STNRGPF01 - 3-channel interleaved CCM PFC digital controller

[Mads Barnkob]

From the STMicroelectronics newsletter from today, there was a new interesting PFC controller that is well within the power range for use in fairly large Tesla coils, especially easy to make a design with their eDesignSuite (online circuit designer) that gives you a complete schematic after giving it components values for like 10 key parts. It does seem that you need a ST account to use it though!

The new product: http://www.st.com/content/st_com/en/products/power-management/ac-dc-converters/pfc-controllers/stnrgpf01.html?ecmp=tt4938_gl_enews_mar2017

Digikey got it in stock: https://www.digikey.com/products/en?WT.z_cid=sp_497_0928_buynow&Enterprise=44&lang=en&Vendor=497&mpart=STNRGPF01

The STNRGPF01 is a digital controller designed specifically for interleaved PFC boost topologies and intended for use in high power applications.

The controller is capable of driving up to 3 interleaved channels, generating the proper signals in each condition. Moreover, it implements a flexible phase shedding strategy that enables the correct number of PFC channels based on the actual load condition. With this function, the STNRGPF01 is always able to guarantee the highest power efficiency across a wide range of load current requirements.

The device works in CCM at fixed frequency with average current mode control, and implements mixed signal (analog/digital) control. The inner current loop is performed by hardware, ensuring cycle-by-cycle regulation. The outer voltage loop is performed by a digital PI controller with fast dynamic response.

The controller implements several functions: inrush current control, soft start-up, burst mode cooling management and status indicators.

It also features a full set of embedded protections against overvoltage, overcurrent, and thermal faults.

The STNRGPF01 is configured through a visual dedicated software tool (eDesignSuite) to match a wide range of specific applications. Using eDesignSuite, the user can customize the PFC conversion configuration and all the relevant electrical components. As a result, the tool will automatically generate a full schematic which includes a complete list of material and the final binary object code (FW) to be downloaded to the STNRGPF01.

Key Features

    Interleaved boost PFC
    Up to 3 interleaved channels
    CCM, fixed frequency
    Average current control, cycle-by-cycle
    Inrush current control
    Burst mode support
    Overcurrent and thermal protection
    Soft start-up
    Flexible phase-shedding strategy
        High operating frequency with small PFC inductor, suitable for high power-density applications
        Low ripple current (input/output)
        Simpler integration with other applications
    Flexible design customization to meet specific customer needs
    Firmware
        Turnkey solution for quick design
        eDesign Suite graphical user interface (GUI) for application configuration
    Embedded memory data retention 15 years with ECC
    Communication interfaces
        UART asynchronous protocol for bootloader support
    Operating temperature: -40 °C to 105 °C

[TCWilliamson]

From the STMicroelectronics newsletter from today, there was a new interesting PFC controller that is well within the power range for use in fairly large Tesla coils, especially easy to make a design with their eDesignSuite (online circuit designer) that gives you a complete schematic after giving it components values for like 10 key parts. It does seem that you need a ST account to use it though!

The new product: http://www.st.com/content/st_com/en/products/power-management/ac-dc-converters/pfc-controllers/stnrgpf01.html?ecmp=tt4938_gl_enews_mar2017

Digikey got it in stock: https://www.digikey.com/products/en?WT.z_cid=sp_497_0928_buynow&Enterprise=44&lang=en&Vendor=497&mpart=STNRGPF01

The STNRGPF01 is a digital controller designed specifically for interleaved PFC boost topologies and intended for use in high power applications.

The controller is capable of driving up to 3 interleaved channels, generating the proper signals in each condition. Moreover, it implements a flexible phase shedding strategy that enables the correct number of PFC channels based on the actual load condition. With this function, the STNRGPF01 is always able to guarantee the highest power efficiency across a wide range of load current requirements.

The device works in CCM at fixed frequency with average current mode control, and implements mixed signal (analog/digital) control. The inner current loop is performed by hardware, ensuring cycle-by-cycle regulation. The outer voltage loop is performed by a digital PI controller with fast dynamic response.

The controller implements several functions: inrush current control, soft start-up, burst mode cooling management and status indicators.

It also features a full set of embedded protections against overvoltage, overcurrent, and thermal faults.

The STNRGPF01 is configured through a visual dedicated software tool (eDesignSuite) to match a wide range of specific applications. Using eDesignSuite, the user can customize the PFC conversion configuration and all the relevant electrical components. As a result, the tool will automatically generate a full schematic which includes a complete list of material and the final binary object code (FW) to be downloaded to the STNRGPF01.

Key Features

    Interleaved boost PFC
    Up to 3 interleaved channels
    CCM, fixed frequency
    Average current control, cycle-by-cycle
    Inrush current control
    Burst mode support
    Overcurrent and thermal protection
    Soft start-up
    Flexible phase-shedding strategy
        High operating frequency with small PFC inductor, suitable for high power-density applications
        Low ripple current (input/output)
        Simpler integration with other applications
    Flexible design customization to meet specific customer needs
    Firmware
        Turnkey solution for quick design
        eDesign Suite graphical user interface (GUI) for application configuration
    Embedded memory data retention 15 years with ECC
    Communication interfaces
        UART asynchronous protocol for bootloader support
    Operating temperature: -40 °C to 105 °C

Oooh, now that sounds pretty interesting... Is this only for single phase or do they offer an idea for 3 phase too?

[Mads Barnkob]

Looking at the schematics from their 3kW evaluation board: http://www.st.com/content/ccc/resource/technical/document/data_brief/group1/ae/c8/2d/88/84/ab/4d/02/DM00379111/files/DM00379111.pdf/jcr:content/translations/en.DM00379111.pdf

On page 2 there is a ZVD (zerocrossing voltage detection) taken from one of the AC inputs, all 3 interleaved PFC stages are run from the same DC bus, so as it can only detect zero crossing on one phase, looking on page 5, the current sensing resistor is placed on the grounded output, as with many other PFC IC solutions, so they can not just be paralleled.

[futurist]

It looks great! They even give you gerbers and BOM for the 3kW eval board
I'll try to import BOM to Mouser to see total part cost, but I suppose it won't be cheap as STW40N60M2 MOSFETs alone are 8€ each

[Hydron]

I had a quick look at that chip and the reference design the other day. Didn't price it up but did notice that it uses SiC diodes - when I eventually build a PFC I intend to do the same so I can basically ignore diode switching losses.

I'm personally more interested in the ADP1047/8, similarly they are digital, but with an I2C interface that allows for current/voltage/power measurement during operation and (I believe) changing control parameters (e.g. output voltage) on the fly.

[Mads Barnkob]

I found the STNRGPF01 interesting because it comes in a 3 kW development board, I can only find a few notes about the ADP1047 that says its a 300 Watt rated controller and ADP1048 is 600 Watt.

[Hydron]

300W and 600W are simply the available demo boards, rather than max output. The chip itself is just a controller so is capable of significantly higher power - in fact both the ST and ADI parts need external MOSFET drivers, and the ST part even needs some external logic (flip-flops) to generate PWM for channels 2 and 3.
I looked at the 3kW board (and the ST part) because the power level is about right to use as a start for a DIY design. Even if you don't use the same chip, the design is a good place to start - there won't be significant differences in topology between PFC controllers unless you go to bridgeless or similarly exotic configurations.

[Mads Barnkob]

300W and 600W are simply the available demo boards, rather than max output. The chip itself is just a controller so is capable of significantly higher power - in fact both the ST and ADI parts need external MOSFET drivers, and the ST part even needs some external logic (flip-flops) to generate PWM for channels 2 and 3.
I looked at the 3kW board (and the ST part) because the power level is about right to use as a start for a DIY design. Even if you don't use the same chip, the design is a good place to start - there won't be significant differences in topology between PFC controllers unless you go to bridgeless or similarly exotic configurations.

I have tried, and failed, to scale up a PFC IC to 2kW that was not made for it, mostly it messes up the feedback loops and prevents it from even running or or melts down fairly fast. I do want to give my 2kW unit another try soon, with a change to the voltage and current feedbacks. I will make a separate thread for that

[Hydron]

I have 2x 800W and 2x 875W output Dell workstation power supplies which I'm thinking of tearing down and using for PFC experiments. Each PSU has a 2 board construction, with separate PFC and DC-DC PCBs, and thus are relatively easy to re-purpose. I'm hoping that I can remove the existing PFC chips and rewire for operation as a 1600W+ interleaved PFC converter with a new ADP1048 controller.

As for feedback loops, one of the reasons I'm looking at the ADP104x parts is that it can all be setup in software then flashed into the chip, in a nice GUI with graphs etc no less! Hopefully if I'm cautious and use a bulb in series or something to limit fault current I'll get away without blowing up too much silicon

[TCWilliamson]

I have 2x 800W and 2x 875W output Dell workstation power supplies which I'm thinking of tearing down and using for PFC experiments. Each PSU has a 2 board construction, with separate PFC and DC-DC PCBs, and thus are relatively easy to re-purpose. I'm hoping that I can remove the existing PFC chips and rewire for operation as a 1600W+ interleaved PFC converter with a new ADP1048 controller.

As for feedback loops, one of the reasons I'm looking at the ADP104x parts is that it can all be setup in software then flashed into the chip, in a nice GUI with graphs etc no less! Hopefully if I'm cautious and use a bulb in series or something to limit fault current I'll get away without blowing up too much silicon

Please let me know how that comes along... could be quite nice.

[Hydron]

I will be making stuff public, but I cannot guarantee how good the documentation will be (I don't have the best record), but I will try my best!

I've bought one of the ADP1048 chips to test, but realistically the PFC design is behind a few other things on the to-do list.

[futurist]

Do all Dell server/workstation PSUs have separate PFC?
I have 3x 750W Dell PE 2950 server PSUs which became useless when server died

[Mads Barnkob]

Futurist, did you ever get a BOM made? and found a total price?

Hydron, you have to make a good documentation! I got 6 of these 1000 Watt server power supplies, which actually was meant to be pit in series for 3x 12VDC for a royer IH.

[futurist]

I didn't manage to import BOM, mouser asks for manufacturer part numbers which aren't included in pdf
So you need to search for ~120 parts manually

Here is pdf bom converted to excel if someone wants to make project at mouser
pfc_bom.zip

[alexhanyuan]

Any updates on using either ICs? The STNRGPF01 is being discontinued, but can be found on Arrow for ~$4. Curious to try it out in a ~25kW design.

[Mads Barnkob]

Any updates on using either ICs? The STNRGPF01 is being discontinued, but can be found on Arrow for ~$4. Curious to try it out in a ~25kW design.

Back in March 2022 I posted another ST reference design, a 15 kW PFC frontend full design that they released: https://highvoltageforum.net/index.php?topic=1960.0

[alexhanyuan]

I stumbled across that design as well. Looking to run it off of single phase though.

Any guesses on why the STNRGPF01 is discontinued? Is everything moving to MCUs? The design tool seemed to be a big selling point for it.