Linear regulator

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi folks,

Im trying to project a linear regulator for +-47V and current limit of 20A pulse (protect the transistors).
Mains devices will be mj15003/4
Transformer is 42+42AC >1KVA

This is the circuit that I think and simulate. Is it OK?
 

Attachments

  • regulator.png
    regulator.png
    31.8 KB · Views: 557
Learn about "foldback" current limiting and "SOA" (safe operating area) protection. Constant current limiting is not usually suited for linear circuits because it can still lead pass transistors to huge power dissipation and instantaneous failure. The trick is to make the current limiting threshold dependent on the voltage across the pass transistor, time, and on its temperature if overcurrent conditions are frequent or sustained.

Bipolar power transistor datasheets contain SOA curves.
 
thanks Eva,

I was see previously the SOA of MJ15003
DC current is 20A below 10V Vce
Pulse is 30A
Second breakdown is above 60V Vce


The only think that I need to avoid is 60Hz peak capacitors charge current that exceed 30A (I adopt 20A threashold) at full load or at start.
I will operate whit fuses of 15A at mains (will open in case of short or overcurrent).

The principal objective is to drop peak voltage from +-58V to +-48V to feed Ucd modules. I want the minimum possible drop at BJTs, minimum dissipation.

YES, my last post circuit can instantaniously damage MJ15003/4 wich a peak power at start capacitor charge but current not ultrapasses 20A.
:dead:

Is this true? I will always be a series NTC at 220V Mains

I need a example of foldback, without opamps
 
From the cct. U posted, both output transistor cannot deliver 20A, why? this due to not enough Ib, by R1, and R6 to supply the base current !!!. Suppose mj1500X Beta = 20, Ib must have 1A, another problem is if the current limited tr. turns on, can it stand 1A, otherwise it will breakdown and burnt. the Zener diode connected to the error amp. tr. E and B term. ??? mistake.

SOA, the BVce > 60V, it seems not enough, usually we only use 1/3 of the BVce, U have also consider heat dissipation, continue operation--- many factors.
 
Have you considered getting the amp working with a conventional transformer + rectifier + smoothing caps?
Then learn how to build a regulator for low current/low voltage duty.
Expand your knowledge to include high voltage, then try high current + high voltage and finally go back to investigating current limiting when you have some knowledge of temperature de-rated SOAR.
Once you have all that knowledge look to adding in a current limited HiV HiI PSU for your amp.

Oh,
and keep in mind that the final 4.7mF cap will provide a very Hi current to the amp for a very short time using up all or most of the transient capability of your output devices.
 
janneman said:
I think the original poster was talking about the SOA of the supply series transistor, while you guys are talking about the SOA of the amp that is supplied by this circuit. If so, this can develop into an interesting discussion
Hi,
I was specifically referring to the regulator SOAR, but my comment
back to investigating current limiting when you have some knowledge of temperature de-rated SOAR.
could equally well apply to the amplifier SOAR.
Understanding both allows the design capability to be optimised economically rather than throwing resources at it and hoping to achieve reliability.
 
AX tech editor
Joined 2002
Paid Member
AndrewT said:

Hi,
I was specifically referring to the regulator SOAR, but my comment could equally well apply to the amplifier SOAR.
Understanding both allows the design capability to be optimised economically rather than throwing resources at it and hoping to achieve reliability.

Andrew,

Correct, but in the reg the pass transistor normally has only a few volts Vce so SOA would not be an issue there. In the amp it is, of course.

Jan Didden
 
Temperature de-rated SOAR is always an issue for an economic and fast and current limited PSU regulator.
If it weren't them fold back current limiting would not have needed to be invented.
High capacitance on the output of the PSU puts the amplifier at risk and can make the design more troublesome. Reduce the output capacitance to keep those amplifer SOAR risks low and you find that the regulator must meet the peak current demand. Now the foldback becomes even more important.
Consider what happens if the mains is running high through periods of low demand. PSU Vce (and dissipation) can easily exceed double what might occur during periods of high power demand.
 
AX tech editor
Joined 2002
Paid Member
AndrewT said:
Temperature de-rated SOAR is always an issue for an economic and fast and current limited PSU regulator.
If it weren't them fold back current limiting would not have needed to be invented.
High capacitance on the output of the PSU puts the amplifier at risk and can make the design more troublesome. Reduce the output capacitance to keep those amplifer SOAR risks low and you find that the regulator must meet the peak current demand. Now the foldback becomes even more important.
Now consider what happens if the mains is running high through the night. PSU Vce (and dissipation) can easily exceed double what might occur during periods of high power demand.

Andrew,

I don't disagree on these issues, but even with a steady state 10V across the series pass xsistor, SOA is no issue, especially since those 20A loads will be few and long between. And if Vce is more, you have the wrong transformer ;)

With foldback limiting you are basically looking at a single transition from low Vce and high Ic to low Ic and high Vce, unlikely that SOA would be a problem.
I agree it has to be looked at carefully to be sure, but I would first worry about the amp SOA.

Jan Didden
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.