In this section of the tutorial you create an initial inductor design for the Buck converter. Then you will perform post-processing in MDM after simulating the schematic to obtain a detailed inductor loss estimate.
In this topic:
This topic addresses the following key concepts:
In this topic, you will learn the following:
Remember, you want an inductor with an inductance in between 600 and 700 nH. To create an initial inductor design for the Buck converter, follow these steps:
Check the Write MDM generated model parameters to all lower model levels? checkbox.
You have now saved an MDM physical model of the inductor to the symbol L1. From this physical model, a Level 2 model for L1 has been extracted. To see this, double-click on L1. You should see the following:
You can see that the model status box is green and that model status is Ok: A Model Has Been Created For This Inductor.
Make sure that the checkbox Power probe calculates detailed losses is unchecked. Then click OK to close the dialog.
You can see now that the ESR of the Level 2 model generated by MDM is displayed on the schematic above L1. In order to measure losses, you must add a power probe, as is standard procedure in SIMPLIS when you want to measure the power loss in a component. To do so,
Next you will add the mean measurement to the power probe. This measurement will be performed after every simulation, with the mean value for the power dissipation output to the waveform viewer.
To add the mean measurement to the Power(L1) probe, follow these steps:
Now you will run the simulation using the power probe to calculate inductor losses based on the Level 2 model, but without using MDM for post-processing. This was previously the only way to calculate losses in SIMPLIS. In the previous section you added a mean value measurement to the power probe. In this section you will see that this measurement only measures the loss due to the RMS current in the inductor.
You will see that the power probe has measured an average loss in the inductor of 7.39mW.
However, it is very important to note that the ESR of the Level 2 model that is shown on the schematic is based on the DC resistance of the inductor winding ONLY. Therefore, the loss measurement which you just performed does NOT include core losses, AC winding losses, and losses due to proximity effect.
In order to obtain those detailed losses, you must run MDM post-processing after the circuit simulation.
Now you will run set up and run MDM post-processing after the circuit simulation to obtain a detailed calculation of inductor losses - including core losses, AC winding losses, and proximity effects, as well as the inductor temperature, none of which you can obtain by just using a regular SIMPLIS power probe as in the previous sub-section.
In post-processing, MDM calculates average inductor loss over a specified section of the circuit simulation waveform. An instantaneous power loss waveform, as given by the regular power probe, is not provided. Therefore, for the results to be meaningful, you want to make sure that MDM uses an integer multiple of the converter's steady-state cycle for the calculation. The more cycles are included, the longer the calculation will take, so the ideal number of cycles is one. Calculating average power loss over one steady-state cycle will give you the average power loss in the inductor when the converter is in steady-state.
Note that MDM assumes that you have set-up the post-processing to produce meaningful results. It will calculate the power over a quarter- or half-cycle, over 3.37 cycles or during converter startup if you so instruct it to. While average power loss in steady-state is what you are usually most interested in, if you run post-processing on a different type of waveform, it's up to you to make sense of the results properly.
For the inductor, MDM needs both the current and voltage waveform of the inductor. You do not need to necessarily plot these explicitly to the Waveform Viewer in SIMPLIS. In this example, you will notice that while the inductor current is plotted, the inductor voltage is not. MDM will automatically acquire the waveforms it needs regardless of what you plot to the Waveform Viewer.
To set up MDM post-processing for the inductor in the Buck converter,
By default, the MDM Results window is open to the Results Overview tab. To the left, it shows the initial inductance, a textual breakdown of the different types of losses in the inductor, and also the core and winding temperatures, as well as the boxed volume of the inductor. Note that the DC Bias Current Losses are given as 0.007W - the same as you previously measured with the power probe. However the actual total losses are much higher - 0.044W. You can see that MDM post-processing can give you a much better insight into what the inductor losses will be than you can obtain by just setting an ESR in the schematic.
You might ask yourself if you could not obtain the same results by just setting the ESR of L1, in a Level 1 model, to the required value to obtain a loss of 44mW. While this is of course possible, there are two limitations with this approach. First, this would not provide a detailed loss analysis which you can use to improve your inductor design. Second, you do not know this value of ESR ahead of time. While it is easy to calculate the DC winding resistance once a physical model of the inductor is created in MDM, the AC winding losses and the core losses depend on many factors that are operating-point dependent: frequency, flux density inside the core, temperature, the duty ratio of the waveforms, and DC current. Thus the "ESR" for each operating point would be different. MDM post-processing will give you accurate loss estimates for different operating points without the need to adjust the ESR of the inductor in the schematic.
There are some other important things to note:
Now you can analyze the MDM inductor results in more depth. You can see that the DC winding losses and core losses are quite low at about 7mW and 5mW, respectively. However the total winding losses are quite high, with proximity losses dominating at 19mW. To gain more insight into why proximity losses are so high, switch to the Losses By Winding tab of the MDM Results window:
Here the loss density of each turn (in W/cm3) of the winding is displayed. The more red the turn, the higher the losses inside it. Note that the scale to the right is relative and differs for each inductor and operating point. The most intensely red turn is the one with the most losses relative to the others - it does not necessarily mean that it's very hot.
You can see that the turns closest to the air gap have the highest loss density. This is the result of fringing flux field produced by the air gaps of the core. As the magnetic flux in the core travels across the air gap, it fringes and a field is generated inside the winding window which creates proximity losses in the winding. So despite using a large bifilar wire to reduce the winding resistance, the way the turns are arranged relative to the air gap produces significant winding losses.
Now switch to the Waveforms tab of the MDM Results window. Here the flux density inside the inductor core (in Tesla), the inductor current (in Amperes) and the inductor voltage (in Volts) are shown:
Note that the peak flux density is about 25mT. If you were to examine the N87+ material used for this inductor inside MagDB, you would see that the peak flux density of this material is given as 360mT. Also note that the peak current in the inductor is about 6.5A.
Now switch to the L vs. Current tab. Here the inductance of the inductor as a function of the current is shown:
The inductance starts to drop off only after 48A, with another significant drop after 75A, and the inductor saturates only above 105A. This is many times higher than the peak inductor current in this converter. You can conclude from this, as well as from the peak flux density, that this inductor is using a core which is too large for this application. Indeed, if you go back to the Results Overview tab, you will see that the boxed volume is 13.12 cm3. A 7A, 600nH inductor can be much smaller. Additionally, you can see by looking at the core and winding temperatures that the inductor is quite cool, heating up less than 3 degrees above the ambient temperature of 25 degrees.
Therefore, there is a lot of room to improve this inductor design. You will do this in the next section: 2.4 Refine the Inductor Design Using MDM
Minimize or close the MDM Results window. Save the schematic as 2_my_buck_E30_core.sxsch.
A complete schematic with the inductor design developed in this section, set up for MDM post-processing, is available as 2.3_SIMPLIS_MDM_tutorial_buck_converter_E30core.sxsch in the zip archive of schematic files: