Every figure in this article comes from one source video — Lithium Solar's Building a 3.5kWh DIY Solar Generator for $650 — Start to Finish, which has been watched more than 6.8 million times. The host documents the project from a friend's question about solar generators all the way to a finished toolbox unit that runs power tools, charges from a single 270W panel and cost about $641 in parts. If you would rather watch the build before reading the breakdown, the video is right here:
The core question the video answers is practical: can a normal hobbyist build a real 3.5kWh solar generator at home, and what does it actually cost? The answer is roughly $641 and a weekend of careful work. Below is the full breakdown, every number taken from the on-camera build. If you are comparing this to a store-bought unit or sizing any battery project, the same three pillars decide everything — capacity, discharge rate and recharge source — which is exactly how we evaluate our own battery packs at Dajiu Energy.
A pre-built 3.5kWh-class portable power station with a 1,500W inverter typically costs significantly more than the $641 this build came to, and you give up the ability to choose every component. The DIY route has three advantages the video makes clear: you control cell quality (every one of the 390 cells was individually tested), you control the wiring and BMS, and you can service or upgrade any part later. The trade-offs are the time, the tools (a spot welder is required) and the safety discipline — with recycled lithium cells, mistakes are expensive. For many people a quality commercial unit is the better call; for someone comfortable with electronics, a build like this delivers more capacity per dollar than almost anything off the shelf.
The heart of the build is 130 modem battery packs bought from Battery Hookup at $1.50 per pack — 390 individual 18650 cells total, rated about 2,600mAh each. The host opens each pack by snipping the two corners away from the terminals with heavy wire cutters, then pulls the PCB out with long-nosed pliers to free the three cells inside. The result is surprising: out of 390 cells, not one single cell was bad. Almost all tested in the 2,500-2,700mAh range on an individual capacity tester, and only 11 fell below the 2,500 threshold — still usable, just slightly lower capacity.
The gray cells tested slightly higher than the purple ones, so the host separates them to keep an even distribution across the series packs. This is the single most important habit when working with recycled cells: test everything individually, sort by capacity, and never trust a cell's printed label. The same discipline applies to new cells too — voltage matching before assembly prevents the pack from developing imbalances that the BMS then has to fight for its whole life. When we build LiFePO4 packs at the factory, every cell is graded and matched before a single weld is made.
The host builds two packs in a 14S26P configuration — 14 cells in series, 26 cells in parallel per series step. Each group of 26 parallel cells shares one voltage, and the 14 groups are wired in series to reach the system voltage. The math from the video: one cell is 3.7V nominal × 2.6A = 9.62Wh per cell; × 364 cells = 3,501Wh of storage. With 14 cells in series, the pack runs at 51.8V nominal — slightly above a nominal 48V system, which is ideal for 48V inverters and charge controllers.
The host deliberately leaves one lane of each cell holder empty and keeps the total at 364 cells rather than pushing to the 4.5-5kWh that honeycomb-style holders could fit, for two reasons: the standard holders and nickel strip are cheaper, and a 4.5-5kWh battery in that toolbox would simply be too heavy to move. That is a real design decision every builder faces — energy density versus practicality — and it is the same reason portable and stationary packs have different form factors.
The packs are welded with 0.15mm pure nickel strip in a 2P width, using a Sunco 709A spot welder (the 240V version, run inside the house on a dedicated 20A circuit) set to 7.5 current and 12 pulse. Each cell gets four welds through the slot in the nickel — the slot forces current to pass down through the cell and back up the other side rather than arcing straight across. One important detail: the 2P strip means the host has to bridge every pair of parallel sets with short nickel pieces to make the series connections, which he notes would have been cleaner with 4P-wide nickel had it been available locally.
The safety moment every builder should copy: as soon as series connections start, the whole pack is live. The host covers finished areas with Kapton tape so a dropped screwdriver or a misplaced strip of nickel cannot short across terminals, and he stresses double- and triple-checking every connection before welding. This is non-negotiable practice — a shorted 48V pack with this much energy behind it is dangerous. The same caution applies to any high-voltage battery bank, whether DIY or factory-built.
Each pack is protected by a Daly 14S 48V BMS rated for 60A continuous discharge, in the common-port version — one negative lead comes out to the load, rather than separate charge and discharge negatives. The wiring harness runs from the main negative up through each series point to the main positive; the host solders each balance lead onto the nickel tabs at the top of the pack, being careful not to overheat the plastic cell holders underneath.
A useful detail from the video: one viewer's comment from a previous project convinced the host to keep all balance wires the exact same length, so they all have equal resistance and the BMS reads every cell group evenly. After wiring, the balance leads are taped down with two layers of Kapton, the pack is wrapped in oversized heat-shrink (cut for a 15S pack, shrunk onto a 14S one), and an insulating plastic sheet goes between the two packs before they are strapped together.
Power delivery inside the toolbox: a Reliable Electric 48V 1,500W pure sine wave inverter feeds four recessed 12A/125V outlets (each with two USB ports), the main on/off is a Blue Sea Systems 50A circuit breaker, and a DC-DC buck converter steps 48V down to a 12V rail. The 12V side runs two fused cigarette-lighter sockets (10A fuses) plus a temperature-controlled fan system: a small PCB with a remote sensor switches two 80mm computer case fans to ventilate the box and keep it cool.
Two engineering details worth copying: the inverter is mounted on an insulating plastic sheet above the battery with airflow on all sides, and the breaker question was verified with the manufacturer — the 50A breaker is UL-rated to 32V DC but confirmed good to 65V DC, which covers the 58V fully-charged pack. Always verify your protection components against your actual system voltage. This is exactly the kind of spec check we run on every commercial battery system before it ships.
For charging, the host uses a generic MPT-7210A controller — a DC boost converter rather than a true MPPT buck controller. Its key advantage for a portable build: it accepts an input lower than battery voltage, so you can plug in a 12V or 24V panel, or even a 130W laptop charger as an AC-to-DC source, and it boosts up to charge the 48V bank. The trade-off is that with a boost-type controller the panel's operating voltage must stay at or below battery voltage, so panels must be wired in parallel, not series. In the video's test, a single 270W SunEdison panel delivers about 180-185W on a partly cloudy day — enough to recharge a meaningful chunk of the 3.5kWh bank per day.
The video also highlights an often-missed feature: this class of controller does not require a solar source at all. Any DC source within range works, which is how the host charges from a laptop brick when there is no sun. Matching the controller to the battery chemistry and voltage is the same conversation we have with every customer who asks about solar-ready battery packs — the charge source has to fit the bank, not the other way around.
The finished generator is tested against real tools. A 6.5A Black & Decker hammer drill runs easily; a 7A Craftsman shop vac pulls about 19A while running — and, more impressively, 70A at the moment it starts. That inrush spike is the classic reason battery banks need headroom: motors and compressors briefly draw many times their running current. The 1,500W inverter and 48V pack handle it without breaking a sweat, which shows why matching the inverter surge rating to your tools matters more than the continuous wattage alone.
The video ends with two safety points that apply to any DIY battery enclosure. First, if you do not cut ventilation holes and run fans, you must remove the rubber gasket along the lid — otherwise a sealed watertight toolbox becomes a pressure vessel if a cell ever vents, which can turn a fault into an explosion risk. Second, active cooling is mandatory if the generator will sit in the sun: the video shows the fans spinning up from internal heat during outdoor solar charging. Ventilation, insulation and fusing are what separate a safe pack from a liability, and they are the same three things we bake into every factory-built battery.
For someone comfortable with a soldering iron, a spot welder and careful assembly, this build delivers 3.5kWh of 48V storage with a 1,500W pure sine inverter for about $641 — genuinely hard to match from a retail power station of the same capacity. For everyone else, the video is still a masterclass in what goes into a battery system: cell testing, pack math, BMS selection and thermal management. Those are the same fundamentals that determine the quality of any battery you buy, which is why understanding them helps even if you never build your own. If you need the storage without the DIY time, a drop-in or custom LiFePO4 pack engineered to your voltage and capacity is what we do every day.
And if you're not hunting for a consumer brand but need the battery itself — a custom LiFePO4 pack at a specific voltage and capacity, an OEM or ODM project, or wholesale supply for your own power station line — that's what we do. Dajiu Energy is a Chinese manufacturer building LiFePO4 batteries for RVs, homes, golf carts and portable applications since 2017. Our product range covers LiFePO4 drop-in replacement batteries, EV and golf cart batteries, wall-mounted home storage, portable power stations, commercial energy storage and battery swap cabinets — explore the full lineup in our product center, or browse more buying guides. Tell us your voltage, capacity and size, and we'll engineer the pack to hit it.
These are the most useful of the video's top comments, with our practical answers. The original creator did not reply to these threads, so the answers below are ours — written the way we'd answer a customer on the shop floor. Thanks-only comments were left out on purpose; these are the ones with actual field experience in them.
Source video: Lithium Solar — "Building a 3.5kWh DIY Solar Generator for $650 — Start to Finish" (youtube.com/watch?v=PVnQ87Fvsk4). Comment excerpts are quoted verbatim from the video's top comments; the original creator did not reply to these threads, so all answers are Dajiu Energy's own. Build figures are transcribed from the video; prices are as stated at filming time and change with the market.
Tell us your voltage, capacity, size and application. We will get back to you with a professional quote and technical spec as soon as possible.