hi yef
Added text.
E

 

What is this? What function?

 

Its away to create a current limited voltage... The current through V4 can never exceed I1 or I2, steered by ideal diodes D4/D5.

Unfortunately I've not been able to duplicate Eric's example, with or without the current limiter. @ericgibbs, can you post your .ASC file please.

 

Its away to create a current limited voltage

I don't do current limit supplies because in SPICE parts don't break when you apply 1,000,000 amps or 20,000 volts. I might put a 1 ohm resistor in the voltage source.

 

Hi @Irving
Re-run the same asc file this morning, showing the plot for Vload.
Works OK. LTS XVII (x86) Win 10.
E

 

Ah, it looks like its working but it isn't. The acid test is not the voltage wrt GND, which is a 10Meg resistor away (tho that resistor doesn't exist IRL) but the voltage across R6 , or better, the current through R6. Plot that and you'll see what I mean...

I went back to the UCC5304.txt file to see if I could work out why it doesn't like VGND (or VSS as was) floating. Suffice to say, its a horrible collection of voltage & current sources and switches to implement startup,brownout UVLO, OVLO,etc, functionality. However, the key thing we are interested in is the output function, and conveniently, this is broken out as a seperate subcircuit which I've tested independently and this works as expected:



Here is the output driver schematic I recreated from the subckt: Note the definition of the output impedance R1 and the behavioural current source B1 which implements a current lookup interpolation table for the voltage across R1 which is then passed through a parallel 1ohm resistor to give a voltage Vt. The numerical value of Vt is then used to give an instantaneous value of R1 per iterative cycle which in turn,together withthe on resistance of the high and low switches simulates the dynamic current limits defined in the data sheet.

 

Ah, it looks like its working but it isn't.

Hi @Irving,
I agree, but the TS's problem that he raised, is that it would not run in simulation.
So I suggested the options parameters to be added so that it would run.

The TS did say he has a 48Vdc 4.5A PSU, but requires a 48Vdc pulse for a few uSecs at 75Amps
I believe he is using the 5304 source/sink device as Gate driver for the MOSFET, which will discharge a high value capacitor, which has been charged from his 48V 4.5A PSU, through the R6 for the uSec period.

But he is showing a R6 load value of 1.5R, which as you know will result in a maximum current of 32Amps.

IMHO, the circuit design requires a redesign.

E

 

Hi @ericgibbs

I hear you... it's one thing not running, but another making it run but not producing meaningful outputs. This is one at least 3 or 4 threads on this issue, and the TS's requirements have expanded and changed over the last few weeks. I'm still not clear of the end goal - originally it was a requirement to sequence 3 power rails (+5, -2 then +20v) to a Ga-As 10GHz RF amp, then it morphed to switching the +20 at 1MHz 50% duty cycle for no obvious reason to two parallel amps 20v @ 1.5ohm total (and not sure if that would damage the amps.) And now a new 48v requirement...