[time-nuts] Anybody want a Thunderbolt power supply?
nsayer at kfu.com
Wed Aug 31 01:56:26 EDT 2016
> On Aug 30, 2016, at 9:26 PM, Mark Sims <holrum at hotmail.com> wrote:
> The -12V is used to drive the RS-232 signals... it also eventually gets to the EFC dac so it can swing below ground. Also, the +12V gets to the DAC. Pay close attention to generating noise on these lines.
I’ve designed a prototype. I’ll provide noise and ripple measurements when I get the first boards back. I’m going to shoot for <= 35 mV P-P. If you need better than that, then it probably turns into a hybrid switching+linear system, which would be more expensive. It’s been a while since I looked, but I believe that’s what I was able to achieve with the supplies for the FE-5680A GPS discipline board.
> Also, when I did Lady Heather's temperature control PID (with great help from Warren Sarkison) I did some measurements on the effects of temperature on the system with only the tbolt under temperature control and also with both the power supply (the brick supplied with the original TAPR orders) and the tbolt temperature stabilized. It seems about 1/4 to 1/3 of the temperature sensitivity of the system could be attributed to the power supply and the rest came from the oscillator/tbolt. Some attention to the tempco of the power supply would be a good thing.
I’ve not made any efforts along those lines to start with. For one, the input is +12 V and is simply passed through as the +12 volt output (there is a TVS and cap on the input, however). Given that the Thunderbolt that I have on loan came with what I can only guess was once a laptop power supply, I started off thinking the bar was pretty low.
If you want the hybrid approach, then adding the temperature compensation to the linear regulators would clearly be the way to go.
More information about the time-nuts