Ok i was looking at base stations etc and antenna placement, and found this ground plate, option I suppose to boost the strength of signals for base stations, but I suppose it works for non metal cab roof installs.
As for this threads topic, I go for mid tractor, so about half way from the fixed pivot axil to the steering axil. This for me is a great compromise for visibility of shifts in heading from the steering axil but still representing the implement and rear axils position on a standard tractor.
This for me goes away if you go duel antenna, because then you can do both, for me a centre line duel antenna will give you a great setup, you can have one over the axil that best represents the implements position and have another as close to the front axil on a tractor as you can to see deviations, side tracking etc as soon as it happens. Itâs recommended to keep them around 2 meters apart, but less if needed. Also at the same height so keeping setup as simple as possible.
With the new duel gnss chips UM982 with 2 antennas on one chip, and lots of variants out there at simplify prices to a signal antenna UM980 I think duel antenna will go mainstream.
I guess accurate heading is also an advantage of the dual antenna approach but accurate roll is very important. Much more important than accurate pitch.
The comment on using position data from both antennas sounds odd. Iâm not aware of any auto-steer application that uses a single antenna position (as part of dual setup) for a specific purpose. Is that even any advantage since the geometry is known as well as accurate roll data?
You are correct you get better roll detection from side by side, but better heading or yaw detection front to back.
So It comes down to whatâs best, use an IMU sensor for roll and have great yaw heading data or use 2 antenna just to get better roll data?
I say fantastic yaw heading data wins, but thatâs me.
Knowing where the implement is and if the tractor front end is on coarse or not, seems the better thing to know, tilt sensors are already part of the Opengps system so why not keep that then refine the tractors heading better?
heading data in a front and back setup for me wins as that centre line data it creates can be projected back to the implement. And in the end implement position is often more important.
It would be interesting to test what gives the better results knowing the tractors centreline and heading, and relying on tilt sensor data or using a side by side antenna setup and have great roll data. But poorer heading data.
You cannot have both so whatâs more important?
Has anyone considered if duel tilt sensors, improve tilt data?
Tilt sensor positioning as this is a thread giving better roll data can the possition of a tilt sensor improve roll data?
Ok an Ai answer.
To obtain the highest quality data from an IMU on a tractor, the sensor must be positioned to minimize mechanical noise while maximizing its ability to capture true vehicle rotation.
Optimal Mounting Location
Center of Gravity (CG): Mount the IMU as close to the tractorâs center of gravity as possible. This is the point about which the vehicle naturally rotates; placing it here captures genuine motion without the âcentrifugalâ or exaggerated accelerations that occur at the edges of the vehicle.
Rigid Attachment: The IMU must be fixed to a rigid part of the chassis. Avoid mounting it on flexible panels, plastic trim, or thin sheet metal, as these amplify vibrations and cause âringingâ in the data.
Wheelbase Midpoint: If the exact CG is inaccessible, position the unit firmly to the body mid-way along the wheelbase.
Proximity to Other Sensors
Near GPS Antenna: For high-precision auto-steering (GNSS/INS), mount the IMU as close to the GPS antenna as possible, preferably on the same rigid structure (e.g., the cab roof over the A or B pillar).
Lever Arm Calibration: If they cannot be placed together, you must accurately measure the distance (the âlever armâ) between the IMU and the antenna with millimeter precision to correct for position offsets during turns.
Orientation and Alignment
Principal Axes: Align the IMUâs internal axes with the tractorâs primary axes:
X-axis: Pointing straight forward.
Z-axis: Pointing straight down (or up, depending on the sensorâs default).
Level Mounting: Ensure the sensor is mounted perfectly level with the tractorâs frame. Any initial tilt will be interpreted as a constant roll or pitch offset unless zeroed out through software calibration.
Environmental Protection
Vibration Isolation: Use thick double-sided foam tape or specialized damping mounts if the tractor experiences extreme engine vibration.
Magnetic Interference: Keep the IMU away from large ferrous metal masses (like the engine block) or high-current wires (like the alternator or battery) to avoid distorting the magnetometer (compass).
Shielding: Use a protective casing to guard against temperature fluctuations and mechanical shocks, which can cause sensor drift over a long working day.
This for me says the IMU is on the cab roof and over the rear axil centre point in a normal tractor as that sees the least rotational movement.
And in a front back setup to get great yaw data itâs the same place I would put the rear antenna. So that meets another suggested best place near a gps antenna.
So yes side by side gives the best roll data but is it the best use of duel antenna for the system? I think not. I think yaw is better than roll as we have roll data from the IMU. And duel antenna is fantastic for yaw.
If anyone has the gear to test what gives the cleaner pathing data from either setup side by side better roll data or front back better yaw data using the old IMU for roll data, and can test them and prove out both setups it would be great.
I asked Ai over the IMU it suggested that close to the antenna was best so cab roof and over the rear axil to reduce rotation motion
Now my brain says as you, low down for the IMU, so I now the question is did the Ai have it right? I think that as long as we can tell the system where it is the less stress it gets the better so lower near the rear axil. Or CG the better they suggest rubber mounts to cut down on vibrations so I would go Cab over the rear axil. So over the area of the tractors that twists and moves the least.
I also note in the link they use duel antenna setups to get yaw data, and IMU to get tilt still.
I understand what you mean, your correct you get the best roll data from side by side, and yes it helps, but in my mind itâs not clear which data is more important, the Yaw data from a front back setup so you have an accurate fix on the tractors rear axil and can tell the pathing from the front antenna. Then Picking up tilt data from the IMU as normal. ( it seems that you could calculated all types of tilt as well, but as you point out this is not the normal way to do this)
Or clearer ( maybe) roll data from side to side antenna but get less accurate or zero Yaw data, and take out the IMU.
In my brain it says, knowing the heading better, and having ok roll data from the IMU, mounted far lower on the tractor, ideally under the roof antenna over the rear axil seems a better setup. We are then adding Yaw data you didnât have rather than just improving roll data.
As far as I can tell the system cannot detect yaw data in either setups suggested now, so in a side by side duel antenna setup or signal antenna setup. It just sees if it comes off the path it cannot see if the tractors front end is off the path just if the fixed calculated gps location is off the path.
With yaw you can use Yaw data to see if the roll data matters so filter the roll dataâs influence on the pathing. And see if the tractor is on path and if the heading of the front of the tractor is aiming for path or away from path earlier.
This is maybe my wrong thinking but if it is this is why we have discussion pages.
What is better, adding Yaw data, or improving on the IMU Roll data?
We also have to consider whatâs the best placement for an gps antenna what we want to know is where the implement is in relation to the last pass, the best place is over the fixed rear axil of the tractor, but that is terrible for detecting heading as itâs the last point on the tractor to move off the heading line so sees the smallest shift in GPS. Locations so we move it up forwards to see those heading shifts earlier, but with duel setup in a front to back setup we get the perfect antenna placement over the rear axil and the best yaw data from the second being far further forwards as the further forwards you place it the easier to detect the actual tractor heading is to see in an ideal world, one antenna over the rear axil one over the front axil. And the IMU detecting roll as it normally does. I see this as the better use of a second antenna. As knowing better pathing data for me is better than improved roll data. In some respects the system uses wheel angle data to make up for no Yaw data. Which works but again would cleaner yaw data or just having Yaw data be better than cleaner roll data.