Thanks for the manuals, guys, though I appreciate you sending them I have them all including one from the parent company. You’ll notice in the manuals that the CSA isn’t detailed. As Charles pointed out in a note to me, the ECG220 as a replacement could be the issue. There is a guy online that has a couple of the original MFE3002’s and after I replace both ECG220s at once to make sure both are good, if that doesn’t correct the problem I am going to order the originals. Failing that, I am going to sub in an LTC2057 with a higher current output using the 2n2219/2n2905 complementary pair that was used in the original. I’ll have to setup a nulling circuit for zero. This unit was very, very stable for it’s age and reference. It has a .01% KVD as well so I believe it is worth saving as a secondary standard. The Fluke 332D I picked up at the same time has a pretty amazing voltage reference varying less than 8uv over 24 hours with stretches of less than 5uv over 10 hours on a reference of 15V. It is running in a room that varies no more than 1 or 2 degrees F. So having these two get me closer than I’ve been for a while. The issue with the 332D is the power consumption. The DAS 46A was around 2PPM after stabilizing, they recommend it stay on 24X7, and I didn’t get a plot of its stability prior to the failure. On another note, I picked up a Keithley 616 electrometer the other day. Between the 332D live subframe and the high voltage in the 616 while measuring ohms, I have’t been shocked so many times since working as an electrician with my father in high school. He used to find the correct wire by running 110V through it. I still change light switches and receptacles live. I’ll report back after swapping the ECG220s. Thank you, Jerry