Battery, Charging & Starting Diagnosis

90 min read Training Guide

OCV vs SOH, conductance vs load, alternator output, parasitic draw, starter amp draw, voltage drop. The most common under-hood diagnosis in a shop.

Table of contents

Battery, Charging & Starting Diagnosis

"My car won't start" is the single most common ticket in a service department. Half the time it is the battery. A third of the time it is the charging system. The rest is split between starters, cables, parasitic draws, and one-in-a-hundred oddities like a fried BCM. If you run the sequence in this guide every time, you will fix the right thing and not replace a starter on a car that needed a 150 dollar battery.

Battery Chemistries


| Type       | Where it shows up                           | Notes                         |
|------------|---------------------------------------------|-------------------------------|
| Flooded    | Most older vehicles, many economy models    | Cheap, needs maintenance check|
| EFB        | Some light start-stop applications          | Enhanced flooded, tougher     |
| AGM        | Start-stop, European, high-accessory loads  | Required on most start-stop   |
| Lithium    | Specialty, performance, some motorsport     | Not interchangeable with lead |

Never install a flooded battery in a start-stop car that was factory AGM. The car will not register correctly, charging will be wrong, and the battery will die in months. Always match chemistry to the factory spec, or step up, never down.

Key Ratings

  • CCA (Cold Cranking Amps at 0 F): how much current the battery can deliver for 30 seconds at 0 F while holding 7.2 V. Typical 600-800 on passenger cars, 700-1,000 on trucks.
  • RC (Reserve Capacity): minutes the battery can deliver 25 A before dropping to 10.5 V. Important for accessory draw when the engine is off.
  • Ah (Amp-hours): more common on AGM and European batteries.
  • Group size: physical case dimensions (35, 48, 65, 94R, H6, H7, etc.).

Always match or exceed OEM CCA and match group size. An undersized battery will crank but the charging system will over-work and the next one will die sooner too.

State of Charge vs State of Health

These are different. You can have a fully charged battery with terrible health, and a depleted battery with perfect health.

State of Charge (SOC)

Measured by Open Circuit Voltage (OCV) after at least 2 hours sitting with no loads or charging.


| OCV    | Approx SOC |
|--------|------------|
| 12.65 V| 100%       |
| 12.45 V| 75%        |
| 12.24 V| 50%        |
| 12.06 V| 25%        |
| < 12 V | deeply discharged |

A battery sitting at 12.1 V for a week will sulfate. A battery sitting below 11.5 V is likely damaged.

State of Health (SOH)

Requires a real battery tester: a conductance tester (Midtronics, Snap-on EECS, Bosch) or a carbon-pile load tester. The tester reports measured CCA vs rated CCA.


| Measured CCA vs rated | Action                         |
|-----------------------|--------------------------------|
| 90-100%               | Good                           |
| 75-90%                | OK, monitor                    |
| 60-75%                | Replace on the next service    |
| < 60%                 | Replace now                    |

Most shops replace below 75% measured. A 700 CCA rated battery that tests at 320 CCA is failing regardless of what OCV says.

Load Test vs Conductance Test

Old-school carbon pile load test:

  1. Battery at full charge or near it.
  2. Apply a load equal to half the CCA rating for 15 seconds.
  3. Watch voltage. At 70 F, voltage should stay above 9.6 V for the 15 seconds.

Valid, still used, still teaches the concept. The downsides: takes time, needs a resistor bank, and struggles on deeply discharged batteries.

Conductance tester:

  1. Clamp to the posts.
  2. Enter battery rated CCA and chemistry.
  3. Press test. Get measured CCA, voltage, internal resistance, and a pass/fail in about 10 seconds.

Conductance testers work on partially discharged batteries and give a cleaner answer. Either approach is defensible. Most shops run conductance testers because they are fast and documented with a printout.

Alternator Output Spec

Most passenger vehicles run 13.5 to 14.5 V at 2,000 rpm with accessories off. 14.2 V is a comfortable target. Below 13.0 V is undercharge. Above 15.0 V is overcharge and will cook a battery. With headlights, HVAC blower high, rear defroster, and radio on, voltage should hold 13.0 V or above at 2,000 rpm.

AC Ripple (Diode Test)

Alternators rectify AC from the stator through a diode pack. A failed diode leaks AC. With the DMM set to AC volts at the battery terminals, engine at 2,000 rpm with accessories on:

  • Under 50 mV AC: healthy
  • 50-100 mV AC: marginal, watch for other symptoms
  • Over 100 mV AC: failing diode pack, replace alternator

AC ripple will kill a new battery in months even when DC charge voltage reads "correct." Always check ripple when a battery failure pattern looks premature.

Parasitic Draw

With the key off, doors closed, trunk closed, hood switch taped if needed (or use the jumper), the vehicle needs about 30 to 40 minutes to fully "sleep" (all modules shut their receivers). Only then measure draw.

Measurement options:

  • DMM in 10 A DC, in series with the negative battery cable
  • Low-current amp clamp (0.1 mA or 1 mA resolution, around the negative cable)

Acceptable draw:


| Vehicle vintage                  | Acceptable draw |
|----------------------------------|-----------------|
| 1990s, minimal electronics       | under 50 mA     |
| 2000s-early 2010s                | under 50 mA     |
| Late model with telematics       | under 30 mA     |
| Heavy-telematics luxury, EV      | under 20 mA     |

A 100+ mA draw will run down a 70 Ah battery in under a week of no driving. To isolate, pull fuses one at a time, wait a minute between pulls so modules resettle, and watch for the drop. When the number drops by a noticeable amount, the circuit on that fuse is the culprit. From there you bisect the circuit.

Common draw culprits: aftermarket radios, aftermarket alarms, door courtesy switches stuck closed, trunk lights stuck on, glove-box lights, BCMs that fail to sleep.

Starter Amp Draw

A clamp meter on the positive cable during a crank gives you the truth.


| Engine type                | Peak inrush | Steady crank |
|----------------------------|-------------|--------------|
| 4-cyl gas                  | 150-250 A   | 100-150 A    |
| V6 gas                     | 200-300 A   | 150-200 A    |
| V8 gas                     | 250-400 A   | 200-300 A    |
| V6/V8 diesel               | 400-700 A   | 250-400 A    |

Low draw with slow crank = weak battery or a poor cable connection. The starter is not getting the current it needs.

High draw with slow crank = bad starter, worn brushes, shorted armature, or a mechanical drag (seized accessory, hydrolocked cylinder, timing-related lock).

Good battery, normal draw, normal speed, still no start = fuel or ignition problem, not a starting-system problem.

Voltage Drop Tests

Voltage drop is the proper cable test. Visual "looks clean" lies. The test:

  1. Big positive cable, battery positive post to starter B+ stud: under 0.5 V during crank.
  2. Big negative, battery negative post to engine block: under 0.2 V during crank.
  3. Battery negative post to chassis: under 0.1 V during crank.
  4. Battery positive clamp to battery positive post (across the clamp connection itself): under 0.1 V during crank.

High drop on any leg points directly at a corroded or loose connection, a broken strand inside a cable, or a failing cable-end crimp. A corroded negative cable at the engine block is the single most common "mystery no-start" that passes a battery test.

Start-Stop and Smart Charging

Modern cars monitor battery state with a current sensor on the negative cable (IBS: Intelligent Battery Sensor). The BCM or gateway calculates state of charge, state of health, and commands alternator output accordingly.

Do not just bolt on a new battery and walk away on these cars:

  • BMW, Mercedes, Audi, Porsche, Volvo, Land Rover: require battery registration (code in the new battery to the BCM) via a factory or capable aftermarket tool. Without it, the car keeps charging to the old battery's tired profile and the new battery either never fully charges or cooks.
  • Many Ford and GM late models: require a BMS reset or initialization.
  • Most EV 12 V systems: same rules plus their own quirks.

Scan tools that do this: Autel, Launch, Snap-on, OEM factory tools, FCOM, Foxwell, ISTA. Confirm the tool supports the specific brand you are working on before you sell the job.

Worked Example

Customer: "Intermittent no-start, clicks."

  1. OCV: 12.1 V (low, below 50% SOC). Charge the battery.
  2. After a full charge, rest 2 hours. OCV climbs to 12.45 V only. Suspicious.
  3. Conductance test: 320 measured CCA on a rated 700. Fail.
  4. Inspect cables. Negative battery cable post shows green corrosion. Clean both posts with a terminal brush, apply no-ox grease.
  5. Replace battery with OEM-spec or better, register/relearn if required.
  6. Verify charge voltage at idle: 14.2 V. At 2,000 rpm with headlights and HVAC on: 13.9 V. Good.
  7. Verify AC ripple: 25 mV. Good.
  8. Clear codes. Start-stop fault clears on the second key cycle.
  9. Verify parasitic draw after 40 min sleep: 22 mA. Good.

Document everything on the RO: OCV before, conductance result, charge voltage at idle and loaded, AC ripple, parasitic draw, cables cleaned and torqued, battery replacement date, registration/relearn performed.

What to Document

Every battery or charging job finishes with these notes on the RO, in this order:

  • OCV before test
  • Conductance result (measured CCA vs rated CCA)
  • Charge voltage at idle and at 2,000 rpm loaded
  • AC ripple
  • Parasitic draw after sleep
  • Cable/terminal condition and any cleaning or torque performed
  • Battery part number installed, date code, chemistry
  • Registration or BMS relearn performed on cars that require it
  • Customer-facing recommendation (next check interval, etc.)

That record is your defense when a customer comes back in 11 months claiming "you replaced it wrong." It is also how you learn what a healthy reading looks like on ten different makes so that next year you are the one the apprentice is asking.