Aquaero with PowerBooster overheating powering DDC?

[eternal_fantasy]

Few hours after plugging in my DDC to channel 1, which is set to adjust the pump's power between 50% and 100%, I smelt something burnt coming from the Aquaero (my case was still open from installation), then after uninstalling it from my pc and took off the heatsink, this is what I found;









There seems to be brown marks around the FET and some sort of solder-like substance coming out between the FET and the PCB...

Also in the picture the 2 resistors increasing channel 1 power are clealy visible, so there is no doubt wether this is the factory-modified power-boosted version (which is a little brownish itself).

Any thoughts? It heats up as it reduces the power going to the pump, then overheats itself... I thought it was modified so that it can handle the DDC's load?

[Ketzer7]

Did you have anything connected to the other 3-pin headers on there?

From what Shoggy clarified for all of us, the total power output of the powerboosted aquaero is 45W, with up to 25W available on channel 1 and up to 15W available on the other 3.

The problem is really the sum wattage of all your devices plugged in there.

If you had the DDC plus a bunch of other high draw fans or devices off the other channels, is it possible you exceeded the 45W total limitation?

If not, then it sounds like somehow your DDC was sucking up more than 25W off that one channel and was burning out the power regulator.. Was your DDC basically running restriction free when you powered it up? The pump will actually consume more current and power the less restriction there is for it to work against oddly enough.

Otherwise, hopefully Shoggy will chime in and help you out. Sorry.

[eternal_fantasy]

A combind total of less then 7W of fans over the remaining 3 fan channels. The loop has been running for months off the PSU's Molex cable fine at full 12V, with a flowmeter in the loop just for fun. The thing is the Aquaero runs nice and cool at 50C when powering the pump, its when it tries to undervolt it that it overheats. I think the problem is that the small heat sink was not able to dump the heat being so small and with a large peice of steel blocking it's airflow that the heat just kept building up over the hours till the solders under it melted...

Look at the MiniNG heatsink that is designed to power a DDC:



I had those before and they get HOT!! not sure how the tiny sink on the Aquaero was designed to cope with a higher load then the MiniNG...

[mozis]

There seems to be brown marks around the FET and some sort of solder-like substance coming out between the FET and the PCB...

Yeh it has to be solder and it also means those transistors got really hot!

Any thoughts? It heats up as it reduces the power going to the pump, then overheats itself... I thought it was modified so that it can handle the DDC's load?

It can handle more current due to the heatsink, I dont think there r any other modifications. Judging from the pics you posted - my aquaero looks the same.

Maybe the heatsink was somehow mounted badly but it seems there was even some thermal paste on those transistors?
Or as someone has stated already, if you havent connected anything else, maybe somethings wrong with the pump?

[eternal_fantasy]

It can handle more current due to the heatsink, I dont think there r any other modifications. Judging from the pics you posted - my aquaero looks the same.

Maybe the heatsink was somehow mounted badly but it seems there was even some thermal paste on those transistors?
Or as someone has stated already, if you havent connected anything else, maybe somethings wrong with the pump?

Quote from the AC Rep:

like Brodholm said, it is not only the cooling block, you will also have to add two small SMD resistors to the board. If you only add a heatsink you can reach up to 15W on each channel (keep the 45W overall limit in mind).

you can see the 2 larger resistors from my last picture, 00RE and 15RO, they are the modified SMDs.

And like I said on my second post, the pump has been running fine on 12V for months, and it is now fine again plugged back into the PSU's Molex. Anyone with a 18W DDC would know that it get very hot on full power. The heat shifts to the Aquaero once the DDC is undervolted, and itself becomes much cooler.

The white stuff is something I used to clean the FETs with, they had a large peice of thermal pad there and caused everything around the area to become greasy... The actual heatsink was at around 90 degrees celsius(!), measured with an infrared surface thermometer, before I shut the system down... so I think the heat had no problem travelling to the heatsink from the FET, but then kida just stayed there...

[mozis]

Ahh OK, you're right about that powerbooster thingy.

Is it still working?

Maybe you wanna try giving the thing some air flow over that heatsink?

But then again i dont know if that much heat is allright at all.

[eternal_fantasy]

Yeah it's still working but the smell is still there. It looks damaged though so don't think it's a great idea letting it continue powering a 18W pump that my other hardware depends on for cooling...


Posting here so that all info thats needed by the Aqua Computers people can be located in one place, incase I missed anything... At the rate of being only able to reply to 1-2 of my e-mails per day, I thought I can save both parties some time by pointing them to this thread.

[Brodholm]

Any news yet? I have 2 aquaeros with powerbosters waiting to get installed. This worries me a bit :\

[Diverge]

I'm interested to know what went wrong too. I just got an aquaero and was looking at the components used, and creating a schematic to see what is going on with the driver circuit. The transistors used on each channel are rated for 8A, 50W http://www.onsemi.com/PowerSolutions...o?id=MJB45H11G

From looking at your PCB, it doesn't look like any other components gave out, or overheated, just the MJB45H11G

I'd be curious to see under the transister, to see if the PCB was damaged at all from the heat.

edit: woops, looks like I made a mistake. the package is a DPAK, not a D2PAK, so the data sheet is http://www.onsemi.com/PowerSolutions...do?id=MJD45H11 and it's only rated for 20W.

My plan is to remove the transistor from the PCB, and use a D45H11, which is a TO-220 making it easy to mount to a heatsink, and it's rated for 70W. I'm gonna run my D5 off it

This is the transistor I plan to use: http://www.onsemi.com/PowerSolutions...t.do?id=D45H11 I'm gonna remove the stock one, and mount this one off the PCB to a nice heatsink, and use nice gauge wires for the collector and emitter on the transister, so no worries of blowing traces up

[dejanh]

It could also just be a bad part...

[ohms]

what do the MiniNG and the Aquaero do anyway?

[Nickel020]

Reduce the voltage and thus lower the noise

[ohms]

ah ok, but won't that affect the performance of say a pump?

[Brodholm]

ah ok, but won't that affect the performance of say a pump?

yes it will. But when you are just idle you can have your pumps on low. And when the loop start to heat up the aquaero will increase the pump speed. If you programmed it that way.

Was your unit very isolated and having bad airflow eternal_fantasy?

[scamps]

I run a "new" Aquaero since it is available. And I have no problems at all, neither with Laing, nor with Laing @+. And I have not heard of any other user problems, yet.

I guess this could be just a bad part and no need to hype this damage.

And where did you mount it? Is there any air flow?

[Jam]

fwiw, my Aquaero with pre-assembled PowerBooster has no issues undervolting my DDC.