|Thread Links||Date Links|
|Thread Prev||Thread Next||Thread Index||Date Prev||Date Next||Date Index|
I believe that any specification technique and test technique needs to accommodate the parameters I have suggested either directly or indirectly. For the PSE at least the test method for the parameters I think is fairly straight forward because we can vary the load. The PD is more interesting but still needs to have a set of parameters that can be measured and the most obvious knob we have for testing is changing voltage at the PI.
I think that Norfolk is a good venue to compare and contrast various specification techniques and propose conditions for testing them as time is short and I am on travel this week.
Some first comments:
2nd bullet. Not clear why you mention SELV etc.?
- Try to specify unbalance with absolute values so positive or negative voltage differences is not important.
-The fact that lower or higher V2 that can compensate higher/lower R2 is true however the test should address it inherently to simplify the test
-To cover both points try to define mathematical desired _expression_ and then design what the test setup need to be to test it. For example; I would start to play with the following: RUNB=dV/dI. What is need to be done so for a given requirement of worst case RUNB limit, the dV will not exceed di or equivalent method.
-I would try to replace the V1 with some fixed resistive load that represents the current during normal operation so P2PRUNB between ANY two pairs can be extracted . (P2PRUNB is defined between any two pairs).
-No need to test pair unbalance. No value added. Its effect on P2PRUNB is negligible.
Here is my proposed way to create balance specifications at the PSE and PI. I welcome any comments.