13.8v vs 24v TEC PSU

[Joe Camel]

THIS is the 24v - 25Amp PSU


HERE is the 13.8v - 25Amp PSU



and this is *about* the setup im testing:


this is done @ stock settings: 520/540 (pencil V modded X850 XT card... id tell ya what V the core is running, but the backplate of my cooler blocks the measure point )

this is with:
13.8V PSU with a 172w TEC, DD Maze4, Mag3, Caprice core with 2x Nidec Beta V fans on full

min........Room.......Water.......GPU
----------------------------------
3 I.........18.9.........20.6........-13
5 I.........19.0.........20.7........-13
3 L.........19.1........21.3........+11
5 L.........19.2........21.5........+13
10L........19.3.........21.7.......+14

this is with:
24v PSU with a 172w TEC, DD Maze4, Mag3, Caprice core with 2x Nidec Beta V fans on full

min........Room.......Water.......GPU
----------------------------------
3 I........17.5.........22.9.........-17
5 I........17.7.........23.2.........-17
3 L........17.9........23.9..........+7
5 L........18.1........24.2..........+8
10L.......18.6.........24.7..........+9

I=idle
L=load (ATi tool running "show 3D")
temps are in C


so far, i can say that the 24V does show an improvement, but that my cooling needs to be upgraded...

please feel free to post questions/comments as to how i can improve this post ...THANKS!


.

[craig588]

24V is more efficiant. The higher you go with voltage the greater your effeciancy becomes. That's why high HP machines run on 480 or higher volts.

[Holst]

More fans or bigger fans on the rad should help you get your water temps down to 21-20*c

This should gain you 3-4*c lower temps. 6*c loaded is pretty good.

I dont understand what you mean about efficency craig. I would say that 13v is more efficent than 24. But 24v is more effective.

What are the specs for your 172watt TEC?

[Joe Camel]

172w TEC

Imax = 7.9 Amps
Qmax = 120 Watts
Vmax = 24.6 Volts
Delta Tmax = 69 (C)
Size = 40mm X 40mm X 3.4mm

as for the cooling, these are already 120+ CFM fans...ill probably need to go with my 2x single pass Fedco's (in series) and i might swap the Mag3 for my D4 (i have a 16v PSU for the pump )

[n00b 0f l337]

Wait, same tec both times? Then your difference was that with 13.8 volts you were running half power, get a 12volt 172watt tec and check your results. Or get a 350 watt 24volt tec and run it at 12volts and it will be much more efficient.

[Joe Camel]

sorry i dont know of any 172w 12v TECs
the 226+ (350-400+) are all 50x50 (or BIGGER) which wont fit "in" the WB i have.

[n00b 0f l337]

Yeah but the current testing is running the tec on the 13.8v at around half power, of course it performs better at full power.

[craig588]

Yeah, this testing is meaning less.

[n00b 0f l337]

But if your sticking with the meanwell, how much for that radioshack thingy? I can get a 15amp one on ebay for $10....

[Holst]

Yeah, this testing is meaning less.

No its not.

I think this is WAY more interesting than just compaing two 172watt tec with different Vmax.

This goes to show what sort of performance you will get if you undervolt a TEC, and how running closer to Vmax increased water temp.

[Joe Camel]

Yeah but the current testing is running the tec on the 13.8v at around half power, of course it performs better at full power.

congradulations ...you win the master of the obvious award...

what im trying to do is to show HOW MUCH (or how little as the case turned out to be) better.

Yeah, this testing is meaning less.

not if your trying to decide between a 12 and 24 volt PSU
or what kind of cooling to use for a 172w (or higher) TEC @ full volts

plus, now i get to play around and make a new cooling loop...CPU's dont really put out enough watts (delta T) for a benefit to my dual (2x 120x240) rads but this thing might just put them to good use...



thanks for the input

.

[n00b 0f l337]

Well it could have been a test of two things, the way you did it is option 1. Most of us were thinking otherwise. What you could have done is get a 172 24 volter, and a 350 24volter, and run the 350 at 13.8 volts for a near equal wattage and see how much better the undervolted one works.

[ls7corvete]

24V is more efficiant. The higher you go with voltage the greater your effeciancy becomes. That's why high HP machines run on 480 or higher volts.

Um, not with pelts.....

What you could have done is get a 172 24 volter, and a 350 24volter, and run the 350 at 13.8 volts for a near equal wattage and see how much better the undervolted one works.

that would be interesting.

Yeah, this testing is meaning less.

Eh? So you already knew that this setup would cool 4-5c worse when using 200w less energy to the TEC?

[n00b 0f l337]

Yeah pretty much, and it was like 85 watts less energy.

[ls7corvete]

Yeah pretty much, and it was like 85 watts less energy.

Ya, my bad, all points still valid though.

[saratoga]

24V is more efficiant. The higher you go with voltage the greater your effeciancy becomes. That's why high HP machines run on 480 or higher volts.

Higher voltage saves you money on copper wire and not much else. So if his 12v supply was in another town, you might have a point. If its in his house, its not an issue.

[craig588]

Most people are running really thin wires, I still hear about people running warm wires while running TECs.

[craig588]

Eh? So you already knew that this setup would cool 4-5c worse when using 200w less energy to the TEC?

Yes, but it's not 200 watts less, probably around 100. It's offensive if someone didn't know that, it's almost common sense.

[ls7corvete]

Yes, but it's not 200 watts less, probably around 100. It's offensive if someone didn't know that, it's almost common sense.

Ya, my bad, all points still valid though.

Offensive is not reading the whole thread.

[craig588]

I did read the entire thread. (Unless you edited a post)