I've been sourcing power electronics for system integration projects for 9 years. I've made and documented 7 significant mistakes, totaling roughly $48k in wasted budget. Now I maintain our team's pre-check list.
When someone asks me 'What's the best power supply for a home battery storage system?' I can't answer without asking three questions. The right architecture depends on your scenario. Here are the three I see most often.
What was best practice in 2020 may not apply in 2026. The fundamentals haven't changed, but the execution has transformed.
First, the scenario split
Almost every home battery storage or small DC microgrid project falls into one of these three buckets:
- Scenario A: 5-15 kWh, 3-6 kW, retrofit AC-coupled. Cost-sensitive, indoor, limited space.
- Scenario B: 15-50 kWh, 8-20 kW, DC-coupled hybrid. Whole-home backup or small commercial. Efficiency and cycling matter.
- Scenario C: 50+ kWh, 20+ kW, modular or outdoor DC microgrid. High density, harsh environment, or future expansion.
If you pick the wrong bucket, you'll either overspend or underspec. I've done both.
Scenario A: The AC-coupled retrofit (5-15 kWh)
For most existing homes, you're not rebuilding the electrical panel. You're adding a battery next to an existing solar inverter or as a standalone backup. Here, a standard ac dc psu or integrated AC-coupled inverter is usually the right call. You want simple, certified, and serviceable.
In my first year (2017), I made the classic oversimplification error: I assumed a higher-efficiency DC-coupled design would always pay back faster. It didn't. On a 10 kWh retrofit, the extra conversion stage and custom wiring ate the savings. The home battery storage cost was 18% higher than the AC-coupled quote, and the payback stretched by two years.
What I tell people now: if you're retrofitting and your daily cycle is shallow (under 8 kWh), don't over-engineer. A quality AC-coupled PSU with a 10-year warranty is often the better economic choice. It's tempting to chase the latest topology. But the 'always go DC-coupled' advice ignores install labor, permitting, and existing equipment.
Scenario B: The DC-coupled hybrid (15-50 kWh)
Once you're designing a whole-home backup or a small commercial system, DC coupling starts to make sense. You'll be cycling deeper and running more efficiency-sensitive loads. This is where soft-switching llc and a half bridge bidirectional dc dc converter become worth the complexity.
Why? A soft-switching LLC resonant converter reduces switching losses and EMI. In a bidirectional design, it can move energy efficiently in both directions: solar to battery, battery to loads. The half-bridge bidirectional DC-DC stage handles the voltage matching between battery bank and DC bus. In 2022, we switched a 20 kW commercial backup design from a hard-switched full-bridge to a soft-switching LLC half-bridge. Idle losses dropped by 22%. The client didn't care about the topology name, but they cared about the $340/year saved in standby consumption.
That said, this scenario has a trap: not every integrator needs bidirectional. If your battery only charges from solar and discharges to a dedicated inverter, a unidirectional LLC plus separate charger can be cheaper and easier to certify. I learned that after a $6,800 redesign because we over-specified bidirectional capability for a system that never used it.
Scenario C: High-density or outdoor DC microgrid (50+ kWh)
This is where liquid cooled power supply units and serious dc microgrid supplier partnerships come in. When you're stacking 50+ kWh in a garage, basement, or outdoor cabinet, heat density becomes the limiting factor. Air cooling stops scaling around 15-20 kW in a compact enclosure.
In September 2022, we deployed a 60 kWh modular system in a coastal location. We used air-cooled PSUs to save $4,200 upfront. Big mistake. Humidity, salt air, and dust clogged the heatsinks. After three months, two units derated. The fix cost $9,500 in replacements plus a one-week delay. That's when I learned: in harsh environments, a liquid-cooled power supply isn't a luxury. It's a reliability requirement.
For DC microgrids, you also need to think about source compatibility. A good DC microgrid supplier will provide not just hardware but also protection coordination, grounding, and communication protocols. Don't treat them like a catalog. Treat them like a design partner. Ask for thermal simulations, MTBF data, and field failure rates. If they can't provide that, walk away.
How to tell which scenario you're in
Use these three questions. Don't guess.
- What's your daily cycle depth? Under 8 kWh and retrofit? Scenario A. Over 15 kWh and efficiency-sensitive? Scenario B or C.
- Where does the equipment live? Climate-controlled indoor? Air cooling may be fine. Outdoor, garage, or dusty? Go liquid-cooled or sealed with active thermal management.
- Will you expand? If you might add more batteries or DC loads later, invest in a bidirectional DC-DC stage and a supplier who supports modular expansion.
One more thing: don't let home battery storage cost be your only metric. I've seen $1,200 savings on a power supply turn into $5,000 in service calls. The most frustrating part of sourcing: the same thermal issue recurring despite clear specs. You'd think datasheets would prevent it, but interpretation varies wildly.
Prices as of 2026; verify current rates and certifications. For safety and grid interconnection, check UL 9540A, UL 1741, and local NEC requirements. Regulatory information is for general guidance only. Consult official sources for current requirements.
So glad I started using a pre-check list. Almost approved another air-cooled unit for an outdoor cabinet last year. Would have repeated the same $9,500 mistake. Now we run every project through the three-scenario filter before requesting quotes.
