Characteristics of 9 NJ132L JFET dies measured by D. Dowell, 1999 March


The sample: 9 dies, dated 3-2-93, lot #011402-2. There was no pre-selection before mounting them in the carrier. Two devices were unacceptable for any instrument. The other 7 were good enough for SHARC2.

Voltage noise

Here are all of the noise spectra at T = 115 K, Vd = 5V, Ids = 24 uA, T = 115 K. Note that I am limited by the preamp noise in making precise measurements of the JFET noise.
  • JFET #1 en JFET#1
  • JFET #2 en JFET#2
  • JFET #3 en JFET#3
  • JFET #4 en JFET#4 BAD!!!
  • JFET #5 en JFET#5
  • JFET #6 en JFET#6
  • JFET #7 en JFET#7
  • JFET #8 en JFET#8
  • JFET #9 en JFET#9 BAD!!!

    For the 7 decent FETs, the maximum JFET noise (with the preamp contribution subtracted in quadrature) is approximately 10 nV/sqrt(Hz) at 3 Hz and 6 nV/sqrt(Hz) at 30 Hz (JFETs #2 and #8).

    Transconductance

    Instead of the transconductance g, I give the output impedance Z of the JFETs. They are reciprocals: Z = 1/g.

    The two very noisy JFETs had an output impedance at room temperature which was far in excess of the other 7. The noisy JFETs had Z of 3470 and 3550 ohms, while the quiet ones had Z of 1160-1350 ohms. JFETs with a high value of Z at room temperature should probably be discarded. From now on, only the properties of the 7 quiet JFETs are discussed.

    The Z at 110 K was somewhat lower and correlated with the Z at room temperature:
    JFET Z

    Source voltages

    The sample of 7 FETs had a range of 0.46V in the source voltages at 110 K, which is about a factor of 10 greater than allowable for SHARC2. The cold source voltage is tightly correlated with the room temperature source voltage:
    JFET offsets
    Last modified Monday, 17-May-1999 06:02 PDT
    cdd@socrates.caltech.edu