Want dynamic tx-eq tuning

We have a static set of default values that we ship with Dendrite for determining what the transmit equalization settings should be for a given transceiver. However, experience has shown that these do not always work.

For transceivers that do not have explicit tx-eq settings override from the operator, we should use these values as initial conditions rather than a static specification. When a link does not come up due to tx loss of lock or tx loss of signal, we should have an algorithm to tune the link dynamically. Ideally this could be a feedback-based control if we can get said feedback from the transceiver module. We had an issue recently where I manually did this based on intuition on how the tx-eq circuit works and the fact that we were seeing loss of lock. But to the extent possible this should be automatic based FIR circuit models and feedback based observations.

Start with dpd/misc/xcvr_defaults.csv and trace how Dendrite uses these transceiver transmit-equalization defaults. Determine whether the transceiver exposes feedback for tx loss of lock or signal, then define and validate a dynamic tuning approach that preserves explicit operator overrides and replaces static values with initial conditions.