Technique makes or breaks weighing. One thing to consider is absolute accuracy, i.e., My 30 grain charge is an accurate weight, maybe 30.00 grains +/- 0.01 grains, or my 30 grain charge is the same as last time, the same as last year, and the year before. Might be 30.1 grains, but it is the SAME.
Sameness makes groups.
The beam scale usually offers long term accuracy but is slow compared to dispensers.
Some of the cheaper digital reloading scales are just relabeled Chinese scales that have gone through a Quality Control sampling and are good when sold.
Good scales are not cheap. Since this is a Budget based thread, you guys using (relatively) expensive scales and dispensers won't be much help, but go ahead and post.
In addition to full scale accuracy (@ calibration weight) digital scales can have linearity errors, roll over accuracy, and auto zero capture range.
Additional errors can be temperature drift, short term electronic drift. The trick is to use technique to identify, monitor and correct for these error sources.
I use a super cheap scale but employ checks to monitor the accuracy @ the weight I am tossing.
For those on a BUDGET, technique if generally FREE.
First thing to do is determine how accurately you NEED to charge. The general rule of "+/- 0.1 grain" works, sometimes.
The "I weight to 0.002 grain" is likely a little ridiculous and not in the scope of BUDGET weighing.
Percentage of charge weight is a better metric. +/- 1% is fine for general plinking.
E.g.,
7 grains for a pistol load, 1% is a little better than +/- 0.1 grain.
30 grains for a small rifle cartridge, +/-1% is +/- 0.3 grains.
60 grains for a larger rifle cartridge, +/- 1% is +/- 0.6 grains.
Anything beyond 0.1% accuracy is borderline anal
When you read about load development in 1/2 grains steps, or 0.2 grain steps, +/- 0.6 grains just won't cut it.
If you are going for 0.2 grains, a scale that displays to 0.1 grain advertised, but only displays EVEN numbers just won't cut it.
For most cheap digital scales monitoring ZERO between each charge sometimes allows the scale electronics to capture anything close to zero as an automatic zero. Sounds good but scale drift can contribute to zero errors that will not be seen from charge to charge.
With a Zero/tare of your powder cup/pan each time you remove the cup you get a free check of the tare value. If that value changes, your charge may be wrong.
Calibrating a scale at full scale value might be convenient, but is does not tell you much about the linearity, or sensitivity at lower values.
A 500 grain digital scale is likely a 50 gram scale with a different label. Tossing 30 grain charges from a scale calibrated at 50 grams is putting a lot of trust in the scale. I suggest a check weight somewhere close to your charge. Check weights of suitable accuracy (0.1%) are cheap.
Not the really cheap ones you sometimes find on ebay, something just a little better.
A
Class M1 or better, 2 gram (30.86 grains) check weight has a tolerance, when NEW of 1.2 mg ( little better than 0.02 grains) and puts you well into the anal range for small rifle charges

A
5 gram M1 check weight wound be great for 50 to 60 grain charges. The cheaper Chrome plated magnetic steel weights are fine for reloading, with the non-magnetic Stainless better. Monitoring the negative pan value when removed, occasionally 'Tossing' the check weight instead of powder is a free way to monitor your charge process.
The cheap scale I use (Much less than $50) has a 50 gram full scale.
It reads to 0.02 grains but it's best resolution is in carats. Displays to the nearest 0.005 carats.
That's better than 0.02 grains (by a little).
Mine is very repeatable if I keep room temperature constant (between HVAC cycles) and monitor performance during a charging session.
I have no doubt believing my target charges are better than +/- 0.1 grain.
My vendor (in China) will not stand behind this scale that is a couple YEARS out of warranty.
Here's a short video of Charging on the CHEAP
EDIT, I just powered it on, let it sit for a minute and checked a 2 gram, 10 carat, M1 weight, without recalibration.
View attachment 7034031