Present criteria :: device is regarded as balanced if power is within NN db of the target.

Present criteria :: device is regarded as balanced if power is within NN dB of the target.

• Prime focus receivers -- 0.8 db (i.e., 20%)
• IF Rack -- 0.5 db

• Prime Focus Receivers -- 0.8 dB (i.e., 20%)
• IF Rack -- 0.5 dB

• Spectrometer -- 0.5 db
• Spectral Processor -- 2 db

• Spectrometer -- 0.5 dB
• Spectral Processor -- 2 dB

This is where we want input from the staff. Do we want different criteria when RFI is present?

This is where we want input from the staff. Do we want different criteria when RFI is present?

In the following scheme, the criterion depends on the backend that is in use. Here are some suggestions, dependent on backend.

In the following scheme, the criterion depends on the backend that is in use. Here are some suggestions, dependent on backend.

(we do not propose to implement RFI rejection for the Spectral Processor because of the difficulties inherent in the SP.)

(We do not propose to implement RFI rejection for the Spectral Processor because of the inherent difficulties.)

These numbers are not based on any rigorous analysis. The VLBA and BCPM have their own leveling mechanisms, so if their input levels are within 2-3 db of the ideal, they are just fine. The Spectral Processor has a high dynamic range, so the level is not very critical.

These numbers are not based on any rigorous analysis. The VLBA and BCPM have their own leveling mechanisms. So if their input levels are within 2-3 dB of the ideal, they are just fine. The Spectral Processor has a high dynamic range, so the level is not very critical.

One problem with the dependence on back end is that the managers for the prime focus, IF Rack, and (to be written) converter rack do not know what the back end is. They would have to have an additional parameter giving them the db tolerance criterion.

One problem with the dependence on backend is that the Managers for the Prime Focus, IF Rack, and (to be written) Converter Rack do not know what the backend is. They would have to have an additional parameter giving them the dB tolerance criterion.

This psuedo-MR is part of a series of MRs aimed at creating a production Balancing API in Python. For more information regarding this effort, please refer to the BalancingMetaMR.

back to balancing overview

This isn't really an MR, but it is being used to document the balancing criteria for the various back ends.

This isn't really an MR, but it is being used to document the default balancing criteria for the various back ends.

%META:TOPICMOVED{by="FrankGhigo" date="1149261355" from="Software.ModificationRequest11C506" to="Software.BalanceCriteria"}%

%META:TOPICINFO{author="FrankGhigo" date="1148579220" format="1.0" version="1.1"}% %META:TOPICPARENT{name="Main.MyObservingIssues"}%

Balancing criteria

1. Introduction

2. Background

• Prime focus receivers -- 0.8 db (i.e., 20%)
• IF Rack -- 0.5 db
• Converter Rack -- not implemented yet
• Spectrometer -- 0.5 db
• Spectral Processor -- 2 db

3. Requirements

In the following scheme, the criterion depends on the backend that is in use. Here are some suggestions, dependent on backend.

(no idea as yet what to use for GASP or CGSR2)

(we do not propose to implement RFI rejection for the Spectral Processor because of the difficulties inherent in the SP.)

These numbers are not based on any rigorous analysis. The VLBA and BCPM have their own leveling mechanisms, so if their input levels are within 2-3 db of the ideal, they are just fine. The Spectral Processor has a high dynamic range, so the level is not very critical.

One problem with the dependence on back end is that the managers for the prime focus, IF Rack, and (to be written) converter rack do not know what the back end is. They would have to have an additional parameter giving them the db tolerance criterion.

We could just adopt criteria for the device with the most stringent requirements, i.e., the Spectrometer/Spigot, and adopt them for all devices, thus eliminating the need for knowing the backend.

-- FrankGhigo - 25 May 2006