In Reply to: Weigh counting tolerances posted by Patrick Rider on 03/16/01 at 1:03 PM:
This is a very common problem. As far as I am aware there are no standards regarding this issue.
The accuracy of counting scales varies greatly.
The internal resolution of a counting scale is what really determines its accuracy (1/1,000,000 is a very accurate counting scale).Unfortunately, highly accurate counting scales can cost over $1000 USD, which many end users are not willing to spend. Low cost scales(with internal resolution of 1/30,000 to 1/100,000) can cost approximately $600 USD and can be a viable alternative if the capacity is low enough to weigh very small parts.
Capacity of a scale is the maximum amount of weight that can be read by the scale without damaging it.
The consistancy of the manufactured parts is critical(especially when counting small parts). Inconsistant parts will yield inconsistant results.
The amount of parts sampled in order to calculate the average piece weight is also very important. I say, the more parts sampled the better. A sample of 100 parts is much better than a sample of 10 parts. The problem is that many times this is too time consuming to do.
You need to try to negotiate with your customer(I do understand the difficulty of this) on a mutually agreeable tolerance. Your scales will almost never agree.
If possible, discuss obtaining the same type of scales, calibrated by the same scale company at regular intervals.
Otherwise the only alternative is to hand count (and we all know that if you bring the human element into this process, your better off using a counting scale :)
Hope I have been some help, let me know if your have any more specific questions as I have really generalised this.