AMC3306EVM: AMC3306M05 and C2000

Colin Bryant

Part Number: AMC3306EVM
Other Parts Discussed in Thread: AMC3306M25, AMC3306M05

Tool/software:

New to using Sigma Delta Modulators, need sanity check.

Have some basic questions, questions I guess, are related to both the AMC3306 and the F28379D MCU.

Using AMC3306EVM (eval board) along with a LaunchXL_F28379D (C2000 MCU); MCU is providing the Sigma Delta Filters for the 3306 Sigma Delta Modulators. Have a question regarding the input data range vs output of Sigma Delta Filters (MCU_output)

AMC3306EVM comes with AMC3306M25 Sigma Delta Modulators installed so, signal input range is ± 250 mV.

I replaced the AMC3306M25 with AMC3306M05 (± 50 mV; input signal range)

3306EVM

Using a 3 channel arbitrary waveform generator to input sinusoidal signal into INP and INN ports of J2 on 3306EVM.

Below is part of simulink model using LaunchXL_F28379D as target

Using F28379D for Sigma Delta Filters
Using F28379D ePWM for Sigma Delta Clk
- clk to Sigma Delta Modulator (AMC3306EVM)
- clk to Sigma Delta Filters (F28379D)

FIGURE 1 - Simulink Model deployed to F28379D

To see full (rail to rail) waveform out of Sigma Delta Filters:

Set Input (signal source) DC Offset to +27 mV
Set input (signal source) Amplitude to approx 63 mV pk to pk (approx ± 31 mV ref +27mV dc offset)
With these settings below is image of what I see on output of Sigma Delta Outputs
- apologize for the small hard to see image
  - Top set of waveforms are being measured towards the end of the model shown in FIGURE 1.
    - Signal is centered about a count of 16384
    - high side pk is very close to 32768
    - low side pk is very close to 0
  - Bottom set of waveforms is directly out of MCU SD Filters
    - Identical to the waveforms described above.

Questions / comments:

Am using Sin3 filter setting on the MCU SD Filters (Filter has option for 16bit or 32 bit representation); 3306M05 datasheet says 16 bits of resolution can be achieved. (2^16 = 65635; 2^15 = 32768)

Did not see difference in SD Filter output range related to filter resolution option; 16 or 32 bit

From what I am seeing.. max signal I can get out of SD Filter is 0 to 32768) why?

Should it be 0 to 65635 or is msb supposed to be an indicator of positive / negative excursion

My signal into the SD modulator is approx 60 mV pk pk.. with a 27 mV dc offset; This is a signal of ± 30 mV pk, right? Should it not be at least ± 50 mV?

Lowering my input signal amplitude to approx 50 mV pk to pk (± 25 mV) then lowering DC offset to approx 22 mV I get the following waveforms:

Again sorry for the hard to see image: (again top and bottom plots are practically identical

Bottom set of waveforms are being measured at output of SD Filter blocks shown FIGURE 1.
- since DC offset was lowered so was the center point of waveforms
- low side pk... green waveform is just starting to touch a zero count; then looks like it wraps up a max count of 65635.

What am I missing?

Is it how I am driving Modulator?

Are these output waveforms correct?

Am I using the MCU Filters correctly?

3 months ago

0 Gustavo Fonseca 3 months ago

TI__Prodigy 361 points

Hi Colin,

Sorry for the delayed response,

The way you are driving the modulator seems correct, as in the ranges you are driving the modulator are within spec.

I believe the main reason you are only getting from 0 to 32768 is due to you never going below zero except in your second example. 0 to 32768 represents positive values when 32769 to 65635 represent negative.

However, I am concerned that you are getting max values for your first set up, this would imply that the SD is measuring maximum voltage (higher than +- 50 mV). Is this due to how you set up your SDFM?

Looking at the second waveform, this looks more correct as you are driving the modulator at 50% of its capabilities thus the 50 of the effective range is used. The main reason for the values that dip below 0 going to 65635 is negative

Please let me know more about your setup, I am very willing to help more but a bit more information about your setup would be appreciated.

Thanks so much for your patience,

Gustavo Fonseca

0 Colin Bryant 3 months ago in reply to Gustavo Fonseca

Prodigy 60 points

Hello Gustavo Fonseca,

Thank you for taking the time to look at my issue.

The first graph was intended to show max signal I could input into sd_modulator (amc3306m05) and not clip the output of sd_filter (MCU F28379D).

Looking at the datasheet for the AMC3306M05, it is saying Diff input voltage before clipping is ± 64 mV nominal; The specified linear diff full scale is the ± 50 mV

So my signal was approx 60 mV pk to pk (maybe @ 63 mV pkpk clipping occurred) or ± 30 mV with dc offset of approx 27 mV.

Here is more information about my set up:

3 Sigma Delta Modulators eval boards(3306EVM) replaced 3306M25's (± 250 mV) with 3306M05 (± 50 mV)
- DC supply to Modulators is +3.3V; supplied from LAUNCHXL_F28379D
- SD_clk is provided by one of the LAUNCHXL_F28379D's ePWM blocks running at approx 20 MHz

1 LAUNCHXL_F28379D (C2000 Delfino; sysclk = 200 MHz; 4 core floating point MCU (CPU_1, CLA_1, CPU_2, CLA_2)
- SD_clk for SDFM (SD_Filters in MCU) is also supplied by same ePWM block running at approx 20 MHz

Input into the 3306EVM's is provided by 3 ch Arbitrary Waveform generator. (signal source)

Signal Source to SD_modulator (3306EVM) originally had generator at 0V dc offset and no intelligible signal was notices at output of MCU SD_Filters.

As i increased the signal source's dc offset I stared to see the desired signal out of SD_Filters.

The waveforms you are seeing in the plots from first post are logged signals tied to output of SD_Filters peripheral SDFM's in MCU. (this is a Simulink Model)

Believe may main issue is due to the LAUNCHXL_F28379D. There is no negative rail available (as far as I know).

This would make sense to me.. having a range of 0 to 32768 for 0 to 3V; then 0 to -32768 for 0 to -3V; total diff being 65536 steps.

This would also explain the seemingly required dc offset; if I only have single ended operation range available.

So naturally I guess the next question would be, can I get MCU SFDM's to operate differentially?

Best and Thanks for your time,

Colin

0 Gustavo Fonseca 3 months ago in reply to Colin Bryant

TI__Prodigy 361 points

Hello Colin,

The AMC3306M05 can handle up +-64mV but this is not recommended as more than 50mV the response of the device becomes non-linear. Thus, it is recommended to stay in the 50mV range.

Hmm it is strange that you need this DC offset to get a readable signal, I believe this would be more to do with the actual setup of your circuit rather than the SDFM.

Can you try to input a DC signal going from -full scale to full scale to verify your setup?

Thanks for the very in-depth breakdown,

Gustavo Fonseca

0 Colin Bryant 2 months ago in reply to Gustavo Fonseca

Prodigy 60 points

I did DC from - to +, ckt behaved the same. I am currently working with a Mathworks Consultant. Have asked him this question. He is checking on how to operate the F28379D SDFM in a differential mode. Will update when I get his comments, should be today 7/9/2025.

0 Gustavo Fonseca 2 months ago in reply to Colin Bryant

TI__Prodigy 361 points

Hi Colin,

No worries, let us know when you have some more clarity on the Simulink side of things.

0 Colin Bryant 2 months ago in reply to Gustavo Fonseca

Prodigy 60 points

Issue Resolved. It was a data type issue.

Data type out of Simulink SDFM block was uintXX, could be uint16 or uint32 based on "Data representation:" of block settings in SDFM block parameters in Simulink.

Adding a data conversion, like intXX, corrected the issue.

Best,

Colin

0 Gustavo Fonseca 2 months ago in reply to Colin Bryant

TI__Prodigy 361 points

Ah interesting,

Thanks for the update,

Gustavo Fonseca

Data converters

Data converters forum

AMC3306EVM: AMC3306M05 and C2000