This is how we test our devices

New firmware for the Quad devices is now available for customers. Below is test data showing the performance.

First, few things about GPS: There is no absolute truth what comes to GPS and the same situation can never be repeated again. This is because GPS satellite coverage changes during the day, weather and atmosphere changes.

Best thing what can be done is to test all devices at the same time. Still, not the same but closest we can get.

Devices used

Vehicle BMW E90  
Logger MoTeC C125 display logger  
CAN Bus 500k BMW PT_CAN Wheel speed, steering angle
Reference GPS One typically used in MoTeC installations GPS+GLONASS
QUAD devices April 2022 FW 12V version in Aluminum casing
Antenna Standard GPS  

 

Logger setup

DEV1 channels: GPS Channels. When reference is tested, it is always in this channel.

DEV2 channels: GPS Channels ending ...L20.

Standard MoTeC GPS setup, 38.4k baud rate.

 

Antenna setup

 

Tests are done in 2 antenna positions:

- On the roof, antenna is mounted on the passenger side. See logs ending with ...roof

- Side window, antenna tilted 50degrees to the side. This is to simulate motorcycle lean angle. Logs with ending ...50deg. This method is quite tough for the GPS but is easy to repeat. Antennas are 40cm apart, this is important to remember, we'll come back to this later...

 

Testing method

Testing is done using 2 methods. We call it Rounds and Laps.

Rounds

Antennas are mounted either on the roof or tilted 50deg. Test is done in open sky location by driving 4 rounds with steering wheel turned to full lock. Result should always be very close to a circle and both devices should have the same form as shown below.

Laps

Laps are done by driving few laps using the same route, antenna on the roof and tilted 50deg. Again, laps should overlap since laps are driven at the same time and the antenna is in the same location for both devices.

Results

Rounds

Reference vs QUAD...L20 channels, Antenna on the Roof.

Noticeable here is the location error of the reference and altitude is below the sea level. Track of the QUAD device is almost perfect. 

Caption
Reference vs QUAD...L20 channels, Antenna 50deg tilted.

Similar results, reference is not really 

Quad vs Quad, location error, roof

Below is Google earth track from both QUAD devices, tested at the same time. 4 circles. Circles are almost perfect, location error between devices is 71.34cm as measured with Google Earth.

If you remember above, antennas are 40cm from each other. Due to this loops can't be overlapping and the final location error between 2 devices is about 31cm (71.34cm - 40cm = 31.34cm)

 

Quad vs Quad, location error, antenna 50deg tilted

Below is Google earth track from both QUAD devices. 4 circles. Circles are almost perfect, location error between devices is 55.85cm as measured with Google Earth.

Interesting here is that error is smaller when leaning.

 

So, how did QUAD fair against the reference?

First, on the roof: Reference Red, QUAD Blue

Error between devices is 214.25cm.  

Then, antenna tilted 50deg: Distance between devices is 240.24cm.

Laps

2 laps were driven for each devices, on the roof and antenna tilted 50deg

Quad vs Quad, antenna roof

Tracks looks ok, altitude and speed is correct. OK.

 

Quad vs Quad, antenna 50deg tilted

Tracks looks ok, speed is correct. Altitude takes a while to correct when satellites are found. OK.

Reference vs Quad, antenna roof

Looks OK. Reference will have a larger speed latency, more green is starting to show....

 

Reference vs Quad, speed latency

Note the GPS Speed trace in the middle. Green is reference, Red is QUAD. Wheel speed is shown as reference, without any corrections for turning circle etc...

Speed latency was one issue we fixed in the April release. Firmware has a special dynamics mode for car and motorcycle racing. 

Quad vs Quad, speed latency

Note the GPS Speed trace in the middle. Green is QUAD, Red is QUAD. Wheel speed is shown as reference, without any corrections for turning circle etc...

 

Ref vs Quad, track of laps, antenna on the roof

Reference is Red, QUAD is blue. 

Ref vs Quad, track of laps, 50deg tilted

Reference is Red, QUAD is blue. Note how positional accuracy of the reference suffers when the antenna is tilted.

 

Quad vs Quad, antenna 50deg tilted

Quad is Red, QUAD is blue.

Quad vs Quad, antenna 50deg tilted

Quad is Red, QUAD is blue.