317 CCS on 60Vdc

[AndrewT]

Hi all,
is there a way to configure a 317 as a CCS that can under near short circuit conditions allow it to continue limiting at well above it's in-out 40V limit?

I want to use dual polarity supplies of between +-45V to +-60V with adequate smoothing AND to pass bias currents upto 400mA.
The two CCS to be set to a limit just above normal bias current.

I do not want excessive volts drop across the CCS when within the "normal" bias range.

In normal operation the 317+resistor does this duty quite well on it's own.
But in short circuit accident mode the 317 will go overvoltage.

[analog_sa]

Have a look at the CCS portion of Allen Wright's supershunt.

[leadbelly]

Why not use the LM317HV?

[analog_sa]

Because it will leave no safety margin? Or because he wants to use whatever he's got at hand?

[jackinnj]

If you put a resistor before the regulator it will IR drop V...or is this too simple -- it's what's done with the Maida to do some of the heavy lifiting.

TL783 will work but only 700mA, so will a pair of depletion MOSFETs.

fwiw, the TL783 datasheet was updated in April of last year.

[AndrewT]

Originally posted by jackinnj
If you put a resistor before the regulator it will IR drop V...or is this too simple --

too simple.
If I am drawing 100mA of bias @ +-50V.
Set the CCS to 120mA.
Now short the output (a slipped probe).
The resistor has to drop ~10V to keep the 317 below 40V.
The resistor will require to be ~83r.
Now back to the "normal operation with CCS passing 100mA.
The +-50V input (from the PSU) will be ~ 50 - 83*0.1 - 1.4V(317Vdrop) - 1.25V ~=+-39Vdc.
Most of that loss is down to the 83r protection resistor.

I'll download the TL783 datasheet, it may help.

[AndrewT]

I've tried Googling the Allen Wright shunt reg. I cannot find a schematic of the CCS.
Can anyone post a link to it?

[analog_sa]

Lets say this bears some similarities

[AndrewT]

I tried a two transistor CCS,
2n5401 & 2sc1943.
When passing ~80% of CCS rated current it drops only 700mV.
As it approaches 100% of CCS rated current the voltage drop rises rapidly.

This is far lower than the Vdrop across a 317+resistor.

That two transistor version is what I will build up to insert into the PSU output leads, after the smoothing caps.

Ensure you have sufficient current bypassing the load to keep the 2n5401 well away from saturation. Even Ic=30mA @ 1.3Vce is not a problem for the 2n5401.

Why, you are asking?
Because I blew yet another output stage when a probe shorted the driver's C to E legs and my lab supply only goes upto +-31Vdc

[AndrewT]

It works!

The 51.5Vdc supply loses 0.5V through the protective CCS when the load is 200r (255mA)
Short out one of the 50r resistors leaving 150r and the CCS starts to limit, 45V across the load (~300mA) and 5.5V across the CCS.
Short out the whole load and the CCS drops all 51V.
The 8C/W sink, over a couple of seconds, does not even get warm.
No sparks, no overheating.
All returns instantly to normal on removing the short.

I wish I had built this up a long time ago.
Now for the negative supply, using 2n5551 & 2sc5200