[time-nuts] TADD-1 and Spectracom 8140T Phase Noise

[John Ackermann N8UR]

I had a chance today to take a TADD-1 and one of the Spectracom 8140T
amplifier modules over to a friend who has an HP 3048 phase noise
measurement system.  We got some very encouraging results.

First, as a point of reference the HP 5087A distribution amp is spec'd
at 5MHz to have phase noise less than -145dBc(1 Hz) >1kHz from the
carrier.

We did our tests at 10MHz, so the results are precisely comparable --
I'm not sure how residual phase noise like this would scale with input
frequency.

At 1kHz, the TADD-1 is about -142dBc.  That's its noise floor (at least
out to 2MHz), and it's better than -140dBc from 100Hz on out.  At 1Hz,
it's around -107dBc and at 10Hz, it's -130dBc.

The Spectracom 8140T performs similarly.  It's spec'd to be at -130dBc
at 1kHz, and it's almost exactly that. At 1Hz, it's at -115dBc, at 10Hz,
-122dBc, and at 100Hz, -125dBc.  Its noise floor out past 10kHz is a bit
lower than the TADD-1 at -145dBc.  Therefore, it's a bit better than the
TADD-1 at the extreme ends, but in the middle (between 10Hz and 10kHz
offsets), it's as much as 15dB worse.

It's not surprising that the Spectracom noise floor is a bit lower than
the TADD-1 since it has a 10MHz filter of some sort, while the TADD-1 is
broadband out to at least 40MHz.

Screenshots are at http://www.febo.com/time-freq/hardware/dist_amps/

John