Why do power banks need to be tested in both directions?

A power bank charges in and discharges out, and the two paths fail for different reasons. The discharge side governs output behaviour — voltage under load, protocol handshake, ripple, over-current protection. The charge side governs the input stage: whether the pack accepts the input power it advertises and whether the charging curve behaves. A pack with a healthy output and a weak input stage passes an output-only bench and still underdelivers in the field.

What does a two-instrument bench cost in practice?

Floor space for a second instrument, separate fixtures and cabling for charge versus discharge testing, re-wiring steps where reversed connections damage units, two result sets that have to be reconciled by serial number, and two procedures for the operator to keep straight. None of it appears as a line item, but it shows up in takt time and in skipped steps.

What do the GTI6310H and GTI6320H cover?

Both carry a load side rated 0–240W and a source side adjustable across 0–20/30/50V, so output testing and charge/discharge testing run from one bench, one fixture and one result sheet. Accuracy is quoted at ±0.1% for voltage and current, typical, and a factory calibration report is available per unit on request.

Should I choose the GTI6310H or the GTI6320H?

The models share the load and source ranges. The GTI6320H adds over-current protection (OCP) testing, so pick it when the packs you build ship with protection circuits you are expected to verify before shipment. If OCP is not on your test list and protocol coverage plus load power are the deciding factors, the GTI6310H covers the same ranges for less.

What is a source side and a load side on a tester?

The load side sinks current — it is the electronic load that the product's output is tested against. The source side supplies current — it is the programmable DC supply that charges the product or drives it through a discharge cycle. On the GTI6310H and GTI6320H both sides sit in one unit, so output testing and charge/discharge testing run from one bench.

Can one tester charge and discharge a power bank?

Yes. Because the same unit carries a source side (0–20/30/50V) for supplying current and a load side (0–240W) for sinking it, one bench runs the charge test and the discharge test on the same fixture, with one result sheet against one serial number.

Does testing both directions slow the line down?

It is faster than the two-instrument setup it replaces. You avoid re-cabling between charge and discharge fixtures, and you avoid the second data set that has to be reconciled by serial number. The saving shows up in takt time and in skipped steps, not in the number of measurements.

How do I size test capacity for a power-bank line?

Work from takt time rather than from daily output: units per shift divided by available seconds gives seconds per unit, and that decides whether a single bench or a multi-station system is the right shape. The arithmetic is covered in the post 'How to size charger test capacity before you buy'.

Line Setup

Power banks fail in two directions — most benches only test one

5 min read · Xie Yang · 2026-09-13 · For power-bank and storage-power production lines

Quick answer. A power bank takes charge in and gives charge out. Most QC benches verify the output side and assume the input side is fine — which is why a pack can pass your line and still deliver less than the label says. Running both directions usually means two instruments and two setups: a supply for the charge test, an electronic load for the output test. A tester with a source side and a load side does both from one bench, one fixture, one result sheet.

A power bank is the only product on your line that has to work in both directions. While it charges, it is a load. While it feeds a phone, it is a supply. Same enclosure, same cell, same board — two circuits behaving under two completely different sets of conditions, and they do not fail for the same reasons.

Testing one direction and assuming the other is a habit. It is not a test plan.

Why do power banks fail in two directions?

On the discharge side you are watching output behaviour: does the pack hold voltage under load, does the protocol handshake complete, how much ripple rides on the output, does over-current protection trip where it is supposed to.

On the charge side you are watching the input stage: does the pack accept the input power it advertises, does the charging curve behave, does the cell reach the right end point before the charger cuts off. That input stage is where the quiet failures live. A weak input path does not show up as a dead unit — it shows up as a pack that takes four hours to fill when the customer expected two, or one that reports the wrong capacity after a full charge.

Nobody writes that on a complaint form. They write "capacity is not as advertised", which is why the first thing most teams do is re-test the discharge side — the side they already tested.

What a two-box bench actually costs you

Running both directions with separate instruments is the traditional setup, and on paper it is cheaper than a combined unit. In practice it costs things that never appear as a line item:

Floor space. Two instruments per station, or two stations and a handoff between them.
Setup time and fixtures. Charge test and discharge test need different cabling and different fixtures. Re-wiring between tests is where reversed connections damage units — and it is the step that gets skipped when the line is behind.
Two data sets. Someone has to reconcile the charge result and the discharge result by serial number. If they are recorded in two places, they sometimes are not reconciled at all.
Operator load. Two instruments means two procedures, which means two chances to run the wrong one on the right unit.

GTI6310H vs GTI6320H at a glance

Model	Load side (sink)	Source side	OCP test	Pick it when
GTI6310H	0–240W	0–20/30/50V	No	Standard power banks; protocol coverage and load power decide
GTI6320H	0–60V / 0–10A / 0–240W	0–20/30/50V	Yes	Packs ship with protection circuits you must verify before shipment

Both models: 20+ fast-charge protocols (QC2.0/3.0, PD3.0/3.1, PPS, PE, Huawei/Samsung/OPPO/VIVO), ±0.1% typical V/I accuracy, factory calibration report on request. Protocol requirements follow the USB-IF USB PD specification — usb.org/usb-charger-pd.

None of that shows up as takt time on a quote sheet. All of it shows up on the line.

What one bench changes

The GTI6310H and GTI6320H carry both sides in one unit. There is a load side rated 0–240W that sinks current for the output test, and a source side adjustable across 0–20 / 30 / 50V that supplies current for the charge and discharge test. Same fixture, same operator, same result sheet, one serial number against both directions.

The source range is the part people under-read. 0–20V covers the common single-cell packs. 0–30V and 0–50V cover multi-cell packs and the higher-voltage storage products. If your line switches between 1S power banks and multi-spec storage, that span is the difference between one bench and two — and it is usually cheaper than the second bench.

Accuracy on both models is quoted at ±0.1% for voltage and current, typical. If you need the calibration paperwork for an incoming-inspection team, the factory calibration report is available per unit on request.

6310H or 6320H — the difference is one test

The two models share the load and source ranges. The practical difference is that the GTI6320H adds over-current protection testing, and the GTI6310H does not.

Packs that ship with protection circuits you are expected to verify → GTI6320H.
Standard power banks where protocol coverage and load power are what matter, and OCP is not on your test list → GTI6310H, same load and source range, lower cost.

If you are sizing for a line rather than a bench, the same family runs to 4×4 multi-station parallel testing, which is a different decision — covered in 8 test ports or 1.

When two instruments still make sense

Splitting charge and discharge across two boxes is not automatically wrong. It holds up when your charge-side and discharge-side volumes are wildly different, when you already own a calibrated supply you trust, or when a single-direction station is dedicated to one product for months. The case for one combined bench is strongest where operators rotate between stations, where floor space is tight, and where every unit has to be tested both ways in a single pass — which is most power-bank lines once volume picks up.

The honest limit

One bench does not make the test better. It makes it happen in one place, on one unit, with one set of numbers. If your charge-side test procedure is weak, combining the instruments will just produce weak results faster. Fix the procedure first — then decide whether the second box is worth its space and its cabling.

Where this comes from The load rating (0–240W), the source range (0–20/30/50V), the ±0.1% typical accuracy, the protocol coverage and the presence of OCP on the GTI6320H are taken from the GTI6310H and GTI6320H product pages on this site. The failure-mode descriptions — input stage weakness presenting as a capacity complaint — are standard behaviour for packs with separate charge and discharge paths, not measurements of a specific batch. Your own returns log is the dataset that matters.