Hi, maybe my answer will be useful:

when it comes to upgrading a basic SEM or TEM, the choice really depends on the research focus. Detectors are usually the most critical: in SEM this means EDS for elemental mapping and EBSD for crystallography, while in TEM it’s typically EDS or EELS, which provide access to chemical composition and electronic structure. The next priority is specialized holders and in-situ systems (heating, cooling, gas or liquid cells), as they allow you to study dynamic processes in real time and greatly expand the instrument’s capabilities. For biological and soft-matter samples, cryo systems combined with high-sensitivity cameras are essential.

Ion sources, FIB attachments, and gas injection systems offer unique possibilities for sample preparation and local modification, but they are usually more expensive and application-specific. In practice, the three upgrades most often considered essential are: an analytical detector (EDS/EBSD/EELS), in-situ holders tailored to your experiments, and a modern camera (for TEM) or an additional BSE/EBSD detector (for SEM). Everything else depends on whether you’re working with nanomaterials, biological systems, corrosion processes, or preparing TEM lamellae.

I concur with M. I. Sulatsky . It absolutely makes a difference what field you are in. I am in materials using an SEM.

I have told our clients that if someone offered me a free SEM, I would decline it unless it had both BSE and EDS. I look for various phases and both tools are important for that. If BSE shows a brightness difference, there is a compositional difference to be found. EDS will help to characterize the difference.

Segmented BSE is better than a single-element detector. It allows for shaded or topographic imaging.

I find BSE is terribly underused. I see plenty of secondary electron images in papers but relatively few backscattered electron images. The failure to use BSE can mislead researchers. (I had one client looking to characterize InP. Because they were using SE and ignoring the EDS signal, they were taking pictures of the carbon paint background and didn't know any better.)

I will also say that beam control is very important for EDS in order to do area analysis and mapping.

We do have heating stages, and they are useful but on rare occasion.

So far, we have gotten along without EBSD. There is another system on campus that has it. That lab also has FIB and gas injection capabilities for specialized research.