Dummy Load

[v1nd]

So I'm working on a project that requires me to put a 14A load on some batteries at 12V, I was just planning on throwing some resisters down with a large heat sink, but I neglected to take into account the voltage drop over time, changing the amount of amps the load is drawing.

I'm really not advanced enough to understand a lot of the circuit that was in the video that Dave put up, although more than some did make sense. Nevertheless, I still don't feel quite comfortable changing the setup on my own.

Instead, I was wondering if there's someplace I could buy a reliable dummy load at 14A. If it costs too much, I'll think about making my own...but that's once I check some prices.

Any replies would be helpful, Thanks!

[batee]

I've seen some awesome discrete current sources come from the DIY audio crowd.

Nelson Pass uses these in his open amp projects:
http://passdiy.com/pdf/zen-ver2.pdf

The part of Fig 1 that's the current source is:
R1, R6, R7, R11, C3, Q3, and Q2.

To change the current, change the value of R1.

He also shows some mods to the current source, detailed in the papers.

http://passdiy.com/projects.htm

Bryan

[batee]

Followup:

I noticed 14A as a requirement.  The FET shown in the design I linked to is rated for 11A.  Ideally the FETs should be run at <30% of their rated values, so it might be necessary to choose a different P channel FET or to use multiple FETs in parallel.

Here's the design, with the exception that it uses multiple N-channel FETs.  You should have no problems finding a single FET that'll handle 14A, and it shows how they can be connected in parallel to increase current-handling capacity.

[Zero999]

I noticed 14A as a requirement.  The FET shown in the design I linked to is rated for 11A.  Ideally the FETs should be run at <30% of their rated values, so it might be necessary to choose a different P channel FET or to use multiple FETs in parallel.

Here's the design, with the exception that it uses multiple N-channel FETs.  You should have no problems finding a single FET that'll handle 14A, and it shows how they can be connected in parallel to increase current-handling capacity.

I think you'll need to parallel the MOSFETS.

The problem is not the current rating but the power dissipation, the circuit dissipates 168W, more if the battery voltage is on the high side. You need at least four MOSFETS, five would be ideal.

[mikeselectricstuff]

Multiple fets can help spread the heat out over a heatsink, but 168W in a single  part is probably doable with good thermal coupling (i.e. no insulating washer!).
However a  cheaper solution may be to dump the majority of the power in a resistor and the remainder in the fet.
For a battery discharge application, you're not going to care about maintaining constant curent down to a low input voltage, so connecting a power resistor in series with the FET calculated to draw a bit more than 14A at whatever is your minimum input voltage of interest  will divide the power between the R and the Fet, as well as protecting the fet from transients. You could also use the resistor for current sensing.

[Zero999]

Yes, dumping most of the power in a resistor is a sensible approach.

I doubt you'll be able to safely dissipate 168W with a single part, of course there are MOSFETs with ratings like that but it's normally not practical when the heatsink, operating temperature and the thermal resistance of the part to heatsink are taken into account. Certainly nothing in a TO-220 package is going to hack it.

For example, here's a link to a MOSFET 'rated' to 600W, with a perfect heatsink and an ambient temperature of 25°C, in reality you need to design for a less than perfect heatsink and a higher ambient temperature than 25°C,typically 30°C or 40°C, unless you're sure it's only going to be used in an air conditioned room. The thermal resistance from junction to the tab is 0.208°C, add another 1°C to account for a good fan/heatsink and the thermal resistance between the tab and heatsink and the maximum dissipation is under 100W, at an ambient temperature of 30°C.
http://www.farnell.com/datasheets/39854.pdf

Perhaps your best bet is to solder some high powered SMT MOSFETs to a piece of copper busbar mounted on a large heatsink with a fan. The resistance between the device and copper bus bar will be negligible and the resistance betweent he busbar and the heatsink should be much less than the device to heatsink would've been - I've done this before.

[alm]

To answer the original question, yes, you can buy commercial off-the-shelf dummy loads, but they're not as common as power supplies, so expect to pay more. I'd expect at least a few hundred $, even from cheaper manufacturers. Of course, Agilent makes them, for a price. Cheaper brands charge at least €500 for them (nothing special about this German distributor, it was just the only source I could recall off the top of my head, 'Unser Preis zuzügl. MWSt.' means 'Our price excluding VAT'). So if it's for private use, DIY is probably the way to go.

I agree with Hero999 that you'll probably need to spread the load over multiple MOSFETS (or resistors), purely to keep the junction temperature low enough (below 150?C or so) without extreme cooling (that's what those max dissipation graphs are about, the junction-case thermal resistance and the junction temperature).

[batee]

In the current sources, the max power dissipation won't be 12V * 14A = 168W.  Some of that energy will be dissipated in the source resistors.

Rsource = 0.65 V / Iload = 0.65/14 = 0.0464 Ohms

Prsource = Iload^2 * Rsource = 14*14*.0464 Watts = 9.09W.

To reduce the voltage dropped by the FET, and therefore the amount of power the FET has to dissipate, it would be possible to insert more resistors in the circuit (as the "load" shown in the current source diagrams). 

An array of Power BJTs would be infinitely more adjustable. 

[Kiriakos-GR]

I would use car lamps H4 , as many the load needs ..

By having them on,  for three minutes the load stabilizes it self , and thats it.