LM317 Votage Regulator Instability?

C

Cursitor Doom

Guest
Hi all,

I'm having serious difficulty trying to get this issue resolved and was
wondering if anyone here has come across the same problem.

I've got a linear PSU that uses a LM317 (TO-220 case) to drop 37V down to
12V for a small 555 timer board that's set up as a pulse generator (at
about 0.1Hz). The thing is, as soon as I connect the board, the 12V from
the reg drops to between 6V and 7V and wanders around this level (like as
if I'd put a really heavy load on it). But the 555 board's resistance is
about 22k ohms so it only needs a very small current. When I swap the
board for a 10W 50 ohm power resistor drawing 0.25A, the reg output is
totally fine, staying rock steady at 12V and the resistor and reg get
warm as you would expect.
I can only think there may be something about the timer board that
'upsets' the LM317 and causes it to incorrectly partially shutdown,
because nothing on the board is getting warm and trying spare identical
timer boards causes the voltage to drop by the same amount also.

I can't recall this ever happening before! Any suggestions?
TIA.
 
On Thu, 14 Jul 2016 11:53:30 -0400, Phil Hobbs wrote:

Too small an output cap, or too low an ESR? Three terminal regulators
can oscillate. Try hanging 100 uF aluminum electro on the output and
see if it starts behaving better.
Click to expand...

The datasheet says output caps are "optional" although I do always use
'em. The one I'd fitted was 1nF which seemed a bit small, so I subbed it
for a 100n but it's made no difference. 100uF seems HUGE! I might try it,
though.
 
On Thu, 14 Jul 2016 11:49:42 -0400, Ralph Mowery wrote:

Try loading it down some. The data sheet calls from about 3 to 10 ma as
the minimum load.
Click to expand...

Tried it before posting but no joy, I'm afraid.
 
On Thu, 14 Jul 2016 18:40:16 +0100, Lee wrote:

JOOI What version of the 555? If it's a bipolar version they can exhibit
a "crowbar effect", and the 100uf cap is usually suggested to
ameliorate it.
Click to expand...

The ones I'm using here are all NE555CN.
 
On Thu, 14 Jul 2016 13:29:38 -0400, Bob Engelhardt wrote:

On 7/14/2016 11:40 AM, Cursitor Doom wrote:
... But the 555 board's resistance is
about 22k ohms so it only needs a very small current. ...

Can you measure what the board is actually drawing?
Click to expand...

When connected to an external bench power supply at 12V., the board draws
9.3mA. The bench PSU doesn't have any issues powering this board.

When connected to PSU that uses the LM317, however, it draws 3mA and the
voltage across it drops to around 6V.
 
On Thu, 14 Jul 2016 10:35:32 -0700, Jeff Liebermann wrote:

Maybe you blew up the
LM317?
Click to expand...

Don't see how that's possible, Jeff, since it's still happy to supply
passive loads that are much heavier, and at a rock solid stable output
voltage.
 
In article <nm8bok$rig$1@dont-email.me>, curd@notformail.com says...
Hi all,

I'm having serious difficulty trying to get this issue resolved and was
wondering if anyone here has come across the same problem.

I've got a linear PSU that uses a LM317 (TO-220 case) to drop 37V down to
12V for a small 555 timer board that's set up as a pulse generator (at
about 0.1Hz). The thing is, as soon as I connect the board, the 12V from
the reg drops to between 6V and 7V and wanders around this level (like as
if I'd put a really heavy load on it). But the 555 board's resistance is
about 22k ohms so it only needs a very small current. When I swap the
board for a 10W 50 ohm power resistor drawing 0.25A, the reg output is
totally fine, staying rock steady at 12V and the resistor and reg get
warm as you would expect.
I can only think there may be something about the timer board that
'upsets' the LM317 and causes it to incorrectly partially shutdown,
because nothing on the board is getting warm and trying spare identical
timer boards causes the voltage to drop by the same amount also.

I can't recall this ever happening before! Any suggestions?
TIA.
Click to expand...

Try loading it down some. The data sheet calls from about 3 to 10 ma as
the minimum load.


---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus
 
On 07/14/2016 11:40 AM, Cursitor Doom wrote:
Hi all,

I'm having serious difficulty trying to get this issue resolved and was
wondering if anyone here has come across the same problem.

I've got a linear PSU that uses a LM317 (TO-220 case) to drop 37V down to
12V for a small 555 timer board that's set up as a pulse generator (at
about 0.1Hz). The thing is, as soon as I connect the board, the 12V from
the reg drops to between 6V and 7V and wanders around this level (like as
if I'd put a really heavy load on it). But the 555 board's resistance is
about 22k ohms so it only needs a very small current. When I swap the
board for a 10W 50 ohm power resistor drawing 0.25A, the reg output is
totally fine, staying rock steady at 12V and the resistor and reg get
warm as you would expect.
I can only think there may be something about the timer board that
'upsets' the LM317 and causes it to incorrectly partially shutdown,
because nothing on the board is getting warm and trying spare identical
timer boards causes the voltage to drop by the same amount also.

I can't recall this ever happening before! Any suggestions?
TIA.
Click to expand...

Too small an output cap, or too low an ESR? Three terminal regulators
can oscillate. Try hanging 100 uF aluminum electro on the output and
see if it starts behaving better.


Cheers

Phil Hobbs


--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510

hobbs at electrooptical dot net
http://electrooptical.net
 
On 07/14/2016 11:49 AM, Ralph Mowery wrote:
In article <nm8bok$rig$1@dont-email.me>, curd@notformail.com says...

Hi all,

I'm having serious difficulty trying to get this issue resolved and was
wondering if anyone here has come across the same problem.

I've got a linear PSU that uses a LM317 (TO-220 case) to drop 37V down to
12V for a small 555 timer board that's set up as a pulse generator (at
about 0.1Hz). The thing is, as soon as I connect the board, the 12V from
the reg drops to between 6V and 7V and wanders around this level (like as
if I'd put a really heavy load on it). But the 555 board's resistance is
about 22k ohms so it only needs a very small current. When I swap the
board for a 10W 50 ohm power resistor drawing 0.25A, the reg output is
totally fine, staying rock steady at 12V and the resistor and reg get
warm as you would expect.
I can only think there may be something about the timer board that
'upsets' the LM317 and causes it to incorrectly partially shutdown,
because nothing on the board is getting warm and trying spare identical
timer boards causes the voltage to drop by the same amount also.

I can't recall this ever happening before! Any suggestions?
TIA.

Try loading it down some. The data sheet calls from about 3 to 10 ma as
the minimum load.
Click to expand...

Too light a load causes the output voltage to go too high, though.

Cheers

Phil Hobbs


--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510

hobbs at electrooptical dot net
http://electrooptical.net
 
On Thu, 14 Jul 2016 15:50:43 -0400, Phil Hobbs wrote:

On 07/14/2016 01:12 PM, Cursitor Doom wrote:
On Thu, 14 Jul 2016 11:53:30 -0400, Phil Hobbs wrote:

Too small an output cap, or too low an ESR? Three terminal regulators
can oscillate. Try hanging 100 uF aluminum electro on the output and
see if it starts behaving better.

The datasheet says output caps are "optional" although I do always use
'em. The one I'd fitted was 1nF which seemed a bit small, so I subbed
it for a 100n but it's made no difference. 100uF seems HUGE! I might
try it,
though.


100nF is guaranteed to make a 317 oscillate. Microfarads, me lad,
microfarads.
Click to expand...

I'll certainly give it a try first thing tomorrow, Phil. But I'm not
optimistic given this confident pronouncement which I quote from the
datasheet:

"Typically, no capacitors are needed unless the device
is situated more than 6 inches from the input filter
capacitors, in which case an input bypass is needed.
An optional output capacitor can be added to improve
transient response. The adjustment terminal can be
bypassed to achieve very high ripple rejection ratios
that are difficult to achieve with standard 3-terminal
regulators."

So the manufacturers don't seem to think it's of any great importance at
all. BTW, on the input side, the filter capacitor is less than one inch
from the the reg's Vin terminal.
 
On Thu, 14 Jul 2016 21:46:53 +0100, Lee wrote:

It might be interesting to see if you have the same problem with an
ICM7555, but it's much easier just to add the cap, as suggested :) :)
Click to expand...

Quite! And I shall certainly do so first thing tomorrow. IF it works it
will highlight an astonishing and appalling omission from the datasheet
IMHO.
 
On 7/14/2016 11:40 AM, Cursitor Doom wrote:
... But the 555 board's resistance is
about 22k ohms so it only needs a very small current. ...
Click to expand...

Can you measure what the board is actually drawing?
 
On Thu, 14 Jul 2016 15:40:04 -0000 (UTC), Cursitor Doom
<curd@notformail.com> wrote:

I've got a linear PSU that uses a LM317 (TO-220 case) to drop 37V down to
12V for a small 555 timer board that's set up as a pulse generator (at
about 0.1Hz).
Click to expand...

The data sheet at:
<http://www.ti.com/lit/ds/symlink/lm317.pdf>
says that 37V is the maximum output voltage with an input-output
voltage difference of 40V absolute max. With 37v in and 12v out,
you're within spec, but seems to me too close for comfort. Maybe you
blew up the LM317? Try another LM317 and watch your input voltage
carefully. Or maybe add a big dropping resistor between your 37v
supply and the LM317 input.



--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
 
On 14/07/2016 18:12, Cursitor Doom wrote:
On Thu, 14 Jul 2016 11:53:30 -0400, Phil Hobbs wrote:

Too small an output cap, or too low an ESR? Three terminal regulators
can oscillate. Try hanging 100 uF aluminum electro on the output and
see if it starts behaving better.

The datasheet says output caps are "optional" although I do always use
'em. The one I'd fitted was 1nF which seemed a bit small, so I subbed it
for a 100n but it's made no difference. 100uF seems HUGE! I might try it,
though.
Click to expand...

JOOI What version of the 555? If it's a bipolar version they can exhibit
a "crowbar effect", and the 100uf cap is usually suggested to
ameliorate it.
 
On Thursday, July 14, 2016 at 4:39:48 PM UTC-4, Cursitor Doom wrote:
On Thu, 14 Jul 2016 15:50:43 -0400, Phil Hobbs wrote:

On 07/14/2016 01:12 PM, Cursitor Doom wrote:
On Thu, 14 Jul 2016 11:53:30 -0400, Phil Hobbs wrote:

Too small an output cap, or too low an ESR? Three terminal regulators
can oscillate. Try hanging 100 uF aluminum electro on the output and
see if it starts behaving better.

The datasheet says output caps are "optional" although I do always use
'em. The one I'd fitted was 1nF which seemed a bit small, so I subbed
it for a 100n but it's made no difference. 100uF seems HUGE! I might
try it,
though.


100nF is guaranteed to make a 317 oscillate. Microfarads, me lad,
microfarads.

I'll certainly give it a try first thing tomorrow, Phil. But I'm not
optimistic given this confident pronouncement which I quote from the
datasheet:

"Typically, no capacitors are needed unless the device
is situated more than 6 inches from the input filter
capacitors, in which case an input bypass is needed.
An optional output capacitor can be added to improve
transient response. The adjustment terminal can be
bypassed to achieve very high ripple rejection ratios
that are difficult to achieve with standard 3-terminal
regulators."

So the manufacturers don't seem to think it's of any great importance at
all. BTW, on the input side, the filter capacitor is less than one inch
from the the reg's Vin terminal.
Click to expand...

I think Phil is suggesting uF's on the output. (I always put 4.7- 10 uF there.)
I'm going with Jeff L's idea, that you've blown it up. Try another.
They also make a HV version of the lm317.. which is the only version we stock.

George H.
 
On Thu, 14 Jul 2016 22:44:43 +0100, Lee wrote:

Noted, Lee, and many thanks for the info. It's late here now so I won't
be posting again until morning whereupon I'll report back with my
findings!
 
On Thu, 14 Jul 2016 15:42:48 -0700, John Robertson wrote:

Jeff's point is you are at the maximum recommended input voltage,
anything changing there and the device will go south,( east, west, or
north - ) which may prove catastrophic.
Click to expand...

Yeah, there's not much headroom for sure. I'd sooner have used a LM317AHVT
but I don't have any 'in stock' here - but that's what I'd have used if
I'd had one.
 
On 07/14/2016 01:12 PM, Cursitor Doom wrote:
On Thu, 14 Jul 2016 11:53:30 -0400, Phil Hobbs wrote:

Too small an output cap, or too low an ESR? Three terminal regulators
can oscillate. Try hanging 100 uF aluminum electro on the output and
see if it starts behaving better.

The datasheet says output caps are "optional" although I do always use
'em. The one I'd fitted was 1nF which seemed a bit small, so I subbed it
for a 100n but it's made no difference. 100uF seems HUGE! I might try it,
though.
Click to expand...

100nF is guaranteed to make a 317 oscillate. Microfarads, me lad,
microfarads.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510

hobbs at electrooptical dot net
http://electrooptical.net
 
On 14/07/2016 19:40, Cursitor Doom wrote:
On Thu, 14 Jul 2016 18:40:16 +0100, Lee wrote:

JOOI What version of the 555? If it's a bipolar version they can exhibit
a "crowbar effect", and the 100uf cap is usually suggested to
ameliorate it.

The ones I'm using here are all NE555CN.
Click to expand...

It might be interesting to see if you have the same problem with an
ICM7555, but it's much easier just to add the cap, as suggested :) :)
 
On 14/07/2016 22:19, Cursitor Doom wrote:
On Thu, 14 Jul 2016 21:46:53 +0100, Lee wrote:

It might be interesting to see if you have the same problem with an
ICM7555, but it's much easier just to add the cap, as suggested :) :)

Quite! And I shall certainly do so first thing tomorrow. IF it works it
will highlight an astonishing and appalling omission from the datasheet
IMHO.
Click to expand...

IMO, The problem isn't so much the omission from the LM317 datasheet,
it's more an omission from the NE555 one.

Interestingly the ICM7555 datasheet says:
"The ICM7555 and ICM7556 devices are, in most instances,
direct replacements for the SE/NE 555/556 devices. However,
it is possible to effect economies in the external component
count using the ICM7555 and ICM7556. Because the bipolar
SE/NE 555/556 devices produce large crowbar currents in the
output driver, it is necessary to decouple the power supply
lines with a good capacitor close to the device. The ICM7555
and ICM7556 devices produce no such transients"
 
You must log in or register to reply here.

Welcome to EDABoard.com

Sponsor

Back
Top