0

I am designing the following circuit to limit current inrush to a device on startup (to about 1.5 amps over about 3 seconds). Rload is just for testing, the actual load is a bunch caps at 56v, but I have not tested with those yet. Everything simulates fine. I have bread-boarded the circuit and if I apply 12vdc everything matches the simulation. If I apply 48vdc, it fails. I see about 4 amps, which is the overload limit of my power source and stuff gets hot pretty quick. It doesn't appear to be damaging anything, because I can go back to 12vdc and it goes back to about 1.5 amps. I assume it some kind of thermal runaway, but everything I have tried to fixed it has had little effect. The base resistors do not seem to make a difference. Before I added the emitter (ballast) resistors, I was blowing up transistors, so those seem to work.

So I expect this circuit to limit the current in the load to 1.5a when I apply 56v. It doesn't, and I don't know why. It works fine when I apply 12v.

Any ideas what is going wrong here?

limiter

DGB
  • 51
  • 5
  • 2
    Is this an actual design? If so, let's see how you approached the design. If not, and it's just a "borrowed and adapted by hunt and peck" design, then I think we need to hear a lot more about what you are actually trying to achieve. – jonk Jun 04 '20 at 18:04
  • How are you heatsinking those transistors? Even at 1.5A (0.5A each) they are dissipating 20W each, What thermal resistance (degrees/watt) is the heatsink rated for? What junction temperature does that give, and do they work at that temperatuce? –  Jun 04 '20 at 19:51
  • Can you expand on "it fails" - fails how? does what? doesn't do what ? What is the end current once charged? - you have (3 x 1R //) + 1 R = 1.333 Ohm series and a 3.85 Ohm load so about 25% losses in the series R's. Is that acceptable. || At 48V are the transistor safe areas violated? – Russell McMahon Jun 05 '20 at 02:58
  • Yes it is a design. – DGB Jun 05 '20 at 15:12
  • I originally started with 12 C/W, but the transistors barely heated during a test, so I changed to 26 C/W so I could feel if the device had passed any current during a test (a test last about 10 seconds). – DGB Jun 05 '20 at 15:22
  • Failure is that it should be limiting the current to 1.5-ish amps, but it acts as a short. The final current will end up being around 1/4 amp. This circuit is to replace a set of fixed resistors. I thought everything was in a safe area, but apparently I missed something. – DGB Jun 05 '20 at 15:23
  • 26C/W times 20W = dead transistors. –  Jun 05 '20 at 19:34
  • Yes for the steady state, but for a 10 second test, they only get warm with the 26s (no temp change with the 12s). Also after a test with the 26s at 48v, I can run a test with 12 volts and everything works fine (exactly the same response as before the test with 48v). – DGB Jun 05 '20 at 20:06

1 Answers1

0

I think I found it. It looks like I am violating the the forward bias safe operating area. The graph shows a current limit of about 0.3 Amps at 48 V (about 15W), well less than what I was designing for and more in line with what I was observing. Thanks Russell McMahon.

DGB
  • 51
  • 5