Maker Pro
Maker Pro

Noob / peer circuit review - LED Power Source

kickit2

Nov 20, 2011
14
Joined
Nov 20, 2011
Messages
14
I originally posted a thread reguarding making an LED power source to drive 84 LEDs at 10v for a solar installation. https://www.electronicspoint.com/driving-leds-var-power-source-efficiency-t241211.html

I have heeded the advice given there and moved to a boost/buck regulator and brought up the overating voltage to 25.7v and moved to 80 LEDs instead (8x10 LEDs with a 4Ohm resister on each run).

Being that I am still very much a noob (last time playing with circuit design was a highschool class) and I do not have any components (will be ordering them later), I decided to lay the circuit out in LTSpice and run the simulation. The simulation took over 12 hours and when all was said and done, I'm not seeing what I think I should have seen.

My concern is this:
This circuit uses PWM (~100Hz) to allow LED dimming and yet adjusting the duty cycle seems to have no affect on the circuit's power consumption. The power consumption simulated to jump (AC style) very quickly (10 microsecond range) with an average of about 7.5W. There is no diffrence in simulated power draw between the LEDs being switched on or off. I would assume that with the simulation being as indepth as it is, there would be a differance shown, but nada.

Also - when the MOSFET is switched to turn power on or off there is a jump and or dip in current and voltage. What is a good way to smooth this out for the load to the LEDs are not getting such a blast on switching or is this just to be expected and is it damaging to the LEDs at all?

So here's my plea - 97% of people here are LESS noob than I and could probably shine some light on this in some manner (Maybe I'm just not looking at this right?) I have attached the schematic and some of the plots.

Thanks
 

Attachments

  • FULL_CIRCUIT-LTSpice.zip
    4.8 KB · Views: 90
  • circuit-battery_power.png
    circuit-battery_power.png
    6 KB · Views: 169
  • circuit-resister_load.png
    circuit-resister_load.png
    6.2 KB · Views: 189
  • circuit-voltage_from_regulator-before_load.png
    circuit-voltage_from_regulator-before_load.png
    5.6 KB · Views: 144
  • circuit-voltage_from_regulator-after_load.png
    circuit-voltage_from_regulator-after_load.png
    5.3 KB · Views: 154
  • circuit-schematic.png
    circuit-schematic.png
    43.7 KB · Views: 175
Last edited:

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Can you show us where you got your circuit from. I don't understand it and I'd like to read an explanation (if available)
 

kickit2

Nov 20, 2011
14
Joined
Nov 20, 2011
Messages
14
Well I didn't "get it" from anywhere - its more of a culmination of successful bits put together. Ill try to break it down..


The LT1171 is the boost/buck regulator and its layout is in accordance with the datasheet to give 25.7V output. I upped the ammount of cap on the input and output of it to try and stablize the ripple a bit more than the design showed. I also increased the rating of the inductor as 50uH inductors seem quite hard to find at a decent price. BTW - I realize that the resisters used with the regulator are not real values in all cases - game plan is to build them via series with a trimmer. This output runs to the load.

The LT1030 is LDO voltage regulator giving a 9v output. This will be replaced by a non-adjustable unit - but it was just easier to simulate with this one as Linear Tech does not have a fixed 9v LDO and I didn't want to spend the time to find a spice model for another 9v LDO. Anyways - this output runs to power the 555 as well as though two voltage dividers to supply conparison values. Note that the power to the 555 runs first though the relay so it can be switched.

The 555 is setup to be astable but instead of pulling the output from its output pin, I am dumping it to ground and instead pulling the oscilations off the trigger and threshold. This sawtooth waveform is fed to one of the comparitors for use in making an adjustable duty cycle while maintaing the specific frequency. Again I realize that these resister values used are not real in all cases and I will adjust them in the future to fit real values (exactly 100Hz is not needed, just close.. ie +10% or so)

The comparitor that is connected to the 555 also connects to a voltage divider which also has a 10k pot so I can adjust the negative referance voltage. Note that LTSpice does not have pots so I used R10 and R9 to represent a 10k pot set at 70%. In this way I can, as I said before, increase or decrease the duty cycle without messing up the frequency (100Hz). The comparitors output runs to the Mosfet to switch current for the main load.

The other comparitor is setup with voltage dividers so that its output goes low if the voltage from the battery falls below 10.6V. A high value from this comparitor is required to turn the npn transister "on", and in doing so, allow current though the relay powering the rest of the LM1171 and the 555. So basically, if the power is below 10.6 volts, the LT1171, 555 and the load will recieve no power preventing the battery from being damaged due to being drained too low. The "off" positions of the relay run though 10MegOhm resisters to ground to make LTSpice stop complaining.

And thats pretty much it - I hope it makes sence now.
 
Last edited:

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
If I understand you, you are using a buck/boost converter to get 25.7 V and then running that through a linear regulator to get 9V to run the LEDs? That makes no sense at all. The idea of a DC/DC converter is to avoid the losses of a linear regulator, and you are basiclaly gauarnteeing a loss of 65% instead. You should either using a convert to ouptut 9V or running the LEDs in longer strings off the 25.7V.

Bob
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
I would also look at whether the LT1171 can be used to create a current limited supply. If you do this then you will not need resistors to limit current (although you will still need them to balance the current between strings).

Looking at the datasheet, you can set a current limit.

This could also be used for your dimming control, reducing complexity significantly. I'm not sure if the external current limit allows the limit to be set all the way to zero though.

You can probably do something with the current limit to shut things down when the battery voltage falls, reducing complexity still more.
 

kickit2

Nov 20, 2011
14
Joined
Nov 20, 2011
Messages
14
Bob: No - the 9v regulator powers only the 555 and the voltage dividers for use in PWM as well as switching the relay for battery protection. Although I was thinking about this and think I can do it a bit diffrently.

Steve: Your post actually got me thinking about using a current controlled device. IF they work the way that I think they would work (still noob here - gotta do more reading) then I definatly understand how you say they can make the circuit less complex. So it looks at though I am not done yet!


Thanks
 
Top