Success story: CMM with custom signal health monitor solved long-standing cable internet issues

[ccpk1]

I wanted to share a quick Show & Tell of a custom signal health dashboard I built using the Cable Modem Monitor (CMM) integration, and how it finally helped me track down a phantom physical-layer issue I’ve been fighting for years.

The Struggle: "Everything is within spec"

For the past few years, I’ve noticed massive spikes of uncorrectable errors on my modem—sometimes quiet for a few hours, but then racking up into the millions. Because the interference would unpredictably bleed across multiple channels, I couldn't pin it down. I even had my ISP send line techs out twice. But because my speed tests were fine (~500 Mbps), they just shrugged, told me "signals and everything else is within spec," and said that's just how cable internet is.

I tried manually tracking the signal levels a year ago, but the noise and varying patterns was too intermittent to catch a pattern by hand.

The Dashboard: Automating and Understanding the Data

I stumbled on the CMM integration while I was doing some other overdue home assistant tech monitoring cleanup. Getting it up and running was quick and easy, but that’s only half the battle. You still need to understand the key data points, the DOCSIS specs, and how they behave over time.

Interference isn't always a uniform blanket of noise. Depending on the frequencies involved, it behaves unpredictably—sometimes isolating to just one channel, but more often splashing across the spectrum. This cross-channel bleed pushes the signal levels of neighboring channels out of spec, causing a cascading mess of errors. I built a custom dashboard to track my overall signal health and plot these error rates continuously so I could finally see the patterns.

Example where uncorrected errors might go quiet for a few hours after reboot, then re-appear.

The Smoking Gun

Having granular, 24/7 historical data changed everything. The dashboard revealed that while the noise occasionally bled across the spectrum to cause those million-error spikes, the epicenter of the interference was consistently hitting one specific downstream channel (Channel 12 @ 435 MHz).

That frequency was the exact clue I needed. It overlapped with a USB SD radio antenna I use for my rtl_433 water leak sensors (operating at 433.92 MHz). The antenna was mounted right in my server rack, just 18 inches away from the cable modem, blasting interference which, surprise to me, was getting picked up directly into the coax and modem even though they are hard wired!

The Fix
I moved the SDR antenna a few feet away using an extension cable and used a wrench to properly torque the coax connector on the back of the modem.

As you can see in the screenshots, the uncorrectable error rate immediately plummeted to an absolute flatline of 0.0/min and the SNR across all channels immediately improved.

Interested in setting up your own Signal Health Monitoring sensor?

NOTE: These are custom sensors and dashboard code, NOT supported by the CMM maintainers. If you need help, feel free to post in the comments below and I'll do my best to help you. DO NOT CREATE ISSUES FOR THESE CUSTOM CONFIGURATIONS.

The signal health sensor is based on the following thresholds. Of course you could go into the template and adjust them if you have good reason too, but these should be fairly standard defaults for most.

• [D] = Official DOCSIS 3.1 Hardware Specification Limit.
• [I] = Industry Best Practice / Warning Tier (Used universally by ISPs to roll trucks before a complete [D] failure occurs).

Metric	Good (Normal Operation)	Fair (Edge of Spec)	Poor (Hardware Failure)
Downstream Power (dBmV)	± 0 to ± 7.0	± 7.1 to ± 15.0 [I]	> ± 15.0 [D]
Upstream Power (dBmV)	35.0 to 50.0	30.0 to 34.9 and 50.1 to 55.0 [I]	< 30.0 or > 55.0 [D]
SNR / MER (dB)	≥ 33.0	25.0 to 32.9 [I]	< 25.0 [D]
DS Power Delta (dB) (Highest minus Lowest)	0 to 3.0	3.1 to 4.0 [I]	> 4.0 [I]
Error Rate (Uncorrectables per min)	≤ 10	11 to 50	> 50

💡 A quick note before you copy/paste: While this YAML looks a bit intimidating, it is entirely dynamic. If you paste this and get errors, check your entity names! This code uses the default sensor.cable_modem_* naming pattern. If your entities are named using a different pattern from setup, just do a quick Find/Replace in the code below to match your specific sensors.

Derivative Sensor to measure errors per minute

sensor:
    ## Cable Modem Uncorrectable Error Rate ##
  - platform: derivative
    source: sensor.cable_modem_total_uncorrected_errors
    name: Cable Modem Uncorrectable Error Rate
    unit_time: min
    time_window: "00:10:00" #Use larger window to smooth the per minute rate
    max_sub_interval: "00:01:00" # Keeps the sensor alive at 0.0 when no errors occur
    round: 1

Signal Health Sensor

template:
  - sensor:
        ## Cable Modem Signal Health ##
      - name: "Cable Modem Signal Health"
        icon: mdi:signal
        availability: >
          {{
            states('sensor.cable_modem_ds_channel_count') not in ['unknown', 'unavailable', 'none']
            and states('sensor.cable_modem_us_channel_count') not in ['unknown', 'unavailable', 'none']
          }}
        state: >
          {% set entities = device_entities(device_id('sensor.cable_modem_status')) %}
          {% set docsis = states('sensor.cable_modem_docsis_status') | lower %}
          {% set ds_power_sensors = expand(entities) | selectattr('entity_id', 'search', 'cable_modem_ds_.*_power$') | rejectattr('state', 'in', ['unknown', 'unavailable']) | list %}
          {% set ds_snr_sensors = expand(entities) | selectattr('entity_id', 'search', 'cable_modem_ds_.*_snr$') | rejectattr('state', 'in', ['unknown', 'unavailable']) | list %}
          {% set us_power_sensors = expand(entities) | selectattr('entity_id', 'search', 'cable_modem_us_.*_power$') | rejectattr('state', 'in', ['unknown', 'unavailable']) | list %}
          {% set ns = namespace(ds_abs_worst=0.0, ds_values=[], us_grade='good', us_deviation=-1.0, snr_min=none) %}

          {% for s in ds_power_sensors %}
            {% set lock_status = s.attributes.lock_status | default('', true) | string | lower %}
            {% if lock_status not in ['not_locked', 'partial_lock', 'unlocked'] %}
              {% set value = s.state | float(0) %}
              {% set ns.ds_values = ns.ds_values + [value] %}
              {% if value | abs > ns.ds_abs_worst %}
                {% set ns.ds_abs_worst = value | abs %}
              {% endif %}
            {% endif %}
          {% endfor %}

          {% for s in ds_snr_sensors %}
            {% set lock_status = s.attributes.lock_status | default('', true) | string | lower %}
            {% if lock_status not in ['not_locked', 'partial_lock', 'unlocked'] %}
              {% set value = s.state | float(0) %}
              {% if ns.snr_min is none or value < ns.snr_min %}
                {% set ns.snr_min = value %}
              {% endif %}
            {% endif %}
          {% endfor %}

          {% for s in us_power_sensors %}
            {% set lock_status = s.attributes.lock_status | default('', true) | string | lower %}
            {% if lock_status not in ['not_locked', 'partial_lock', 'unlocked'] %}
              {% set value = s.state | float(0) %}
              {% set grade = 'poor' if value < 30 or value > 55 else 'fair' if value < 35 or value > 50 else 'good' %}
              {% set deviation = 35 - value if value < 35 else value - 50 if value > 50 else 0 %}
              {% if grade == 'poor' and ns.us_grade != 'poor' %}
                {% set ns.us_grade = 'poor' %}
                {% set ns.us_deviation = deviation %}
              {% elif grade == 'poor' and deviation > ns.us_deviation %}
                {% set ns.us_deviation = deviation %}
              {% elif grade == 'fair' and ns.us_grade == 'good' %}
                {% set ns.us_grade = 'fair' %}
                {% set ns.us_deviation = deviation %}
              {% elif grade == 'fair' and ns.us_grade == 'fair' and deviation > ns.us_deviation %}
                {% set ns.us_deviation = deviation %}
              {% endif %}
            {% endif %}
          {% endfor %}

          {% set ds_power_grade = 'poor' if ns.ds_abs_worst > 15 else 'fair' if ns.ds_abs_worst > 7 else 'good' %}
          {% set ds_delta = (ns.ds_values | max - ns.ds_values | min) if ns.ds_values | count > 1 else 0 %}
          {% set ds_delta_grade = 'poor' if ds_delta > 4 else 'fair' if ds_delta > 3 else 'good' %}
          {% set snr_grade = 'good' if ns.snr_min is none else 'poor' if ns.snr_min < 25 else 'fair' if ns.snr_min < 33 else 'good' %}
          {% set err_rate_state = states('sensor.cable_modem_uncorrectable_error_rate') %}
          {% set err_rate = err_rate_state | float(0) %}
          {% set error_grade = 'good' if err_rate_state in ['unknown', 'unavailable', 'none'] else 'poor' if err_rate > 50 else 'fair' if err_rate > 10 else 'good' %}

          {% if docsis not in ['operational', 'locked'] %}
            poor
          {% elif 'poor' in [ds_power_grade, ns.us_grade, snr_grade, ds_delta_grade, error_grade] %}
            poor
          {% elif 'fair' in [ds_power_grade, ns.us_grade, snr_grade, ds_delta_grade, error_grade] %}
            fair
          {% else %}
            good
          {% endif %}

        attributes:
          health_data: >
            {% set entities = device_entities(device_id('sensor.cable_modem_status')) %}
            {% set docsis = states('sensor.cable_modem_docsis_status') | lower %}
            {% set ds_power_sensors = expand(entities) | selectattr('entity_id', 'search', 'cable_modem_ds_.*_power$') | rejectattr('state', 'in', ['unknown', 'unavailable']) | list %}
            {% set ds_snr_sensors = expand(entities) | selectattr('entity_id', 'search', 'cable_modem_ds_.*_snr$') | rejectattr('state', 'in', ['unknown', 'unavailable']) | list %}
            {% set us_power_sensors = expand(entities) | selectattr('entity_id', 'search', 'cable_modem_us_.*_power$') | rejectattr('state', 'in', ['unknown', 'unavailable']) | list %}
            {% set err_rate_state = states('sensor.cable_modem_uncorrectable_error_rate') %}
            {% set err_rate = err_rate_state | float(0) %}
            {% set ns = namespace(
              ds_values=[],
              ds_evaluated=0,
              us_evaluated=0,
              ds_power_observed=none,
              ds_power_channel=none,
              us_power_observed=none,
              us_power_channel=none,
              us_power_grade='good',
              us_power_deviation=-1.0,
              snr_observed=none,
              snr_channel=none,
              decisive_metric='none',
              governing_reason='good',
              final_grade='good',
              metrics={},
              non_good_metrics={},
              omitted_metrics=[]
            ) %}

            {% set docsis_metric = {
              'grade': 'good' if docsis in ['operational', 'locked'] else 'poor',
              'observed': docsis,
              'comparison': 'enum_status',
              'thresholds': {
                'good_values': ['operational', 'locked'],
                'poor_values': ['partial_lock', 'not_locked', 'unknown']
              }
            } %}
            {% set ns.metrics = dict(ns.metrics, **{'docsis_status': docsis_metric}) %}
            {% if docsis_metric.grade != 'good' %}
              {% set ns.non_good_metrics = dict(ns.non_good_metrics, **{'docsis_status': docsis_metric.grade}) %}
              {% set ns.final_grade = 'poor' %}
              {% set ns.decisive_metric = 'docsis_status' %}
              {% set ns.governing_reason = 'docsis_' ~ docsis %}
            {% endif %}

            {% for s in ds_power_sensors %}
              {% set lock_status = s.attributes.lock_status | default('', true) | string | lower %}
              {% if lock_status not in ['not_locked', 'partial_lock', 'unlocked'] %}
                {% set value = s.state | float(0) %}
                {% set ns.ds_values = ns.ds_values + [value] %}
                {% set ns.ds_evaluated = ns.ds_evaluated + 1 %}
                {% if ns.ds_power_observed is none or value | abs > ns.ds_power_observed %}
                  {% set ns.ds_power_observed = value | abs %}
                  {% set ns.ds_power_channel = s.entity_id | replace('sensor.cable_modem_', '') | replace('_power', '') %}
                {% endif %}
              {% endif %}
            {% endfor %}

            {% if ns.ds_evaluated > 0 %}
              {% set ds_power_grade = 'poor' if ns.ds_power_observed > 15 else 'fair' if ns.ds_power_observed > 7 else 'good' %}
              {% set ds_power_metric = {
                'grade': ds_power_grade,
                'observed': ns.ds_power_observed | round(1),
                'unit': 'dBmV',
                'worst_channel': ns.ds_power_channel,
                'comparison': 'absolute_distance_from_zero',
                'thresholds': {
                  'fair_max': 7.0,
                  'poor_max': 15.0
                }
              } %}
              {% set ns.metrics = dict(ns.metrics, **{'downstream_power': ds_power_metric}) %}
              {% if ds_power_grade != 'good' %}
                {% set ns.non_good_metrics = dict(ns.non_good_metrics, **{'downstream_power': ds_power_grade}) %}
                {% if ns.final_grade != 'poor' and ds_power_grade == 'poor' %}
                  {% set ns.final_grade = 'poor' %}
                  {% set ns.decisive_metric = 'downstream_power' %}
                  {% set ns.governing_reason = 'downstream_power' %}
                {% elif ns.final_grade == 'good' and ds_power_grade == 'fair' %}
                  {% set ns.final_grade = 'fair' %}
                  {% set ns.decisive_metric = 'downstream_power' %}
                  {% set ns.governing_reason = 'downstream_power' %}
                {% endif %}
              {% endif %}

              {% set ds_delta = (ns.ds_values | max - ns.ds_values | min) if ns.ds_values | count > 1 else 0 %}
              {% set ds_delta_grade = 'poor' if ds_delta > 4 else 'fair' if ds_delta > 3 else 'good' %}
              {% set ds_delta_metric = {
                'grade': ds_delta_grade,
                'observed': ds_delta | round(1),
                'unit': 'dB',
                'comparison': 'spread',
                'thresholds': {
                  'fair_max': 3.0,
                  'poor_max': 4.0
                }
              } %}
              {% set ns.metrics = dict(ns.metrics, **{'downstream_power_delta': ds_delta_metric}) %}
              {% if ds_delta_grade != 'good' %}
                {% set ns.non_good_metrics = dict(ns.non_good_metrics, **{'downstream_power_delta': ds_delta_grade}) %}
                {% if ns.final_grade != 'poor' and ds_delta_grade == 'poor' %}
                  {% set ns.final_grade = 'poor' %}
                  {% set ns.decisive_metric = 'downstream_power_delta' %}
                  {% set ns.governing_reason = 'downstream_power_delta' %}
                {% elif ns.final_grade == 'good' and ds_delta_grade == 'fair' %}
                  {% set ns.final_grade = 'fair' %}
                  {% set ns.decisive_metric = 'downstream_power_delta' %}
                  {% set ns.governing_reason = 'downstream_power_delta' %}
                {% endif %}
              {% endif %}
            {% else %}
              {% set ns.omitted_metrics = ns.omitted_metrics + ['downstream_power', 'downstream_power_delta'] %}
            {% endif %}

            {% for s in us_power_sensors %}
              {% set lock_status = s.attributes.lock_status | default('', true) | string | lower %}
              {% if lock_status not in ['not_locked', 'partial_lock', 'unlocked'] %}
                {% set value = s.state | float(0) %}
                {% set ns.us_evaluated = ns.us_evaluated + 1 %}
                {% set grade = 'poor' if value < 30 or value > 55 else 'fair' if value < 35 or value > 50 else 'good' %}
                {% set deviation = 35 - value if value < 35 else value - 50 if value > 50 else 0 %}
                {% if grade == 'poor' and ns.us_power_grade != 'poor' %}
                  {% set ns.us_power_grade = 'poor' %}
                  {% set ns.us_power_deviation = deviation %}
                  {% set ns.us_power_observed = value %}
                  {% set ns.us_power_channel = s.entity_id | replace('sensor.cable_modem_', '') | replace('_power', '') %}
                {% elif grade == 'poor' and deviation > ns.us_power_deviation %}
                  {% set ns.us_power_deviation = deviation %}
                  {% set ns.us_power_observed = value %}
                  {% set ns.us_power_channel = s.entity_id | replace('sensor.cable_modem_', '') | replace('_power', '') %}
                {% elif grade == 'fair' and ns.us_power_grade == 'good' %}
                  {% set ns.us_power_grade = 'fair' %}
                  {% set ns.us_power_deviation = deviation %}
                  {% set ns.us_power_observed = value %}
                  {% set ns.us_power_channel = s.entity_id | replace('sensor.cable_modem_', '') | replace('_power', '') %}
                {% elif grade == 'fair' and ns.us_power_grade == 'fair' and deviation > ns.us_power_deviation %}
                  {% set ns.us_power_deviation = deviation %}
                  {% set ns.us_power_observed = value %}
                  {% set ns.us_power_channel = s.entity_id | replace('sensor.cable_modem_', '') | replace('_power', '') %}
                {% elif grade == 'good' and ns.us_power_observed is none %}
                  {% set ns.us_power_observed = value %}
                  {% set ns.us_power_channel = s.entity_id | replace('sensor.cable_modem_', '') | replace('_power', '') %}
                {% endif %}
              {% endif %}
            {% endfor %}

            {% if ns.us_evaluated > 0 %}
              {% set us_power_metric = {
                'grade': ns.us_power_grade,
                'observed': ns.us_power_observed | round(1),
                'unit': 'dBmV',
                'worst_channel': ns.us_power_channel,
                'comparison': 'inclusive_range',
                'thresholds': {
                  'good_min': 35.0,
                  'good_max': 50.0,
                  'poor_min': 30.0,
                  'poor_max': 55.0
                }
              } %}
              {% set ns.metrics = dict(ns.metrics, **{'upstream_power': us_power_metric}) %}
              {% if ns.us_power_grade != 'good' %}
                {% set ns.non_good_metrics = dict(ns.non_good_metrics, **{'upstream_power': ns.us_power_grade}) %}
                {% set us_reason = 'upstream_power_low' if ns.us_power_observed < 35 else 'upstream_power_high' %}
                {% if ns.final_grade != 'poor' and ns.us_power_grade == 'poor' %}
                  {% set ns.final_grade = 'poor' %}
                  {% set ns.decisive_metric = 'upstream_power' %}
                  {% set ns.governing_reason = us_reason %}
                {% elif ns.final_grade == 'good' and ns.us_power_grade == 'fair' %}
                  {% set ns.final_grade = 'fair' %}
                  {% set ns.decisive_metric = 'upstream_power' %}
                  {% set ns.governing_reason = us_reason %}
                {% endif %}
              {% endif %}
            {% else %}
              {% set ns.omitted_metrics = ns.omitted_metrics + ['upstream_power'] %}
            {% endif %}

            {% for s in ds_snr_sensors %}
              {% set lock_status = s.attributes.lock_status | default('', true) | string | lower %}
              {% if lock_status not in ['not_locked', 'partial_lock', 'unlocked'] %}
                {% set value = s.state | float(0) %}
                {% if ns.snr_observed is none or value < ns.snr_observed %}
                  {% set ns.snr_observed = value %}
                  {% set ns.snr_channel = s.entity_id | replace('sensor.cable_modem_', '') | replace('_snr', '') %}
                {% endif %}
              {% endif %}
            {% endfor %}

            {% if ns.snr_observed is not none %}
              {% set snr_grade = 'poor' if ns.snr_observed < 25 else 'fair' if ns.snr_observed < 33 else 'good' %}
              {% set snr_metric = {
                'grade': snr_grade,
                'observed': ns.snr_observed | round(1),
                'unit': 'dB',
                'worst_channel': ns.snr_channel,
                'comparison': 'minimum_floor',
                'thresholds': {
                  'fair_min': 33.0,
                  'poor_min': 25.0
                }
              } %}
              {% set ns.metrics = dict(ns.metrics, **{'snr': snr_metric}) %}
              {% if snr_grade != 'good' %}
                {% set ns.non_good_metrics = dict(ns.non_good_metrics, **{'snr': snr_grade}) %}
                {% if ns.final_grade != 'poor' and snr_grade == 'poor' %}
                  {% set ns.final_grade = 'poor' %}
                  {% set ns.decisive_metric = 'snr' %}
                  {% set ns.governing_reason = 'snr' %}
                {% elif ns.final_grade == 'good' and snr_grade == 'fair' %}
                  {% set ns.final_grade = 'fair' %}
                  {% set ns.decisive_metric = 'snr' %}
                  {% set ns.governing_reason = 'snr' %}
                {% endif %}
              {% endif %}
            {% else %}
              {% set ns.omitted_metrics = ns.omitted_metrics + ['snr'] %}
            {% endif %}

            {% if err_rate_state not in ['unknown', 'unavailable', 'none'] %}
              {% set error_grade = 'poor' if err_rate > 50 else 'fair' if err_rate > 10 else 'good' %}
              {% set error_metric = {
                'grade': error_grade,
                'observed': err_rate | round(1),
                'unit': 'errors_per_min',
                'comparison': 'rate',
                'window_seconds': 600,
                'baseline_established': true,
                'provisional': false,
                'source': 'derivative_sensor',
                'thresholds': {
                  'fair_max': 10.0,
                  'poor_max': 50.0
                }
              } %}
              {% set ns.metrics = dict(ns.metrics, **{'error_rate': error_metric}) %}
              {% if error_grade != 'good' %}
                {% set ns.non_good_metrics = dict(ns.non_good_metrics, **{'error_rate': error_grade}) %}
                {% if ns.final_grade != 'poor' and error_grade == 'poor' %}
                  {% set ns.final_grade = 'poor' %}
                  {% set ns.decisive_metric = 'error_rate' %}
                  {% set ns.governing_reason = 'error_rate' %}
                {% elif ns.final_grade == 'good' and error_grade == 'fair' %}
                  {% set ns.final_grade = 'fair' %}
                  {% set ns.decisive_metric = 'error_rate' %}
                  {% set ns.governing_reason = 'error_rate' %}
                {% endif %}
              {% endif %}
            {% else %}
              {% set ns.omitted_metrics = ns.omitted_metrics + ['error_rate'] %}
            {% endif %}

            {% set payload = {
              'grade': ns.final_grade,
              'decisive_metric': ns.decisive_metric,
              'governing_reason': ns.governing_reason,
              'non_good_metrics': ns.non_good_metrics,
              'channel_counts': {
                'downstream_evaluated': ns.ds_evaluated,
                'upstream_evaluated': ns.us_evaluated
              },
              'metrics': ns.metrics,
              'omitted_metrics': ns.omitted_metrics,
              'evaluated_at': now().isoformat()
            } %}

            {{ payload | to_json | from_json }}

The Dashboard Setup: Custom Signal Health Card

To get the clean, color-coded dashboard look from my screenshots, you'll need to use a custom YAML card. Because the Cable Modem Monitor (CMM) integration packs so much detailed diagnostic data into the attributes of the sensor.cable_modem_signal_health entity, standard Home Assistant cards can't quite display it all dynamically.

Required Custom Cards (via HACS):
Before pasting the code, make sure you have installed these two excellent frontend plugins by thomasloven via HACS:

1. template-entity-row: This is the magic that allows us to use Jinja2 templates to dynamically change the icon colors (Green/Orange/Red) and extract the "worst channel" data directly into the secondary info text.
2. fold-entity-row: This keeps the dashboard clean by tucking the granular metrics (Downstream, Upstream, SNR, etc.) into a collapsible dropdown under the main "Signal Health" header.

How to add it:
Go to your Lovelace Dashboard, click Edit, click Add Card, scroll down to Manual, and paste the following YAML:

type: vertical-stack
title: Cable Modem
cards:
  - type: entities
    show_header_toggle: false
    entities:
      - type: custom:template-entity-row
        entity: sensor.cable_modem_status
        name: Modem Status
        icon: mdi:check-network-outline
        state: "{{ states('sensor.cable_modem_status') | title }}"
        secondary: "Uptime: {{ states('sensor.cable_modem_system_uptime') }}"
        color: >-
          {{ '#4CAF50' if states('sensor.cable_modem_status') | lower in ['online', 'operational', 'up'] else '#F44336' }}
      
      - type: custom:fold-entity-row
        head:
          type: custom:template-entity-row
          entity: sensor.cable_modem_signal_health
          name: Signal Health
          state: "{{ states('sensor.cable_modem_signal_health') | title }}"
          icon: >-
            {% set state = states('sensor.cable_modem_signal_health') %} 
            {% if state == 'good' %} mdi:signal-cellular-3
            {% elif state == 'fair' %} mdi:signal-cellular-2
            {% elif state == 'poor' %} mdi:signal-cellular-1
            {% else %} mdi:signal-cellular-outline {% endif %}
          color: >-
            {% set state = states('sensor.cable_modem_signal_health') %} 
            {% if state == 'good' %} #4CAF50
            {% elif state == 'fair' %} #FF9800
            {% elif state == 'poor' %} #F44336
            {% else %} #9E9E9E {% endif %}
          secondary: >-
            {% set hd = state_attr('sensor.cable_modem_signal_health', 'health_data') %} 
            {% if not hd %} Calculating...
            {% elif hd.governing_reason == 'error_rate' %} Error rate {{ hd.metrics.error_rate.observed | round(1) }}/min
            {% elif hd.governing_reason == 'snr' %} SNR {{ hd.metrics.snr.observed | round(1) }} dB on {{ hd.metrics.snr.worst_channel }}
            {% elif hd.governing_reason == 'downstream_power' %} DS power {{ hd.metrics.downstream_power.observed | round(1) }} dBmV on {{ hd.metrics.downstream_power.worst_channel }}
            {% elif hd.governing_reason == 'downstream_power_delta' %} DS delta {{ hd.metrics.downstream_power_delta.observed | round(1) }} dB
            {% elif hd.governing_reason in ['upstream_power_low', 'upstream_power_high'] %} US power {{ hd.metrics.upstream_power.observed | round(1) }} dBmV on {{ hd.metrics.upstream_power.worst_channel }}
            {% elif hd.governing_reason.startswith('docsis_') %} DOCSIS {{ hd.metrics.docsis_status.observed | replace('_', ' ') }}
            {% else %} {{ hd.decisive_metric | replace('_', ' ') | title }} {% endif %}
        
        entities:
          - type: custom:template-entity-row
            entity: sensor.cable_modem_signal_health
            name: Downstream Power Grade
            state: >-
              {% set hd = state_attr('sensor.cable_modem_signal_health', 'health_data') %} 
              {{ hd.metrics.downstream_power.grade | title if hd and hd.metrics is defined and hd.metrics.downstream_power is defined else 'Unknown' }}
            icon: >-
              {% set hd = state_attr('sensor.cable_modem_signal_health', 'health_data') %} 
              {% set metric = hd.metrics.downstream_power if hd and hd.metrics is defined and hd.metrics.downstream_power is defined else none %} 
              {% if not metric %} mdi:signal-cellular-outline 
              {% elif metric.grade == 'good' %} mdi:signal-cellular-3 
              {% elif metric.grade == 'fair' %} mdi:signal-cellular-2 
              {% elif metric.grade == 'poor' %} mdi:signal-cellular-1 
              {% else %} mdi:signal-cellular-outline {% endif %}
            color: >-
              {% set hd = state_attr('sensor.cable_modem_signal_health', 'health_data') %} 
              {% set metric = hd.metrics.downstream_power if hd and hd.metrics is defined and hd.metrics.downstream_power is defined else none %} 
              {% if not metric %} #9E9E9E 
              {% elif metric.grade == 'good' %} #4CAF50 
              {% elif metric.grade == 'fair' %} #FF9800 
              {% elif metric.grade == 'poor' %} #F44336 
              {% else %} #9E9E9E {% endif %}
            secondary: >-
              {% set hd = state_attr('sensor.cable_modem_signal_health', 'health_data') %} 
              {% if hd and hd.metrics is defined and hd.metrics.downstream_power is defined %}
              {{ hd.metrics.downstream_power.observed }} {{ hd.metrics.downstream_power.unit }} worst {{ hd.metrics.downstream_power.worst_channel }}
              {% else %}No data{% endif %}

          - type: custom:template-entity-row
            entity: sensor.cable_modem_signal_health
            name: Downstream Delta Grade
            state: >-
              {% set hd = state_attr('sensor.cable_modem_signal_health', 'health_data') %} 
              {{ hd.metrics.downstream_power_delta.grade | title if hd and hd.metrics is defined and hd.metrics.downstream_power_delta is defined else 'Unknown' }}
            icon: >-
              {% set hd = state_attr('sensor.cable_modem_signal_health', 'health_data') %} 
              {% set metric = hd.metrics.downstream_power_delta if hd and hd.metrics is defined and hd.metrics.downstream_power_delta is defined else none %} 
              {% if not metric %} mdi:signal-cellular-outline 
              {% elif metric.grade == 'good' %} mdi:signal-cellular-3 
              {% elif metric.grade == 'fair' %} mdi:signal-cellular-2 
              {% elif metric.grade == 'poor' %} mdi:signal-cellular-1 
              {% else %} mdi:signal-cellular-outline {% endif %}
            color: >-
              {% set hd = state_attr('sensor.cable_modem_signal_health', 'health_data') %} 
              {% set metric = hd.metrics.downstream_power_delta if hd and hd.metrics is defined and hd.metrics.downstream_power_delta is defined else none %} 
              {% if not metric %} #9E9E9E 
              {% elif metric.grade == 'good' %} #4CAF50 
              {% elif metric.grade == 'fair' %} #FF9800 
              {% elif metric.grade == 'poor' %} #F44336 
              {% else %} #9E9E9E {% endif %}
            secondary: >-
              {% set hd = state_attr('sensor.cable_modem_signal_health', 'health_data') %} 
              {% if hd and hd.metrics is defined and hd.metrics.downstream_power_delta is defined %}
              {{ hd.metrics.downstream_power_delta.observed }} {{ hd.metrics.downstream_power_delta.unit }} spread
              {% else %}No data{% endif %}

          - type: custom:template-entity-row
            entity: sensor.cable_modem_signal_health
            name: Upstream Power Grade
            state: >-
              {% set hd = state_attr('sensor.cable_modem_signal_health', 'health_data') %} 
              {{ hd.metrics.upstream_power.grade | title if hd and hd.metrics is defined and hd.metrics.upstream_power is defined else 'Unknown' }}
            icon: >-
              {% set hd = state_attr('sensor.cable_modem_signal_health', 'health_data') %} 
              {% set metric = hd.metrics.upstream_power if hd and hd.metrics is defined and hd.metrics.upstream_power is defined else none %} 
              {% if not metric %} mdi:signal-cellular-outline 
              {% elif metric.grade == 'good' %} mdi:signal-cellular-3 
              {% elif metric.grade == 'fair' %} mdi:signal-cellular-2 
              {% elif metric.grade == 'poor' %} mdi:signal-cellular-1 
              {% else %} mdi:signal-cellular-outline {% endif %}
            color: >-
              {% set hd = state_attr('sensor.cable_modem_signal_health', 'health_data') %} 
              {% set metric = hd.metrics.upstream_power if hd and hd.metrics is defined and hd.metrics.upstream_power is defined else none %} 
              {% if not metric %} #9E9E9E 
              {% elif metric.grade == 'good' %} #4CAF50 
              {% elif metric.grade == 'fair' %} #FF9800 
              {% elif metric.grade == 'poor' %} #F44336 
              {% else %} #9E9E9E {% endif %}
            secondary: >-
              {% set hd = state_attr('sensor.cable_modem_signal_health', 'health_data') %} 
              {% if hd and hd.metrics is defined and hd.metrics.upstream_power is defined %}
              {{ hd.metrics.upstream_power.observed }} {{ hd.metrics.upstream_power.unit }} worst {{ hd.metrics.upstream_power.worst_channel }}
              {% else %}No data{% endif %}

          - type: custom:template-entity-row
            entity: sensor.cable_modem_signal_health
            name: SNR Grade
            state: >-
              {% set hd = state_attr('sensor.cable_modem_signal_health', 'health_data') %} 
              {{ hd.metrics.snr.grade | title if hd and hd.metrics is defined and hd.metrics.snr is defined else 'Unknown' }}
            icon: >-
              {% set hd = state_attr('sensor.cable_modem_signal_health', 'health_data') %} 
              {% set metric = hd.metrics.snr if hd and hd.metrics is defined and hd.metrics.snr is defined else none %} 
              {% if not metric %} mdi:signal-cellular-outline 
              {% elif metric.grade == 'good' %} mdi:signal-cellular-3 
              {% elif metric.grade == 'fair' %} mdi:signal-cellular-2 
              {% elif metric.grade == 'poor' %} mdi:signal-cellular-1 
              {% else %} mdi:signal-cellular-outline {% endif %}
            color: >-
              {% set hd = state_attr('sensor.cable_modem_signal_health', 'health_data') %} 
              {% set metric = hd.metrics.snr if hd and hd.metrics is defined and hd.metrics.snr is defined else none %} 
              {% if not metric %} #9E9E9E 
              {% elif metric.grade == 'good' %} #4CAF50 
              {% elif metric.grade == 'fair' %} #FF9800 
              {% elif metric.grade == 'poor' %} #F44336 
              {% else %} #9E9E9E {% endif %}
            secondary: >-
              {% set hd = state_attr('sensor.cable_modem_signal_health', 'health_data') %} 
              {% if hd and hd.metrics is defined and hd.metrics.snr is defined %}
              {{ hd.metrics.snr.observed }} {{ hd.metrics.snr.unit }} worst {{ hd.metrics.snr.worst_channel }}
              {% else %}No data{% endif %}

          - type: custom:template-entity-row
            entity: sensor.cable_modem_signal_health
            name: Error Rate Grade
            state: >-
              {% set hd = state_attr('sensor.cable_modem_signal_health', 'health_data') %} 
              {{ hd.metrics.error_rate.grade | title if hd and hd.metrics is defined and hd.metrics.error_rate is defined else 'Unknown' }}
            icon: >-
              {% set hd = state_attr('sensor.cable_modem_signal_health', 'health_data') %} 
              {% set metric = hd.metrics.error_rate if hd and hd.metrics is defined and hd.metrics.error_rate is defined else none %} 
              {% if not metric %} mdi:signal-cellular-outline 
              {% elif metric.grade == 'good' %} mdi:signal-cellular-3 
              {% elif metric.grade == 'fair' %} mdi:signal-cellular-2 
              {% elif metric.grade == 'poor' %} mdi:signal-cellular-1 
              {% else %} mdi:signal-cellular-outline {% endif %}
            color: >-
              {% set hd = state_attr('sensor.cable_modem_signal_health', 'health_data') %} 
              {% set metric = hd.metrics.error_rate if hd and hd.metrics is defined and hd.metrics.error_rate is defined else none %} 
              {% if not metric %} #9E9E9E 
              {% elif metric.grade == 'good' %} #4CAF50 
              {% elif metric.grade == 'fair' %} #FF9800 
              {% elif metric.grade == 'poor' %} #F44336 
              {% else %} #9E9E9E {% endif %}
            secondary: >-
              {% set hd = state_attr('sensor.cable_modem_signal_health', 'health_data') %} 
              {% if hd and hd.metrics is defined and hd.metrics.error_rate is defined %}
              {{ hd.metrics.error_rate.observed }} errors/min
              {% else %}No data{% endif %}

Huge thank you for this integration! Having continuous, frequency-specific data finally gave me the proof I needed to fix an issue my ISP couldn't find, and proved that sometimes the interference really is coming from inside the house!

[sfennell]

@ccpk1 - Thanks for this! Exactly what drew me to this integration.

[kwschulz]

You guys are awesome and it's contributors like you who'll help elevate this project to the next level. There's a real community forming around this idea and it's really great!

Just wanted to point out, in case it helps readers looking for a simpler starting point, that the integration also ships a dashboard generator service that produces a baseline Lovelace YAML tailored to your modem's channel layout. Doesn't replace what Chad built here. Those serve different needs, but pair well together.

Two of many principles on this project are reducing developer friction and making the UX as smooth as possible. Granted I've had a few detours, but the goal remains. Chad, this makes me wonder if there's space for the dashboard generator to absorb some of what you've learned. The card patterns, the worst-channel callouts, the way you surface error rates over time. Even without baking aggregate health into core yet, there might be a contribution path that gets your diagnostic experience to other users without them having to rebuild it from scratch. Worth thinking about if it interests you.

[ccpk1]

Thanks Ken, I really appreciate that. I'm definitely interested in looking at how we can fold some of this into the dashboard generator to make it easier for people to use out of the box without having to rebuild it from scratch.

I've been thinking through the mechanics of executing this without baking aggregate health into the core yet. Keeping this logic strictly in the frontend (Lovelace/custom cards) comes with a few hard limits:

1. No History: Frontend template calculations don't go into the HA Recorder. We wouldn't be able to provide the history graphs of the "Signal Health" grade or the rolled-up SNR.
2. No Derivative Rates: Calculating things like "errors per minute" requires tracking state changes over time. We can't natively calculate that rate in a dashboard card without a backend sensor keeping track of the deltas.
3. No Automations: Users wouldn't be able to write automations against these diagnostics (e.g., "If aggregate Signal Health drops to Poor, send a notification") because that state would only exist in their browser when the dashboard is actively open.

Knowing those limits, I see two paths for how we could approach the dashboard generator:

• Option 1: The UI-Heavy Approach (Frontend Only)
  We lean heavily into UI-side Jinja and custom template cards to do the math on the fly. It gives users a fantastic real-time status view, but we'd lose the errors-per-minute metric, historical graphs, and automations. I'd need to test the limits to see how heavy the UI processing gets, but it could work for a great "quick glance" dashboard.
• Option 2: The "Bring Your Own Backend" Approach
  The dashboard generator service could output the Lovelace YAML plus a block of YAML containing the necessary template and derivative sensors. We just instruct the user to copy/paste those sensors into their own configuration.yaml. It adds a manual step for the user, but it gives them the fully functional, historical dashboard without adding bloat to the integration's core code.

I need to do a little more POC for option 1 to see what I haven't thought of. Let me know your thoughts, and if there's a pathway you think I'm missing. Thanks!