Which is better circuit …?
Short answer: In my opinion, neither circuit is OK. I saw the circuits and checked the links. In general, I did not like the choices and concepts I could get from them. One of the reasons is the configuration is unnecessarily convoluted and confused. Even the symbology, the diagram and way of using the Op Amps (in LM358) of the first are confusing. The second diagram uses a relay which role, behavior and its wiring are not clear.
My suggestion, is based on the circuit found in the LM350 datasheet by Texas, replaced here by the LM317. I constructed it a few times, and including trimpot for fine adjustment. It worked well and essentially behaves as a 3 stage charger as described in BU-403 and summarized here:
- Constant current charge: Responsible for up to 70% of total charge. Charger works initially Works at Maximum current, about 2A, being limited by the LM317 itself and then by its built in thermal overheating protection to a lower value, about 1.0~1.5A, depending of the heat sink being used.
- Constant Voltage Topping charge: Responsible to reach 100% of charge, its charging voltage is higher than the final voltage. This stage terminates when the charging current goes below a given threshold. Definition of the topping voltage is made by voltage divider in LM317 ADJ terminal with (R2+R3) and R1. If R1 is replaced by a fixed resistor and a trimpot, this topping voltage can be fine-tuned. Minimum current limit is around 5% of nominal AH rating is defined by voltage drops that occur in R6 and R2, being compared/measured by the Op Amp (circuit: LM301)
- Floating Voltage stage: When charging is considered complete, the final voltage should be reduced to avoid gassing and dry up of battery’s electrolyte. Resistor R4 is the main responsible one to reduce the initial Topping voltage.
When I made this circuit, I did some changes, highlighted in colored lines (replacing Red by Blue when using LM741, or LM358):
Additional information: (A) Protection diodes This circuit works nicely, but protection diodes around the LM350/LM317 are a safer approach - see Figure 17 of that LM350 datasheet. (B) LM350➡️LM317: Instead of LM350, you can use an LM317, much easier to find. It will be able to provide 1.5A, that is more than enough for the 7AH SLA battery. (C) LM301 alternatives: If positive supply line of LM301 is moved to a higher voltage as Vin, this IC can be replaced by the easy to find LM741 or by one OpAmp of LM358 (dual OpAmp). Probably a direct replacement to the LM301 there (maintaining V+ in Vout of LM317) would be TL081 for similar (TL071).
Counterfeit semiconductors: There are reports of fake TL081, and even LM317 & LM350 that have been sold thru Oriental web-based platforms. I have seen some “LM317” that do not withstand more than 25V~30V in their input. Reprintings of TL082 or TL084 are indeed anonymous sandblasted LM358 and LM324, all this with sad results. So, if voltages, response speeds or power dissipation are not close to the maximum/rated official limits, someone may be able to use a fake component without major problems. As LM317 would need less than 25V in this circuit, you might not have problems, even with fake regulators. So if problems do occur, consider re-checking if all components are original (or not).
Voltage adjustments and Current threshold: While key components for voltage and current limits were explained above, in my opinion the current threshold is highly dependent of battery size (in AH) and should be checked more carefully. To not change current and voltage settings (when R2 is trimmed) an alternative would be to fine-tune R6 - replacing the 0.2R resistor by 10x 2R2 1W resistors in parallel and eventually cutting gradually each one while monitoring at which current the change from stages 2 ↔️ 3 occurs. The typical is set the change at ~5% of AH. In this case, about 350mA.