FRA using TopSPICE; Simulation Tips

[@RobertBolanos]

In this video, I show the FRA feature of TopSPICE and show how to initialize the capacitors and inductor in order to speed up the simulation and reduce simulation time. Expain how to set the FRA transient parameters and also to use the "CRTL" key to have the cursor follow your mouse to make measurement. Questions, comments can be sent to [e-mail]. "May your simulations always converge". [e-mail]

[@user-power_electronics]

Robert, Thank you for the detail explanation of using Top Spice FRA analysis.

[@RobertBolanos]

No problem Thanh

[@futuretech8871]

Great!! Thanks

[@merden102]

Hİ sir. Firstly, thanks for smps and other electronic tutorial videos. I have a problem. We use zero-pole cancellation method for smps feedback. When I research "pole-zero cancellation is theoretically okay, but practically unobtainable. Even if a perfect pole-zero cancellation were possible, the system would not be practically implementable because of internal stability issues" what would you say about this situation? Thanks sir. ctms.engin.umich.edu/CTMS/index.php?aux=Extras_PZ

[@RobertBolanos]

Hi Power_Man_24, the zero-pole cancellation method has been used for many years in SMPS. In fact you can use it to compensate most well behaved system. If you have the Sanjaya Maniktala book "Power supplies A-Z", he used the same technique. As long as the flyback is in DCM, you will not have the RHZ and use be able to use the zero-pole cancellation method. Most people design fkybacks using DCM because zero-pole cancelation is very easy. Now if you operate the flyback in CCM, you will have a RHZ that you have to deal with. Again you have to cancel the main pole from the output filter with a zero in your error amp. The key here is that you can not cancel the RHzero. Once you calculate the RHZero, you need to put the crossover at least one decade below the RHzero frequency. For example, lets say that the switching frequency is 100KHz, it is common to set the crossover frequency at 1/10 of the switching frequency. So in a regular DCM, the Fc(crossover) would be 10KHz. But in a CCM where a RHZ is at 30KHz, the RHZ and the FC are too close to each other. The solution is to set the Fc to at least one decade below the RHZ. This means that I would have to set the FC at 3KHz or lower. I would personally set it to 1ms, that way your transient reaction is 1ms which is still fast enough for most supplies. But the zero-pole is common used method although there are slight variations being used and in fact is a practical implementation. Not saying that it is not practical would be incorrect. Otherwise the power supplies that I have designed (MMS, IBEX, New Horizons), and shown below, for outer space would not be working. So it is a proven and very practical method. MMS Spacecraft kzbin.info/www/bejne/jnXVZ5KhpZJoqs0 IBEX Spacecraft kzbin.info/www/bejne/q4nJoHSPps16jKc New Horizen kzbin.info/www/bejne/onSsg2Nuib16e7M

[@merden102]

@@RobertBolanos Hi Robert Bolanos . Thank you very much for your revealing answer. The information you provided really helped me to understand the feedback design. Thank you very much, sir.

[@merden102]

I also have one more question. In DCM Mode fsw=100kHz but Fcrossover=200Hz is these values suitable?

[@RobertBolanos]

power_man_24 A crossover of 200Hz is achievable. You would have a phase margin of 90 degrees. The transient to a step function would react in 1/200Hz or 5 ms

[@merden102]

@@RobertBolanos I get the answer exactly as you said, sir. PM=90 Degrees GM=-82db ibb.co/fGPRx8K