Total harmonic distortion (THD) is the most-cited amplifier spec and the most misread. The CTA-2006-D power rating tolerates up to 1% THD+N at full rated output, but a competent modern car amp measures 0.01% to 0.05% at the 1 W operating point where music actually sits. You can measure that in your shop with a $650 USB analyzer and free software.
Key Takeaways
- CTA-2006-D rates car amplifier RMS power at ≤1% THD+N into 4 Ω at 14.4 V, with a 1-minute preconditioning step at 1/8 rated power. It is a power-rating method, not a sound-quality threshold.
- Modern competent amplifiers measure 0.01% to 0.05% THD+N at 1 W/4 Ω/1 kHz on the bench. SQ-grade builds target the lower end of that range.
- Meaningful in-shop THD measurement can be done with a $650 QuantAsylum QA403 and free REW software. Audio Precision lab gear is not required.
- In-vehicle, ground integrity, charging system condition, and gain structure cause most of the excess distortion we trace. The amp itself is rarely the problem.
What THD Actually Measures and What's Audible
The Audio Engineering Society places typical-listener audibility on music at about 1% THD, and as low as 0.1% to 0.3% for trained listeners on simple tones (AES). THD is the ratio of the RMS sum of an amplifier's harmonic components to its fundamental output, expressed as a percentage. THD+N adds the noise floor in quadrature. We measure it because it puts a number on the part of the output that wasn't in the input, the distortion the amp added on its way through.
Audibility on music depends on the listener and the harmonic profile. Even-order harmonics (2nd, 4th) sound musically related and are more forgiving. Odd-order harmonics (3rd, 5th, 7th) sound harsh and are what gives clipped amplifiers their character.
In a vehicle, engine and road noise mask a lot of distortion. On quiet passages, on a car parked with the engine off, or on competition tracks, you'll hear it. That's why SQ judges listen to the system in a quiet lot, not on the highway.
What Does CTA-2006-D Actually Certify (and What Does It Miss)?
CTA-2006-D, the Consumer Technology Association's car-amplifier power-rating standard, certifies continuous output at ≤1% THD+N into 4 Ω at 14.4 V with a 1-minute, 1/8-power preconditioning step (CTA). The full conditions:
- 14.4 V DC supply, regulated
- 4 Ω resistive load (non-inductive)
- RMS continuous output at the rated frequency or band
- Distortion ceiling: ≤1% THD+N at the rated power
- Preconditioning: 1 minute at 1/8 rated power before the rated-power test
That's a power-rating spec. It establishes how a manufacturer is allowed to publish "100 W RMS x 4 channels." It is not a sound-quality threshold. An amp can hit its CTA-2006-D rating at 1% THD+N at full power and still measure 0.01% THD+N at 1 W into 4 Ω, which is the operating point most music actually lives at.
| Budget amp | Mid-tier amp | SQ-grade amp | |
|---|---|---|---|
| At 1 W (typical music output) | 0.5% | 0.1% | 0.05% |
| At rated power (CTA-2006-D) | 1.0% | 0.3% | 0.01% |
Source: Audio Intensity bench measurements; representative published manufacturer specs.
A CTA-2006-D rating tells you about the right edge of this table. The 1 W column on the left is where music actually sits. The standard also doesn't cover full-band distortion, distortion at lower impedances, or distortion under realistic dynamic music load. That's why we measure on the bench instead of just trusting the data sheet.
Equipment You Actually Need
You can measure THD+N down to -120 dB residual with a $650 QuantAsylum QA403 and free REW software, no lab budget required (QuantAsylum). Three tiers, depending on how serious you are about measurement.
Pro Bench
- Audio Precision APx515 or APx555 ($15k to $50k+). The reference. Audio Precision is what most amp manufacturers publish their numbers from (Audio Precision).
- Calibrated 4 Ω non-inductive dummy load, 200 W minimum.
- Bench DC supply capable of 14.4 V at the amp's full current draw.
Serious Enthusiast / Shop Bench
- QuantAsylum QA403 (around $650). USB analyzer with residual THD+N below -120 dB at line level, more than enough headroom to characterize any car amp.
- Free measurement software: REW or ARTA.
- Same dummy load and DC supply requirements as above.
- Resistive attenuator (10:1 or 100:1, depending on the power level under test) so the analyzer's input never sees more than about 2 V.
Minimum Viable
- A 24-bit USB audio interface with a measurement-grade ADC (Focusrite Scarlett, MOTU M2, Cosmos APU front end). Residual THD+N around -100 dB is the practical floor.
- REW (free).
- Dummy load. A power-resistor bank in a fan-cooled aluminum housing works. Speakers don't, because their impedance varies with frequency and they add their own distortion to the measurement.
- A True-RMS DMM for cross-checking output voltage.
Things you don't need: a smartphone app, an oscilloscope by itself (you can see clipping, but you can't isolate harmonic content), or a "step-down transformer" of any kind. That last one shows up in a lot of internet guides and it's wrong. Speaker-level to analyzer-level uses a resistive attenuator, not a transformer.
Bench Procedure: Measuring an Amplifier in Isolation
A bench-isolated CTA-2006-D-style run takes seven steps, including a mandatory 1-minute preconditioning step at 1/8 rated power before the rated-power measurement. This is how we characterize an amp before it ever sees the inside of a vehicle.
- Connect the amp's RCA inputs to the analyzer's signal generator output. Set the generator to whatever level produces 200 mV at the amp's input (the typical line-level reference).
- Connect the amp's speaker output to the 4 Ω dummy load. Take the analyzer's input from across the load through your attenuator.
- Set the amp's gain to its CTA-2006-D test position (the manufacturer's published gain setting, or the position that produces rated power from a 200 mV input).
- Power up the bench supply at 14.4 V. Verify under load with a DMM.
- Run the preconditioning step: 1 minute at 1/8 rated power, 1 kHz tone. This warms the output stage to a representative operating temperature.
- Capture three measurements:
- 1 W into 4 Ω, 1 kHz, single tone. This is the SQ-relevant number.
- 1/2 rated power, 1 kHz. Mid-range headroom check.
- Rated power, 1 kHz. Confirms the CTA-2006-D claim.
- Run a sweep from 20 Hz to 20 kHz at 1 W. Note any frequency-dependent rise in THD+N.
A solid Class AB or Class D car amp will measure under 0.05% THD+N across the band at 1 W. We've seen Arc Audio and Tru Technology amps come in under 0.02% at this operating point on our bench. If your number is above 0.1% at 1 W and the amp isn't visibly damaged, suspect your test setup before you suspect the amp.
In-Vehicle Procedure: Measuring the Real Chain
The delta between bench and in-vehicle is everything: head unit, DSP processing in the signal chain, signal cabling, ground integrity, charging system, and gain structure all add measurable distortion that the amp alone never sees. The bench tells you what the amp is capable of. The vehicle tells you what's actually happening at the speaker terminal.
- Load a USB stick with calibrated test tones. 1 kHz at -10 dBFS and a 20 Hz to 20 kHz log sweep, both at known reference level.
- Disconnect the speaker(s) on the channel under test. Connect the amp output to a portable dummy load (a 4 Ω rig will fit in a small toolbox case).
- Probe across the load through your attenuator into the analyzer. Run with engine off, ignition on (alternator not charging).
- Run the same three measurement points from the bench procedure: 1 W, 1/2 rated, rated. Run the sweep.
- Repeat the entire sequence with the engine running. The delta between engine-off and engine-on is the charging system's noise contribution.
If you don't want to disconnect speakers, you can probe across the speaker terminals with the speaker connected, but the impedance variation across frequency will distort the sweep. The dummy load gives you a clean number that's directly comparable to the bench.
How Do You Read a THD vs Frequency Sweep?
Five fault-pattern shapes account for nearly every excess-distortion reading we trace in customer systems, and the THD+N vs frequency curve tells you which one you're looking at. A single 1 kHz tone misses all of them.
- Rising HF distortion above 5 kHz. Often output-stage slew-rate limitation or transistor cutoff behavior. Normal in small amounts; a sharp rise is a problem.
- 60 Hz, 120 Hz, 180 Hz spikes. Alternator whine bleeding into the signal path. Trace to grounding or to a noisy charging system.
- Rising LF distortion under load. Power-supply sag. The amp's rail voltage is dropping when bass hits, and the output stage is leaving its linear region. Big-three upgrade or a larger alternator.
- Asymmetric harmonic content (strong 2nd or 3rd, others absent). DC offset or a biased output stage. Could be a failing component. This one shows up in the harmonic spectrum at a single tone, not in the THD+N sweep shape, which is why it isn't plotted above.
- Broadband floor rise across the whole sweep. Ground loop or shielding problem in the signal cabling.
You can't get this kind of diagnosis from a single 1 kHz tone. The sweep is the difference between knowing the amp is dirty and knowing why.
Where Does Excess Distortion Come From in an Installed System?
From our bench logs across years of tuning and warranty work, five install-side causes account for nearly all of the excess distortion we measure at the speaker terminal. The amp itself is at the bottom of the list.
- Improper gain setting. Most installers we see set gains by ear, by feel, or by head-unit volume position. Under normal-volume music, the amp doesn't clip. Under transients, it does. An oscilloscope or a clip-detection tool is the only way to set gain structure properly.
- Inadequate big-three upgrade. Factory ground straps and B+ wiring are sized for stock current draw. A 1000 W system pulling 80 A will sag the rail voltage if the wiring hasn't been upgraded. Distortion follows the sag.
- Bad ground at the amp chassis. Paint not scuffed off, a loose chassis bolt, or a bolt threaded into a coated captive nut. Resistance in the ground path shows up as 60 Hz/120 Hz harmonic content under engine load.
- Driver over-excursion mistaken for amp distortion. Subwoofers especially. If a 10" sub is at its 18 mm one-way Xmax and you're feeding it past that, the THD coming off the cone is the driver, not the amp. Worth measuring at the amp output before you blame the amp.
- Cheap LOCs and signal converters. Pulling a high-level signal off a factory head unit through a budget LOC adds its own THD. We've measured 0.5%+ THD+N from converters costing under $40. Active LOCs with a clean output stage solve it.
Where to Go From Here
Spec-sheet numbers are a starting point. The actual distortion in your system is the bench number plus everything the install adds. We've spent enough time on the analyzer to know where the install losses come from, and we sell amps from Arc Audio, Tru Technology, and Audiocircle that bench cleanly enough that the install becomes the limiting factor instead of the amp. Goldhorn DSPs handle the signal path with the same low-distortion target. THD measurement is one step in a longer car audio DSP tuning workflow that also covers gain structure, time alignment, and in-cabin EQ.
If you're building a system aimed at SQ-grade THD numbers and want help speccing the chain, contact us and tell us what you're starting with. We'll work backwards from the target.
