:: keithley 2015 THD. low noise modification and noise log

continuation of mod tinker tweak looky look (long nagging version of http://www.eevblog.com/forum/beginners/o-qn-on-funny-nplc-multislopenoise/60/)

overview of logs

main problems encountered, stability of thermal EMF, containment of 1uV spike noise, high sampling rate with long RS232 cable and reduction in self generated noise.

previous re-cap picture with notes

FLIR test run with 2N3904 TO92.

FLIR of original BJT in SOT23, showing excessive temperature. estimated junction temp with approx 100mW dissipation = 95C (Tja=357C/w). with power derating, this temp is way too near max operating envelop.

similar scenario here with the VREG regulation

SOT23 changed to TO92

replacement with TO92 BC5xx, however, these TO92 seem like just as sensitive to heat as SOT23.
DMM partial start up plot

input shorted at R113-AGND

front rear switch bypassed hard solder shorting.

before this, front/rear switching exhibits a difference of

(and yet again, manipulating picture position in blogger is a xxxking pain in the ass !! you hear me google?)

where do these 1uV spikes come from ?

PPM range vs NPLC summary