New to solar and need a sanity check for off-grid system using 48V Yixiang DIYs.

Been down the DIY battery rabbit hole myself when I was kitting out the van, so happy to share what I've learned.

48V Yixiang cells can be decent value but the quality control is very inconsistent — some folks get great cells, others get lemons straight out of the box. First thing I'd do before anything else is capacity test every single cell individually. Don't skip this step, seriously.

A few things worth sanity checking on your system:

• BMS sizing — are you giving yourself enough headroom on continuous discharge? A lot of people underspec this
• Cable sizing — 48V systems still pull serious current under load, don't go cheap on your busbars and interconnects
• Victron kit — if you're not already planning around a Victron ecosystem (MPPT, MultiPlus, Cerbo GX etc.) I'd strongly recommend it for monitoring and integration. Pricey upfront but worth every penny for an off-grid setup

What's your intended use case? Van, cabin, static installation? Makes a difference to how you'd approach the design — a van setup has very different constraints to something sat on a concrete base.

Also worth checking out the CATL/EVE alternatives if you haven't committed to Yixiang yet. Fogstar sell tested cells with decent support if you'd rather not gamble on aliexpress lottery.

What inverter are you pairing with the batteries? That'll probably be where your next set of questions come from 😄

@CoveMick makes a solid point on QC — I'd add that the inconsistency often shows up specifically in internal resistance spread across a batch, which absolutely hammers your pack's usable capacity even if every cell tests at nominal voltage out of the box.

When I built the 48V bank for the shepherd's hut, I spent a weekend doing full charge/discharge cycles on individual cells with a decent capacity tester before even thinking about grouping them. The spread on my Yixiang batch was surprisingly wide — I ended up setting aside roughly 15% of cells that were outliers.

Worth grabbing a YR1035+ or similar IR tester. Match cells by both capacity and internal resistance before assembly, not just one or the other. Your BMS will thank you later, and so will your Victron MPPT when it's not fighting an unbalanced pack.

@SprinterProject @CoveMick — when you were testing your Yixiang cells initially, were you running a proper capacity test on each one before building the pack?

I'm looking at something similar for my motorhome build and I'm wondering whether a cell-level capacity test is worth the hassle before assembly, or if most people just trust the BMS to sort out weak cells over time.

Also — what kit are you using to actually test them? I've seen people mention the Junsi iCharger but wondering if there's a more budget-friendly option that still gives reliable results for 48V cells. Presumably you'd need to test them at the cell level (3.2V nominal) rather than pack level?

Trying to figure out how much time to budget for this before I commit to purchasing.

@Moorey44 capacity testing is non-negotiable with budget Chinese cells — I'd strongly recommend a proper discharge cycle before you even think about assembling your pack.

I use a ISDT battery meter paired with a decent load tester on every new cell batch. For 48V builds specifically, you want each cell within ±2mAh ideally before grouping them — anything wider and your BMS will be fighting imbalance from day one.

Worth noting: top-balance before first use, not bottom-balance. Gets all cells singing from the same hymn sheet.

If you're running a Victron system (which I'd highly recommend for 48V off-grid), the MPPT charge data will also expose weak cells over time — sudden voltage sag on one group is a dead giveaway.

Don't rush the commissioning phase. Painful I know, but those early hours of testing save you from discovering problems when you're miles from civilisation.

Capacity testing is only half the story. I'd also run them through a few full cycles before you commit to building the pack — cells that look fine on first discharge can drift badly by cycle 3 or 4. Seen it on my boat build with a different brand, cost me a weekend of rebusbarring.

Also worth noting: whatever BMS you're pairing these with, make sure it can log individual cell voltages over time. A Victron Cerbo setup makes this fairly straightforward if you're already in that ecosystem. Spotting a rogue cell early is far less painful than diagnosing one after the pack's been in a confined space for six months.