Low-loss circuit powers solar lantern
Edited by Bill Travis
Ramsesh Kumar, Bangalore, India -- EDN, April 1, 2004
The solar-lantern circuit in Figure 1 is a low-loss configuration that uses a 7W, four-pin CFL (compact fluorescent lamp) and a 12V, 7-Ahr, sealed, maintenance-free battery. The inverter features greater-than-85% efficiency, less-than-2-mA quiescent current, and a shunt-charge controller with deep-discharge and overcharge protection for the battery. The low quiescent current and the deep-discharge and overcharge protection ensure long life for the battery. The preheating feature in the inverter avoids the blackening of the end of the CFL, thereby ensuring long life. The circuit finds application in rural areas as a reliable, compact, portable light source and in urban areas as an emergency-lighting system. The shunt charge-controller circuit comprises IC1, a low-current, voltage-reference 2.5V LM385, and IC2, an LM324 comparator. IC2A, with resistors R1 through R8 and transistor Q1, provides protection against deep discharge of the battery.
The circuit switches off the load, including the inverter and the lamp, when the battery voltage falls below 10.8V and thus protects the battery from deep discharge. Under a no-load condition, the discharged battery voltage is approximately 12.2V. Hence, the circuit provides a deep-discharge reset level of 12.3V to avoid oscillations. Red LED1 indicates a low-battery condition. IC2B with resistors R9 through R14 and transistor Q2 provides protection against overcharging the battery. Q2 switches on and shunts the solar array when the battery voltage exceeds 14.8V and thus protects the battery from overcharging. Q2 turns off when the battery voltage drops below 12.5V and thus enables battery charging. D2 is a reverse-blocking diode. It prevents the discharge of the battery through the solar cells when the cells are not generating electricity. Amber LED2 indicates that the battery is in full-charge condition. Green LED3, along with IC2C and resistors R15 through R20, provides an indication of charging.
Table 1, Table 2, and Table 3 give core and winding details for the magnetic components in the circuit. The inverter uses a Class D, push-pull, force-driven topology with MOSFETs as switching devices. IC3, an SG3524, drives the inverter. The force-driven topology ensures trouble-free start-up in all environmental conditions. The switching frequency is approximately 26 kHz. Q6, along with resistors R29, R30, and R31 and capacitor C10, forms the preheating circuit. In addition to the 12V, 7-Ahr sealed, maintenance-free battery, the circuit uses a 10W, 12V single-crystalline-silicon solar-cell panel. The recorded backup time is approximately eight to 10 hours for a fully charged battery with a light output of 370 lumens using a 7W, four-pin CFL.
Check out our Best of Design Ideas section!
-
Prices for low cost solar lamps, heater, lanterns, fans and other products available at www.altenergy.in/solarpricelist.html
deena - 2007-10-3 19:38:00 PST -
Its a nice circuit. However in Figure 1, pin number 13 appears twice. Due to this its not very clear as to whether two E's [Ea and Eb] should be used or Ea and Cb is to be used. I have not wired this circuit, but am doubtful of the polarity of T1. If someone could shed some light on this.
Santosh B - 2006-6-6 06:19:00 PDT -
Beautifull Design for " Low Loss Circuit powers Solar Lantern" EDN 4/1/2004 saw one Typo, "Q6" in the text should be actually "Q5".. Someone probably pointed that out already...
Tony:>)
Tony DiGiesi - 2006-14-1 09:28:00 PST -
I agree, this looks interesting, but needs to be corrected. I am also wondering of the CFL tube and pre-heat relay are shown correctly connected. Normally, C7 would be shorted out by the pre-heat relay. Is this a differenty way to achieve the same preheating effect?
Mike - 2004-21-4 07:34:00 PDT -
There are lots of omissions on this idea. On figure 1, the terminal point above IC2A unlabelled, R32 not specified, C5 and C6 couldn't be in ufd., S1B is floating, terminal near S1B unlabelled, etc. It looks like an interesting circuit, if it could be built. Please correct the drawing.
Alan Crawford - 2004-7-4 14:44:00 PDT
