Robert - Thank you for posting this. This is exactly what I do (but without the Bode 100). I watch your videos and can immediately apply it to what I do for work.
Hi Robert, I've been working in the industry for around 3 years now (extra low voltage, embedded systems designs mostly) and would really appreciate your insight into something that's plagued me for a little while. The power supplies I incorporate into my designs are almost always non-adjustable output voltage targeting a well known load. I usually expedite designs of power supplies with tools like TI WEBENCH, or by simply following the design guidelines in the manufacturer datasheets and (if available) associated app notes for the IC that I boil down to. I've never once tested stability of the supply (or even been urged to by my peers or superiors) beyond testing with basic adjustable load equipment before I link it in to the system as a whole. This bothers me, because I see so much information out there from experienced engineers such as yourself and Steve Sandler that propose methods for far more detailed analysis of power supplies. I'm becoming more and more curious about hidden problems in my past and current designs that I just haven't been looking for because we've been following the paradigm of "look, it's working. No more info needed". Are these analysis methods more targeted towards power supply designs intended for use with an unknown load? Or do you think it's beneficial to add these tests for all power supplies (or atleast all major rails) in a system such as that which I described above?
Hi Brett. Thank you for your question. I work in the SPACE community where the requirements are very rigorous. We get our requirements from both NASA and ESA (European Space Agency) and make them into a set of specifications and they always require that we measure both the Gain and Phase Margins over the operating input voltage range and load range. This is done to ensure that the supply will stay stable. In addition, we also do a WCA (Worst Cast Analysis) to estimate how the supply will perform as the parts change due to aging, radiation, and temperature changes. If your supply is critical, then I would recommend doing these measurements. Keep in mind that the output filter and the Error Amp are performed to verify your design while the total loop is the actual measurement that the customer is really interested in having. So in my opinion, I would recommend that you do these types of measurements because it gives you and your customers confidence that your supply will perform and remain stable. I hope this makes sense. If you have any questions, post them here are you can email them to rbola35618@aol.com. Best regards, Robert Bolanos
Typically the signal from the OPTO goes to the FB input of the PWM. So the comp pin would be the pin that feeds the FB of the PWM. I hope that makes sense.
Robert - Thank you for posting this. This is exactly what I do (but without the Bode 100). I watch your videos and can immediately apply it to what I do for work.
Thank you Tim for the comment!
Nice analysis Robert, greetings from Mexico.
Saludo’s a mis paisanos en Mexico!
Love your spelling of PI. (PIE - hehe)
very good information. thanks a lot
Thanks for the comments. Robert
Long awaited
Hi pzil er. Finally felt good enough and had some free time to do a video. Nice hearing from you!
very helpful. Thank you
Thank you for the comments!
Hi Robert,
I've been working in the industry for around 3 years now (extra low voltage, embedded systems designs mostly) and would really appreciate your insight into something that's plagued me for a little while. The power supplies I incorporate into my designs are almost always non-adjustable output voltage targeting a well known load. I usually expedite designs of power supplies with tools like TI WEBENCH, or by simply following the design guidelines in the manufacturer datasheets and (if available) associated app notes for the IC that I boil down to. I've never once tested stability of the supply (or even been urged to by my peers or superiors) beyond testing with basic adjustable load equipment before I link it in to the system as a whole. This bothers me, because I see so much information out there from experienced engineers such as yourself and Steve Sandler that propose methods for far more detailed analysis of power supplies. I'm becoming more and more curious about hidden problems in my past and current designs that I just haven't been looking for because we've been following the paradigm of "look, it's working. No more info needed".
Are these analysis methods more targeted towards power supply designs intended for use with an unknown load? Or do you think it's beneficial to add these tests for all power supplies (or atleast all major rails) in a system such as that which I described above?
Hi Brett. Thank you for your question. I work in the SPACE community where the requirements are very rigorous. We get our requirements from both NASA and ESA (European Space Agency) and make them into a set of specifications and they always require that we measure both the Gain and Phase Margins over the operating input voltage range and load range. This is done to ensure that the supply will stay stable. In addition, we also do a WCA (Worst Cast Analysis) to estimate how the supply will perform as the parts change due to aging, radiation, and temperature changes. If your supply is critical, then I would recommend doing these measurements. Keep in mind that the output filter and the Error Amp are performed to verify your design while the total loop is the actual measurement that the customer is really interested in having.
So in my opinion, I would recommend that you do these types of measurements because it gives you and your customers confidence that your supply will perform and remain stable. I hope this makes sense. If you have any questions, post them here are you can email them to rbola35618@aol.com. Best regards, Robert Bolanos
Respected sir, Can you please confirm Comp pin is 1. Cathode terminal of optocoupler or 2 collector terminal of optocoupler
Typically the signal from the OPTO goes to the FB input of the PWM. So the comp pin would be the pin that feeds the FB of the PWM. I hope that makes sense.