Understanding the function of an LT1083

[si2030]

Hi There,

I have a problem understanding the function of a circuit I have created in LT spice.

Background: My intention was to create a constant current limit with the linear regulator LT1083. This regulator has current limiting built in and it can handle current up to 7.5Amps. If you gang two together (as per the datasheet including 0.015Ω resisters for load balancing) it will current limit if more than 15 amps is reached.

I have used an AD8211 current sense IC (model from Analog) and am sensing the input of the LT1083s intending to limit the current sensed before the LT1083s. I got this idea from a series of video blogs on a power supply by David L. Jones on his blog EEBlog.

The problem I am facing and which has been alluded too in a reply to a post on the LT forum I made last night is that the LT1083s limit current is based on input voltage to output voltage. I don't understand this properly and how it impacts what I am trying to do.

You can view my circuit working and see that I have placed a load of 1Ω. I have an input of 30 volts and I am setting the LT1083s to 28.8 volts via the input of a voltage on the ADJ pin of 27.5V+1.3V=28V.

However, in the second trace from the bottom you can see that i_select and i_measure instead of being together here (i_select should limit i_measure using the BC547) i_measure is way lower than i_select because the LT1083s are doing their own current limit and I dont know why. What am I missing here?

Could someone please explain why this circuit isnt working and what I need to do to make the constant current limit sub circuit take control of this..

Kind Regards

Simon

 

[alec_t]

the LT1083s are doing their own current limit and I dont know why

The voltage across the 1Ω load is initially 0 whereas the '83s' input voltage is 30. Therefore each '83 is being asked to pass 30/2 = 15A. So the internal current limit (7.5A) is exceeded and the '83 shuts down. As the datasheet explains, it may be necessary to remove the supply voltage or output 'short-circuit' (1Ω) before the '83 can reset itself.

 

[alec_t]

Try increasing R2 to ~15k and making the load at least 4.5Ω. The current control by the sub-circuit then works as you want, up to the point where the '83 internal current limiter takes over.

 

[si2030]

Is this the fact that the BC547 is not removing the voltage on the ADJ pin? That is, the opamp is not being shorted when the BC547 is being turned on?

 

[si2030]

The fundamental fact remains though that this circuit has current traveling into the regulator and out and it should not matter whether you test the current before or after the regulator. If two lt1083s together can handle up to 15 amps and you set the current limit to 10amps then the current limit in the Lt1083s should never be tripped.... so why is the current limit in this circuit totally ineffective with a load below 4 ohms?

anybody?

 

[alec_t]

If you look at the graph of the short-circuit current on the '83 datasheet you will see that if the voltage difference between input and output is > ~22V the current will limit at ~1A. At start-up there is a 30V difference in your circuit, so the '83 probably starts and stays tripped.

 

[si2030]

Double crap.... Would you believe I had this working with a rather less impressive attempt at current limiting only in that simulation I had a feedback system - TL494 pre-regulator - that kept the input only a couple of volts above the input... of course it worked a treat but when I knocked up a sub set of that simulation to test a better method of constant current limiting I just thought to throw in 30 volts input not ever knowing of this little trick for young players... I immediately dropped the input voltage down to 20 volts in the sim and it works perfectly... many thanks for this... saved me a lot of head scratching...