[Martinsliquidlab.org] DDC3.2 Pump Heat Scoping

[Martinm210]

Detail Blog

Back from Halibut fishing and ready to play with some more WC parts..

Getting ready to test some heatsinks and cooling options, so I figured I'd start by exploring heat on these pumps a bit more. This is what I found:

First, numbers at the 2GPM mark:


But...what happens at 3GPM???


Heat Vs. Flow Rate


Conclusion

It appears the FETs and coils run the hottest on the pump and the IC is of little interest. Also the center of the pump base is not very representative of the hottest areas. I will see what I can do about placing a probe or two near the FETs and coil area inside the pump casing. A high restriction system at 1GPM only measured roughly a temperature of Ambient +35C while the very low restriction 3GPM condition resulted in almost twice the delta of Ambient +55C. I’m beginning to think heat removal via a heatsink is more important than most people think when running DDC pumps with tops but it really depends the most on flow rate. The factory Laing top will restrict that max flow rate, but once opened up via a pump top…the heat is ON!

This is a bit preliminary, but I think if heat can be a problem with pump life, designing a system that is well balanced between pumping power and restriction is a good thing. You probably want to shoot for 1-1.5GPM to keep heat levels in check or seriously consider more cooling.

More testing to come….I will be exploring heatsinks and air flow scenarios soon. I just need to find some fine wire to solder a few probes inside my DDC for testing enclosed.

Cheers!
Martin

[zalbard]

That's with housing removed... Wonder how hot it gets inside the stock closed box!
I might have to buy a heatsink after all...

[rge]

After Gabe showed drop in temps at pump bottom with heatsink/fan... what your doing is what I wanted to know, the temp on hotter parts of PCB with and without a heatsink/fan, ie how effective the heatsink cools PCB through the plastic housing. I have a fan blowing under mine...but may switch to heatsinks in near future. Interested to see what you find, if you can get it rigged up with probes without frying the pump.

[Serpentarius]

Martin is the heat necessary for the coils? in your blog, it's almost 100°c ... omg

if the heat is removed, will cause the decline in performance?

[Martinm210]

No, I think it's just a byproduct.

I'm not sure what Tjmax is on all these parts nor what they all really do. I know there is an IC motor controller, several FETs for switching power, and those coils that I believe act as mini transformers. How the circuit operates in detail, I'm not sure (I only know very basic electronics).

From the little bit of playing with the DIY PCB stuff, the ICs were typically rated for 150C, so I'm guessing some of these parts are at similar rating levels? Plastic pump casing is rated for around 200C I think.

If I fry the pump in this experiment, I'll spend some time taking the PCB off to see if we can get some numbers off the FETs and IC...

I purposely picked an older pump that I won't miss too much in the name of science!..

[Boogerlad]

I wonder if non-reference pcbs are cooler...

[zeropluszero]

I love my koolance sink

[Serpentarius]

transformer? i tot the coils is to generate the alternate magnetic field? causing the rotator to spin?

[Martinm210]

transformer? i tot the coils is to generate the alternate magnetic field? causing the rotator to spin?

You are probably right, I'm not sure what they do other than they get hot!

[dasa]

that is surprisingly hot considering its in open air so it will be interesting to see how hot they get with the cover on and fan\heatsink vs a foam pad on the bottom

[CrazyNutz]

I love my koolance sink

Yes I think Koolance is the best solution thus far.

They include a thermal pad that makes contact with most/all hot sections.

[zalbard]

I love my koolance sink

If only it wasn't so overpriced...

[CrazyNutz]

If only it wasn't so overpriced...

http://www.koolance.com/water-coolin...roduct_id=1058

Look's like a pretty nice deal for $35.00

I'd be interested in knowing the water temp difference between pump stock, and pump with this heat sink, i.e. in a loop with just res & pump, so the pump is the only source of heat.

[zalbard]

http://www.koolance.com/water-coolin...roduct_id=1058

Look's like a pretty nice deal for $35.00

Come on, man, that thing is worth $10, the rest is just margins. That's half of pump's price for a tiny chunk of metal...

I'd be interested in knowing the water temp difference between pump stock, and pump with this heat sink, i.e. in a loop with just res & pump, so the pump is the only source of heat.

I somehow can't imagine there would be any difference.

On a side note... I know it sounds ridiculous, but can't we water cool our pumps?

[bmaverick]

Martin, what is the temp reading for a DDC-1 or DDC-1T in your pump portfolio? How much cooler do they run?

[Martinm210]

Martin, what is the temp reading for a DDC-1 or DDC-1T in your pump portfolio? How much cooler do they run?

Should be much better for the 12w flavors. I could probably modify my DDC3.2 to be switchable 3.1 or 3.2 by breaking the solder bridge and wiring that to a 12v switch.

I'll look into that. It won't be exact..but a good indicator of DDC1 heat levels.

[[AK]Zip]

A DDC1 should be significantly cooler since they run on significantly lower since they are 9-10watt versus 18watt of the DDC3.2

[Waterlogged]

Come on, man, that thing is worth $10, the rest is just margins. That's half of pump's price for a tiny chunk of metal...

I somehow can't imagine there would be any difference.

On a side note... I know it sounds ridiculous, but can't we water cool our pumps?

Water cooled pumps are so 2010.

http://www.xtremesystems.org/forums/...7&postcount=80

[phantomferrari]

hey martin, did you do a little mini test on the laing 3.25 pump a little while back when it came to temps or am i just imagining things?

[NaeKuh]

After Gabe showed drop in temps at pump bottom with heatsink/fan...

i kinda showed this a very long time ago..

I was the first one who made sinks soley for the DDC though Bei Fei..

Then Koolance followed expanding on the idea, and the swiftech copied me but made the base sink larger..

Yes I think Koolance is the best solution thus far.

They include a thermal pad that makes contact with most/all hot sections.

Noticed small issues with them tho.
First off they dont Fit DDC-2's.
So any DDC with old PCB wont work.

Come on, man, that thing is worth $10, the rest is just margins. That's half of pump's price for a tiny chunk of metal...

I somehow can't imagine there would be any difference.

On a side note... I know it sounds ridiculous, but can't we water cool our pumps?

:\

Actually my sink was around what u calculated in raw parts... lol..

The koolance's is more and its worth it.

I have yet to see someone who has a shell saying it was a bad purchase.

[Martinm210]

hey martin, did you do a little mini test on the laing 3.25 pump a little while back when it came to temps or am i just imagining things?

No, this was the first heat measurement I've done on the PCB area. I did some measuring on the MCP35X with my laser on the plastic base and came up with nearly 60C, but it was just informal.

[Gamekiller]

The koolance sinks won't fit on the pmp 400 attached to the 402x2s when they're in a normal drive bay right? I'm a bit worried about my PMP 400.

[NaeKuh]

And martin do you recall when i would complain like crazy and not to mention how many melted petra gel stuff we had? :P

The koolance sinks won't fit on the pmp 400 attached to the 402x2s when they're in a normal drive bay right? I'm a bit worried about my PMP 400.

rev. 1.1 will

[Martinm210]

And martin do you recall when i would complain like crazy and not to mention how many melted petra gel stuff we had? :P

We've all known they get hot, no real surprise. I just didn't think the PCB areas would be that much hotter nor did I think that flow rate made such an impact on the heat level.

[Martinm210]

rev. 1.1 will

FYI, you might consider turning your heatsinks 90 degrees if you can...that would convect vertically much better if you don't have fan flow.