National Semiconductor's literature usually refers to a 2 volt overhead for the LM317. Other manufacturers may be different. There are low dropout versions (LDO) that operate with less than 1 volt overhead. I've considered using one of those LDOs but I've heard there were disadvantages like problems with ocsillation. How much of a problem is it? I could easily add a small capacitor if that is all it takes. What else might be needed or would work? -RussH
Zener + transistor to regulate voltage
- Thread starter Josey
- Start date
Help Support Candle Power Flashlight Forum
MrAl
Flashlight Enthusiast
Hello there,
Thanks for looking that up Russ, and since you mentioned the
LDO version i decided to take a quick look at those myself.
I found the LM1117, which has a dropout voltage of about
1.2 volts. This means it could be used as constant
current to drive an LED with one resistor and would
operate down to about 4.7 volts before current decreased.
Too bad they dont make a zero volts dropout version
Im going to look at these a little more but so far
it looks like in order to meet stability requirements
an external cap has to be connected across the input
of the device of about 10uf. I guess that's not too
bad.
Take care,
Al
Thanks for looking that up Russ, and since you mentioned the
LDO version i decided to take a quick look at those myself.
I found the LM1117, which has a dropout voltage of about
1.2 volts. This means it could be used as constant
current to drive an LED with one resistor and would
operate down to about 4.7 volts before current decreased.
Too bad they dont make a zero volts dropout version
Im going to look at these a little more but so far
it looks like in order to meet stability requirements
an external cap has to be connected across the input
of the device of about 10uf. I guess that's not too
bad.
Take care,
Al
Ordin_Aryguy
Newly Enlightened
- Joined
- Feb 17, 2004
- Messages
- 131
Amen Brother Al! /ubbthreads/images/graemlins/thumbsup.gif (Once again, you're spot on)
Current regulation is the only way to fly with LED's. Every one that I've ever come into contact with hasn't behaved itself too much in the presence of heat. The hotter they get, the better they conduct, the better they conduct, the hotter they get... pretty soon your solder is melting.
Maitaining a constant current prevents that ugly run away. Constant voltage assures disaster if you prefer to run all your goods up near the edge of the performance envelope... and that would pretty much be all of us.
Ordin
Current regulation is the only way to fly with LED's. Every one that I've ever come into contact with hasn't behaved itself too much in the presence of heat. The hotter they get, the better they conduct, the better they conduct, the hotter they get... pretty soon your solder is melting.
Maitaining a constant current prevents that ugly run away. Constant voltage assures disaster if you prefer to run all your goods up near the edge of the performance envelope... and that would pretty much be all of us.
Ordin
Ordin_Aryguy
Newly Enlightened
- Joined
- Feb 17, 2004
- Messages
- 131
Mr. Al, 10uF?? What's a few uF's among friends anyhow? Awww, they come in 0603 pkgs with X5R dielectrics these days. They aren't as stable over temperature as an X7R or NPO would be. As long as it's sized so that at the temp extremes the device is likely to be exposed to it still has enough uumph to keep the LDO from oscillating, it will be just fine.
Those new-fangled high capacitance ceramics are wonderous little devices. Cheaper than tantalums, smaller than tantalums, none of the "surge surprises" that tantalums are so famous for, and miles less ESR. Better'n sliced bread, they are.
Ordin
Those new-fangled high capacitance ceramics are wonderous little devices. Cheaper than tantalums, smaller than tantalums, none of the "surge surprises" that tantalums are so famous for, and miles less ESR. Better'n sliced bread, they are.
Ordin
9X25
Newly Enlightened
Josey,
look in your inbox(email).
9X25
look in your inbox(email).
9X25
MrAl
Flashlight Enthusiast
Hi there Ord_Guy,
Very good point about the possible thermal problems!
Also good point about the 10uf caps not being as big
as they used to be No longer need a big fat
cylindar sticking up towering over the tiny SM parts.
Another good point about using constant current over
voltage regulation that i forgot to mention is that
it takes care of variations from the manufacturing of
the LED too. Any variation in voltage at the given
current (such as 20ma) wont matter using current
regulation. This means each circuit doesnt have to
be tweeked individually for a particular LED.
I guess it would be worth mentioning that this would
handle every color too without any changes to the
circuit.
Take care and thanks for the interesting points,
Al
Very good point about the possible thermal problems!
Also good point about the 10uf caps not being as big
as they used to be
No longer need a big fatcylindar sticking up towering over the tiny SM parts.
Another good point about using constant current over
voltage regulation that i forgot to mention is that
it takes care of variations from the manufacturing of
the LED too. Any variation in voltage at the given
current (such as 20ma) wont matter using current
regulation. This means each circuit doesnt have to
be tweeked individually for a particular LED.
I guess it would be worth mentioning that this would
handle every color too without any changes to the
circuit.
Take care and thanks for the interesting points,
Al
MrAl and OrdMan, thanks for the info. I'll have to look for some of those 10 uF ceramics. I'll be ordering some 317s again soon & I'll get some 1117s (or cousin LMS8117A- I need about 1 amp) to play with, too. I'll have to get one of the larger packages. -RussH
Josey
Flashlight Enthusiast
Thanks Christian:
I appreciate that design and will be trying your circuit.
Thanks Doug:
Your filed down LEDs look a lot better than I had imagined. I'll be copying your work. Current regulation it is.
And thanks everyone. I've got a ton of ideas running through my head.
Josey
I appreciate that design and will be trying your circuit.
Thanks Doug:
Your filed down LEDs look a lot better than I had imagined. I'll be copying your work. Current regulation it is.
And thanks everyone. I've got a ton of ideas running through my head.
Josey
[ QUOTE ]
Josey said:
Thanks Doug:
Your filed down LEDs look a lot better than I had imagined. I'll be copying your work. Current regulation it is.
[/ QUOTE ]
You're very welcome.
I know it sounds crude, but the results are excellent. The 12 LED array ends up with just over 90 degrees of beam angle, that is it will light up two feet of wall (or newspaper) from a foot away. Even, artifact free light. Plenty bright enough for reading, at least so far.....
Doug Owen
Josey said:
Thanks Doug:
Your filed down LEDs look a lot better than I had imagined. I'll be copying your work. Current regulation it is.
[/ QUOTE ]
You're very welcome.
I know it sounds crude, but the results are excellent. The 12 LED array ends up with just over 90 degrees of beam angle, that is it will light up two feet of wall (or newspaper) from a foot away. Even, artifact free light. Plenty bright enough for reading, at least so far.....
Doug Owen
idleprocess
Flashaholic
[ QUOTE ]
Doug Owen said:
[ QUOTE ]
idleprocess said:
A 9V battery is a proper battery. There's either a stack of raw flat cells encased in plastic, or 6 "AAAA" batteries.
[/ QUOTE ]
Here I would have to disagree. We're starting out at a nominal 12 Volts, easily capable of running three LEDs in series, and you're suggesting regulating it down to 9 Volts, which while 'doable' will only drive two LEDs. This automatically means this system will use 50% more current (3 strings needed at 9 Volts to do what two at 12 could have done).
[/ QUOTE ]
I was just addressing the tangential issue of "battery" vs "cell" in that statement.
I only mentioned voltage regulators because I'm familiar with them and they're easy to use. I don't know how efficient they are, but I'm fairly certain they're not as bad as resistors (if they are terribly inefficient, I'll stand corrected). Since this isn't a portable operation, there could be a case for simplicity at the expense of some efficiency. 1-2 amp voltage regulators are dirt-cheap and don't take a lot of know-how to use. Also, because they're usually "power" components, they're a bit sturdier (mechanically and electrically) than many of the components that I've seen used in more complicated power supplies.
A voltage regulator p/s also assumes some sort of current regulation - likely a small resistor that's dropping a fraction of a volt due to a convenient output voltage.
Again - I'm shooting from the hip here, not really knowing a whole lot about voltage regulator mechanics.
A current supply is a good way to go, since LEDs regulate their own voltage handily if current is restricted.
Doug Owen said:
[ QUOTE ]
idleprocess said:
A 9V battery is a proper battery. There's either a stack of raw flat cells encased in plastic, or 6 "AAAA" batteries.
[/ QUOTE ]
Here I would have to disagree. We're starting out at a nominal 12 Volts, easily capable of running three LEDs in series, and you're suggesting regulating it down to 9 Volts, which while 'doable' will only drive two LEDs. This automatically means this system will use 50% more current (3 strings needed at 9 Volts to do what two at 12 could have done).
[/ QUOTE ]
I was just addressing the tangential issue of "battery" vs "cell" in that statement.
I only mentioned voltage regulators because I'm familiar with them and they're easy to use. I don't know how efficient they are, but I'm fairly certain they're not as bad as resistors (if they are terribly inefficient, I'll stand corrected). Since this isn't a portable operation, there could be a case for simplicity at the expense of some efficiency. 1-2 amp voltage regulators are dirt-cheap and don't take a lot of know-how to use. Also, because they're usually "power" components, they're a bit sturdier (mechanically and electrically) than many of the components that I've seen used in more complicated power supplies.
A voltage regulator p/s also assumes some sort of current regulation - likely a small resistor that's dropping a fraction of a volt due to a convenient output voltage.
Again - I'm shooting from the hip here, not really knowing a whole lot about voltage regulator mechanics.
A current supply is a good way to go, since LEDs regulate their own voltage handily if current is restricted.
[ QUOTE ]
idleprocess said:
I was just addressing the tangential issue of "battery" vs "cell" in that statement.
..........
Since this isn't a portable operation, there could be a case for simplicity at the expense of some efficiency.
..........
A current supply is a good way to go, since LEDs regulate their own voltage handily if current is restricted.
[/ QUOTE ]
idleprocess,
Howdy. I'm sorry for misunderstanding your intent, you're absolutely right, the common '9 Volt battery' is indeed an excellent example of the 'battery made up of cells' idea. So few folks agree with me, I'd be shortsighted indeed not to agree!
If you check back to the very start of this good stuff, you'll see that efficiency is indeed very important. The original chap lives off grid and is dependent on solar power (which costs about $10/Watt for average five hours a day, less in winter). Although it is often quite true that lots of available energy changes the efficiency/cost/simplicity balance.
Also, I think all are in agreement that current regulation is better than voltage regulation. For that matter, probably all but a few would agree that regulated circuits in general are superior to either resistored or DD set ups. FWIW, the sorts of regulators we're talking about (as opposed to switching *converters* like in the ARC AAA or MM) in general have 'exactly' the same efficiency as a resistored light would running at the same point. The regulator is, if you will, a 'smart resistor' basically. That is it's a variable conductor (resistor) who's value is automatically varied by a control circuit. The circuit is 'looking' at something (like the actual current or voltage delivered to the load) and trying to adjust things to match it's own idea of the ideal value (which is hopefully the one we've told it.....).
Again, sorry for the misunderstanding, thanks for adding to the discussion.
Doug Owen
idleprocess said:
I was just addressing the tangential issue of "battery" vs "cell" in that statement.
..........
Since this isn't a portable operation, there could be a case for simplicity at the expense of some efficiency.
..........
A current supply is a good way to go, since LEDs regulate their own voltage handily if current is restricted.
[/ QUOTE ]
idleprocess,
Howdy. I'm sorry for misunderstanding your intent, you're absolutely right, the common '9 Volt battery' is indeed an excellent example of the 'battery made up of cells' idea. So few folks agree with me, I'd be shortsighted indeed not to agree!
If you check back to the very start of this good stuff, you'll see that efficiency is indeed very important. The original chap lives off grid and is dependent on solar power (which costs about $10/Watt for average five hours a day, less in winter). Although it is often quite true that lots of available energy changes the efficiency/cost/simplicity balance.
Also, I think all are in agreement that current regulation is better than voltage regulation. For that matter, probably all but a few would agree that regulated circuits in general are superior to either resistored or DD set ups. FWIW, the sorts of regulators we're talking about (as opposed to switching *converters* like in the ARC AAA or MM) in general have 'exactly' the same efficiency as a resistored light would running at the same point. The regulator is, if you will, a 'smart resistor' basically. That is it's a variable conductor (resistor) who's value is automatically varied by a control circuit. The circuit is 'looking' at something (like the actual current or voltage delivered to the load) and trying to adjust things to match it's own idea of the ideal value (which is hopefully the one we've told it.....).
Again, sorry for the misunderstanding, thanks for adding to the discussion.
Doug Owen
9X25
Newly Enlightened
Josey,
you're welcome.
I plan on going off-grid someday, too.
I recommend you tie all the L.E.D.'s in the same lamp assembly in parallel, that way if one fails , you're still operational. I assume your house is 12VDC.
Just remember that all L.E.D.'s are "current-driven" devices. Multiply current per L.E.D.(usually 20ma+) times the number in parallel and you'll know what to set your current regulator at. If you need help with the math, don't be afraid to ask; it's a real simple formula.
Solar cells and L.E.D.'s are a match made in heaven.
My mom has 2 solar-powered flashlights and loves them both!
9X25
you're welcome.
I plan on going off-grid someday, too.
I recommend you tie all the L.E.D.'s in the same lamp assembly in parallel, that way if one fails , you're still operational. I assume your house is 12VDC.
Just remember that all L.E.D.'s are "current-driven" devices. Multiply current per L.E.D.(usually 20ma+) times the number in parallel and you'll know what to set your current regulator at. If you need help with the math, don't be afraid to ask; it's a real simple formula.
Solar cells and L.E.D.'s are a match made in heaven.
My mom has 2 solar-powered flashlights and loves them both!
9X25
Josey
Flashlight Enthusiast
Thanks 9X25:
Since I'm my own power company, I have multiple redundancies in everything. I do have a 12V system, but my panels are wired either at 12V or 24V and then dumped into my batteries at 12V.
Some of my loads (and all of my LEDs) are DC (6V, 12V or 24V) but most are run through an inverter to 120VAC. My well pump is 240VAC (which originates from my 12V battery bank).
If you're in Florida, you can pump 12VDC solar power though an inverter and run your meter backwards. I think that is legal in Florida; it is in Washington, although it doesn't do me any good.)
You'll love solar once you get into it. All the town people (30 miles away) lose power quite often, or suffer from poor quality power. I never lose power.
Josey
Since I'm my own power company, I have multiple redundancies in everything. I do have a 12V system, but my panels are wired either at 12V or 24V and then dumped into my batteries at 12V.
Some of my loads (and all of my LEDs) are DC (6V, 12V or 24V) but most are run through an inverter to 120VAC. My well pump is 240VAC (which originates from my 12V battery bank).
If you're in Florida, you can pump 12VDC solar power though an inverter and run your meter backwards. I think that is legal in Florida; it is in Washington, although it doesn't do me any good.)
You'll love solar once you get into it. All the town people (30 miles away) lose power quite often, or suffer from poor quality power. I never lose power.
Josey
idleprocess
Flashaholic
Most LED arrays out there are series-parallel, so you balance redundancy against current efficiency.
With a nominal 12V setup, strings of 3 LEDs in parallel would maximize current efficiency using existing voltage.
Of course, a current regulator might make series-parallel arrays irrelevant if it balances the nominal watts on input and output sides.
With a nominal 12V setup, strings of 3 LEDs in parallel would maximize current efficiency using existing voltage.
Of course, a current regulator might make series-parallel arrays irrelevant if it balances the nominal watts on input and output sides.
