Author Topic: High-Wattage Dummy Load  (Read 22591 times)

0 Members and 1 Guest are viewing this topic.

Offline orion242

  • Supporter
  • ****
  • Posts: 746
  • Country: us
Re: High-Wattage Dummy Load
« Reply #25 on: October 11, 2013, 02:13:27 am »
all mosfets must be insulated from each other by a sil-pad or similar, you will want a fan performing active cooling, but without selecting a capable heatsink, cannot calculate air flow rate,

I’m working on the same thing myself.  A dummy load that can handle ~1KW @ 12v.

Currently I’m using 10x 100W 1ohm resistors (all parallel for 0.1ohm effective) for the current sense.  The low sense resistance allows me to test the 5 & 3v3 rails at high current.
http://www.ebay.com/itm/151062972593?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1439.l2649

The resistors are pretty cheap IMO.

Anyhow, do the MOSFETs really require isolation from each other and the heat sink?  They are all sharing the same power input and path to the sense resistor.

you take the op amp, mosfet and load resistor, and parrellel as many of those segments as you need to reach your rated current at 12V, (like i said in the chat box)

I currently have a single op amp looking at the sense resistor bank and driving 6 MOSFETs.  It seems to work well at medium power, but at low loads it oscillates to no end.  I’m having been working this out in my free time and was thinking it was interference or more likely the offset voltage of the op amp not allowing it to drive low enough with single rail supply.  Now you have me questioning my single amp design.  Is this really required and why if the MOSFETs will naturally share the load nicely?

Thanks in advance
« Last Edit: October 11, 2013, 02:22:55 am by orion242 »
 

Offline Harvs

  • Super Contributor
  • ***
  • Posts: 1202
  • Country: au
Re: High-Wattage Dummy Load
« Reply #26 on: October 11, 2013, 03:13:11 am »
Is this really required and why if the MOSFETs will naturally share the load nicely?

No they won't, not even close to share nicely!

The Ids/Vgs is a very non-linear curve, which they're not all matched, and the gain has a strong dependence on temperature in the wrong direction.  If one mosfet starts hogging current and getting hotter, it will proceed to hog the current even further.

Have a look at the schematics for the HP 6030B for a good design of paralleling multiple mosfets in a load.  There's a few clever things they've done in that design.
 

Offline poorchava

  • Super Contributor
  • ***
  • Posts: 1672
  • Country: pl
  • Troll Cave Electronics!
Re: High-Wattage Dummy Load
« Reply #27 on: October 11, 2013, 07:13:29 am »
Water cooling. I have once wandered onto some forums where guys were building monster audio amplifiers (several kW RMS). Whatever you do, you can increase cooling efficiency by throwing whole setup into water or mineral oil. This is not a solution for a 24-hour test, but for an hour - easily.
I love the smell of FR4 in the morning!
 

Offline johansen

  • Super Contributor
  • ***
  • Posts: 1081
Re: High-Wattage Dummy Load
« Reply #28 on: October 11, 2013, 08:17:50 am »
I have a kilogram of A1 nichrome left over... handles 40 amps in my furnace at 1300C just fine. You would ne
ed about 10 $ worth of it
 

Offline eevblogfan

  • Frequent Contributor
  • **
  • Posts: 569
  • Country: 00
Re: High-Wattage Dummy Load
« Reply #29 on: October 11, 2013, 08:52:40 am »
Hey

for that high power , the best solution IMHO is very big resistor , with say 150W load in A control loop ,

the electronic load will keep constant current , until the resistor warm up , then the resistor is warming hence voltage drop thus lower current right ? . well , the CC is going to dissipate less power in order to "move" the required power to the resistor , 150W is easy , safe and cheap , you can either buy or get for free one of those 2000W heaters ( fan cooled ) ,  open it up , hack it in order to be one huge resistor ( depending on your load ) , if 1 Ohm needed , cut it every  7 Ohm or so , you'll get 7 resistors , parallel them up and you have got 2KW 1Ohm resistor :)  , you may want 0.12 Ohm right ? (100A 12 V ) , well , 20 pieces of 2.4Ohm is going to do the trick :)

It is going to be very course , Hence if you need like 1~5 % accuracy , add some active load to react in order to draw the necessary current ,

that is my take on crazy ass KW loads , I hope someone will find it helpful :)

Cheers :)
 

Offline orion242

  • Supporter
  • ****
  • Posts: 746
  • Country: us
Re: High-Wattage Dummy Load
« Reply #30 on: October 30, 2013, 01:42:33 am »
Is this really required and why if the MOSFETs will naturally share the load nicely?

No they won't, not even close to share nicely!

The Ids/Vgs is a very non-linear curve, which they're not all matched, and the gain has a strong dependence on temperature in the wrong direction.  If one mosfet starts hogging current and getting hotter, it will proceed to hog the current even further.

Have a look at the schematics for the HP 6030B for a good design of paralleling multiple mosfets in a load.  There's a few clever things they've done in that design.

Thanks.  Made the changes and its humming along at 250w smoothly and keeping cool.  Parts coming to to scale to 1kw (x2 500w).
« Last Edit: October 30, 2013, 01:46:36 am by orion242 »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf