Conspit’s QR Test Methodology

Methodology developed with Ernest

Goal

After spending considerable time with both CDP and CDR, I noticed CDP felt cleaner and more composed. CDR felt punchier, with certain frequencies feeling amplified and others absent. The question wasn’t which one is better, it was simpler than that. Does what I felt actually show up in the data? And if it does, what separates them?

Equipment

Conspit Ares Platinum Wheelbase
Conspit CDP Quick Release
Conspit CDR Quick Release
ADXL335 3-Axis Analog Accelerometer x 2
OWON SDS1102 Oscilloscope
9V battery pack
MotorBuster

Setup & Mounting

The wheelbase was secured by 4xM6 screws on the deck of the sim rig used for daily driving. Both accelerometers were mounted via heavy-duty double-sided tape. One at the shaft, one at the rim, while making sure they are in the same orientation relative to the axis of rotation. Accelerometers were powered by a 9V battery pack to isolate them from mains electrical noise. Both units were connected directly to the oscilloscope. Yellow trace corresponds to the shaft accelerometer, blue trace to the rim.

Procedure

MotorBuster was configured to send constant magnitude pulses at a single target frequency. Each QR was tested across seven frequencies, that is 30, 60, 100, 150, 200, 250, and 300Hz. During development, each frequency was run a minimum of 7 times to validate consistency. Final recorded runs were conducted three times per frequency, per QR. The QR was then swapped and the full procedure repeated under identical conditions.

Room temperature was recorded at the start of each session. Session one was conducted at 11:30 AM and room temperature was 28°C. Session two at 2:00PM and room temperature was 32°C.

Note – The wheelbase’s shaft was not mechanically locked during testing and the steering wheel remained mounted throughout. Under the vibration frequencies and magnitudes used, the shaft moved in small, rapid oscillations not visible to the eye, but well within the detection range of the accelerometer. Since the goal was to measure vibration transmission through the QR assembly rather than rotational torque response, locking the shaft provides no benefit.

Known Limitations

1. Accelerometer calibration
The ADXL335 units were not cross-calibrated against each other before testing. Readings reflect relative behaviour between the two QRs under identical conditions, not independently verified absolute values.

2. Test frequencies
Not all of them. MotorBuster runs on integer millisecond timing, so some frequencies get rounded. 100, 200, and 250Hz were delivered exactly. 30, 60, 150, and 300Hz were approximations. Since both QRs were tested under the same conditions, this affects them equally and doesn’t change what the comparison tells us.

3. Not testing multiple QR units
Only one unit of each QR was tested. Testing 10 or 15 units from the same batch wouldn’t tell you much either. A bad unit can slip through regardless of sample size. What actually guarantees consistency is the manufacturer’s quality control process. A faulty unit could be anything from a simple tolerance issue to a manufacturing defect to loose QC standards, none of which I have visibility or control over. Quiet revisions are also a known industry practice, meaning a unit purchased significantly later may behave differently. The findings here reflect the specific units tested, nothing more. If you suspect your unit differs, the methodology is documented and the test is replicable.