0
\$\begingroup\$

I have 2 nos 7W LED Modules, each of which are fed from a separate XL6001 based LED Driver. The DC power input to the drivers is from a 12V, 7.5AH SLA Battery. The LED driver takes 12V DC from the battery and boosts it to 22V DC for the LED module. The current consumption at the lamp side is 636mA (for both lamps), while on the battery side is upwards of 1.2A (for both lamps). All this works fine.

Now, in a separate mode of operation, I would like these lamps to light up from AC power. The design of the operating unit (the overall box) is such that power must go through the LED drivers to reach the lamps. So, for the AC power mode, I have been trying to use an LM317 voltage regulator outputting 12V DC which is then fed to the above LED drivers. I am observing that after a few minutes the output of the LM317 drops to 4.5V DC and current falls down to 300mA (for both lamps) on the input of LED driver. This results in "abysmal" brightness from the lamps. I have also tried using an LM350 (3A rated) and found the same sudden voltage drop.

What am I missing? Is it heatsinking...though I am using a regular heatsink which is used with TO-220 packages. Ambient is around 28 C.

Should I configure LM317 as a constant current source and then feed its output to the LED driver ?

Should I configure the output voltage from LM317 to be 22V and then feed it to the LED driver ?

The input to voltage regulator is from the bridge rectified output of a 230V:24V single phase, 50VA transformer.

Thanks and best regards,

Vishal

\$\endgroup\$
6
  • 1
    \$\begingroup\$ why 220Vac to 24Vac by transformer rectified to 30Vdc then LM317 to 12Vdc then boost to 24Vdc . Makes no sense. \$\endgroup\$
    – D.A.S.
    Commented Mar 13, 2017 at 20:16
  • \$\begingroup\$ When operating correctly, what is the measured LED voltage from the LED driver output? What is the max forward voltage for each LED from the datasheet? The LEDs are connect in series? And the measured current through the LEDs is 636mA? \$\endgroup\$ Commented Mar 13, 2017 at 20:17
  • \$\begingroup\$ What about heatsinks on the LEDs? What is the temperature of the LEDs? You cannot run 7W LEDs at 7W without substantial thermal management. \$\endgroup\$ Commented Mar 13, 2017 at 20:26
  • \$\begingroup\$ @Tony Stewart, the main power path is for a battery charger which is used to light up the lamps once battery is fully charged. But now in a second mode i have to light the lamps from AC. Since it's the same unit and same led drivers have to be used for power to reach the lamps....we have to go through this voltage conversions. \$\endgroup\$
    – Vishal
    Commented Mar 14, 2017 at 9:35
  • \$\begingroup\$ Or install a separate power supply unit, may be a 12V, 2A switching ps. But the AC input cannot be changed, unless I change the transformer...!! \$\endgroup\$
    – Vishal
    Commented Mar 14, 2017 at 9:37

2 Answers 2

1
\$\begingroup\$

Disable/Disconnect one of the LED drivers.
Or disconnect the LEDs from one of the drivers.

I want to see if lowering the load on the LM317 helps.

UPDATE

Do you have a schematic?

I'd like to know exactly what series parallel means.
I assume two strings of 7 White LEDs.

Also how is the parallel was done?
You should not drive two strings with one driver. If you do, you should balance the current to the strings.
That would not be the cause here though. Different matter all together.

Now the question is what is causing the failure? The LM317 or Driver?

You said the output of the :LM317 (Vin to Driver) drops to 4.5v. If that is true then the problem must be the LM317 overheating.

How long did you allow it to remain in failure mode? Long enough for it to cool down and recover from the thermal shutdown?
That may take as long as it did to overheat.

FROM THE TI LM317 DATASHEET:
When an overload occurs the device shuts down Darlington NPN output stage or reduces the output current to prevent device damage. The device will automatically reset from the overload. The output may be reduced or alternate between on and off until the overload is removed.

End of Update


Try Doubling the value of RS.

Very likely the XL6001 is going into thermal shutdown.

The datasheet does not specify a max power, it just says "internally regulated".
and:

  • Built in Thermal Shutdown Function
  • Built in Current Limit Function

The datasheet does not specify its drive capabilities other than:

The XL6001 could directly drive 5~8 Series 1W LED units at VIN>12V.

That would indicate the driver can drive about 5-8 watts. Not 14.


-

  1. When operating correctly, what is the measured LED voltage from the LED driver output?
  2. What is the max forward voltage for each LED from the datasheet?
  3. Are the LEDs are connect in series?
  4. Is the measured current through the LEDs 636mA?
  5. XL6001 is configured as only a Boost Driver?
\$\endgroup\$
7
  • \$\begingroup\$ Hi, (1) There are 2 nos 7W modules and 2 nos LED Drivers. One for each. So wattage per driver is within driver capacity. (2) When operating correctly, measured LED (string) voltage from the LED driver output is 22V (21.8 to 22.6) DC. (3) This is a 14W module that we purchase directly, so I am not sure what LEDs are on it...but they are good quality because we have not seen them failing in many years. (4) LEDs are series parallel (14 nos each of 0.5W). \$\endgroup\$
    – Vishal
    Commented Mar 14, 2017 at 1:07
  • \$\begingroup\$ (5) The measured current at the output of the LED driver (multimeter placed in series with the LED module), shows 318mA per module. Since we have two modules running together I have written 636mA. (6) Yes Looks like it is configured only as Boost driver, because it takes input from 12V and sends it to the 7W module. But I have not yet traced the driver circuit to find this out. \$\endgroup\$
    – Vishal
    Commented Mar 14, 2017 at 1:10
  • \$\begingroup\$ This is how it's wired for every 7W module. Two 0.5W LEDs are in parallel, and 7 such units are in series. This string is fed from a single driver. This module-&-driver is replicated in the second lamp also. Lamps have adequate heatsinking. However Voltage Regulator heatsinking is just a TO-220 heatsink. It looks like the LM317 is going into thermal shutdown \$\endgroup\$
    – Vishal
    Commented Mar 14, 2017 at 4:20
  • \$\begingroup\$ As soon as the driver is removed, which means load is removed from the regulator., It's voltage builds back up again. \$\endgroup\$
    – Vishal
    Commented Mar 14, 2017 at 4:20
  • \$\begingroup\$ I asked you to shut down one (1) of the two (2) drivers. And see if it keeps working like it should. I want to see if it will work with one half (50%) of the load with two drivers. \$\endgroup\$ Commented Mar 15, 2017 at 23:09
0
\$\begingroup\$

LM317 at 1.2 amps? That's ambitious. If it shuts down after a short while I'd bet on insufficient heat sinking.

Get a switching 12VDC power supply rated for proper current, like 2 amps.

Another option would be a standard 19V laptop power supply, since XL6001 is rated for this voltage too. You'll need to check input capacitors on your driver can take the voltage, though.

\$\endgroup\$
4
  • \$\begingroup\$ Thanks peufeu. I thought earlier it would be because of current capacity of LM317. However, using an LM350 (which is rated for 3A) also shows the same effect? Any pointers? \$\endgroup\$
    – Vishal
    Commented Mar 13, 2017 at 19:47
  • \$\begingroup\$ Need: regulator input dc voltage, heat sink rating in °C/W. Is heat sink very hot? \$\endgroup\$
    – bobflux
    Commented Mar 13, 2017 at 19:53
  • \$\begingroup\$ Regulator input voltage is about 26V (from the bridge rectifier). I understand the voltage headroom for the regulator is very high in this case...but the result was no different when it was 18V using a different transformer. Heat Sink rating not available. What should it be for this power? How to calculate upper limit of thermal resistance allowed here? Heat Sink is pretty hot...I better not touch with bare hands. \$\endgroup\$
    – Vishal
    Commented Mar 13, 2017 at 19:58
  • \$\begingroup\$ 26-12=14V so it dissipates 17W, your heat sink should be 2°C/W maximum, which is already quite large. Why not use a 12V 2A switching supply? You can get those cheap... \$\endgroup\$
    – bobflux
    Commented Mar 13, 2017 at 20:01

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.