How to Plan an EMD Sample Test to Avoid Wasted Round Trips

A wasted EMD sample round trip is rarely a material defect. It is an under-specified test. When the battery designation, the current formulation, the test list, the sample size and the pass thresholds are not agreed before the powder ships, the result comes back as an opinion instead of a decision, and the clock resets. Fix seven inputs in writing first: application, battery designation, existing formulation, test items, sample quantity, acceptance criteria and feedback cycle. With those agreed, the first build either passes or shows which variable to change, and one trial replaces three.

Why sample tests go back and forth

Round trips usually trace to one of five gaps, and none of them is a material problem.

• The test list is agreed after the sample ships, so the lab runs its standard panel and the answer misses the device question.
• There is no control arm, so a difference cannot be separated from an unrelated line change.
• The formulation is recalled rather than written, so the lot is pressed to a different recipe than production uses.
• Pass thresholds are set after the data arrives, which lets a result be argued instead of judged.
• No one owns the feedback date, so the trial waits in a queue.

Each gap adds a full cycle: the test duration plus the queue time around it. Closing the gaps costs information, not material, which is why planning beats a larger sample.

Seven inputs to fix before the sample ships

The seven inputs below decide whether the first build is a decision or a discovery.

Two of the seven are most often left out: the control arm and the feedback date. When either is missing the trial still runs, but the result cannot be closed, and the next step becomes a second sample.

Match the test list to the cell family

The seven inputs are the same for every primary cell, but the measurements behind them change with the chemistry.

• Carbon-zinc: weight discharge stability and the blend ratio between EMD and a lower-cost manganese powder.
• Alkaline: weight capacity at the reference drain and shelf life, because low metallic impurities hold self-discharge down.
• Button and coin: weight particle-size distribution and moisture, because the electrode is small and the fill is tight.
• Primary lithium: weight residual moisture after heat treatment, because water reacts with the lithium anode and the electrolyte and generates gas in a sealed can.

Build the list from the device load rather than from a generic battery panel. A cell measured at the wrong drain can pass convincingly and still fail in the product.

How much sample to ask for

Quantity is a calculation, not a guess. Estimate the grams consumed per cell build, multiply by the cells needed per lot, double it for two lots, then add the control arm and a retained reference. Two lots matter because a single lot cannot show batch-to-batch consistency. State the calculation in the request, and confirm that sample and bulk share the same production route: a pilot-batch sample proves less than one drawn from routine output.

A stage-by-stage plan

1. Pre-request: write the seven inputs into one dated document, including the control arm and the thresholds.
2. Request: ask for sample from at least two production lots plus a retained reference, and name the documents that should travel with it.
3. Receipt check: confirm lot identity, moisture at dispatch, packing, and that the sample came from routine output.
4. Build: use production tooling, recipe, fill and crimp parameters, and run the control in the same window.
5. Test: run the agreed protocol and report distributions rather than single cells.
6. Feedback: deliver the data on the agreed date and format, state pass or fail against the fixed thresholds, and if it fails, name the variable to change.
7. Close: convert to order, or record the reason the program stops.

Documents that should travel with the sample

• A TDS, an SDS and a batch COA that states the test methods, not only the numbers.
• Heat-treatment or drying guidance where the cell is non-aqueous, and the neutralisation route where residual sodium matters, as in lithium cells.
• Written confirmation that the sample and the bulk share one production route.
• The quantity logic, MOQ and lead time in writing, so the plan can be costed.

Set thresholds before the data

Numbers fixed in advance turn a result into a decision. Fix a minimum capacity, an allowed spread across the lot, a retention figure after storage, and the number of leaking cells that ends the trial. Report the spread rather than the average of a few hand-built cells, because a small shift in the distribution is what moves yield. One constraint is easy to miss: a pass on the wrong protocol still answers the wrong question, so set the drain, pulse and cut-off from the device, not from habit.

The feedback cycle is the part most often forgotten

Name a test owner on each side, agree a data format and a date. Good feedback states pass or fail against the fixed thresholds and, when it fails, identifies whether the next change is a formulation tweak, a drying step or the material. The trade-off is real: a very short deadline adds cost and error risk, while a very long one stalls the program. Set the cycle from the slowest test on the list, since a storage check takes longer than a fresh-capacity check.

Close the first build as a decision

Send the seven inputs through the Qingchong contact page: application, battery designation, current formulation, test list, quantity, thresholds and feedback cycle. Ask for sample from two production lots with a retained reference, plus the EMD product TDS, SDS and batch COA, so the trial is judged against numbers you set before the powder shipped.

FAQ

How much EMD sample do we need?

Enough to build a control and two production lots under your recipe, plus a retained reference from each lot. Calculate grams per build times cells per lot, double it for two lots, and add the control.

Can we skip the control arm to save sample?

No. Without a control on your current material, a difference in the result cannot be separated from a line change, so the trial produces no decision. The control is part of the quantity calculation, not an optional extra, and it usually costs less than the second round trip it prevents.

Should acceptance thresholds be set before or after the test?

Before. Thresholds set after the data arrives let the result be argued rather than judged. Fix minimum capacity, allowed spread, retention after storage and the leaking-cell limit in advance, and report pass or fail against those numbers.

What should happen if the first trial fails?

The plan should already name the variable to change. With fixed thresholds, a failed trial shows whether the problem sits in the formulation, the drying or heat-treatment step, or the material itself, and whether a second round is justified. That is a result, not a wasted sample.