[time-nuts] Resistors in VCO

[Garry Thorp]

Beware of 1/f noise in resistors with DC through them. The attachment shows the close-in phase noise of a low-noise 100MHz oscillator tuned by the circuit shown, with both metal film (Panasonic ERA3A) and thick film (Multicomp) resistors in R1 and R2 positions. Phase noise was measured with the resistors at ambient and +85°C; the oscillator was at ambient temperature all the time.

Beyond ~100Hz offset the metal film measurements are limited somewhat by the test set (Agilent E5052B / 100 correlations), but still clearly show the effect of the thick film resistors.

The oscillator's tune sensitivity was ~215Hz/V at 5V. The phase noise measurements give resistor noise voltage density (at 1Hz) in the region of 2uV/rtHz at 25°C and 3uV/rtHz at 85°C. (Sqrt(4kTR) predicts ~9nV/rtHz white noise for 4.55k.)

There is a nice paper on resistor current noise at: https://dcc-llo.ligo.org/public/0002/T0900200/001/current_noise.pdf

Garry Thorp


Tom Knox wrote: Hi everyone, some VC oscillators I have played with keep trimmer resistor in circuit to adjust the EFC close enough for the PLL to work. I was wondering if changing to more stable resistors could improve ADEV and P/N performance. Also could this add stability if used to divide the loop voltage. It seems foil resistors are exceptional in many ways. Do most high end oscillators already use foil?
-------------- next part --------------
A non-text attachment was scrubbed...
Name: phase noise due to resistor flicker noise.pdf
Type: application/pdf
Size: 55568 bytes
Desc: phase noise due to resistor flicker noise.pdf
URL: <http://www.febo.com/pipermail/time-nuts/attachments/20140217/ba72a768/attachment-0001.pdf>