MAX3485 with USB RS-485

[ppr92]

Hi there!
This it my first post so thank you in advance for your time!

I made a PCB with uC with UART wich I connect to a MAX3485 (3.3V supply). To test it, I tried to comunicate the uC with other devices successfully:
1) ESP32 with external MAX3485 (3.3 levels)
2) ESP32 with external MAX485 (5V levels)

But when I use a USB RS-485 communication doesn't work. I see no response from MAX3485.
And I see that differential signal A-B is always negative (from 0V to -3V aprox.). My guess is that difference A-B is never >200mV so there will be never a change in RO from MAX3485, but I don't know why that's happening.
I saw that RS-485 should work regardless signals level (5V or 3.3V) but I can't understand what is happening and why it doesn't work.
The main goal would be that the uC with MAX3485 can communicate with a PC with USD RS-485 adapter.

I attach 2 photos of oscilloscope. Yellow is signal A, Blue is signal B and Red is A-B
1) MAX3485 communicating with MAX485
2) MAX3485 communicating with USB RS-485. I attached a scheme of this too.

Any info would be appreciated.
Thanks!

 

[KlausST]

Hi

Test each device individually and note the voltage levels.

Btw: I agree that RS485 should work with mixed supply voltage levels of 3.3 V and 5.0 V.

Klaus

 

[doraemon]

Hello!

Just in case: did you set a 120 ohm load?

Dora

 

[FvM]

The waveforms show USB-485 or MAX3485 transmitting? In any case, there's a problem with TxEnable. The sender is apparently switching between "space" (active state) and high Z. This is usually the case with "poor mans" RS-485 converters that are operating TxEnable by a monoflop rather than keeping it active during the full character frame by a dedicated TxEnable signal generated in the USB UART.

Getting the correct "mark" (idle state) level on RS-485 depends on the relation between pull-up/pull-down and terminating resistors. Using termination resistors without strong pull-up/pull-down resistors results in a differential level below receiver hysteresis and accidental received level.

Possible solutions:
- use strong (a few 100 ohms) pull-up/pull-down resistors to enforce idle state during driver high Z.
- for short cables (< 5m) and low or medium speed, remove 120 ohm termination resistors at both ends.
- use a state-of-the-art USB-to-RS485 bridge with active TxEnable. Wrong idle state (equals break signal) may be still a problem, depending on the receiver software.

 

[KlausST]

Hi

@FvM: how they are connected is shown in the picture´s name.

@ppr92:
After reading FvM´s post..
I´ve seen that the USB seems to cause the problem. It seems it´s never disabled.
Maybe there is some option (within PC driver setup, or ort setup) to automatically enable/disable the driver.
--> if there is a RS488 / RS485 option -> set it to RS485
--> if there is a full duplex / half duplex option -> set it to half duplex.

Klaus

 

[FvM]

@FvM: how they are connected is shown in the picture´s name.

Yes, but not data direction of the waveforms.

 

[ppr92]

Hi.

This is great! I tried what @FvM suggest as first possible solution. Connecting two 100ohms resistors (pull-up channel A + pull-down channel B) does the trick.
However, line seems a little bit unbalanced. I attached capture of oscilloscope and new scheme (sorry for the lazy scheme).
1) When neither devices is transmiting, signal A has a iddle state level of 2V and signal B of 1.2V. Plugging or unplugging the USB-485 doesn't affect. Any idea about this?
2) App note AN960 from analog devices (page 6) calculate the value of the pull-up pull-down resistor as a voltage diviver. For VCC = 3.3V the value I get is 930ohm but I get better results with 100ohm. Any comment about this?

Thanks!

--- Updated Oct 26, 2021 ---

Btw I saw that USB-485 has a texas instrument chip 75176B "SNx5176B Differential Bus Transceivers - datasheet"
(in case it helps)

 

[KlausST]

Hi
The shown resistors: 100R | 120R | 100R
Is not suitable.
The total resistance (termination should be 100...120Ohms
And the bias should result in more than +200mV

Try 500R | 120R | 500R

Klaus

 

[ppr92]

Hi.

I tryed combination 220R | 150R | 220R. This allowed to elevate average value for difference A-B approx 400mV and still have an equivalent Rab = 110ohm. USB-485 works fine with this.

However, I think I going to change in my design MAX3485 to MAX3072E wich has true fail-safe implemented with a threshold of -50mV to 200mV.
Hope this works!

Thank you,
Pedro.

 

[KlausST]

Hi,

The functional problem is not a resistor or termination problem.
I still agree with FvM that it´s a driver control problem.

***
I just wanted to point out that 120|100|120 Ohms is way off the expected 100...120Ohms differential resistance.

Klaus

 

[FvM]

I understand post #9 so that the connection is working. Exact termination resistor value doesn't actually matter for low speed, short distance link.

 

[ppr92]

Hi,

The functional problem is not a resistor or termination problem.
I still agree with FvM that it´s a driver control problem.

***
I just wanted to point out that 120|100|120 Ohms is way off the expected 100...120Ohms differential resistance.

Klaus

Right.
Thing is that I don't know how the end user of this will use it. Hopefully will be a PLC or another controller with more sofisticated transceiver.
But should be prepare to work with this "Poor Mans" USB-485.

Anyway, first time working with RS-485.
Thanks!