MAX31855 thermocouple (K) temperature way out - doing my head in.

i working on max31855 based system, , can't right temperature.

i have k probe connecting k socket 1½" board, 0.5" of traces board entry point chip.  chip powered 3.3v, , decoupled, etc.

the temperature probe reads way off real temperature.

i have dmm temperature probe.  @ room temperature (25° today) both read same value.

when max31855 reads 100° dmm reads 147°.  when max31855 reads 150° dmm reads 244°.  have melted solder paste (reflow temperature ~220°) confirm dmm right temperature, , max3155 wrong.

things have tried:

if swap probes (they both k) same results.
as wires link k socket pcb, have used both normal 25awg , proper k thermocouple wire (offcuts old thermocouple).
the chip mounted underside of board 10mm enclosure.  enclosure has ventilation directly below chip, stays @ room temperature (this later addition see if helped problem, no avail).

a couple of degrees difference can tolerate. 100° difference taking wet.

i have scoured data sheet till it's made eyes bleed.  have taken other people's example code (lady ada, rocketscream, etc) , checked , double checked calculations, , it's fine (shift right 18, mask, sign extend, convert float, multiply 0.25 - mean doesn't simpler).  analysing data stream chip on 'scope shows temperature it's reporting program calculating.

so going wrong?!

i don't ask help, ... help...?

seems cold-junction compensation gone wrong when temperature valid if @ room temperature.
do use breakout board or chip ?
can upload photo how thermocouple connected chip ?

Arduino Forum > Using Arduino > Sensors > MAX31855 thermocouple (K) temperature way out - doing my head in.

arduino