STM32H723ZG + DS8500 HART Modem → OCD pin always LOW (No Carrier) when interfacing with 2088 Pressure Transmitter

[Faizan_001]

I am working on establishing HART communication using an STM32H723ZG Nucleo board, a Maxim DS8500 HART modem, and a 2088 Pressure Transmitter (4–20 mA + HART).

Hardware Setup:

• STM32H723ZG

  • USART3 → Debug to PC (115200, 8N1, no parity)

  • USART6 → Connected to DS8500 D_IN/D_OUT (1200 baud, 8 data bits, odd parity, 1 stop bit)

  • GPIOs:

    • PE0 → DS8500 RST (Active Low)

    • PA0 → DS8500 RTS (Low=TX, High=RX)

    • PB0 → DS8500 OCD (Input, carrier detect)

• DS8500 Connections (per Maxim datasheet and app note):

  • MODE = High (for HART mode)

  • 3.6864 MHz crystal connected between XTAL1/XTAL2

  • REF decoupled with 0.1 µF

  • FSK_IN/OUT connected to the HART loop via the recommended RC filters

• Loop: 2088 transmitter powered (24 V, 4–20 mA loop).

Software:

• USART6 configured at 1200 bps, 8E1 (actually 8 data bits + odd parity + 1 stop).

• GPIO configured:

  • PE0 default high (releases reset)

  • PA0 high (sets DS8500 in RX/demodulator mode)

  • PB0 input (read OCD)

• Debug prints OCD state every second on USART3.

Problem:

• OCD pin always reads LOW → No carrier detected, even though the transmitter should respond.

• I confirmed that RTS=High (RX mode) and RST=High (not in reset).

• MODE pin is tied High.

• REF pin shows ~1.23 V (so demod section seems active).

• The loop is wired correctly and the transmitter works in analog (4–20 mA changes with pressure).

What I Tried:

1. Toggling RTS between TX and RX modes. OCD still remains low.

2. Sending a HART preamble (0xFF * 20) via USART6 → FSK_OUT shows modulation, but still no demodulated data.

3. Checked wiring against Maxim’s reference design (filters on FSK_IN/FSK_OUT).

4. Verified 3.6864 MHz crystal oscillation with oscilloscope.

Questions:

1. Is my USART6 configuration correct for DS8500 (1200 bps, 8 data bits, ODD parity, 1 stop)?

2. Does OCD only go high when a valid UART start bit + carrier amplitude >120 mVp-p is present, or should it indicate raw carrier energy on the loop?

3. Could there be an issue with the filter/bias network on FSK_IN preventing detection?

4. Has anyone successfully interfaced STM32H7 with DS8500 + Rosemount/2088 transmitter, and can share working settings/schematics?

Any insights from ST engineers or community members experienced with DS8500 + STM32 would be very helpful.

Thanks!

 

[MHoll.2]

If You don't get the Carrier detect it's probably a problem with Your Sensor or the DS8500, so you should ask Analog Device.

Did You check that the input signal from the Sensor is OK (with an oscilloscope)?

[Chris21]

-> "1200 bps, 8 data bits, ODD parity, 1 stop" this is correct for HART.

However, ST includes the parity bit as part of the data frame, make sure the USART word length is 9:

The 2088 transmitter is not going to respond unless you send a properly constructed HART message (Start byte after preambles, ending with checksum, etc.:

 

 

 

[Faizan_001]

Hey, 

mentioned below is my Usart6 used to communicate with DS8500. 

Now, i will show you the command i am sending. 

now let me show you my code 

and this is what i am receiving on my serial monitor 

Now what could possibly be the issue? 

[Chris21]

Also, your 4–20 mA current loop should have a resistor in it, typically 250 Ω.

[Faizan_001]

Yes, everything is there. Check the above chat 

[Tesla DeLorean]

Not sure it's functionality should be STM32 specific, perhaps look more broadly at other working MCU examples.

Is the HAL layer returning any errors?

If you scope the modem,or use a logic analyzer, can you inspect the modem's  OCD and data output to the STM32's UART RX pin?