Dummy Action

Last week, I finally got a few parts I was waiting for. Namely the insulating pads that allow me to connect the MOSFET to the heatsink without making the entire case the drain! So with that, assembly is complete. Just need a label for the front which I can do at a later date if ever I make it to Staples to buy adhesive transparency sheets.

Properly mounted MOSFET. Note the pad under it, and the nylon bushing isolating the screw

Properly mounted MOSFET. Note the pad under it, and the nylon bushing isolating the screw


Performance

With the unit finished, of course I had to run it through its paces. I was rather chuffed to finally get something done. On the whole, it performed about as well as it ever did, with a couple of exceptions. One thing I noted is the range switch for the display is a bit dicky. On occasion it will display some odd numbers when I switch it to the higher range, but it seems to settle out quick. Might also be some bounce or play in the switch which confuses the display. Not a problem, practically.

The other exception is a concern. I ran it through some runs drawing current from my craptacular battery charger and monitored the temperature. My inclusion of a big heatsink and a fan was definitely a good idea as this poor MOSFET heats up like crazy. Without having gone through all the calculations to determine my heatsink and active cooling requirements, I had rather hoped for the best that what I was doing was sufficient.

Fortunately, I seem to be able to draw at least 2A at about 12V (the crusty battery charger is not a regulated supply) and the temperature seems to stabilize nicely at about 120°C. The idea being the cooling I provide must at least balance the heat the MOSFET is putting out, even better if it drains it away faster. This prevents a thermal runaway situation and a melted MOSFET. As it is, being able to draw about 24W stable is perfectly fine. I can push it up to 3A or about 36W but the temperature climbs up there dangerously close to the MOSFET’s 175°C operational limit. It’s not a dangerous runaway action and I can get a good minute and a half to two minutes of 36W operation before I need to dial it back so cool the MOSFET off.

Here it is happily drawing 1A at 12V (12W)

Here it is happily drawing 1A at 12V (12W)

Analysis

Really, that’s not bad. 30ish watts is not a bad number for a first crack at it. My ideal load would be able to draw 72W as the power supply intend to build should be stable capable of 24V at 3A operation. What is very apparent is I’m going to need a heck of a lot more cooling to reach that kind of power dissipation. That means a much bigger heatsink and/or a much beefier fan. Also, I do believe the addition of the silicone isolating pad which, while thermally conductive, is not as thermally conductive as it was just on the heatsink with paste. So it goes.

I did try using a much larger and more powerful AC fan and was able to achieve 36W stable operation. Using a mains voltage fan has a number of plusses in my design. First being that the 12V supply only has 200mA available for use so a beefier 12V fan is quite out of question. It is, however, no problem whatsoever to tap off another connection from the mains before that to power an AC fan. 50-60mm AC fans, after a quick poke online, are kind of pricey. The 80mm AC fan I have is quite the beast but could theoretically be bolted to the back of the case. Its power cable would be kind of flapping in the breeze and I’d have to crimp and tie that off nicely to ensure its safety.

The only thing I’m lacking to switch out the fans is appropriate M4 hardware to mount it onto the back. Well bugger, I’m always missing something aren’t I? Fortunately, its not a major concern for now. I could get some M4 bolts next time I’m out, or I could go out and find a smaller AC fan, or I could just leave it for now and keep this in mind for a future upgrade. 30W is good enough to begin my power supply testing and apart from that, I’m very happy with how my electronic load functions.

Here it is assembled and in standby mode. The 5mA there is a lie :) see previous posts.

Here it is assembled and in standby mode. The 5mA there is a lie :) see previous posts.

Comments are closed.