THVD1550: Cause of Output waveform

Ottey-san,

After reviewing the waveforms, I have a few questions. 1) What's the data rate of the application? 2) How long is the cable? Is there any grounding shift between the transmitter side and receiver side? 3) What's the supply? It looks like 5V. What the maximum and minimum voltage level of A and B signal?

Although I haven't figured out the cause of this behavior, I will just try to provide some information. A and B pins are symmetrically designed in the device, therefore the 'dull' part shows in the same place during the transition. Adding termination matching the cable's impedance helps the signal integrity by reducing reflection at end of the cable. The drawback as you pointed out, is smaller amplitude since the load is heavier than the case without termination.

Please provide the information at your convenience. I will help you continue debugging.

Regards,

Hao

Dear Hao

Thank you for your reply.

I answer your question as below.

1) What's the data rate of the application?

I don't know the data rate, but the baud rate is 19200bps

2) How long is the cable? Is there any grounding shift between the transmitter side and receiver side?

The cable length is 2〜3m. There is probably no ground gap between the transmitter and receiver.

3) What's the supply?

The supply voltage is 5V.

Regards,

Y.Ottey

Ottey-san,

Thanks for your information. I'm sorry I have to ask more questions. Is there any load on the bus, like pull up/pull down, or capacitor? If you disconnect the receiving board from the transmitter board, would it make any difference? Where does the 5V supply come from? How much transient current can it provide? Would you be able to measure the bus current/supply current during the transmission? 

Regards,

Hao 

Dear Hao

Thank you for your reply.

Answer the following questions you asked:

" Is there any load on the bus, like pull up/pull down, or capacitor?"

A pull-up resistor (16kΩ) is connected to the A-side load, and a pull-down resistor (16kΩ) is connected to the B-side load.
However, a similar waveform is observed even if the pull-up and pull-down resistors are removed.

We are currently checking the other parts.

Regards,

Y.Ottey

Ottey-san,

Please let me know if you have any updates. I suspect this behavior is application related. Could you please check the basics like ground, load, etc? We haven't seen this in the lab. You can also get the EVM to compare the waveforms.

www.ti.com/.../RS485-HF-DPLX-EVM

Regards,

Hao

Dear Hao

Thank you for your reply.

I have confirmed one of the checks you have requested and will share the information with you.

”If you disconnect the receiving board from the transmitter board, would it make any difference?”

I have confirmed that the waveform of the output signal has the same shape when the measurement is performed with the receiving board removed, or when the receiving board is changed to another board (the driver IC on the receiving side is different).

Regards,

Y.Ottey

Ottey-san,

How close is the 100nF decoupling cap to the IC? Is it possible to try a bigger value? Another experiment you can do is to use a 'clean' supply/ground for the device. I don't know if any of these would help.

Regards,

Hao

Dear Hao

Thank you for your reply.

"How close is the 100nF decoupling cap to the IC? Is it possible to try a bigger value?"

The 100nF capacitor is located close to the IC.
Also, when measuring the voltage waveform of VCC, there is no noise as shown below, and it seems that the output signal does not change due to the fluctuation of the power supply voltage.

* Ch1 (yellow) of the attached waveform is RS485-A, and ch2 (green) is 5V power supply voltage. The voltage waveform is displayed with an offset when measured between the power supply GND of the IC. (50mV / div)

In addition, the same waveform was obtained when the measurement was performed with the cable connecting the receiving board, transmitting board, and receiving board removed.

Are there any other possible causes of the phenomenon being seen this time?

Best Regards,

Y.Ottey

Ottey-san,

This really puzzles me. Can you share your schematic or even layout with me? If you're worried about the privacy, you can send the message to hliu@ti.com. Have you ever tried with other RS-485 parts or different boards? Regards,

Hao

Hao

Thank you for your reply.

I have detailed schematics, so I am thinking of sending you an email.
By the way, even if you change to another receiving board, the same output signal waveform will be obtained.
I do not measure it with other RS-485 parts (or another THVD1550), so I confirm.

Y.Ottey

Ottey-san,

Thanks for your information. Again, is it possible to capture the current waveform during the bus pin ramping? Have you tried with different  supply, like 4.5V, 5V and 5.5V?

I'm looking forward to your email.

Regards,

Hao