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.
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.