**T&M??**

# Online monitoring of resistive power load (Part 2)

**Keywords:MOSFET?
SMD resistors?
decoupling capacitors?
software-automation?
PWM?
**

**Interesting particular cases**

The generic model has a considerable number of parameters, allowing many different implementations. However, there are some particular parameter combinations and restrictions that give some interesting particular cases, on aspects such as easiness of adjustment and low component count. These particular cases are described next. They are ultimately defined by imposing a specific relationship between *Ut1* and *Ut2* at gains-calculation phase. All the discussion carried out in generic model concerning model adjustment and model performance applies to them. The effects of *U* and *Rp* variations that can occur during the monitoring will be discussed for each particular case.

*Ut1 = Ut2 = Ut*

This is maybe the most obvious particular case. Just choose an appropriate *Ut* value and insert it in gains formulae ((2), (3), (4), (5)), replacing *Ut1* and *Ut2* by *Ut*. The variation of *U* doesn´t affects the equality *Ut1* = *Ut2*. However, the variation of *Rp* destroys the equality *Ut1* = *Ut2*. Nevertheless, this particular case can be faced as a simple starting-point for model adjustment.

*Gu1 = Gu2 = Gu.*

The great advantage of this particular case is the reduction of the number of components necessary for model implementation, due to the elimination of one gain stage. Under this particular case, the voltages *Ut1* and *Ut2* must be chosen so that their ratio respects (6). Note that *Ut1* Ut2 and, since *Gu1* = *Gu2* = *Gu*, *Uu1* = *Uu2* = *Uu*. In (6), the *Rp* value should be the same that will be used for model-gains calculation.

Being *Ut1* and *Ut2* chosen, the model-gains can be calculated as described in the general model discussion. Note that *Gu* can be calculated either from (2) or (3), because these formulae give the same result. Regarding *U* and *Rp* variations, there are no special considerations beyond those already addressed in general model discussion.

*Gi1 = Gi2 = Gi*

The previous approach can be applied in a similar way to gain stages *Gi1* and *Gi2*, yielding the same advantages. Under this particular case, the voltages *Ut1* and *Ut2* must verify the ratio (7). Note that *Ut1* > *Ut2* and, since *Gi1* = *Gi2* = *Gi*, *Ui1* = *Ui2* = *Ui*.

The considerations for model adjustment and performance of the previous particular case apply completely to this one. Regarding *Gi* calculation, note that it can be calculated either from (4) or (5), because these formulae give the same result.

*Ud1 and Ud2 with equal slope moduli on δi and δs.*

In the particular cases presented so far, the modulus of the slope of *Ud1* on *δi* is different from that of *Ud2* on *δs*. As a consequence, the threshold-points are affected differently for a given comparator IOV. If desired, the model can be adjusted to avoid this situation, by equating the moduli of the slopes of *Ud1* and *Ud2* on the respective threshold-points. It can be done by imposing the ratio (8) between *Ut1* and *Ut2*. Note that *Ut1* Ut2.

Related Articles | Editor's Choice |

Visit Asia Webinars to learn about the latest in technology and get practical design tips.