haven't considered which chart should i use

The first step is to determine whether you should or should not subgroup your data.

You stated that your process is continuous. To my understanding, that would mean a process similar to extrusion where there is a likelihood that parts are autocorrelated, meaning that measurements for part 2 depend on the results for part 1. If autocorrelation exists, do not subgroup your data but use an individual's (IMR) control chart. The frequency of parts selected should consider both the duration of the autocorrelation and the sources of variation in the process, such as material changes.

If you intended a different meaning by continuous process, such as the process is always running, you may consider forming rational subgroups of consecutive parts. The frequency of collecting subgroups should consider the sources of variation in the process, such as material changes or tooling adjustments.

 

Welcome! I think you may have a misconception about control charts. Almost all charts have only a single measurement resul per part. BUT many charts plot the average of several parts (with only 1 result per part). This is called subgrouping. As Miner said.

I also think that instead of a UCL LCL limits you are referring to USL USL (lower and upper specification limits) as true (statistical/SPC control limits for control charts are not calculated until a chart type is determined and the calculation can be performed using the subgroup average and range/Standard deviation as well a sthe appropriate factors based on the subgroup size.

I commend you for coming here to ask questions and would advise that you read a good book on SPC. Donald Wheeler has written a great starter book “Understanding Statistical Process Control”. Alternatively you can visit his website SPCPress.com to read free articles on control charts (and other statistical topics).

In the meantime you can plot your data in a simple run chart and explain what the process is and how the data are sampled. Then we can potentially provide more assistance on how to move forward.

 

There may be some language problems here but: YES you need to get training. Read Donald Wheeler as I mentioned in my previous post. This is the fastest and best training. SPC is NOT an easy thing - no one should attempt it without training and knowledge and understanding. Would let a high school student build a mile long bridge over a mile deep gorge? Would you let a first year college student perform heart surgery on your child? No you wouldn’t.

Again a couple of warnings: the data and the process tell you what chart to use. Every part has only ONE measurement per characteristic. Subgroups are formed by measuring several sequentially produced parts one time each. (There are exceptions to this but they are advanced knowledge) You do not form a subgroup by measuring one part many times. (Except when performing a MSA)

And this just scratches the surface of what you need to know.

The most important things to understand: what is the characteristic, how is it measured, what process creates the characteristic, what are the normal changes of the inputs (material change, set-up or lot change, operator change, different machines, allowed adjustments..) and when do they occur. There is no way we can answer any of your questions without knowing these things.

You can’t just find some data and try to do statistics on it…and hope to learn anything….

 

Without knowing more about your process, we cannot help you determine the most rational type of subgrouping. It is NOT measuring the same part multiple times. Depending on your process, it MIGHT be measuring 5 consecutive parts to form a subgroup then waiting a period of time determined by the known sources of longer-term variation, such as a material change, tooling change, setup, etc. I agree with Bev's recommendation for getting training, and Dr. Wheeler is the best source. Read his articles.

 

Output current of what? What is the process? Why are you measuring this ‘output’. What does it do?

We have asked many times now for what the proces is and ho it is sampled - yet you do not tell us. We cannot help unless you answer our questions. This is like paying whack a mole without a mole or a mallet…