Led "Regulation"

[Forrest Christian]

I seem to quite often need to run a "voltage present" LED on a very wide
range of voltages...  For instance, in one case the input voltage can
run from 5 to 60VDC, or more.

I need to power this LED from the power source, but need to regulate the
current so that the LED is pretty much constant output across the
voltage range.

I'm aware of the standard schematic using a PNP, a Zener, and a couple
of resistors, but would prefer something with lower part count.

I'm also aware of the LM317HV, but the cost there is prohibitive (I
would need the HV becuase of the higher voltage range input).

Of course there's always the option of putting a regular voltage
regulator and resistor in, but again, that gets somewhat costly.

I'm hoping I can find something just marginally more expensive than
putting a current limiting resistor in place.   Is there some magic
circuit I've missed...

-forrest
--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Marcel Birthelmer]

Forrest,
what about a basic current mirror circuit? a few resistors and two
transistors should do nicely. Maybe one even exists as an integrated
package?
- Marcel

On Wed, Jul 16, 2008 at 5:50 PM, Forrest W Christian <forrestc@...>
wrote:

> I seem to quite often need to run a "voltage present" LED on a very wide
> range of voltages...  For instance, in one case the input voltage can
> run from 5 to 60VDC, or more.
>
> I need to power this LED from the power source, but need to regulate the
> current so that the LED is pretty much constant output across the
> voltage range.
>
> I'm aware of the standard schematic using a PNP, a Zener, and a couple
> of resistors, but would prefer something with lower part count.
>
> I'm also aware of the LM317HV, but the cost there is prohibitive (I
> would need the HV becuase of the higher voltage range input).
>
> Of course there's always the option of putting a regular voltage
> regulator and resistor in, but again, that gets somewhat costly.
>
> I'm hoping I can find something just marginally more expensive than
> putting a current limiting resistor in place.   Is there some magic
> circuit I've missed...
>
> -forrest
> --
> http://www.piclist.com PIC/SX FAQ & list archive
> View/change your membership options at
> http://mailman.mit.edu/mailman/listinfo/piclist
>

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Richard Prosser]

Forrest

You're looking for some simple sort of constant current controller?

How about a small light bulb?  :-)

Otherwise you _might_  be able to find a PTC resistor that works as-is
or in combination with a standard resistor.

One problem is going to be heat. At 10mA & a 5V supply, you're going
to drop about 3.8V across the "controller" & dissipate 38mW (R=380ohm)
 . At 60V, the dissipation is ~0.6 watt. (R=5k8). Any active component
dissipating this much is probably going to want a heatsink of some
kind unless there is good airflow.

Another alternative would be to use a simple switcher (eg the "Black
regulator") But at increased cost & component count.

Richard P


2008/7/17 Forrest W Christian <forrestc@...>:

> I seem to quite often need to run a "voltage present" LED on a very wide
> range of voltages...  For instance, in one case the input voltage can
> run from 5 to 60VDC, or more.
>
> I need to power this LED from the power source, but need to regulate the
> current so that the LED is pretty much constant output across the
> voltage range.
>
> I'm aware of the standard schematic using a PNP, a Zener, and a couple
> of resistors, but would prefer something with lower part count.
>
> I'm also aware of the LM317HV, but the cost there is prohibitive (I
> would need the HV becuase of the higher voltage range input).
>
> Of course there's always the option of putting a regular voltage
> regulator and resistor in, but again, that gets somewhat costly.
>
> I'm hoping I can find something just marginally more expensive than
> putting a current limiting resistor in place.   Is there some magic
> circuit I've missed...
>
> -forrest
> --
> http://www.piclist.com PIC/SX FAQ & list archive
> View/change your membership options at
> http://mailman.mit.edu/mailman/listinfo/piclist
>

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Marcel Duchamp]

JFET, I suppose...
http://www2.eng.cam.ac.uk/~dmh/ptialcd/csink/csink.htm

You need to find one with
a) high enough voltage rating
b) adequate current spec (Idss)
c) sufficient power dissipation
c) cost
d) etc.

Forrest W Christian wrote:

> I seem to quite often need to run a "voltage present" LED on a very wide
> range of voltages...  For instance, in one case the input voltage can
> run from 5 to 60VDC, or more.
>
> I need to power this LED from the power source, but need to regulate the
> current so that the LED is pretty much constant output across the
> voltage range.
>
> I'm aware of the standard schematic using a PNP, a Zener, and a couple
> of resistors, but would prefer something with lower part count.
>
> I'm also aware of the LM317HV, but the cost there is prohibitive (I
> would need the HV becuase of the higher voltage range input).
>
> Of course there's always the option of putting a regular voltage
> regulator and resistor in, but again, that gets somewhat costly.
>
> I'm hoping I can find something just marginally more expensive than
> putting a current limiting resistor in place.   Is there some magic
> circuit I've missed...
>
> -forrest

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Apptech]

>I seem to quite often need to run a "voltage present" LED
>on a very wide
> range of voltages...  For instance, in one case the input
> voltage can
> run from 5 to 60VDC, or more.

> I'm hoping I can find something just marginally more
> expensive than
> putting a current limiting resistor in place.   Is there
> some magic
> circuit I've missed...

Here is Richard Prosser's version of "my" G.S.R.* switching
regulator.
D8 may not be needed.
Despite apparent high parts count it can be very small and
cheap.

        http://others.servebeer.com/misc/gsr_rp.gif

Roman Black's "Black Regulator" is perhaps slightly cheaper.
While it has poorer regulation it may well be adequate for
your needs.

I have had a version of this operating  from 12V to 200 V
in.

Here is an independently developed version which is
essentially identical.

        http://sound.westhost.com//appnotes/an006.htm

A simple change to a constant current output drive (suited
to LED use) is seen here

        http://sound.westhost.com//appnotes/an003.htm

If a P Channel MOSFET is used for the high side switch
remember to zener clamp the gate to avoid death at higher
Vin's.

This is a "hysteretic converter". Operation as shown is
marvellously chaotic. Operating frequency and PWM ratio
varies vastly, which allows it to achieve the vast max:min
Vinput ratios that it does. Hysteresis DOS exist despite
some refusing to believe it. Clue: Ripple on output cap is
an essential part of the operation. Adding a series RC
(location left to the student) will provide extra and more
normal hysteresis and a more determinate waveform. MAY help
efficiency somewhat.

For an IC solution a MC34063 would be cheap and goodish.
Vinmax = 40V so a cheap external transistor probably
required for drive.



        Russell


GSR = God's Switching Regulator.
Story offlist on request.














--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[enkitec]

        Attached is a similar circuit I found on the net.

        It seems to have a lower quiescent current without load.

        Mark Jordan


On 17 Jul 2008 at 13:39, Apptech wrote:

> Here is Richard Prosser's version of "my" G.S.R.* switching
> regulator.
> D8 may not be needed.
> Despite apparent high parts count it can be very small and
> cheap.
>
>         http://others.servebeer.com/misc/gsr_rp.gif
>


The following section of this message contains a file attachment
prepared for transmission using the Internet MIME message format.
If you are using Pegasus Mail, or any other MIME-compliant system,
you should be able to save it or view it from within your mailer.
If you cannot, please ask your system administrator for assistance.

   ---- File information -----------
     File:  Switcher.gif
     Date:  25 Jun 2008, 10:40
     Size:  10530 bytes.
     Type:  GIF-image


--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Apptech]

GSR simulacrum.

Enki said:
> Attached is a similar circuit I found on the net.
> It seems to have a lower quiescent current without load.

I'm not sure how you would determine that, but you may be
correct.
I see that that circuit (or an exact facsimile) appears here

        http://www.techlib.com/electronics/regulators.html

The main difference from that and the GSR is the addition of
formal hysteresis for the 'comparator" by using a 10 r
emitter resistor and 10k feedback from the top (hot side) of
the inductor. May well be a good idea. I've been using a
similar arrangement of late to add mild hysteresis / Schmitt
trigger action to other circuits with good results. In this
case it would probably markedly tidy up the chaotic
switching action but may not improve measured performance.
Note that they specifically note the 40 mV pp output ripple.
Ripple here is an essential requirement with the original
design. The added feedback resistors as above would make
this less important.

Apart from that the circuit should have very similar
characteristics and it's not obvious that it's low current
operation would be much different. Component values vary
with design requirements.

Before the Olin-Savant starts throwing wet or dead fish at
the Extralytic capacitors, I note that they note that this
is a brand name for a low loss electrolytic capacitor.


        Russell McMahon


--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Alan B. Pearce]

>I'm hoping I can find something just marginally more expensive than
>putting a current limiting resistor in place.   Is there some magic
>circuit I've missed...

I would be tempted to use Russells transistor switcher, which has a wide
voltage range input. In view of the limited current output it may be
possible to tailor it to use less components than the 'generic' version he
has published here in the past.

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Xiaofan Chen]

On Thu, Jul 17, 2008 at 1:13 PM, Apptech <apptech@...> wrote:
> I'm not sure how you would determine that, but you may be
> correct.
> I see that that circuit (or an exact facsimile) appears here
>
>        http://www.techlib.com/electronics/regulators.html
>

This is basically the circuit I saw for a pre-1992 SICK
photoelectric sensor design (universal AC/DC input,
typically 20V-253V AC or DC). I am not so sure if they
are still producing the sensor. Later I saw a better
design from SICK which uses a P-MOS in place of the
PNP transistor. I think they are still producing the
new sensor.

SICK (http://www.sick.de) is one of the leading
photoelectric sensor manufacturer. I believe they
are Number 1 in  the market.

Xiaofan
--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Xiaofan Chen]

On Thu, Jul 17, 2008 at 8:50 AM, Forrest W Christian <forrestc@...> wrote:
>
> I'm aware of the standard schematic using a PNP, a Zener, and a couple
> of resistors, but would prefer something with lower part count.
>
> I'm also aware of the LM317HV, but the cost there is prohibitive (I
> would need the HV becuase of the higher voltage range input).
>

How about Supertex product? They may have something for you.
But I am not so sure about the pricing.
http://www.supertex.com

Xiaofan
--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Xiaofan Chen]

On Thu, Jul 17, 2008 at 4:54 PM, Xiaofan Chen <xiaofanc@...> wrote:

> On Thu, Jul 17, 2008 at 1:13 PM, Apptech <apptech@...> wrote:
>> I'm not sure how you would determine that, but you may be
>> correct.
>> I see that that circuit (or an exact facsimile) appears here
>>
>>        http://www.techlib.com/electronics/regulators.html
>>
>
> This is basically the circuit I saw for a pre-1992 SICK
> photoelectric sensor design (universal AC/DC input,
> typically 20V-253V AC or DC). I am not so sure if they
> are still producing the sensor.

The SICK design adds a PNP transistor, a zener
and a resistor to form a current source to drive
either one of the bottom two NPN transistors.

The main problem with this type of circuits
is conducted/radiated emission. Efficiency is
also quite low in the high end of the input voltage
due to the very low duty ratio.

Regards,
Xiaofan
--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Apptech]

> The SICK design adds a PNP transistor, a zener
> and a resistor to form a current source to drive
> either one of the bottom two NPN transistors.

> The main problem with this type of circuits
> is conducted/radiated emission. Efficiency is
> also quite low in the high end of the input voltage
> due to the very low duty ratio.

I'd love to see the SICK cct if available.

On my 12V-200V version the efficiency at 200 Vin / 12Vout
was not much over 50%.
BUT that's 'rather better' [tm] that the 12/200 = 6% you'd
get with a linear regulator.
eg for 1 Watt out my cct would dissipate 1 Watt of heat. A
linear regulator would dissipate about 188/12 =~ 16 Watts of
heat.

Emission is my circuit is much relieved by the horrendously
chaotic nature of the switching - well on the way to white
noise. Adding more formal resistive hysteresis as they do
would make emissions worse.




        Russell

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Rolf-4]

Hi Forrest,

How about a current limiting circuit (borrowed from 'the art of
electronics') which I am still reading....

It seems like it will do what you want, but, instead of a Zener and PNP,
it is a PNP and a PMOSFET. May not help with the part count.

Still, I have re-drawn the schematic and it should supply 7mA to the LED
at any line voltage the PMOS can handle....

Basically, the 100R resistor, with the 0.7Vbe voltage on the PNP will
set a current limit of 7mA.

Any variances in the current will be due to the Vbe changes over temp.
Using a MPSA92 as a 'typical' PNP transistor it gives a range of 0.5V
through 0.8V for Vbe over -55 to 125 degrees, which would correlate to 5
through 8mA through the LED over the entire temp range.

Obviously, there is probably something I have missed, but, I would love
to know what because it would be a good learning exercise for me....

Rolf



Forrest W Christian wrote:

> I seem to quite often need to run a "voltage present" LED on a very wide
> range of voltages...  For instance, in one case the input voltage can
> run from 5 to 60VDC, or more.
>
> I need to power this LED from the power source, but need to regulate the
> current so that the LED is pretty much constant output across the
> voltage range.
>
> I'm aware of the standard schematic using a PNP, a Zener, and a couple
> of resistors, but would prefer something with lower part count.
>
> I'm also aware of the LM317HV, but the cost there is prohibitive (I
> would need the HV becuase of the higher voltage range input).
>
> Of course there's always the option of putting a regular voltage
> regulator and resistor in, but again, that gets somewhat costly.
>
> I'm hoping I can find something just marginally more expensive than
> putting a current limiting resistor in place.   Is there some magic
> circuit I've missed...
>
> -forrest
>  

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Olin Lathrop]

Forrest W Christian wrote:
> I seem to quite often need to run a "voltage present" LED on a very
> wide range of voltages...  For instance, in one case the input
> voltage can run from 5 to 60VDC, or more.

Do you have some other more regulated power voltage available?  If so, sense
the 5-60V but then run the LED off the regulated rail.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
[phone].  Gold level PIC consultants since 2000.
--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Tom Wehn]

If you can tolerate the voltage drop and the current is less than a few amps
why not install a 2 or 3 Volt Zener in the line and add a resistor and LED
in series with it? 2 added parts and even a power Zener is under $.25.

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Nino Benci]

The attached diagram is a cct that I have used for many "Constant LED
Current" applications with Vin from 5V to 24V. Your mileage may vary.

Antonio Benci

Rolf wrote:

> Hi Forrest,
>
> How about a current limiting circuit (borrowed from 'the art of
> electronics') which I am still reading....
>
> It seems like it will do what you want, but, instead of a Zener and PNP,
> it is a PNP and a PMOSFET. May not help with the part count.
>
> Still, I have re-drawn the schematic and it should supply 7mA to the LED
> at any line voltage the PMOS can handle....
>
> Basically, the 100R resistor, with the 0.7Vbe voltage on the PNP will
> set a current limit of 7mA.
>
> Any variances in the current will be due to the Vbe changes over temp.
> Using a MPSA92 as a 'typical' PNP transistor it gives a range of 0.5V
> through 0.8V for Vbe over -55 to 125 degrees, which would correlate to 5
> through 8mA through the LED over the entire temp range.
>
> Obviously, there is probably something I have missed, but, I would love
> to know what because it would be a good learning exercise for me....
>
> Rolf
>
>
>
> Forrest W Christian wrote:
>> I seem to quite often need to run a "voltage present" LED on a very
>> wide range of voltages...  For instance, in one case the input voltage
>> can run from 5 to 60VDC, or more.
>>
>> I need to power this LED from the power source, but need to regulate
>> the current so that the LED is pretty much constant output across the
>> voltage range.
>>
>> I'm aware of the standard schematic using a PNP, a Zener, and a couple
>> of resistors, but would prefer something with lower part count.
>>
>> I'm also aware of the LM317HV, but the cost there is prohibitive (I
>> would need the HV becuase of the higher voltage range input).
>>
>> Of course there's always the option of putting a regular voltage
>> regulator and resistor in, but again, that gets somewhat costly.
>>
>> I'm hoping I can find something just marginally more expensive than
>> putting a current limiting resistor in place.   Is there some magic
>> circuit I've missed...
>>
>> -forrest
>>  
>
>
> ------------------------------------------------------------------------
>

[nino_benci.vcf]

---

begin:vcard
fn:Antonio Benci
n:Benci;Antonio
org:Monash University;School of Physics
adr:;;PO Box 27;Monash University;VIC;3800;Australia
email;internet:nino.benci@...
title:Professional Officer / E&IS Manager
tel;work:[phone]
tel;fax:[phone]
x-mozilla-html:FALSE
url:http://www.physics.monash.edu.au
version:2.1
end:vcard

---

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Tom Wehn]

OK I reread this and I could have been clearer. The series resistor and LED
are in parallel with the Zener. Sorry for the ambiguity.

-----Original Message-----


If you can tolerate the voltage drop and the current is less than a few amps
why not install a 2 or 3 Volt Zener in the line and add a resistor and LED
in series with it? 2 added parts and even a power Zener is under $.25.

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Olin Lathrop]

Antonio L. Benci wrote:
> The attached diagram is a cct that I have used for many "Constant LED
> Current" applications with Vin from 5V to 24V.

So now we've got 3 terminal zener diodes to add to the managerie of odd
components, along with the positive electrons, negative resistors, and NPN
and PNP FETs?


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
[phone].  Gold level PIC consultants since 2000.
--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist