Post Go back to editing

ADMT4000 Multi-Turn GMR Counter Behaviour after 46-Turn Reset

Thread Summary

The user is evaluating the ADMT4000 for multi-turn encoder applications and observes unexpected behavior after a 46-turn reset. The final answer clarifies that the 46-turn reset sets the GMR counter to the max turns, and D13 (TURN_CROSS) may become active in Continuous mode if the GMR and AMR turn counts do not align. To ensure reliable operation, the user should reset and start the conversion sequence in Continuous mode after a GMR reset.
AI Generated Content
Category: Hardware
Product Number: ADMT4000

ADMT4000 Multi-Turn GMR Counter Behaviour after 46-Turn Reset

Hello,

I am evaluating the ADMT4000 for an industrial absolute multi-turn encoder application.

I have successfully implemented:

  • Angle reading
  • LUT linearization
  • Fault monitoring
  • Radius monitoring
  • Multi-turn reading using register 0x04 (ABSANGLE)

I am currently investigating the behavior of the GMR multi-turn counter and the D13 (TURN_CROSS) fault.

Test setup

  • ADMT4000
  • Diametrical magnet
  • Magnetic field within specification (approximately 24 mT)
  • D14 (AMR_RADIUS) = 0
  • D9 (GMR_FALSE) = 0
  • D13 (TURN_CROSS) monitored continuously

Test 1: 46-turn reset procedure

I rotated the shaft more than 46 turns CW.

After power-up:

WHOLE_TURNS = 46
D13 = 0

This appears to be the expected reset state.

Test 2: One turn CCW while powered

Starting from:

WHOLE_TURNS = 46
D13 = 0

I rotated exactly 1 turn CCW while the device remained powered.

Result:

WHOLE_TURNS = 45
D13 = 1

The turn counter changed correctly from 46 to 45, but TURN_CROSS became active.

Test 3: Conversion restart

I implemented a software command that:

  1. Aborts conversion
  2. Restarts conversion
  3. Clears faults

After executing this procedure:

WHOLE_TURNS = 46
D13 = 0

The device returned to 46 turns.

Test 4: One turn CCW with power removed

Starting from:

WHOLE_TURNS = 46
D13 = 0

I completely removed:

  • VDD
  • VDRIVE

Then I rotated the shaft exactly 1 turn CCW while the device was unpowered.

After restoring power:

WHOLE_TURNS = 46
D13 = 0

The expected value would have been:

WHOLE_TURNS = 45
D13 = 0

but the counter remained at 46.

Questions

  1. Is the "46-turn CW reset" procedure intended only to clear the GMR counter, or should it also establish a valid multi-turn reference for subsequent operation?
  2. After a successful 46-turn reset, should a CCW rotation performed while the device is unpowered be reflected in the WHOLE_TURNS value after power-up?
  3. Under which conditions is D13 (TURN_CROSS) expected to become active when the WHOLE_TURNS counter remains inside the valid 0...46 turn range?
  4. Is an additional magnetic reset procedure (315° position with >60 mT field) required after the 46-turn reset procedure to obtain correct true power-on multi-turn operation?

Any clarification regarding the proper commissioning sequence for reliable GMR multi-turn operation would be greatly appreciated.

Thank you very much.

Parents
  • The GMR reset of the ADMT4000, whether achieved magnetically or by overturning by >46 turns will reset the state of the turn count sensor to the max turns as you have observed.  On the first conversion of the state of the GMR spiral in a measurement sequence, the GMR is used as the absolute position of the device, in subsequent conversion we use the AMR angle to report the absolute position, the GMR derived turn count is used as the counter cross check: the source for the D13 bit in the FAULT register.  You will there for see a slight difference in behaviour when in One-Shot conversion mode compared to Continuous conversion mode.

    • One Shot – for every conversion the GMR is used for the source of the absolute position. D13 will always show no fault since the AMR derived turn count will match the position of the GMR.
    • Continuous – following the first conversion in the sequence the turn count is derived from the GMR, subsequently the turn count is derived from the AMR. Following a GMR reset the GMR turn count will not necessarily align with the AMR derived turn count and a bit D13 error may be generated.

    If you are operating in Continuous acquisition mode (the default measurement sequence mode) you should always reset start the conversion sequence to ensure that the AMR and GMR derived turn counts match.

Reply
  • The GMR reset of the ADMT4000, whether achieved magnetically or by overturning by >46 turns will reset the state of the turn count sensor to the max turns as you have observed.  On the first conversion of the state of the GMR spiral in a measurement sequence, the GMR is used as the absolute position of the device, in subsequent conversion we use the AMR angle to report the absolute position, the GMR derived turn count is used as the counter cross check: the source for the D13 bit in the FAULT register.  You will there for see a slight difference in behaviour when in One-Shot conversion mode compared to Continuous conversion mode.

    • One Shot – for every conversion the GMR is used for the source of the absolute position. D13 will always show no fault since the AMR derived turn count will match the position of the GMR.
    • Continuous – following the first conversion in the sequence the turn count is derived from the GMR, subsequently the turn count is derived from the AMR. Following a GMR reset the GMR turn count will not necessarily align with the AMR derived turn count and a bit D13 error may be generated.

    If you are operating in Continuous acquisition mode (the default measurement sequence mode) you should always reset start the conversion sequence to ensure that the AMR and GMR derived turn counts match.

Children
No Data