THVD8000: Power line communication support

Also Do we need extra filter at load side when we are communicating on powerline bus?

Hi Ashutosh,

Ashutosh Shinde said:

Our system voltage will be from 48V to 60V. with module of ~100. Can we develop by Using THVD8000 IC bus.

How fast are you communicating and how long is your power line cable going to be?

Is size going to be a concern for your application (you may need to choose large inductors).

Is each node going to be transmitting and receiving data? 

I'm not as well verse in power but I assume your power line cable is 48V to 60V DC?

Do you know your peak current?

Ashutosh Shinde said:

will this support addressing of ~100 Modules?

Our device handles the physical layer for communication. It doesn't involve the software portion. If you choose a software protocol that allows for 100x addresses then it should work from a software perspective. 

Ashutosh Shinde said:

Kindly suggest basic micro controller as well to drive THVD with one ADC to read current & 1 GPIO to control on off action.

I can reassign this thread later once we figure out if THVD8000 is a suitable choice for your power line communication.

Ashutosh Shinde said:

Also Do we need extra filter at load side when we are communicating on powerline bus?

Usually the inductors you choose will filter out the AC communication from our device. But adding some additional capacitors to help smooth out the line can help. I don't initially think you need to add additional filters on the actual power line load after the inductors.

-Bobby

Hi Bobby,

Thanks for attending this query. kindly see below answer to proceed further support.

How fast are you communicating and how long is your power line cable going to be?

---> We have one system each communication module will have distance of 5mm only. so, bus bar total distance not more than 2ft. like this we have 3 rows. one rows will be having 25 modules. so minimum we can consider 75 modules will communicate. see picture of system want to upgrade.

Is size going to be a concern for your application (you may need to choose large inductors).

--->i hope from above picture you clear this system and size.

Is each node going to be transmitting and receiving data? 

--->yes, each node will be going to transmit current status of load & want to control critical load on off operation.

I'm not as well verse in power but I assume your power line cable is 48V to 60V DC?

---> this is the DC rectifier system which has different load of 48 V to 60V with 56A continuous current & peak may vary with load.

Do you know your peak current?

---> right now, we have 125A MCB value, it may go around 100A.

Usually, the inductors you choose will filter out the AC communication from our device. But adding some additional capacitors to help smooth out the line can help. I don't initially think you need to add additional filters on the actual power line load after the inductors.

---> Load will be connected directly so i don't know the filter position how we can handle for now but if all communicated then is it required? as you have idea of above system now.

So THVD8000 will work with this application, or do you have another solution for it?

Ashutosh Shinde said:

---> We have one system each communication module will have distance of 5mm only. so, bus bar total distance not more than 2ft. like this we have 3 rows. one rows will be having 25 modules. so minimum we can consider 75 modules will communicate. see picture of system want to upgrade.

It looks like the THVD8000 should be able to support this distance since not very long. But what is the max data rate you need to support? 

Ashutosh Shinde said:

--->yes, each node will be going to transmit current status of load & want to control critical load on off operation.

Do you have a way to make the node communicate one at a time? (Essentially, do you have a way in software to ensure there isn't two devices communicating at the same time?) I ask because THVD8000 doesn't have a way to arbitrate the bus and you also don't want the data to collide since that would cause signal integrity issues to the receivers.

Ashutosh Shinde said:

I'm not as well verse in power but I assume your power line cable is 48V to 60V DC?

---> this is the DC rectifier system which has different load of 48 V to 60V with 56A continuous current & peak may vary with load.

Do you know your peak current?

---> right now, we have 125A MCB value, it may go around 100A.

The voltage is likely too high for our device to support without risk of damage. Even though the supply power is DC, any transcient like the power supply turning on could result in the device's pins seeing a voltage larger than what it is rated for.

This will end up complicating the design. You likely will need a transformer to try protect the device from a potential over voltage. Since THVD8000 likely can't drive through the transformer by itself, you will need an additional line driver infront of it (THS6222). 

I've linked a reference design below. 

https://www.ti.com/lit/ug/tiduf48c/tiduf48c.pdf?ts=1752100388516

Ashutosh Shinde said:

---> Load will be connected directly so i don't know the filter position how we can handle for now but if all communicated then is it required? as you have idea of above system now.

An inductor is required infront of the load to stop the AC modulated signal from entering the load/supply line. After the inductor, caps may be optional to clean up the signal. 

Ashutosh Shinde said:

So THVD8000 will work with this application, or do you have another solution for it?

So far, I think it can work but are you trying to save cost with this solution? Because this system looks like it may end up with a higher cost than just using additional cables solely for the communication. I'm assuming each node is going to require a THVD8000+THS6222+transformer+inductors which is increase size and cost of the overall solution. 

We may need to set up a call and discuss this a bit more. If I'm misunderstanding the connection set up (where all nodes need these circuits) then I'm not sure if it makes sense to go with power line communication. If all the nodes can somehow share a communication line (maybe on the back plane?) then maybe this would simplify the amount of circuits to 4 nodes that each sit on the back planes instead.... 

-Bobby

can we connect on teams on Monday to understand further.

my email id is ashutosh.shinde@vertiv.com