LTC3108-1

I'm having a similar problem, though in my case the LTC3108 is running, but the output voltage isn't close to the datasheet value. 

I'm using this dev board:

https://www.amazon.com/dp/B076H28XRY?psc=1&ref=ppx_yo2ov_dt_b_product_details

I'm using a Coilcraft LPR6235-752RMRC.  With VS1 = GND and VS2 = VAUX the datasheet output voltage specification is 4.018 to 4.182V.  I'm seeing a stable 3.71V with no load and 1.5ma of short circuit current with a delta-T of 50C across my TEG.  Rising temperature delta raises the current but the output voltage stays the same. This part isn't useable with any lithium chemistry battery because VOUT is so far off spec.  Is this a fault in the LTC3108 design or an error in the datasheet?

I noticed that there are several forum posts for the LTC3108 going back 5 years with no replies.  Is this forum monitored?  Has anyone  been able to use the LTC3108 in a working design?

qis,

I think I know why your output voltage is near zero.  I did a retest and accidentally reversed the thermal flow, and the output was then ~20mV just like yours.  TEG modules are standardized regarding the orientation of the junctions.  One must apply the cold surface to the side of the TEG which has the identification lettering.  The hot side is the surface without lettering.

FYI, with the test set described in my previous message with VS1=VS2=VAUX I'm getting about 4.5V when I should be getting 5V.  With VS1=GND and VS2=VAUX I'm getting 3.7V instead of 4.1V.  Either way it's about 0.5V off.  The LTC3108 system is working well enough for me to proceed.  Either the datasheet is simply wrong or there's something which isn't clear in the datasheet.  The energy conversion works, it's just grossly out of spec.

Steve