The current Booster RF modules were designed primarily for low power dissipation. The target use case was things like laser cooling where one wants a single-tone pulsed CW signal. For context, before we developed Booster, some of our labs had >1kW of Doppler cooling amps which took a chunk of our AC capacity up.
Booster works really nicely for this application (it's now the only thing we use for cooling/state-prep/etc AOMs at Oxford). However, there is a reason why Booster takes about a factor of 10 less power than a comparable output power amp from Mini-Circuits: it has significantly worse linearity. That's not necessarily an issue for applications like driving gates -- we have done high-fidelity quantum gates using Booster (the phase noise from Booster is incredibly low, which is nice) -- but it can be problematic in some cases.
The aim of this issue is: (a) to flag this aspect of Booster so potential users are not surprised (b) to discuss potential ways of increasing the linearity.
The modular design of Booster makes it really easy to swap out the RF stage with a new design (the bulk of the work was sorting out the mechanics, supplies, thermal management, diagnostics/protection, etc with the RF being relatively simple in comparison). So, if there is interest in increasing the linearity, we can either try upping the FET bias current, or designing a new output stage.
Before we do that, it would be good to get some data on the distortion in various stages of the RF path so we can decide on the appropriate resolution.
cc @gkasprow
The current Booster RF modules were designed primarily for low power dissipation. The target use case was things like laser cooling where one wants a single-tone pulsed CW signal. For context, before we developed Booster, some of our labs had >1kW of Doppler cooling amps which took a chunk of our AC capacity up.
Booster works really nicely for this application (it's now the only thing we use for cooling/state-prep/etc AOMs at Oxford). However, there is a reason why Booster takes about a factor of 10 less power than a comparable output power amp from Mini-Circuits: it has significantly worse linearity. That's not necessarily an issue for applications like driving gates -- we have done high-fidelity quantum gates using Booster (the phase noise from Booster is incredibly low, which is nice) -- but it can be problematic in some cases.
The aim of this issue is: (a) to flag this aspect of Booster so potential users are not surprised (b) to discuss potential ways of increasing the linearity.
The modular design of Booster makes it really easy to swap out the RF stage with a new design (the bulk of the work was sorting out the mechanics, supplies, thermal management, diagnostics/protection, etc with the RF being relatively simple in comparison). So, if there is interest in increasing the linearity, we can either try upping the FET bias current, or designing a new output stage.
Before we do that, it would be good to get some data on the distortion in various stages of the RF path so we can decide on the appropriate resolution.
cc @gkasprow