AS5048 Angle Encoder Magnet Distance Adjustment

[jpanhalt]

The AMS AS5048 magnetic angle encoder has an automatic gain control to compensate for different magnetic field strengths, and the gain setting can be read by the user. Recommended magnet-to-chip distance is 0.5 mm to 3 mm using the recommended magnets. The evaluation board has a simple set screw adjust that allows one to adjust that distance.

For the sensor I am designing, I could incorporate an adjustment feature, but that will add complexity to the machining, and I would just as soon avoid it if possible.

Here is a draft of that device:

For those who have used this or similar chips, is it necessary to adjust that distance, or can one simply get the recommended magnet, set the distance at 1.5 mm (for example), confirm it works and move on? How often is there a problem? If you do check the AGC, what value do you try to get?

Regards, John

 

[Tony Stewart]

The input range is limited from +\- 30 to 70 mT and magnet strength varies considerably with temperature, gap and magnet tolerances, so unless you can centre balance it and adjust gap, then fix it with epoxy, you will have issues. External linearization and averaging from noise is also expected.

 

[BobW]

If you know the correct gap before final assembly then you may be able to use a variation of the installation technique for an encoder that I have used.

The magnetic encoder that I recently bought, came with its own magnet mounted on a sleeve that is intended to be a tight friction fit over a 6mm shaft. The installation method involves placing a shim between the magnet and the sensor. The shim is the same thickness as the required gap. The magnet is slid partway onto the shaft. The housing is then temporarily assembled causing the shim to push the magnet further onto the shaft. The housing is then disassembled, the shim removed and then the housing is reassembled, resulting in the correct gap between magnet and sensor. You may be able to design the housing with a hole in the side so that the shim can be withdrawn without having to disassemble the housing (and then plug the hole after the shim is removed). The shim could even be a piece of wire or a drill bit.

Adhesive can also be used when mounting the magnet, so that once it sets, there's no chance of it moving out of position.

 

[jpanhalt]

Thanks for confirming that. I browsed products and illustrations by Renishaw, Avago, and AMS and came to the same conclusion. It seems adjustable magnet distance is not provided as a routine feature. Getting that distance set accurately will be fairly easy in my design, so all I will need to know is what distance to shoot for. I have a test fixture I can use to determine that. One big variable will be the PCB thickness. Ideally, I would have gotten a single-sided board with solder mask only on the top side. I know how to get that done at OshPark, but it can raise questions. So, for the batch I just ordered, I got double-sided boards with solder mask on both sides. For $9 for 3 boards, it will not be a big loss. Besides, I think I will be changing the lead-out design anyway.

The entire unit is made of 6061 aluminum. The magnet will be in a recess and retained with a low-temp cyanoacrylic adhesive.

John