The Battery Tester Voltage Window: Bought for 1500 V, Caught Out at 36 V

Battery test · 5 min read · 29 cited facts

Battery test systems are specified against maximum pack voltage, and the specification usually stops there. It is the easy half: 800 V architectures push ratings toward 1500 V and the catalogue obliges — the Procyon PTS 2100-20 spans 40V to 1500V. The half that catches programmes out is the bottom. Deep-discharge behaviour, cell-level work and second-life grading all live near the floor of the window, and a supply with a high minimum output voltage cannot go where they happen.

One programme, two windows

An automotive programme rarely lives at one voltage. The organisation qualifying an HV_3 pack — unlimited operating band 520–750 V, circuit limit 800 V pk under the LV123 voltage classes — is often also responsible for 48 V board-net modules, whose LV 148 unrestricted operating band is 36–52 V (VDA 320). These are not scaled versions of one requirement; they are different electrical worlds, each with transient clauses of its own.

Now place the instruments against that span. The CTS4000 cycler runs 0V to 110V — it reaches the whole 48 V world, floor included. The Procyon runs 40V to 1500V — it owns the pack world, but its 40 V floor sits above the bottom of the LV 148 band at 36 V. The machine bought for the 1500 V headline cannot hold a 48 V module at the bottom of the module's own operating range, and no amount of power rating fixes a minimum output voltage.

Voltage range is bought with current

The trade is visible inside a single product family. The ELR 9000 regenerative loads put 10.5 kW in a 3U chassis at either end of the range: the 80 V model, the ELR 9080-510, sinks 510 A, while the 1500 V model, the ELR 91500-30, sinks 30 A. Same chassis, same power — the wider voltage window is paid for in current. This is why one cycler rarely covers both 48 V mild-hybrid modules and 800 V packs well: a machine wide enough for the pack at the top of the window is starved of current where the module needs it at the bottom. Architecture offers partial relief, not exemption — the B2C+ converter reaches 800V and 555A in one chassis and gets to 1500V by placing 2 units in series, and an auto-ranging stage like the Vega racks' holds full power down to 1/3 of the voltage range — but the trade never disappears. It moves.

The floor is not 0 V even when the datasheet says 0 V

Ratings that start at zero still meet physics on the way down. The AN236 load family is rated from 0-150V in its 150 V class, yet the manual sets a minimum operating voltage of 1.8V at rated current (1.8V@200A): drawing full current requires voltage across the instrument, so the achievable floor depends on how hard you are pulling. For deep-discharge characterisation and for second-life grading — incoming cells of unknown history, sorted by capacity and resistance near the bottom of the window — this is the specification line that decides whether the test can be run at all, and it almost never appears in a headline.

Specify the bottom in writing

• State the minimum voltage at which full test current must still flow — not only the maximum voltage
• Ask what current is available at the low end of the range, not just at the rating point
• For cell-level and deep-discharge work, put the CTS4000's class of machine (0V to 110V, 0A to 3000A, ± 0.05% FS) on the requisition alongside the pack tester, not instead of it
• Treat 48 V module coverage as its own requirement — the LV 148 floor of 36 V is a hard number that a supply with a 40 V minimum cannot meet

A programme that spans 48 V modules and 800 V packs is, in practice, two instruments — a low-window, high-current machine and a wide-window pack machine — and pretending one specification line covers both is how the bottom of the window gets discovered during commissioning. The SM series supplies show how far a single DC chassis family stretches — up to 1500V and 450A per chassis, parallelable beyond 1MW — but stretching the top is exactly what they are doing. The bottom stays a separate purchase.