Convert solar LED lights to direct power source

Nov 13, 2009

I bought a set of stainless steel solar LED landscaping lights. They
use a single battery and worked OK during the summer. But now with
the short winter days and low sun, they are not working at all. I've
tried refreshing batteries, getting much higher amp-hr batteries,
etc. So I'm thinking I want to just solder wire leads to them and
hook them up in parallel to a 3v power supply that I have.

I have 10 LED lights, approx 30mA each. The power supply is 300 mA
rated, so I'd be right at the limit. Will wiring them up in parallel
work? Will the 3v power damage the LEDs that are evidently designed
for 1.2V? I've hooked one up and it works, but I assume the life will
be shorter. Anything I should do differently? I also have a 6 V,
1200mA power supply and used a couple resistors to divide the voltage
to get 1.5 volts, but for some reason it would not power the LED.

John Fields · Nov 13, 2009

On Thu, 12 Nov 2009 17:26:22 -0800 (PST), borne@gate.net wrote:

I bought a set of stainless steel solar LED landscaping lights. They
use a single battery and worked OK during the summer. But now with
the short winter days and low sun, they are not working at all. I've
tried refreshing batteries, getting much higher amp-hr batteries,
etc. So I'm thinking I want to just solder wire leads to them and
hook them up in parallel to a 3v power supply that I have.

I have 10 LED lights, approx 30mA each. The power supply is 300 mA
rated, so I'd be right at the limit. Will wiring them up in parallel
work?

---
It might, but wiring LEDs in parallel is never a good idea because the
one with the lowest Vf (forward voltage drop) will hog the current,
overheat and die. If it fails shorted and the supply isn't current
limited, then the supply could be damaged.

If it fails open, then the LED with the next highest Vf will hog the
current, and so on...
---

Will the 3v power damage the LEDs that are evidently designed
for 1.2V?

---
If the LEDs are white, they'll certainly have a higher Vf than 1.2V.
---

I've hooked one up and it works, but I assume the life will
be shorter.

---
Yes.

Since it didn't fail outright, it's Vf is close to 3V.
---

Anything I should do differently? I also have a 6 V,
1200mA power supply and used a couple resistors to divide the voltage
to get 1.5 volts, but for some reason it would not power the LED.

---
That's because the divider couldn't supply the current necessary for the
LED to work.

The Vf of an LED isn't specified as the voltage which must be placed
across the LED to force a certain current through it, it's specified as
the range of voltages which will appear across an LED when a given
forward current is forced through it.

That is, if Vf is specified as 3 to 4V at 30 mA, then when 30 mA is
forced through one LED the voltage across it might be 3V, while another
LED of the same type with 30 mA through it might have a voltage across
it of 4V.

Or anything between 3 and 4V.

The correct way to use your 6V supply is to place a current limiting
resistor in series with each LED and to connect all of the resistor-LED
combinations across the supply in parallel: (View in Courier)

+-------+----//----+
| | |
| [R1] [Rn]
|+ | |
[SUPPLY] | |
| |A |A
| [LED1] [LEDn]
| | |
+-------+----//----+

The resistor's job will be to drop the excess voltage from the supply
while allowing the desired current into the LED, and it's value can be
found from:

V(supply) - Vf(led)
R = ---------------------
If(led)

Assuming that the supply is 6VDC and that the current through the LED
needs to be 30mA with 3V minimum across the LED yields:

6V - 3V 3V
R = ----------- = ------- = 100 ohms,
0.03A 0.03A

which is a standard 5% resistor, and would work.

The power dissipated by the resistor would be:

P = If(led) * (V(supply) - Vf(led))

= 0.03A * 3V

= 0.09 watt

So a 100 ohm +/- 5%, 1/4 watt resistor would be fine.

JF

Nov 13, 2009

Thanks. I found this site... http://metku.net/index.html?sect=view&n=1&path=mods/ledcalc/index_eng#parallel

Could I use a single 11 ohm resistor as calculated there, or would
that be the same as wiring in parallel without individual resistors at
each LED, thus making the one with the highest Vf hog all the current?

On Nov 13, 7:44 am, John Fields <jfie...@austininstruments.com> wrote:

On Thu, 12 Nov 2009 17:26:22 -0800 (PST), bo...@gate.net wrote:
I bought a set of stainless steel solar LED landscaping lights. They
use a single battery and worked OK during the summer. But now with
the short winter days and low sun, they are not working at all. I've
tried refreshing batteries, getting much higher amp-hr batteries,
etc. So I'm thinking I want to just solder wire leads to them and
hook them up in parallel to a 3v power supply that I have.

I have 10 LED lights, approx 30mA each. The power supply is 300 mA
rated, so I'd be right at the limit. Will wiring them up in parallel
work?

---
It might, but wiring LEDs in parallel is never a good idea because the
one with the lowest Vf (forward voltage drop) will hog the current,
overheat and die. If it fails shorted and the supply isn't current
limited, then the supply could be damaged.

If it fails open, then the LED with the next highest Vf will hog the
current, and so on...
---

Will the 3v power damage the LEDs that are evidently designed
for 1.2V?

---
If the LEDs are white, they'll certainly have a higher Vf than 1.2V.
---

I've hooked one up and it works, but I assume the life will
be shorter.

---
Yes.

Since it didn't fail outright, it's Vf is close to 3V.
---

Anything I should do differently? I also have a 6 V,
1200mA power supply and used a couple resistors to divide the voltage
to get 1.5 volts, but for some reason it would not power the LED.

---
That's because the divider couldn't supply the current necessary for the
LED to work.

The Vf of an LED isn't specified as the voltage which must be placed
across the LED to force a certain current through it, it's specified as
the range of voltages which will appear across an LED when a given
forward current is forced through it.

That is, if Vf is specified as 3 to 4V at 30 mA, then when 30 mA is
forced through one LED the voltage across it might be 3V, while another
LED of the same type with 30 mA through it might have a voltage across
it of 4V.

Or anything between 3 and 4V.

The correct way to use your 6V supply is to place a current limiting
resistor in series with each LED and to connect all of the resistor-LED
combinations across the supply in parallel: (View in Courier)

+-------+----//----+
| | |
| [R1] [Rn]
|+ | |
[SUPPLY] | |
| |A |A
| [LED1] [LEDn]
| | |
+-------+----//----+

The resistor's job will be to drop the excess voltage from the supply
while allowing the desired current into the LED, and it's value can be
found from:

V(supply) - Vf(led)
R = ---------------------
If(led)

Assuming that the supply is 6VDC and that the current through the LED
needs to be 30mA with 3V minimum across the LED yields:

6V - 3V 3V
R = ----------- = ------- = 100 ohms,
0.03A 0.03A

which is a standard 5% resistor, and would work.

The power dissipated by the resistor would be:

P = If(led) * (V(supply) - Vf(led))

= 0.03A * 3V

= 0.09 watt

So a 100 ohm +/- 5%, 1/4 watt resistor would be fine.

JF

John Fields · Nov 13, 2009

On Fri, 13 Nov 2009 05:50:30 -0800 (PST), borne@gate.net wrote:

Thanks. I found this site... http://metku.net/index.html?sect=view&n=1&path=mods/ledcalc/index_eng#parallel

Could I use a single 11 ohm resistor as calculated there, or would
that be the same as wiring in parallel without individual resistors at
each LED, thus making the one with the highest Vf hog all the current?

---
Please bottom post.

It's never the one with the highest Vf that hogs the current, it's the
one with the lowest Vf, so any scheme where the LEDs are in parallel
without some form of individual current limiting for each LED is suspect
unless you can choose the resistor so that it would protect the LED with
the lowest Vf, no matter what the Vf spread of the others.

Worst case would be with one LED having a Vf of, say, 3V while all the
others had a Vf of 4V, but what's nasty about that is that, since the
current isn't locked down for the other LEDs with 3V across the, the
current they don't draw will be dumped into the LED with the low Vf.

Might hurt it, might not, but why take that chance?

If you need a parallel connection and want something that's going to
work every time and which will give you maximum life from your LEDs, use
a series resistor for each LED and connect those combinations, in
parallel, across your supply.

JF

Nov 17, 2009

On Nov 13, 5:34 pm, John Fields <jfie...@austininstruments.com> wrote:

On Fri, 13 Nov 2009 05:50:30 -0800 (PST), bo...@gate.net wrote:
Thanks. I found this site...http://metku.net/index.html?sect=view&n=1&path=mods/ledcalc/index_eng...

Could I use a single 11 ohm resistor as calculated there, or would
that be the same as wiring in parallel without individual resistors at
each LED, thus making the one with the highest Vf hog all the current?

---
Please bottom post.

It's never the one with the highest Vf that hogs the current, it's the
one with the lowest Vf, so any scheme where the LEDs are in parallel
without some form of individual current limiting for each LED is suspect
unless you can choose the resistor so that it would protect the LED with
the lowest Vf, no matter what the Vf spread of the others.

Worst case would be with one LED having a Vf of, say, 3V while all the
others had a Vf of 4V, but what's nasty about that is that, since the
current isn't locked down for the other LEDs with 3V across the, the
current they don't draw will be dumped into the LED with the low Vf.

Might hurt it, might not, but why take that chance?

If you need a parallel connection and want something that's going to
work every time and which will give you maximum life from your LEDs, use
a series resistor for each LED and connect those combinations, in
parallel, across your supply.

JF

I wired 9 LED lights in parallel with each getting a 100 ohm
resistor. I upgraded my power supply to 6VDC 1200 mA. However,
something is wrong. The first 5 LEDs on the string work great, but
the last 4 flicker. I've checked the wiring, switched LEDs, and no
matter what, it is always the last 4. I can't figure it out. Any
ideas?

John Fields · Nov 17, 2009

On Mon, 16 Nov 2009 15:44:05 -0800 (PST), borne@gate.net wrote:

On Nov 13, 5:34 pm, John Fields <jfie...@austininstruments.com> wrote:
On Fri, 13 Nov 2009 05:50:30 -0800 (PST), bo...@gate.net wrote:
Thanks. I found this site...http://metku.net/index.html?sect=view&n=1&path=mods/ledcalc/index_eng...

Could I use a single 11 ohm resistor as calculated there, or would
that be the same as wiring in parallel without individual resistors at
each LED, thus making the one with the highest Vf hog all the current?

---
Please bottom post.

It's never the one with the highest Vf that hogs the current, it's the
one with the lowest Vf, so any scheme where the LEDs are in parallel
without some form of individual current limiting for each LED is suspect
unless you can choose the resistor so that it would protect the LED with
the lowest Vf, no matter what the Vf spread of the others.

Worst case would be with one LED having a Vf of, say, 3V while all the
others had a Vf of 4V, but what's nasty about that is that, since the
current isn't locked down for the other LEDs with 3V across the, the
current they don't draw will be dumped into the LED with the low Vf.

Might hurt it, might not, but why take that chance?

If you need a parallel connection and want something that's going to
work every time and which will give you maximum life from your LEDs, use
a series resistor for each LED and connect those combinations, in
parallel, across your supply.

JF

I wired 9 LED lights in parallel with each getting a 100 ohm
resistor. I upgraded my power supply to 6VDC 1200 mA. However,
something is wrong. The first 5 LEDs on the string work great, but
the last 4 flicker. I've checked the wiring, switched LEDs, and no
matter what, it is always the last 4. I can't figure it out. Any
ideas?

---
How are they wired?

In parallel, I know, but like this, with a single pair feeding all of
the LEDs:

+---------+------+-------//-------+------+
| | | | |
| [R1] [R2] [R9] [R10]
[SUPPLY] | | | |
| [LED1] [LED2] [LED9][LED10]
| | | | |
+---------+------+-------//-------+------+

or like this, with separate pairs going from each LED assembly to the
supply?:

+-------------------------------------------+
| |
|+-----------------------------------+ |
|| | |
|| +---------------+ | |
|| | | | |
|| |+-------+ | | |
|| || | | | |
|| || [R1] [R2]. . . . . . .[R9] [R10]
[SUPPLY] | | | |
|| || [LED1] [LED2]. . . . . .[LED9][LED10]
|| || | | | |
|| |+-------+ | | |
|| | | | |
|| +---------------+ | |
|| | |
|+-----------------------------------+ |
| |
+-------------------------------------------+

If it's the second way and you don't have the wires buried, I'd almost
be willing to bet that the longer wires are capturing enough AM radio
for the LEDs to rectify it and modulate the current through themselves
so that the flicker you see is the half-wave rectified audio envelope.

The cure would be to place a capacitor across the resistor-LED
combination so that it would bypass the RF.

I'd suspect something with a reactance of an ohm or less at 500kHz would
be OK, so since:

1 1
C = ----------- = ------------------- ~ 3.18e-7F ~ 0.318ľF,
2pi f Xc 6.28 * 5e5Hz * 1R

something like a 0.33ľF or larger polyester cap would be fine.

It's also possible that there's enough 60 Hz floating around and getting
rectified by the LED so that you see a 30Hz flicker. The cure for that,
I think, would be to twist the wires all the way from the supply to the
LED.

If it's wired the first way, I have no clue... :-(

JF

Nov 17, 2009

On Nov 17, 11:55 am, John Fields <jfie...@austininstruments.com>
wrote:

On Mon, 16 Nov 2009 15:44:05 -0800 (PST), bo...@gate.net wrote:
On Nov 13, 5:34 pm, John Fields <jfie...@austininstruments.com> wrote:
On Fri, 13 Nov 2009 05:50:30 -0800 (PST), bo...@gate.net wrote:
Thanks. I found this site...http://metku.net/index.html?sect=view&n=1&path=mods/ledcalc/index_eng...

Could I use a single 11 ohm resistor as calculated there, or would
that be the same as wiring in parallel without individual resistors at
each LED, thus making the one with the highest Vf hog all the current?

---
Please bottom post.

It's never the one with the highest Vf that hogs the current, it's the
one with the lowest Vf, so any scheme where the LEDs are in parallel
without some form of individual current limiting for each LED is suspect
unless you can choose the resistor so that it would protect the LED with
the lowest Vf, no matter what the Vf spread of the others.

Worst case would be with one LED having a Vf of, say, 3V while all the
others had a Vf of 4V, but what's nasty about that is that, since the
current isn't locked down for the other LEDs with 3V across the, the
current they don't draw will be dumped into the LED with the low Vf.

Might hurt it, might not, but why take that chance?

If you need a parallel connection and want something that's going to
work every time and which will give you maximum life from your LEDs, use
a series resistor for each LED and connect those combinations, in
parallel, across your supply.

JF

I wired 9 LED lights in parallel with each getting a 100 ohm
resistor. I upgraded my power supply to 6VDC 1200 mA. However,
something is wrong. The first 5 LEDs on the string work great, but
the last 4 flicker. I've checked the wiring, switched LEDs, and no
matter what, it is always the last 4. I can't figure it out. Any
ideas?

---
How are they wired?

In parallel, I know, but like this, with a single pair feeding all of
the LEDs:

+---------+------+-------//-------+------+
| | | | |
| [R1] [R2] [R9] [R10]
[SUPPLY] | | | |
| [LED1] [LED2] [LED9][LED10]
| | | | |
+---------+------+-------//-------+------+

or like this, with separate pairs going from each LED assembly to the
supply?:

+-------------------------------------------+
| |
|+-----------------------------------+ |
|| | |
|| +---------------+ | |
|| | | | |
|| |+-------+ | | |
|| || | | | |
|| || [R1] [R2]. . . . . . .[R9] [R10]
[SUPPLY] | | | |
|| || [LED1] [LED2]. . . . . .[LED9][LED10]
|| || | | | |
|| |+-------+ | | |
|| | | | |
|| +---------------+ | |
|| | |
|+-----------------------------------+ |
| |
+-------------------------------------------+

If it's the second way and you don't have the wires buried, I'd almost
be willing to bet that the longer wires are capturing enough AM radio
for the LEDs to rectify it and modulate the current through themselves
so that the flicker you see is the half-wave rectified audio envelope.

The cure would be to place a capacitor across the resistor-LED
combination so that it would bypass the RF.

I'd suspect something with a reactance of an ohm or less at 500kHz would
be OK, so since:

1 1
C = ----------- = ------------------- ~ 3.18e-7F ~ 0.318ľF,
2pi f Xc 6.28 * 5e5Hz * 1R

something like a 0.33ľF or larger polyester cap would be fine.

It's also possible that there's enough 60 Hz floating around and getting
rectified by the LED so that you see a 30Hz flicker. The cure for that,
I think, would be to twist the wires all the way from the supply to the
LED.

If it's wired the first way, I have no clue... :-(

JF- Hide quoted text -

- Show quoted text -

Its wired the first way, with two 90 foot wires and the LEDs connected
at nearly equal intervals along the way. It is 24 ga wire, which is
small I know, but the voltage reading at the end is still around 7.
My 6 V power supply is really putting out 8V. Anyway, I'll be
bringing the entire thing back into the garage and playing around with
it over the weekend - trying larger ga wire, running two more power
wires to power the last LEDs with a different feed, etc.

Thanks for your help.

sandy · Nov 22, 2009

On Nov 12, 7:26 pm, bo...@gate.net wrote:

I bought a set of stainless steel solar LED landscaping lights. They
use a single battery and worked OK during the summer. But now with
the short winter days and low sun, they are not working at all. I've
tried refreshing batteries, getting much higher amp-hr batteries,
etc. So I'm thinking I want to just solder wire leads to them and
hook them up in parallel to a 3v power supply that I have.

I have 10 LED lights, approx 30mA each. The power supply is 300 mA
rated, so I'd be right at the limit. Will wiring them up in parallel
work? Will the 3v power damage the LEDs that are evidently designed
for 1.2V? I've hooked one up and it works, but I assume the life will
be shorter. Anything I should do differently? I also have a 6 V,
1200mA power supply and used a couple resistors to divide the voltage
to get 1.5 volts, but for some reason it would not power the LED.

Hello,
Check for series parallel cobination for connecting external battery.
mail me for detailed information @ info@technoveda.net

Per Jensen · Nov 25, 2009

borne@gate.net wrote:

Will wiring them up in parallel

work? Will the 3v power damage the LEDs that are evidently designed
for 1.2V?

You should be aware that no white LED will work at 1,2 volt, hence there
is a small switch mode power supply in the lamps that up the voltage to
3 to 5 volts to overcome the Vf of the LED's.

If you look closely on the PCB in the lamp, you will see a switching
diode, a coil (it's usually green with stripes on, and looks like an 1/4
watt resusor, just a little fatter.

A small circuit of 2 or 3 transistors drive the oscillator and the
inhibit-circuit with the CdS-sensor.

// Per.

Convert solar LED lights to direct power source

Guest

John Fields

Guest

Guest

John Fields

Guest

Guest

John Fields

Guest

Guest

sandy

Guest

Per Jensen

Guest

Welcome to EDABoard.com

Sponsor

Online statistics

Forum statistics

Convert solar LED lights to direct power source

Guest

John Fields

Guest

Guest

John Fields

Guest

Guest

John Fields

Guest

Guest

sandy

Guest

Per Jensen

Guest

Log in

Welcome to EDABoard.com

Sponsor