Power sensor: 8481D or 8484A

[HalFoster]

I have one of each of these, both with a current calibration but only need one and am trying to decide which one to keep.  The only practical differences I'm seeing is in the maximum average power (8481D: 100mW, 8484A: 200mW) and the worst case linearity from 1uW to 10uW (8481D: 2%, 8484A: 1%) Both of the units are fairly comparable in the CF values.

At first glance the 8484A would be better but only slightly.  Does anyone know of factors that I may be missing or have advice?

Thank you,

Hal

[M0HZH]

There are two versions of the 8484A:
1. the original larger one in black metal casing, with more thermal drift (see attached picture)
2. the revised one, where thermal drift has been fixed so the double-insulated large case has been replaced with the standard (smaller) one

If you have the first one, get read of it. If you have the second one keep that, I think it's slightly newer (and better) than the 8481D.

[HalFoster]

I had no clue that there was a later revision of the 8484A - I actually have two and both are of the evidently earlier series in the larger black cases.  I'm not really concerned with any drift as it will only be used on my home bench with a fairly stable temperature - and I tend to re-zero often as it is so that takes care of any minor drift.  I've been thinking about it and I've pretty much decided to keep one of each, along with a 8482A and just use attenuators as needed.  By keeping both, I have a spare in the event of the probably "Oops" moment ;-) 

Thanks,

Hal