LM317 in a high-voltage circuit.

[pawihte]

The specs for the LM317 says that, since it uses a floating
circuit, it can be used at high voltages as long as the
input-output differential is within max ratings. But practical
loads will place a capacitance at the output of the LM317, with
many having a physical capacitor. That capacitance will
momentarily short the output to ground at switch-on, causing the
LM317's max voltage rating to be exceeded. An example would be a
fast-rising 100V input and a 75V output with a large cap at the
output. Is this likely to damage the LM317?

 

[pawihte]

---
From:

http://www.national.com/ds/LM/LM117.pdf

"Besides replacing fixed regulators, the LM117 is useful in a
wide variety of other applications. Since the regulator is
"floating"
and sees only the input-to-output differential voltage,
supplies of several hundred volts can be regulated as long as
the maximum input to output differential is not exceeded, i.e.,
avoid short-circuiting the output."

One would think so.

I have the NatSem linear databook in both CD and printed forms.
But it doesn't say anything about momentarily exceeding the
differential voltage rating which seems to be inevitable in many
practical applications such as the one I described.

 

[Phil Allison]

"pawihte"

I have the NatSem linear databook in both CD and printed forms. But it
doesn't say anything about momentarily exceeding the differential voltage
rating which seems to be inevitable in many practical applications such as
the one I described.

** Not inevitable at all, if you add some simple protection for the IC.

Never heard of Zener diodes ???

One wired across the input and output pins might help.

Wot a fuckwit ......




..... Phil

 

[pawihte]

"pawihte"

** Not inevitable at all, if you add some simple protection for
the
IC.
Never heard of Zener diodes ???

One wired across the input and output pins might help.

Wot a fuckwit ......

Ah, good ol' Phil. What happens to the zener diode, say a 33V or
39V piece, when you apply +100V (which may have a fairly low
series resistance) at one end and the other end is momentarily
grounded? Zzzzt!?

 

[pawihte]

I was going to say that he might also put a bit of resistance
before
the output capacitance.

George H.

But that will defeat the purpose of having a regulated supply -
at least partly.

 

[Michael Black]

The specs for the LM317 says that, since it uses a floating
circuit, it can be used at high voltages as long as the
input-output differential is within max ratings. But practical
loads will place a capacitance at the output of the LM317, with
many having a physical capacitor. That capacitance will
momentarily short the output to ground at switch-on, causing the
LM317's max voltage rating to be exceeded. An example would be a
fast-rising 100V input and a 75V output with a large cap at the
output. Is this likely to damage the LM317?

And how much capacitance? If you're using high voltage, surely these
are no longer low impedance devices, hence you won't see all the high
value capacitors seen on low impedance circuits. Yes, the circuitry
itself will present some capacitance to the regulator, but it being
incidental, how serious can it be?


Michael

 

[Joerg]

The specs for the LM317 says that, since it uses a floating
circuit, it can be used at high voltages as long as the
input-output differential is within max ratings. But practical
loads will place a capacitance at the output of the LM317, with
many having a physical capacitor. That capacitance will
momentarily short the output to ground at switch-on, causing the
LM317's max voltage rating to be exceeded. An example would be a
fast-rising 100V input and a 75V output with a large cap at the
output. Is this likely to damage the LM317?

The rule is that you can't exceed abs max values, not even for a very
short time. Chances are, something avalanches in there and ... tsk ... *BAM*

 

[Phil Allison]

"pawihte"

Ah, good ol' Phil. What happens to the zener diode, say a 33V or 39V
piece, when you apply +100V (which may have a fairly low series
resistance) at one end and the other end is momentarily grounded? Zzzzt!?

** Applying the +100V DC input should NOT be an instantaneous event and
grounding the output is simply not allowed when using IC regs with excess
input voltage.

Long as the 100V supply takes a few milliseconds to rise, a hefty ( say 5
watt) zener will be fine passing enough current to help charge a 10 uF
electro on the load side.

BTW zeners fail by going short - so the reg IC will survive.



...... Phil