Luiquid cooling solar charge controller

Introductions, general chit chat and off-topic banter.
Post Reply
User avatar
offgridQLD
Senior Member
Posts: 1827
Joined: Tue, 23 Jul 2013, 16:05
Real Name: Kurt
Location: Fleurieu Peninsula, SA

Luiquid cooling solar charge controller

Post by offgridQLD »

It's always been my aim to try and keep all my power electronics as cool as practically possible. Given the general consensus is that a cool component is a happy one and should give a longer service life.

I have installed a 2nd MPPT charge controller in my off grid system. It's identical to the first controller I installed and has a very similar size PV array being fed into it (4000w on the 1st and 4200w on the new 2nd) The only real difference being the later has the panels configured in 3s strings vs 2s strings. So we are now on average down converting from around 95 - 110v to charge a 48v battery.

This higher voltage configuration was to allow for a reasonably long cable run. It was a bit of a choose between losing power in the cable run at a lower voltage (70 - 80v) in 2s configuration or losing some power to heat in the controller stepping down 100v or so volts to 48v.

So last weekend (26 deg C day) I really hammered my system to load it up and keep the two Solar charge controllers pumping out all they can muster for hrs on end. The 1st controller as it has done for the past few year had no trouble at all with heat in fact only the slow speed internal fans cycled on and off in a pattern of 45C - 48C at the fets (trigger off and on temp set points for the slow speed internal fan)

Made a small video while loading the system.
https://www.youtube.com/watch?v=5tB70Ih ... OrasSNz70A


Charge controller one looks like this over the day. I'm happy with that everything is well under control and cool considering its pumping out several thousand watts with only two internal slow speed fans blowing a gentle breeze through the case. cycling on and off.
Image


The 2nd controller was getting hotter and the it struggled to hold the temps in the 45 - 48C range. The internal fans running constantly the temps Drifted up higher above 48c triggering the (turbo fan) small high speed external fan that blows air over the heat sink over the fets. Fet temp has to be 58C to trigger this fan. At about 60C things came under control with all the fans screaming away.

Running my hand over the die cast aluminum case I could feel a real hot spot at the back directly behind the area where the two inductors sit in a molded cradle as part of the case. Considering PCB temps were following very close to fet temps I think a lot of the heat was transferring from the inductors.

Image

The temps are within spec but I just hate things running hot if they don't have to. Particularly when it triggers fans that run all the time. (I live on a dirt road and I want to avoid ingesting more dust than necessary into the components.

So for now I have dialed back the max output of the 2nd reg while I gather up parts for a cooling rig.

I have pieced together a liquid cooling system from PC and industrial Ebay parts.

125mm radiator. (basically a miniature version of a Automotive radiator)

40mm x 120mmx 3.3mm aluminum cooling / chill block. (as used to cool industrial laser cutter heads)

125mm 3 speed case fan .

small pump and 125ml tank unit all in one.

I could have just slapped a huge big chunk of aluminum of the back of the case but where is the fun in that. Big heat sinks are expensive to (I have $50 inverted in the liquid cooling rig) I think liquid cooling will do a way better job of removing heat. The radiator and fan can be placed away from the components to reduce dust.

The things I need to sort are heat transfer. the rear of the case isn't smooth and is embossed with a logo. I could just mill a flat section on it (though would prefer not to for warranty purpose) so perhaps some kind of thermal mat to sandwich between the case and chill block. I have no experience with thermal transfer products other than the run of the mill thermal white thermal past.

Tell me I'm crazy or give me some tips suggestions. I'm open to ideas.
Image

Kurt
Last edited by offgridQLD on Wed, 08 Oct 2014, 03:46, edited 1 time in total.

User avatar
Johny
Senior Member
Posts: 3732
Joined: Mon, 23 Jun 2008, 16:26
Real Name: John Wright
Location: Melbourne
Contact:

Luiquid cooling solar charge controller

Post by Johny »

Hi Kurt. Lacking any contributions to your heat-transfer issues I'll put my two-bob's worth in (showing age here).

First. Heat paste isn't a good filler for voids. You probably know but it's purpose is the fill the tiny gaps between two preferably "flat" surfaces. Tests we have done here (RF final amplifiers) show that it's often counter-productive.

If the rear of the controller isn't flat, there is probably little point in trying heat paste. Then again, if you can get more than 50% metal-metal contact then I'd try it.

My only other idea at this stage is to fit a chunk of metal to the rear with a deliberate void and fill it with light mineral oil - or radiator fluid. It would need gaskets etc. so would probably be over complicated.

User avatar
offgridQLD
Senior Member
Posts: 1827
Joined: Tue, 23 Jul 2013, 16:05
Real Name: Kurt
Location: Fleurieu Peninsula, SA

Luiquid cooling solar charge controller

Post by offgridQLD »

Yes it's very tempting to just run a pass over a section of the back of case in the mill. Then do the same to the cooling block for a nice surface contact.

Image

The controller case die cast aluminum that's powder coated. Though the raised logo will be in the way for my 120mm x 40mm liquid cooled chill block I ordered if I run it vertically. That's why I was thinking of a flexible rubber max to go under it (thermal transfer mat). I did see this product online.

Another option is I could place the chill block just above the fan access plate and have it run horizontally. There is enough space there without a raised logo. Though it would be nice to have it directly behind the two inductors.

That said will the heat transfer well enough through the powder coating. My theory is if it's hot to my hand then the heat is getting through. If I need to remove the powder coating then I my as well machine the logo off and place it behind the inductors where I prefer.

Not much experience with thermodynamics Image

As long as I can keep the temps under 50C no matter how hard I drive the regulator then I am happy.

When fully loaded over the day I'm potentially pushing about 20 + kwh through them each day in summer (40- 50kwh production together)

Kurt
Last edited by offgridQLD on Wed, 08 Oct 2014, 07:04, edited 1 time in total.

User avatar
offgridQLD
Senior Member
Posts: 1827
Joined: Tue, 23 Jul 2013, 16:05
Real Name: Kurt
Location: Fleurieu Peninsula, SA

Luiquid cooling solar charge controller

Post by offgridQLD »

The parts I ordered.

Pump & 120ml tank combo.
Image

Radiator 125mm
Image

Chill block 120x40mm
Image

3 speed fan 125mm
Image

Kurt

User avatar
Gabz
Senior Member
Posts: 580
Joined: Thu, 08 Aug 2013, 03:30
Real Name: Gabriel Noronha
Location: Maitland NSW
Contact:

Luiquid cooling solar charge controller

Post by Gabz »

submerse the whole thing in transformer oil
Corporate Member Recharging NSW Pty Ltd. http://rechargingnsw.com.au/

User avatar
Johny
Senior Member
Posts: 3732
Joined: Mon, 23 Jun 2008, 16:26
Real Name: John Wright
Location: Melbourne
Contact:

Luiquid cooling solar charge controller

Post by Johny »

Going by the look of it I would think that almost anything will help a lot. It's a pretty crappy looking heat get-rid-oferer - not much in the way of fins.

User avatar
offgridQLD
Senior Member
Posts: 1827
Joined: Tue, 23 Jul 2013, 16:05
Real Name: Kurt
Location: Fleurieu Peninsula, SA

Luiquid cooling solar charge controller

Post by offgridQLD »

The other side of the case(that has the fets running along it you can't see in the above pic is all tube spikes of aluminum covering a area about 1/3 the width in and 2/3 the length of the case tip to bottom. A plastic shroud is covering them that has a fan at the top that sucks air flow over the spikes and blows it out the top

The front cover has a open vent where air sucked from the bottom and blown through the case vents.


kurt

User avatar
Johny
Senior Member
Posts: 3732
Joined: Mon, 23 Jun 2008, 16:26
Real Name: John Wright
Location: Melbourne
Contact:

Luiquid cooling solar charge controller

Post by Johny »

offgridQLD wrote: The other side of the case(that has the fets running along it you can't see in the above pic is all tube spikes of aluminum covering a area about 1/3 the width in and 2/3 the length of the case tip to bottom. A plastic shroud is covering them that has a fan at the top that sucks air flow over the spikes and blows it out the top

The front cover has a open vent where air sucked from the bottom and blown through the case vents.

kurt
Yes, I see now. That's better. Right on top of the FETs is where you really want to draw the heat off if you can.

Post Reply