TheMGPremium
join:2007-09-04
Canada
kudos:1 | reply to TheMG
Re: DIY Programmable Load - preliminary design - thoughts? So you think something like this would work best and just try to match the FETs so they are roughly equal?
I assume that with such an arrangement, it will be important for all the FETs to be mounted on the same heatsink relatively close together to avoid problems with thermal runaway?
I have some heatsinks I plan on using which are about 10 inches long. My concern is that the FETs mounted furthest away from the cooling airflow will run much warmer, thus putting a lot more stress on them than the other FETs mounted closer to the cooling airflow. |
|
| Yeah, that's what I was thinking about.
How will the transconductance of the FETs change when they warm up? Are they going to conduct more current, or less?
--
Wacky Races 2012! |
|
public
join:2002-01-19
Santa Clara, CA | reply to TheMG
said by TheMG:So you think something like this would work best and just try to match the FETs so they are roughly equal?
I assume that with such an arrangement, it will be important for all the FETs to be mounted on the same heatsink relatively close together to avoid problems with thermal runaway?
I have some heatsinks I plan on using which are about 10 inches long. My concern is that the FETs mounted furthest away from the cooling airflow will run much warmer, thus putting a lot more stress on them than the other FETs mounted closer to the cooling airflow.
I have a similar analog controlled load.
20W per TO220 fet, 15 total, on a fan cooled heat sink.
The current is measured by summing the top ends of the source ballast Rs. No shunt is needed.
Pot control is preferred since it can be adjusted rapidly, a requirenent for testing.
A panel meter showing the current is adequate.
Also consider adding a constant current, constant resistance mode switch. |
|
public
join:2002-01-19
Santa Clara, CA | reply to aurgathor
said by aurgathor:Yeah, that's what I was thinking about.
How will the transconductance of the FETs change when they warm up? Are they going to conduct more current, or less?
The ballast Rs equalize the current, no matching is needed. |
|
iknowPremium
join:2012-03-25 | reply to TheMG
said by TheMG:So you think something like this would work best and just try to match the FETs so they are roughly equal?
I assume that with such an arrangement, it will be important for all the FETs to be mounted on the same heatsink relatively close together to avoid problems with thermal runaway?
I have some heatsinks I plan on using which are about 10 inches long. My concern is that the FETs mounted furthest away from the cooling airflow will run much warmer, thus putting a lot more stress on them than the other FETs mounted closer to the cooling airflow.
FETS don't suffer from thermal runaway, as the forward voltage drop increases with temperature, making them good candidates for paralleling, as the currents will equalize. the hotter devices will draw less current, until the temperatures equalize. |
|
