• State Resolved
  • Locked Locked
  • Replies 1 reply
  • Subscribers 63 subscribers
  • Views 712 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

UCD3138: UCD3138 Vs C2000 for Totem-Pole PFC

A. Manno
Prodigy 20 points
Part Number: UCD3138
Other Parts Discussed in Thread: POWERSUITE, TMS320F28384D

Hello,

my project consists in implementing a GaN-based CCM Totem-Pole PFC converter. I've seen there are a lot of resources and reference designs about this topic and most of them employ UCD3138 or C2000 MCUs.

Since I am a beginner at TI devices, I'd like to have a feedback from a more expert user. Which microcontroller should I choose?

UCD3138 seems to be the best choice for this power supply thanks to the programmable HW state-machine and the tailored peripherals. Moreover the Fusion Digital Power Studio GUI (FDPS_GUI) could be really useful in the control loop parameter setting. However, I'm a little worried about the programming procedure of this controller. Indeed, just to flash the memory I have to create the firmware in CCS, generate the .x0 file, import it in FDPS_GUI and finally upload it on the DSP through PBMbus. Performing intense debugging would be more complex and time-consuming than doing as usual through JTAG. In addition some issues regarding linker and compiler files could arise and it would be difficult to understand where the problem is located.

On the other hand, C2000 enviroment provides a way bigger amount of resorces, trainings and source code examples than UCD3138 and most of new totem-pole PFC designs employs C2000 controller. CCS is sufficient to upload the firmware and debug it and PowerSuite kit is a useful tool for system characterization. Anyway there are several drawbacks. DSP is in charge of exectuting fast and time consuming calculation for the control loop by itself (in UCD3138 the HW state machine takes care of fast loops and DSP of slow loops) and quite good DSP acrchitecture knowledge is required to manage the right timing between different interrupt routines with different frequency.

Given that, which is the controller suggested for a low-expert user for this AC/DC topology?

It is worth mentioning that, whatever the choice will be, the controller will be integrated in a custom board, so I will not use evaluation boards as control cards or launchpads. In this scenario is it still possible to use FDPS_GUI or PowerSuite to monitor performance and fix PID controller constants? 

Thank you

  • 0
    TI__Genius 17090 points

    I think you've described the trade-offs between the UCD3138 and the C2000 pretty well.  

    You did miss the presence of the CLA(s) in the C2000, which work similarly to the hardware in the UCD to offload the control loop from the main processor(s).

    They are kind of like little DSPs, so you have to be familiar with DSP architecture and time management to program them.  

    You didn't mention what your goals are.  I'm a UCD expert, and I think that if you wanted to do a cost effective power supply for mass production, and in a topology and wattage that we have an EVM for, the UCD would be a good choice.  I'm not an expert on the C2000 offerings, but I suspect the same is probably true there as well.  

    If you want to do research on new topologies or control schemes, why not consider something like the TMS320F28384D?  It's got lots and lots of processing power, so you wouldn't have to worry as much about your code timing.  It seems like it would be a great experimenter's tool. Once you get your new scheme worked out, you can see what smaller and more inexpensive processor it will fit on for production.   Or if it's not as cost sensitive, just stick with the bigger solution.