| T O P I C R E V I E W |
| seh-a |
Posted - 28 Sep 2012 : 17:01:39
I am trying to apply different rules on the same chart and it doesn't seem to work. I am using SPCBatchVariableControlChart and I set UCL, LCL, USL, and LSL to the chart and everything works fine. Now, I want add WE rules in on top and when I do that, the normal rule disappears. I am wondering if it is possible to have multiple rules on the same chart? |
| 3 L A T E S T R E P L I E S (Newest First) |
| quinncurtis |
Posted - 25 Jan 2013 : 15:41:22
Sorry, but we cannot follow what you describe. We specifically state that control limits (UCL and LCL) are NOT tied to any spec limits (LSL and USL). But what you state you understand is exactly the opposite.
Spec limits LSL and USL) are NOT control limits. Control limits are tied to a sigma value, as explained in the previous post. Spec limits have an entirely different origin and are not associated with sigma limits, and cannot be auto-calculated. But you keep referencing them as if that are just another set of control limits.
I suggest that you wait until you have the current revision, which you seem to be ordering. Rather than confuse things further. Then we can reference the extensive new control limit routines only found in the new revision.
Once you get the new revision, you specify in the setup of the chart that you are using Western Electric Rules. Then you can auto-calculated the WE control limits, or you can specify the target mean and sigma of the process - and the WE control limits (+-1,2,3 sigma and target) will be calculated from that. What you shouldn't be doing is specifically setting the value of a specific WE control limit.
Spec limits are independent of the WE control limits. They should only be used for misc. limits NOT associated with the WE Rule limits.
|
| seh-a |
Posted - 25 Jan 2013 : 13:58:26
I am still very confused about this control limit business. BTW, I am using Rev 2.0 (soon to be upgraded to 2.2) What we have used in the past are two sets of control limits: (USL and LSL) and (UCL and LCL) let's say the USL is 20, LSL is 0, UCL is 15, and LCL is 5 (the control limits are tighter than the spec. limits) From what you wrote below, I can not specify UCL and LCL because they are calculated based on USL and LSL values. The target value will be (USL + LSL) / 2 and avg and sigma are calculated based on that. Is that right? Is there a way to specify a different target? Let's say a target of 7.5 (instead of 10 as calculated)
Another scenario, I only use WE rules so I specify USL and LSL but the data are actual much lower than (USL + LSL) / 2 so I got all kinds of WE rules violation because they are way off average though they are perfectly normal. How do I solve this problem? |
| quinncurtis |
Posted - 03 Oct 2012 : 18:25:24
You can have multiple control rules, but not the way you describe.
First, the USL and LSL spec limits are not SPC control limits. They are not tied to a sigma level, and are not auto-calculated. So they are treated separately.
The UCL and LCL SPC control limits (+-3 Sigma) are exactly the same as the WE rule #1. They cannot be different. So when the WE rules are created, they use the limits assigned to UCL and LCL. The WE rules also add rules at the +-2 sigma and +-1 sigma levels.
So if you want the WE rules, and the spec limits, create the WE rules, the add the sigma limits. If there is some other limit you need, please describe in detail exactly what it is, and the basis for it.
The current revision (2.2) of the software has a large number of additional ways to add custom rules to a chart. See the online manual, Chapter 8 - Named and Custom Rule Sets: http://www.quinn-curtis.com/QCSPCChartNetManual.pdf
In any response please specify what revision of the software you are using, i.e. Rev. 2.0, 2.1 or 2.2.
|
|
|
