The whole electronic scale thing is a technology in flux. I'll repeat here a few things I've encountered:
Battery power is the most immune to noise. Most people's AC line power has some noise on it from motors and fluorescent lamps and switching power supplies. This noise can cause a scale to drift. If you have an AC powered unit, it may turn out to be worth your time to get a filter for it. An old computer UPS usually filters pretty well. You can also buy EMI filters on line.
Most scales sold for reloading do not have Faraday shielding. This means RF interference may also bother them, even if they are running on batteries. Keep them away from transmitters of any kind and from fluorescent light fixtures for this reason. I have a nice little battery operated range pocket scale made by CED, but if I set it on my computer it drifts. You just can't get it that close to the RF racket.
Many scales are native gram scales with 0.01 gram resolution. That is the unit European load manuals use. It is equal to 0.1543236 grains. Native gram scales convert their results to grains, but because the unit is bigger than 0.1 grains, it will skip some digits in the 0.1 grain least significant digit place. You want a scale whose native unit is grains or whose gram resolution is higher so it doesn't skip digits in the conversion to grains.
Some scales have a settling detector that locks the reading down once it stops changing much. This makes for a stable looking reading, but makes it impossible to trickle charges because the reading stops following the weight at some point. You don't want that. Brian Enos sells scales that don't do that. Ask the maker of any scale you are looking at if it does that.
Battery operated scales often auto-shutdown after some number of seconds to save the batteries. Again, that can mess you up trickling. It can mess you up if you tare the scale, then get back on it too slowly with whatever you intended to weigh. You want a scale that either lets you defeat this feature or, better yet, lets you program it to not happen until several minutes later or however much time you think you might need. That way you can still save your batteries if you forget to turn the thing off, but won't have your readings messed up.
Some scales drift. That's the problem with my PACT. I found temperature to be a factor. The instructions even tell you to let it warm up for 20 minutes, which my battery scale doesn't need. I don't know what the issue is there, but I wound up buying an inexpensive 9"×12" Granite surface plate and a cover with a door for that thing to sit on so its temperature changed only slowly. I understand from others with the same scale that they don't have that issue. It's a QC issue, I suppose. Just be aware that some scales are like that and you can get a good copy or a bad one. Be prepared to make at least one return. I didn't do that fast enough with mine, so I am stuck with it for the matching dispenser (unless I want to pay for a repair). I have moved on to other scales for most purposes.
Some cheap scales rock. Read through the reviews on Amazon. Some inexpensive 20 gram (308 grain) maximum capacity scales with 0.001 gram resolution for under $30 have received great reviews. In one case a great review came from someone who knew how to test scales properly and did so with it and got results that compared favorably to those from a $5,000 scale, IIRC. I think that scale was about $25. Some are under $20, but I don't recall the reviews. Again, look for yourself. The point is, only a decade or so ago no scale that cheap had that kind of resolution. You don't have to pay a lot for it, but do keep in mind these are not actually lab scales. They are vulnerable to the interference issues I mentioned earlier. I don't know anything about their durability.
Actual lab scales: You can buy minimalist versions of these. I have one made by Acculab. It works pretty well, but I think I should have waited and paid twice as much or a little more for an auto-calibrating scale with with a Faraday shield and bigger wind screen box. I weigh things for other work that actually needs the higher resolution, even though reloading really does not. Still, it's nice to have the extra resolution at times as a sort of double-check.
Check weights: The check weights sold for reloading scales are class 6, IIRC. This means a 75 gram weight can be off by by about 0.1 grains. I have trimmed these weights up on a better scale just to be setting the scale up more accurately so it would drift off more slowly. If you know someone with access to a good lab scale, have them check your check weights. Many pharmacies have an annually calibrated higher resolution scale and can weigh your check weights for you. You can buy higher precision check weights, but they are pricey and ultimately unnecessary for reloading. I am happy just knowing what my check weights actually weigh.
One useful practice is, once you have identified an accurate load, to take a piece of stainless wire and trim it to match the charge weight on your scales as they are reading it at that time. Keep it in a closed container to keep airborne oil and dust off of it. When you want to replicated that load, use this check weight to verify how your scales read it today. If it's not the same, you have the necessary adjustment from reading it.
