SN65HVD10: RS485 Modbus Interface across multiple device without common ground

Hi,

In isolated systems, the digital signal is isolated. RS485 signal can be transmitted with shared ground or without common ground, depending on the ground potential difference. Like you mentioned, if the ground potential difference is between -7V and 12V, the transceivers work with no problem. If the ground is shared, usually it comes from one node. You can refer the section 11 of this app note for the detailed discussion.

https://www.ti.com/lit/an/slla272c/slla272c.pdf

I hope it makes sense to you. If you don't mind, you can share the schematic and I will help review it.

Regards,

Hao

Hi Hao,

Thanks for responding so quickly. Sorry for not being clear in my first post. I meant -7V to +12V is typical SN65HVD10 transceiver rating capable of handling the range of this voltage across the devices. Since ground are not common, do you see any concern here? I never had a failure so far.

Regarding Ground, As per below TI Engg (Thomas), this depends on the receiver internal resistors.

https://www.planetanalog.com/signal-chain-basics-84-why-rs485-does-not-need-ground-wires/#

If i need to measure GPD, which ground to consider if its not shared? I think this need to happen without differential probe and on the receiver side.

Regarding the schematics, two of the components (Fan & VFD) are from 3rd party design. Also i just found out today that fan interface is not truly isolated, likely from the one fly back source inside the fan.

On Fan side TI P/N: SN65HVD72D is used while the VFD side which has isolated side have Renesas P/N: ISL83080EIBZ-T.

Below is from our design from Control Board, which don't have dedicated isolated power supply. But also not sure why shorting all grounds damaged the Fan Modbus Port.

Thanks.

BR,

Aditya

Aditya,

When you measure the GPD, please always refer to the same potential, like ground of a node. It depends on the potential difference and impedance, shorting two ground could generate a high current loop. Again RS485 works with -7V to 12V common mode difference. But you want to create a shared ground between nodes, you need to make sure there is no 'short' created. You can check out the Figure 9 of the app note I shared in last post for these different scenario.

Regards,

Hao 

Hi Hao,

Thanks for your reply. The main concern i have is i know today our system works fine without common ground, may be because they are so close to each other, couple of feet apart. Shorting the ground we know it damages the Fan Modbus port. Can you please let me know, what is down side of running without ground? GPD difference should be more where system are far apart from each other, but in our case its powered on same electrical socket.

If i need to measure GPD, can i measure ground of each part of the system like the VFD, Fan & Control board w.r.t A & B? But that will give me 3 different result, may be then i can consider worst one. Same can be repeated on multiple units. But again, not sure if this test is good test as they don't share ground,

Since i need average of A & B, i like to use two same resistor say 10k and use center node w.r.t ground to measure the potential difference.

Looking forward to hear from you soon.

Thanks.

BR,

Aditya

Aditya,

Yes, I think it's a good idea to reference to the common mode voltage of A/B for the ground measurement. I think you can turn one node on to bias the A/B while doing the measurement.

Regards,

Hao

Hi Hao,

Thanks for your input. Since we don't have common ground in our system between the Fan, VFD & the Control Board. I have tested A & B w.r.t respective grounds of the components and got below data. Test has been repeated on three Units to confirm.

Common Mode Voltage Measurement

A & B had same resistor value of 10k. This A & B is common bus between the Fan, VFD & Control Board.

Measurement was performed with Oscilloscope between Resistor common terminal & w.r.t respective Ground of respective component.

 Additional Notes:

1. During the test no high amplitude noise was observed. Some noise was picked up by long ground wire, but was minimal within above mentioned limits.
2. Tested with system running and ideal, voltage did not change.
3. Unit 1 was only brand new unit. Unit 2 & Unit 3 has passed through lot of HDVP test, DLS & Biotron.
4. Meter reading matched the oscilloscope.
5. Fan Transceiver P/N: SN65HVD72D, VFD Transceiver P/N: ISL83080EIBZ-T, Control Board Transceiver P/N: SN65HVD10
6. All these above Transceiver are capable of operating between -7V to 12V.

Please help to provide your inputs.

Thanks.

BR,

Aditya