Powered by Invision Power Board


Forum Rules Forum Rules (Please read before posting)
  Reply to this topicStart new topicStart Poll

> Comparator + Constant Current With Tl431, an optimization challenge
fenugrec
Posted: November 25, 2011 06:24 am
Reply to this postQuote Post


Sr. Member
*****

Group: Trusted Members
Posts: 488
Member No.: 5,637
Joined: July 13, 2006




Hi all,
I feel what I'm trying to do can be pretty simple, but I just can't figure it out. In short : some equipment runs either from battery or AC power. I want a power LED to light with a constant (say +- 10%) current, but it should turn off when the battery voltage drops below a certain threshold.
I don't have room for a comparator IC, and there's no regulated supply in the circuit.

I figure a good start is a TL431 "programmable" shunt regulator (TL431 datasheet). Making the TL431 work as a constant current source is easy; the low voltage threshold is also straightforward. Making a single TL431 do both is a challenge... I'm sure there's a way to do it simply, but I haven't found the miracle circuit yet. I think I'll have no choice but to add at least a transistor and some diodes... hopefully someone here can do better !

Rough specs:
The V+ rail is about 6-9V when powered by the battery, otherwise 9-13V. Low battery threshold : 7.5V.

The first circuit that comes to mind (and works) is
user posted image

Does anyone see how to improve this ?
PMUsers Website
Top
MacFromOK
Posted: November 25, 2011 08:01 am
Reply to this postQuote Post


Forum Addict ++
*******

Group: Spamminator Taskforce
Posts: 14,206
Member No.: 5,314
Joined: June 04, 2006




QUOTE (fenugrec @ November 24, 2011 11:24 pm)
In short : some equipment runs either from battery or AC power. I want a power LED to light with a constant (say +- 10%) current, but it should turn off when the battery voltage drops below a certain threshold.

Why not just power the LED straight thru a zener of the appropriate voltage (and wattage)? Voltage drops below that level, zener stops conducting, LED goes off. beer.gif


--------------------
Mac *

"Basic research is what I'm doing when I don't know what I'm doing." [Wernher Von Braun]

* is not responsible for errors, consequential damage, or... anything.
PMEmail Poster
Top
Sch3mat1c
Posted: November 25, 2011 09:01 am
Reply to this postQuote Post


Forum Addict ++
Group Icon

Group: Moderators
Posts: 19,613
Member No.: 73
Joined: July 24, 2002




If you don't have room for a comparator, you won't have room for a TL431, transistor and more than two resistors. For that matter, you hardly have room for resistors, if you consider there are comparators in SMT packages under 2mm square.

How sharp should the low battery threshold be, and what range of voltage should be tolerated for that 10% change in current? If both specs are quite loose, a zener will suffice -- the turn-off curve is exponential rather than piecewise linear or hysteretic, so it will dim out very slowly as the battery discharges.

If you want close to ideal performance, it would be very easy to create a circuit which does it in, say, 20 transistors or so, but good luck fabricating the IC to fit it in the small space...

Tim


--------------------
Answering questions is a tricky subject to practice. Not due to the difficulty of formulating or locating answers, but due to the human inability of asking the right questions; a skill that, were one to possess, would put them in the "answering" category.
PMEmail PosterUsers Website
Top
kl27x
Posted: November 25, 2011 09:36 am
Reply to this postQuote Post


Forum Addict ++
*******

Group: Spamminator Taskforce
Posts: 4,101
Member No.: 4,223
Joined: February 06, 2006




You didn't mention your voltage range. Preset 3-pin voltage detector IC's could be one option. They're most commonly available in the 1.5-5V range.

Space = down to SOT-23. It requires fewer connections than a comparator and doesn't need a supply voltage higher than the input = much smaller. Power and voltage input are one and the same, ground, output pin. Output pin can drive an LED, directly. Just need a series resistor.
PMEmail Poster
Top
millwood
Posted: November 25, 2011 12:58 pm
Reply to this postQuote Post


Forum Addict ++
*******

Group: Trusted Members
Posts: 1,823
Member No.: 25,377
Joined: July 26, 2009




you can get a constant current source out of a jfet;

you can get a voltage sensor out of a pot or a resistive divider;

you can get a pretty good comparator out of a npn, mosfet, or even a tl431.

so you can build a much simpler design than you have, out of the components mentioned above.
PMEmail Poster
Top
fenugrec
Posted: November 25, 2011 08:49 pm
Reply to this postQuote Post


Sr. Member
*****

Group: Trusted Members
Posts: 488
Member No.: 5,637
Joined: July 13, 2006




Interesting - not quite the type of replies I expected... I omitted some important details in the original post!

QUOTE
power the LED straight thru a zener

As V+ increases above threshold, so will current. Unless I use the zener as a shunt regulator, which in my tests really didn't have a steep enough response... and it would end up looking like a non-adjustable TL431 circuit anyway.


QUOTE
If you don't have room for a comparator, you won't have room for a TL431, transistor and more than two resistors.

About the "no room for a comparator" - as I forgot to mention, I'm not laying out a new PCB, I'm patching an existing circuit. I know they make comparators in SOT23-6, but I don't have any on hand. Besides, a comparator can't also regulate LED current - I would still need some other active component.

I can find pads for a TO-92. 0603 resistors fit everywhere; I also have some SOT23 transistors that are easily soldered in place. Much easier than trying to layout 8 SOIC or TSSOP pins in mid-air... that's more what I meant by "no room".

QUOTE
you can get a constant current source out of a jfet;

That's probably the simplest solution. I would just need to find a jfet that has a reasonable Idss.

Thanks a lot for the ideas, guys.

This post has been edited by fenugrec on November 25, 2011 08:50 pm
PMUsers Website
Top
MacFromOK
Posted: November 26, 2011 12:42 am
Reply to this postQuote Post


Forum Addict ++
*******

Group: Spamminator Taskforce
Posts: 14,206
Member No.: 5,314
Joined: June 04, 2006




QUOTE (fenugrec @ November 25, 2011 01:49 pm)
QUOTE
power the LED straight thru a zener

As V+ increases above threshold, so will current. Unless I use the zener as a shunt regulator, which in my tests really didn't have a steep enough response... and it would end up looking like a non-adjustable TL431 circuit anyway.

A series resistor for the LED/zener would be required anyway. Current could be safely limited at max V and the LED would still be "on" several volts lower.

You would also have to allow for the LED's forward drop when choosing zener voltage btw (a moot point if you're using a different method). beer.gif


--------------------
Mac *

"Basic research is what I'm doing when I don't know what I'm doing." [Wernher Von Braun]

* is not responsible for errors, consequential damage, or... anything.
PMEmail Poster
Top
Sch3mat1c
Posted: November 27, 2011 12:03 am
Reply to this postQuote Post


Forum Addict ++
Group Icon

Group: Moderators
Posts: 19,613
Member No.: 73
Joined: July 24, 2002




Come to think of it, you might be able to trick a TLV431 into PWMing the LED in such a way that it maintains reasonably constant current above the threshold. It will probably keep osc
PMEmail PosterUsers Website
Top
Sch3mat1c
Posted: November 27, 2011 12:03 am
Reply to this postQuote Post


Forum Addict ++
Group Icon

Group: Moderators
Posts: 19,613
Member No.: 73
Joined: July 24, 2002




Come to think of it, you might be able to trick a TLV431 into PWMing the LED in such a way that it maintains reasonably constant current above the threshold. It will probably keep oscillating for a few fractions of a volt below threshold though. If hysteresis is acceptable, and the amount can be controlled by design, this would be fine.

Tim


--------------------
Answering questions is a tricky subject to practice. Not due to the difficulty of formulating or locating answers, but due to the human inability of asking the right questions; a skill that, were one to possess, would put them in the "answering" category.
PMEmail PosterUsers Website
Top
0 User(s) are reading this topic (0 Guests and 0 Anonymous Users)
0 Members:

Topic Options Reply to this topicStart new topicStart Poll

 


:: support us ::




ElectronicsSkin by DutchDork & The-Force