LM317: Can you kill it with an output short

[L.Chung]

Hi,

I saw a circuit on the web that puts overcurrent protection (via
current sensing resistor and transistor)to the input of the LM317 to
"protect" it from output short circuit.

However, form the data sheet of LM317, if the protection diodes are
fitted, it appears to me that the LM317 can survive an output short
without damage.

So if the LM317 is operating within its input voltage limit, will it
survive an output short? Or, put simply, have you killed one recently?

Regards

L.Chung

 

[SioL]

Hi,

I saw a circuit on the web that puts overcurrent protection (via
current sensing resistor and transistor)to the input of the LM317 to
"protect" it from output short circuit.

However, form the data sheet of LM317, if the protection diodes are
fitted, it appears to me that the LM317 can survive an output short
without damage.

So if the LM317 is operating within its input voltage limit, will it
survive an output short? Or, put simply, have you killed one recently?

Regards

L.Chung

Did this whole arrangement also have an additional "output" series
transistor to increase the current rating of the power supply? Like
the output of 317 going to the base of say 2n3055?
If so, short-circuit protection would not work for the whole circuit.

 

[L.Chung]

Did this whole arrangement also have an additional "output" series
transistor to increase the current rating of the power supply? Like
the output of 317 going to the base of say 2n3055?
If so, short-circuit protection would not work for the whole circuit.

No. Just a plain LM317 on its own.

The question remains is whether a LM317 along with the manufacturer
recommended protection diodes can survive an output short.

L.Chung

 

[Michael Redmann]

However, form the data sheet of LM317, if the protection diodes are
fitted, it appears to me that the LM317 can survive an output short
without damage.

Yes, the spec say output is short-circuit protected. But current limit
is higher than 1.5 A depending on package (TO-3: max. 3.4 A). Perhaps
additional circuit switches off at lower current.

Regards
-
Michael Redmann
"It's life, Jim, but not as we know it." (Spock)

 

[Tony]

AFAIK, the LM317 is short-circuit proof, but there is no reason that
external circuitry can't be added to limit the current to some lower
value (eg to protect some other circuitry). The only "protection
diode" usage I know of is for reverse voltage protection, not
short-circuit protection.
Tony

No. Just a plain LM317 on its own.

The question remains is whether a LM317 along with the manufacturer
recommended protection diodes can survive an output short.

L.Chung

Tony (remove the "_" to reply by email)

 

[Winfield Hill]

Michael Redmann wrote...

Yes, the spec say output is short-circuit protected. But current limit
is higher than 1.5 A depending on package (TO-3: max. 3.4 A). Perhaps
additional circuit switches off at lower current.

The LM317's high intrinsic current-limit value means it may be under
severe thermal stress during current limiting, so that the thermal
temperature-limit circuit is also activated. There's some evidence
the chip's life is shortened if it spends lots of time at the thermal
limit temperature. Furthermore, other components in the power supply
(transformer, rectifiers, etc.) are also more severely stressed when
at the chip's intrinsic current limit. OTOH, if an external current
limit circuit is added, this circuit will be under severe stress.

 

[Spehro Pefhany]

Hi,

I saw a circuit on the web that puts overcurrent protection (via
current sensing resistor and transistor)to the input of the LM317 to
"protect" it from output short circuit.

However, form the data sheet of LM317, if the protection diodes are
fitted, it appears to me that the LM317 can survive an output short
without damage.

So if the LM317 is operating within its input voltage limit, will it
survive an output short? Or, put simply, have you killed one recently?

Regards

L.Chung

With both thermal and current limiting, it should be able to withstand
an output short provided the input voltage is within the limit. It may
not do much for the long-term reliability to depend on the thermal
limit, especialy for more than a momentary short-circuit, as the die
under these conditions will exceed the abs max operating Tj by a
considerable margin.


Best regards,
Spehro Pefhany

 

[SioL]

With both thermal and current limiting, it should be able to withstand
an output short provided the input voltage is within the limit. It may
not do much for the long-term reliability to depend on the thermal
limit, especialy for more than a momentary short-circuit, as the die
under these conditions will exceed the abs max operating Tj by a
considerable margin.


Best regards,
Spehro Pefhany

To sum it up, if its very likely that the circuit will encounter a short, it probably
is better not to rely solely upon current protection inside the LM317.
It is probably manufactured by a multitude of vendors today and you can't rely
on how a particular lot will perform, unless you order from the same vendor every
time.

 

[Ban]

AFAIK, the LM317 is short-circuit proof, but there is no reason that
external circuitry can't be added to limit the current to some lower
value (eg to protect some other circuitry). The only "protection
diode" usage I know of is for reverse voltage protection, not
short-circuit protection.
Tony

Hey, Tony please put your answer below to not disrupt the discussion.

The protection diodes of the OP are against an *Input* short at higher
voltages with big caps on the regulator and output. They prevent the charge
flowing backwards through the regulator and eventually destroying it.

 

[L.Chung]

Thank everyone for their inputs. I am referring to a momentary output
short situation.

I guessed that it should survive but prolonged short would put it in
great stress which is likely to damage it as Winfield Hill stated.

Have anyone ever fried one with a momentary output short?

Regards

L.Chung

 

[Phil Hobbs]

Thank everyone for their inputs. I am referring to a momentary output
short situation.

I guessed that it should survive but prolonged short would put it in
great stress which is likely to damage it as Winfield Hill stated.

Have anyone ever fried one with a momentary output short?

Regards

L.Chung

You can fry one with an output short if there's a big filter capacitor (>= 10
uF) on the adjustment pin.

Another reason for an auxiliary current limit is cost-benefit analysis. The
circuitry powered by the LM317 may be much more valuable than the regulator.
In a $10k system, reducing field failures from 5 per year to 3 per year
might easily justify another $2 in parts. Wearing a belt and suspenders costs
something, but if you do it right, it can really reduce the number of
midnight phone calls.

Cheers,

Phil Hobbs

 

[mike742]

I read somewhere that some 3 terminal regulators could
be killed when they when into thermal limit due
to excessive input voltage from leakage inductance
of the power transformer. The solution to this
was a much larger input capacitor before the
regulator (n*1000uf, not just 10 uf or 100 uf).

 

[[email protected]]

An IC needs safe area protection as well as current limit to be
properly killproof.

NT

 

[Spehro Pefhany]

An IC needs safe area protection as well as current limit to be
properly killproof.

NT

Which the LM317 has. "Included on the chip are current limit, thermal
overload protection and safe area protection."



Best regards,
Spehro Pefhany

 

[Tony]

Hey, Tony please put your answer below to not disrupt the discussion.

The protection diodes of the OP are against an *Input* short at higher
voltages with big caps on the regulator and output. They prevent the charge
flowing backwards through the regulator and eventually destroying it.

I missed that post. The diodes may be the same ones I was thinking of
- across the reg (prevents excessive reverse in-out voltage across the
reg, eg under the conditions you mentioned), and to common (prevents
reverse voltages to common) (should have clarified)?

Tony (remove the "_" to reply by email)