Black Belt Information on the GBT Pulsar Spigot Card

The purpose of this document is to give 'black belt' observers a look into how the I.F. and backends are set-up during a spigot run. Under normal operation, this information should not be needed, and simply following the spigot start-up instructions ( should be sufficient to set up the IF path and take your observations.

The IF Path
The spigot card needs only two IF signals - one for each polarization. However, when the low speed samplers are in use the IF is configured for three signals to reach the spectrometer. This is to allow both of the samplers needed by the spigot card to be properly balanced (see the spectrometer section for more info). The CONFIGGUI also set up the IF so that the BCPM can be run simultaneously with the Spigot. Information on the I.F. path for the BCPM is included here, and info on running the BCPM can be found at

The basic I.F. path for the spigot is simple - you want to configure the I.F. to send a signal of the appropriate bandwidth down to the spectrometer. For high speed sampler modes, you just need to get polarization X to spectrometer sampler J1 and polarization Y to spectrometer sampler J4. For the low speed sampler modes, you actually want to send polarization X to samplers J9 & J17, and polarization Y to sampler J21.

To get the signal simultaneously to the BCPM, you need to route a X and Y polarization signals separately to the BCPM. This can be done through using any of the remaining paths from the converter module. There is one additional trick - the BCPM expects the signal to eventually get downconverted to 400 MHz, which is not the frequency the spectrometer wants. This means that in the converter rack you need to increase the frequency of the signal in the paths going to the BCPM by the difference between the BCPM signal and the spigot signal, as shown below.

Spigot Bandwidth Delta Freq
800 MHz 800 MHz
200 Mhz 500 MHz
50 Mhz 25 MHz
12.5 Mhz 68.75 MHz

The Spectrometer
CLEO and the CONFIGGUI are use to set the spectrometer up so that all desired samplers have a path to the sky and can be balanced. Once the samplers have been balanced, the spectrometer manager is turned off, and commands are sent from earth to load the appropriate Xilinx personalities and otherwise finish configuring the spectrometer for spigot use.

Double Nyquist Modes (200 MHz & 12.5 MHz)
The double Nyquist modes are a work-around to allow observers to have smaller bandpasses. The gist of these modes is that the by taking the data in double-Nyquist mode, you have effectively cut the bandwidth in half. However, if you let all that data into the spectrometer, you will get reflections from the ACF in your data. So instead, we then stick a filter on at the I.F. rack which results in power reaching the spectrometer from only half (again) the total bandwidth. This means that if you are running in 800 MHz double Nyquist mode (a.k.a. 200 MHz mode), you will see a a bandpass of 400 MHz, but with only 200 MHz of that data having power. The CONFIGGUI is smart enough that whatever center frequency the observer request, it puts that as the center frequency through the I.F. filters, resulting in the sky center frequency requested being the c.f. of the resultant 200 MHz which has power. However - the pulsar programs (e.g. all the TEMPO programs) want to know the _true_ center frequency of the band. For the case of the 200 MHz mode, if your bandpass is flipped (lower sideband), then the c.f. of the 400 MHz band is going to be 100MHz higher than the sky c.f. (If your resultant data is not flipped (upper sideband) then the sky c.f. is 100MHz lower than the sky c.f.).

So - is your data flipped or not?

A couple final thoughts

For more info, or to send correction/additions or questions, email Karen O'Neil (koneil)