• State
  • Replies 3 replies
  • Subscribers 46 subscribers
  • Views 73 views
  • Users 0 members are here
Support feedback
Options
Options
Related

AMC1306E05: Decoding Manchester output AMC1306E05

Barish Malshikare
Prodigy 10 points
Part Number: AMC1306E05
Other Parts Discussed in Thread: AMC1306M05

We have AMC1306E05 connected to controller where we using sinc3 filter to decode and filter the manchester output stream of AMC1306E05. We are not seeing the promising result with the given configuration as the resutls varies graltely if we use diffrent decimation factor or the clock frequency (CLKIN for AMC1306E0). 

Since both the AMC1306E05 and the controller’s filtering (decoding and Sinc3 filter) are treated as black boxes in our setup, we want to validate the behavior of the AMC1306E05 independently.

The AMC1306E05 datasheet does not provide detailed information about the bitstream output for different input voltages. Section 8.3.4 mentions the expected stream of zeros and ones for 0 V, +250 mV, and –250 mV, but we need to validate that the actual bitstream matches the applied voltage. What would the exact bitstream look like for these voltages, and if we want to check manually, what window size should we use to calculate the percentage of zeros and ones as indicated in the datasheet?


 image.png


  • 0
    TI__Guru**** 182041 points

    Hi Barish,

    Welcome to our e2e forum!  The 'ones density' would not change between Manchester vs. the traditional coding output.  You need to review Figure 52 to compare the uncoded versus Manchester coded bitstream.  What controller are you using to reconstruct the Manchester coded output?  

  • 0 in reply to Tom Hendrick
    Prodigy 10 points

    Hello Tom, 
    Thank you for the response. 

    We are using STM32U5XX and using the MDF (multifunction digital filter) peripheral/module of controller.
    We also have results for different differential input voltage applied to AMC1306E05 with different filters, decimation factor and the clock as follows:

    1) Filter = Fastsin, Decimation = 512, clock = 7.9Mhz
12mV	      51915264
56.7 mV	    62193561
101.9 mV	  70359859
147.1 mV	  58060902
192.2 mV	  37196953
237.4 mV	  19629875


2) Filter = Sin3, Decimation = 16, clock = 18.1 Mhz
12mV	     38963 
56.7 mV	   187852 
101.9 mV   333824 
147.1 mV   484147 
192.2 mV   629862 
237.4 mV   778547 


3) Filter = Sin3, Decimation = 32, clock = 18.1 Mhz
12mV	     312832 
56.7 mV	   1496012 
101.9 mV   2681036 
147.1 mV   3864064 
192.2 mV   5048832 
237.4 mV   6229555 


4) Filter = Sin3, Decimation = 64, clock = 18.1 Mhz

12mV	      2515865 
56.7 mV	   11981721 
101.9 mV   21450188 
147.1 mV   30917580 
192.2 mV   40390092 
237.4 mV   49858201 

5) Filter = Sin3, Decimation = 128, clock = 9.7 Mhz
12mV	     208135475 
56.7 mV	   271983667 
101.9 mV   279499673 
147.1 mV   186647756 
192.2 mV   106407833 
237.4 mV    50848153  


6) Filter = Sin3, Decimation = 128, clock = 12.7 Mhz
12mV	     209969305 
56.7 mV	   267022694 
101.9 mV   278268057 
147.1 mV   192710246 
192.2 mV   106398361 
237.4 mV    50849433 

7) Filter = Sin3, Decimation = 128, clock = 18.1 Mhz
12mV	      20118732 
56.7 mV	    95853465 
101.9 mV   171593420 
147.1 mV   247348224 
192.2 mV   323098368 
237.4 mV   398857113 


8) Filter = Sin3, Decimation = 128, clock = 25.4 Mhz
12mV	      20159897 
56.7 mV	    95788185 
101.9 mV    171539353 
147.1 mV    247321651 
192.2 mV    322997350 
237.4 mV    398782668


9) Filter = Sin3, Decimation = 256, clock = 18.1 Mhz

12mV	      160983091
56.7 mV	    463869747
101.9 mV	  1069925632
147.1 mV	  1675955251
192.2 mV	  2063188403 WARNING: Filter5 SATF=1 (saturation/error detected)
237.4 mV		2147483392 (WARNING: Filter5 SATF=1 (saturation/error detected))


10) Filter = Sin5, Decimation = 32, clock = 7.9Mhz

12mV	      2147483392 (WARNING: Filter5 SATF=1 (saturation/error detected))
56.7 mV	    2147483392 (WARNING: Filter5 SATF=1 (saturation/error detected))
101.9 mV	  2147483392 (WARNING: Filter5 SATF=1 (saturation/error detected))
147.1 mV	  2132144972 (WARNING: Filter5 SATF=1 (saturation/error detected)) 
192.2 mV	  1925125452
237.4 mV	  1258743910



11) Filter = sinc3, Decimation = 256, clock = 7.9Mhz
12mV	      1040050892
56.7 mV	    1919153587 (WARNING: Filter5 SATF=1 (saturation/error detected))
101.9 mV	  2129002342 (WARNING: Filter5 SATF=1 (saturation/error detected))
147.1 mV	  1805260697
192.2 mV	  1200068198
237.4 mV	  629071411


12) Filter = sinc3, Decimation = 256, clock = 7.9Mhz
12mV	      13024358
56.7 mV	    12988108
101.9 mV	  13065676
147.1 mV	  13119232
192.2 mV	  13061734
237.4 mV	  13109555


    Summary of results: 

    Good: Sin3 @ 18.1 MHz (decimations 16/32/64/128), Sin3 @ 25.4 MHz (decimation 128). These climb smoothly without saturation.

    Marginal: Sin3 @ 9.7 MHz and 12.7 MHz (decimation 128) show mid-range roll-off after ~102 mV.

    Poor/saturated: Sin3 decimation 256 @ 18.1 MHz (hits SATF at high inputs), Sin5 @ 7.9 MHz (saturated almost entirely), sinc3 first set (multiple SATF), sinc3 second set (flat ~13 M across inputs), Fastsin @ 7.9 MHz (non-monotonic).

    Based on the observed results, the filter output shows inconsistencies across different decimation, filter, and clock configurations. While variations in output counts are expected, the linearity should have been maintained. We have an open query with STM to understand the root cause of this behavior.


    Our objective is to understand the expected output behavior of the AMC1306E05 for various input voltage levels


    According to Figure 52, Manchester encoding is illustrated based on the uncoded bitstream. Our question is: which bitstream (uncoded or Manchester-coded) does the AMC1306E05 output for different input voltages?
    For example, If 0 mV results in 50% ones and zeros, will the output be:
    1) A repeating pattern like (010101010101...) continuously?
    2) Blocks of 16 zeros followed by 16 ones?
    3) Larger blocks, such as 127 zeros followed by 127 ones?”

    The clarification on the bit stream will be helpful to debug the MDF implementation. 

    Thank you in advance. 

  • 0 in reply to Barish Malshikare
    TI__Guru**** 182041 points

    The AMC1306E05 uses the Manchester coded output while the M05 would use the uncoded output.  The Manchester decoding done in the CPU would need to be using a clock at least 6 times faster than the datastream.  Have you tried the AMC1306M05 as an alternate?  My understanding is that the STM32U5XX does not have a native Manchester decoder so that might be part of the issues you are having.