AFE031: NB-PLC FCC AFE031

Hi Alex,

1. if I make the design compliant to CENELEC B, will it also be compliant to FCC part 15 (i.e., will it pass FCC since it is within the FCC allowed 9-490kHz band)? 

AFE030/AFE031 has full power BW up to 300kHz as shown in the plot below.  AFE032 is the IC that has full power BW slightly above 500kHz for NB-PLC application. Within the CENELEC B Band, transmitting signals may be attenuated after approx. 200kHz, the PLC application should be in compliance to FCC part 15 requirements.  The  PLC receiver side is typically using passive bandpass filter and it can further attenuate unwanted communication signals.  

Regarding to FCC certification, all our PLC products has passed and met FCC requirements. These products have been on the market and used by our customers in PLC application for years. AFE03x products have been on the market since 2010 periods or so.

Below is AFE031 EVM link. 

https://www.ti.com/lit/ug/sbou223/sbou223.pdf?ts=1654526001134&ref_url=https%253A%252F%252Fwww.google.com%252F

2. will there be any special considerations to be taken to have it pass FCC?  

AFE03x IC products have passed and met PLC requirements in certified FCC tests. If the design is following the PLC recommended requirements, the certification process should not see any surprises. Since I do not know your specific PLC application, I am unable to offer any comments. For instance, if your PLC design is tailored for solar power application, i.e. SunSpec Rapid Shutdown, you should not encounter much of issues in meeting FCC requirements.

If you have additional questions, please let me know. 

Best,

Raymond 

HI Raymond and thanks for the prompt response.

Understood. I guess my initial confusion regarding the FCC questions may have come from seeing that the specs for the AFE031 do not directly call out support to ARIB / FCC bands (vs the AFE032 which does), but that may just be due to the BW of the 031 (300khz) vs. that of the 032 (500khz), correct? i.e., the 032 can go the full FCC bandwidth (9-490kHz). 

As to my applicable PCL application: it will be required to implement a PLC modem for which we will be using B-FSK (130-145kHz) and we would be wrapping messages (of no more than 50 bytes in length) around a proprietary protocol. The medium will be the AC lines (115/240Vac) where several of these modems (nodes) will be connected to and will be required to receive/decipher messages from the master node to then control lighting.

On a related not and based on the MCU recommended for use with the AFE031, and the fact that my packet sizes will be relatively small, would it be safe to assume that I should be able to use F28002x instead of the F28004x? I see both can be driven by the same clock (100Mhz), and both have a TMU/FPU and a similar ADC engine to acquire the samples from the AFE031. Is it also safe to think that CPU Utilization with the 100MHz will be doubled that of the 200MHz part (F28379D)?

thanks again..

Alex

Hi Alex,

Yes, you are correct. AFE031 does not have the full power BW to support ARIB or FCC band application (in terns of BW). However, AFE031 and AFE032 are identical in transmitting or receiving any modulated signals from a PLC application. If you have a communication protocol that is able to communicate between 9kHz - 300kHz range, AFE031 will work for the PLC application. In other words, AFE031 will amplify and transmit/receive any modulated signals for communication application via a power line.       

would it be safe to assume that I should be able to use F28002x instead of the F28004x?

Yes, our customers have been using F280025C or similar low cost DSP MCU (F280025x) for BFSK communication. I was told that it is plenty speed to handle signal modulation and demodulation and DAC conversion rate for FSK application. I do not support C2000 processor related inquiries. If you have any specific questions, I can forward the inquiry to our C2000 supporting team. 

Is it also safe to think that CPU Utilization with the 100MHz will be doubled that of the 200MHz part (F28379D)?

I believe that F28379D is still more powerful DSP vs. the latest F280025x. The initial PLC application using F28379D is tailored for OFDM modulation application, which data rate and conversion time have to be much faster in ARIB, FCC or higher BW application. In addition, the F28379D is chosen back in 2012 era; and other low cost class of DSP processors were not developed yet. 

For FSK protocol, the computation and process time should not be very intensive, and F280025x will work well for the application. When using F280025x, the previous codes may require to be port-in for the compiler, but it should be fairly straight forward (I was told).

If you have any questions in F280025x. I can transfer you to the C2000 supporting engineer. If you have any circuit, hardware and/or AC or DC powerline coupling issues, I can help you to resolve these questions. Please let us know. 

Best,

Raymond  

Hi Raymond - thanks for the feedback. One last question I wanted to ask: I am currently still in a circuit design phase and I see the AFE031 datasheets claim that this part can be transformer and/or capacitively coupled to AC mains. I've been able to see plenty of example circuitry with the transformer coupling, but not one that shows a capacitively coupled example. Assuming isolation from the line is not required, is capacitor coupling not a T.I. recommended approach? Is it due to possible signal degradation issues? I am only asking because 1.my application does not require isolation from AC mains (it is fully enclosed), and 2. my application will be cost sensitive and therefore I would prefer to save the cost of a transformer (if it can be done). Have you seen many clients using the capacitor coupling approach? is there any example circuit that T.I. can point to with the capacitor coupling? Is there any special design considerations to be followed if using this approach?

thanks again.. I appreciate all the valuable feedback.

Alex

Hi Alex,

Assuming isolation from the line is not required, is capacitor coupling not a T.I. recommended approach? Is it due to possible signal degradation issues?

The issues are how to couple PLC modulated signals onto DC or AC powerlines.

In the AC powerline or mains, say the input AC is 120Vrms @50Hz or 160Vpp @50Hz, and the communication carrier frequency at 140kHz. If 100nF capacitor is used to couple the PLC signal, the capacitor's impedance with respect to (w.r.t) 60Hz is 1/(2pifC) or approx. 26.5kΩ; the same impedance w.r.t 140kHz is approx. 11.4Ω. You may argue that this is what you need for the application (in theory). 

Is it the safe to couple AC mains capacitively in this manner for both DC and AC circuits for the given application? On the PLC side transmitting and receiving side, the AFE031 is operating in DC voltage up to 24Vdc range, which it requires up to 24Vdc power supply and analog ground (unless it is floating, it will be tied to chassis/Earth GND somewhere). On the AC main side, typically you have AC Hi, AC low/Neutral and earth ground, where AC current is supposed to return to Neutral terminal, none to the Earth Ground. 

Without the transformer, I am not sure how you would couple PLC modulated signals to AC mains or vice versa inexpensively. Earth Ground tied to equipment's chassis is used for safety measure, where user will not be electrocuted (GFI or GFCI circuit is typically limited Ground Fault current < 4mA range).  In addition, the suggested approach is not going to be in compliance with NEC code (because the entire PLC system is NOT fully enclosed). 

BTW. I have not mentioned about powerline surge (overvoltage, overcurrent),  transient events, indirect lightning protection events etc. from the coupling methods. In any cases, the design is required to protect AFE031's PLC circuitry in the real world application. You may find a way to couple modulated  FSK signals onto a powerline via opto or other optical methods, but it will likely cost more than a transformer.  Transformer may seem to be expensive if you buy a few. Once it is in production in high volume demand, its cost may be reduced significantly, if it is sourced correctly.  

Have you seen many clients using the capacitor coupling approach? 

Below is a PLC example for capacitive coupling via low voltage DC powerline. The coupling approach may be used in low voltage DC powerline communication application, see the attached link below. 

https://www.ti.com/lit/an/sprac94d/sprac94d.pdf?ts=1654879428908&ref_url=https%253A%252F%252Fwww.google.com%252F

If you have additional questions, please let me know. 

Best,

Raymond