM Silent Power M 1000watt Main, OCZ Z850W Slave Custom Mountain Mods Extended-Extended Ascension Logitech G502 Proteus 
Mouse PadAudioAudioAudio
Homemade 16ga 304 Stainless pad, Neoprene backed Aune T1 DAC Various Tubes: Mostly switching from 60's GE "S... AKG K7XX Headphones 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Peltiers / TEC
Overclock.net › Forums › Cooling › Specialized Cooling › Peltiers / TEC › Peltier chilling system