Current-limited LED circuit probe

[Tinkering_Guy]

Over the years I've collected a lot of, er, 'stuff' basically because "I'm sure I can make something useful out of that one of these years."  (The SBH and I have disagreements over whether 'years' is measured in single digits or in centuries.)  One of the items was a sort of probe with a transparent tip; it looks like
but with a clip lead out the backslide like

The guts are missing, and the barrel is just the right diameter for an AA battery.  The obvious use would be to simply add the battery and a LED and turn it into a continuity tester, but I'm thinking of using one of my red/green bidi LEDs with a pair of complementary current-limiting circuits on a PCB in the barrel to allow checking power in a live circuit.  Sounds excessively complicated, perhaps, but I work with single- and bi-polar power supplies from 3V to 24VDC, plus low-voltage transformer secondaries (not to mention the unknown circuits I pick up because they're shiny, on sale, or in a piece of gear [like a printer] that I'm dis-integrating).  3V to 48V is a pretty wide range for a single current-limiting resistor.

No pics nor schematics yet, sorry; I'm still noodling this.

[Brass_Machine]

That would be an interesting and cool project. Keep us informed as you go. Question tho... is SBH like SWMBO? 

Eric

[Tinkering_Guy]

SBH == Significantly Better Half   

[75Plus]

Perhaps a fixed voltage regulator IC in the input would reduce the wide range of voltage to a manageable level. They can stand up to 37 volts on the input.

Joe

[Tinkering_Guy]

Perhaps a fixed voltage regulator IC in the input would reduce the wide range of voltage to a manageable level. They can stand up to 37 volts on the input.

Good idea!  I'll need to look through my stuff to see if I have any in DIP or TO-92 packages.  The only ones I'm sure I have are TO-3 and TO-126..

Thanks!

[Bluechip]

T

I've no idea whether this is applicable to what you propose, but you can get current limiting diodes that have a fairly constant current through them over a range of input voltages.

Made by ATC - Semitec ..

http://www.rapidonline.com/netalogue/specs/47-2600e.pdf

Rather poor graph though.

Not expensive here, so must be less in USA.  

Looks a bit like the E562 or E462 covers sort of 3V to 20+V with a current of circa 5-10mA,  enough for a LED.

??

Dave BC

[John Swift]

just another idea  for two or three terminal bi -colour leds
 
a two transistor constant current circuit plus a few diodes

[Brass_Machine]

SBH == Significantly Better Half   

Ahhh!

SWMBO = She Who Must Be Obeyed....  (my wife loves that one!)

Eric

[madjackghengis]

just another idea  for two or three terminal bi -colour leds
 
a two transistor constant current circuit plus a few diodes

John, those are a couple of interesting circuits, looks like they'd do the job hansomely.  Funny, about thirty years ago, working at an electronic burglar alarm company, we bought 2N2222s literally by the millions, for a factory with about twenty workers, building modules for the alarms we made.  I never thought it would remain the premier transistor it has.  I like your nice, simple and basic circuits that do the job with no fuss, and minimum parts to fail, I'll have to copy that set in one of my notebooks, nice job! mad jack

[Tinkering_Guy]

I echo the sentiments!  I never seem to remember to try a bridge... 

[John Swift]

Hi Mad Jack & Tinkering_Guy ,
                                            not many people are used to thinking in terms of constant current supplies
usually  constant current sources are hidden inside analogue IC's - replacing the resistor tail , in long tail pairs in op amp's for instance .

If you don't want to build a switch mode power supply to power L.E.D lights in a car
constant current supplies to 3 or 4  L.E.D.s in series  will keep the light constant as the battery voltage varies between 12 to 15 volts or so , and variation in the forward volt drop of the L.E.D.s
if one L.E.D.  should go short circuit the current would not change in the others

the 2N3819 n-channel F.E.T. is unusual because it is symetrical - the drain and source leads can be interchanged !!
the 2 F.E.T.s  I tested with the drain connected to + supply  and the source and gate connected to 0 v
    the first F.E.T.  the current was 8.4 mA  at 4v  and  9.1 mA  at 20v - not bad for one component
the second F.E.T   the current was 9.0 mA  at 4v  and  9.9 mA  at 20v

a resistor in the source lead will reduce  the the current - negative bias on the gate turns the F.E.T off

starting of with the 2 terminal bi colour LED I did think of using 2 current limiters and 2 diodes to control
alternate half cycles of an ac voltage .
but with a 3 terminal bi colour  L.E.D., I added 2 more diodes   in inverse parallel with the L.E.D.s,
 realised it looked a bit  like a bridge and then I only need one current control circuit

I expect the 2N2222 was originally made in the sixties for the defence or computing industry
made in vast quantitys , the price would fall making it an attractive choice  for many applications