Commanda's Buck and Ghurd's Dump

[Madscientist267]

Sound sick? It is.



Amanda's buck, ready for boxing. The only thing missing is the "I'm doing something" LED. It will be there, promise.

At 1.7A input (~36V), it pumps 3.85A into a battery, dumping at 14.4V. I'm more than pleased with this. I think it will do more, but have to wait for sufficient sunshine to come to my beckon...





Ugh. Absolutely horrible. I feel your pain, Amanda. 





The new dump controller, ready for boxing.

Capable of 60W @ 14.4V continuous. On the heatsink, those are 10R, 10W resistors in the center, with 0R1 3W On the left.

The trannies are (real deal) IRFZ44N MOSFETS, the board is standard all except for the 'I've got power' LED, which is in the main switch (below).





Here she is 'benching-it', soaking in 4A at 14.4V (~57W). The fan keeps her nice and cool, all around. She works like a champ.





The back side of the dump controller. I finally used one of those SR360 diodes man! Unfortunately, it's hidden (along with the fan cap) behind the main wires going down to the controller itself in this pic...



All in all, I am very pleased with this system. Works very well, and I have not one complaint so far. Other than no sun to really speak of. The 3.8A was a very brief tease provided by mother nature a couple days ago...

I WANT SOME REAL SUN!!! 

Steve

 

[ghurd]

Looks fine to me!

What kind of goop holds the resistors and fets to the heat sinks?

No sun?
At least you can tell the ground from the sky!
Out my window, up, down, left and right, is all the same color, white.
Below is a photo... 
G-

[TomW]

Looks like a Polar Bear in a snowstorm to me but a bit gray.
 
 
 
Tom

[Fused]

It looks the same in SC Glen.
Not supposed to look like that here.
Please take it back! 
Madscientist267 , Looks like a fine job. Very clean set up.


Fused

[Madscientist267]

5 min epoxy.

I have some concerns in hindsight, but it's holding up. That was the primary reason for the fan. I get good heat transfer, but at maximum power I wonder if expansion might cause it to crack. Hopefully it softens just enough to avoid that.

Time will tell.

I screwed up and put the board in the most inaccessible place inside the box, so I have to take the cover off to adjust the threshold. This wouldn't be a problem except someone in design for Radio Shack didn't keep up with QA and the aluminum covers don't fit like they should anymore. It was a %*@#$ to get that thing in there.

It cuts in at ~14.45 and at full power is rising to 14.47V. I may just leave it there. 

The buck is using the plastic cover, so it is a cinch to adjust that one, and the board is in a better position, just need to drill a hole so I can get to the pot.

Steve
 

[ghurd]

If the 2 fets are part of the dump load controller, put a Gate resistor on each.
The fet tab is connected to the Drain, so the loads are effectively combined.
The Gate resistors keeps them switching in time a little more evenly.
G-

[Madscientist267]

One step ahead of ya boss.

Used 2ea 47R resistors for the gates.

The drains are brought out separately to the load, even though they are probably in contact at the heatsink as well.

The sources run over to the two 0R1 ballasts, the other side of which is connected to negative.

The other side of the load is connected to positive.

The fan is connected across a 470uF cap with a Schottky diode charging it, so that the speed is proportional to the power going in.

Doesn't work properly (as found out on the first controller) if the diode and cap are not there. Fried one fan in the first one trying to go directly across the load. Apparently, most brushless fans are designed with the motor finals in close margin to smoke threshold. It's fine once you get them going, but the repeated pulses at startup threshold make the fan's trannies see startup surge continuously in this region.

No problems frying fans with the diode and cap scheme.

Had to do something similar in my charger as I cooked 2 fans before I figured out what the problem was. The only difference there is that the output to the fan is analog and so I used a 25R rheostat to limit current at low speeds.

Figured I would mention all this just in case someone builds one and goes 'compact' and decides to use a fan. It's the easiest way to make the fan only run on demand without baking it.

Steve