Technical Edition v2
v2.0
BatteryLogger — Load-Test Protocol
We don't guess. We stress-test your battery for 10 minutes and let physics answer.
The only honest way to diagnose a lithium-ion battery is under real load. Resting voltage readings look fine on nearly every failing battery — until it is under load.
The fundamental problem
Resting voltage tells you almost nothing
A lithium-ion battery at rest will show close to its rated voltage even when its internal resistance has degraded significantly. It reads 36.1V on a multimeter — looks fine. Under 12A load for 90 seconds, it collapses to 28V. That battery is dead. The multimeter never told you.
"We connected the battery and it showed full voltage." That is not a test. That is a resting observation. A failing battery can hold nominal voltage at rest for months while delivering almost no real-world range under load.
This is why every professional battery test standard — IEC 62660, automotive BMS validation, industrial UPS certification — tests under load. BatteryLogger applies the same principle at workshop scale, at a fraction of the cost.
The engineering principle
Core diagnostic law
Voltage sag under load reveals internal resistance. Internal resistance reveals true battery health.
When a battery delivers current, voltage drops according to: V_drop = I × R_internal. A healthy battery with low R_internal holds its voltage steady under load. A degraded battery with high R_internal shows immediate voltage sag — measurable, repeatable, and conclusive. This is the real test. Everything else is theatre.
The test protocol
10 minutes. Under real load. Every time.
Load applied
10–15A
Electronic dummy load simulating a scooter climbing a gradient — the most demanding real-world scenario for the battery pack.
Golden test window
10 min
Long enough to catch delayed voltage sag (8–12 min). Short enough to run 6 battery tests per workshop hour without disruption.
Reading interval
30s
Voltage, current, and temperature sampled every 30 seconds. 20 timestamped data points per test session, automatically stored.
Data points per test
20
A complete time-series trace from first load application through voltage stabilisation or sag. No gaps. No estimates.
What happens during 10 minutes
0:002:005:008:0010:00
0:00 – 2:00
Immediate sag zone. Severely degraded batteries reveal themselves here. Voltage drops below the 0.75 threshold within the first 4 readings. The report is ready in under 2 minutes.
2:00 – 8:00
Progressive decline zone. Moderately degraded batteries show a gradual voltage curve. Temperature starts climbing above the normal operating range. The slope alone confirms diagnosis.
8:00 – 10:00
Delayed sag zone. Batteries with cell imbalance appear healthy early but show failure in the final phase. Caught only with a 10-minute window — the reason we do not stop at 5.
Healthy vs. failing — what the data shows
The numbers do not lie under load.
| Time |
Healthy battery (36V nominal) |
Failing battery (36V nominal) |
| 0:30 |
35.8V — holds steady, ratio 0.99, temp 22°C |
33.1V — immediate sag, ratio 0.92, temp rising |
| 1:30 |
35.6V — stable under full load, temp 23°C |
30.4V — accelerating drop, temp 31°C, Status: Warning |
| 3:00 |
35.4V — holding curve, Status: OK |
27.2V — below 0.75 threshold, Status: Replace. Report generated. |
| 6:00 |
35.1V — normal decline curve, temp 25°C |
Test concluded at T+3:00. 6 readings logged. Verdict confirmed. |
| 10:00 |
34.9V — 20 readings logged. Final report: OK. Safe for service. |
Complete collapse documented. Customer report printed and signed. |
Before and after
How workshops ran before. How they run now.
Without BatteryLogger
✕Technician reads 36V at rest and says "looks fine"
✕Customer returns: "it still doesn't last on hills"
✕No data. No proof. Dispute begins.
✕Workshop absorbs cost of repeat labour or replacement
✕Diagnosis based on experience, not measurement
With BatteryLogger
✓Battery connected to load station. 10-minute test starts.
✓Voltage sag detected at T+1:30. Dashboard flags: Replace.
✓Report: 20 timestamped readings, S/N, trace, verdict.
✓Customer sees the data. No argument possible.
✓Diagnosis is objective, repeatable, and signed.
System overview
Three components. Zero complexity.
1
Electronic load station
Draws a constant 10–15A from the battery, simulating real scooter load on a climb. The ESP32 sensor unit connects to the battery terminals simultaneously and reads voltage, current, and temperature every 30 seconds.
2
BatteryLogger dashboard
Every reading appears in real time. Status classification — OK, Warning, or Replace — is computed from the voltage ratio under load, not at rest. If voltage falls below 0.75 of rated voltage, the system flags it immediately. If it falls below 0.90, it warns. The technician sees this from across the room.
3
Timestamped service report
Print the full report: battery identity, serial number, brand, 20 readings with exact timestamps, temperature trace, voltage curve, and a clear verdict with the disclaimer that this is a workshop service tool. One click. One page. Legally defensible documentation of what happened under real operating conditions.
By the numbers
10 min
complete diagnosis
Six batteries diagnosed per hour. Each with a printed report. Each dispute-proof. At 100 € total hardware cost, you recover the investment the first time you show a customer their timestamped voltage trace and avoid an argument.
Cost comparison
Professional load-test methodology. Workshop price.
The same principle used in industrial battery validation labs — applied at a fraction of the cost.
Professional lab equipment
800 – 2000
€ — HIOKI BT3554, Cadex C8000, Arbin BT
BatteryLogger
100
€ hardware — no subscription, self-hosted, your data
See a live test run at bat.alamanjo.eu
The demo shows real battery data from actual load tests.
No sign-up required. Works on any device. Print a sample report in 30 seconds.