• State Resolved
  • Locked Locked
  • Replies 10 replies
  • Answers 1 answer
  • Subscribers 35 subscribers
  • Views 605 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

MAX3232E:UART RS-485 communication failure

jeferson paz
Prodigy 10 points
Part Number: MAX3232E
Other Parts Discussed in Thread: MAX3232

We are facing a problem in a recent assembly batch of item MAX3232E, where, in the current batch, an external module is unable to lower the voltage to LOW on the TX line, causing communication failure.

When replacing the MAX3232E with another from another batch, the failure no longer occurs.

NOK item

OK item

Apparently the MAX3232E output current at pin 12 is higher in this batch, which prevents the RS-485 module from lowering the voltage to an acceptable level for LOW recognition.

  • +1
    Guru 317480 points

    Pin 12 is an output; no other device should try to drive this signal.

    The schematic is incomplete. Please show all components that are connected to this signal.

  • 0 in reply to Clemens Ladisch
    Prodigy 10 points

    An RS485 module is connected to the CN8 connector

    And the lines go straight to pins 8 and 10 of a Raspberry Pi

  • 0
    TI__Guru 52552 points

    Hi Jefferson,

    Can you please show  all the connections to pin 12 - as if your VOL level is too high that means some connection to pin 12 is  sourcing that current - the device with no pull-ups attached would output ~0V;  both legacy and newer devices (the one that has reported suspicious behavior is the newer version of the device which is why there is a difference in marking) are rated to the same max and minimums in our datasheet - where we really only anticipate 1.6mA of current being sunk into this pin during a VOL measurement (which should max out at 400mV). Both devices shown imply that more than 1.6mA is being sunk into this pin which mean we don't guarantee the output voltage level, VOL, at that point so most likely this application was designed against what our datasheet would imply a suggested design be.  So that means in a 3.3V system your pullup resistor on R1OUT (I assume there is one there - because in UART applications the ROUT pins are typically pulled up to VCC) that implies that you really shouldn't have a pull-up that is effectively smaller than 1.526k at the minimum - right now the VOL level is too high on both. 


    Essentially here is what my most educated guess is the problem at the moment:

    1. There are two versions of this device - a legacy and newer material; it appears you have both in this test. 

    2. They could have slightly different typical measurements - but we don't guarantee typical values in datasheets - we only guarantee boundary conditions - which both legacy and newer material are bound to the same values.

    3. Based on your results from legacy and newer material I can conclude that the effective pull-up between VCC and ROUT is too small for what we would recommend at 3.3V. Our test is at 1.6mA which yields a max VOL of 0.4V - while the device can handle more current we don't guarantee VOL levels above 1.6mA. Design for more current and what ends up happening is that designer introduces unnecessary risk into system because they are now operating in an undefined range and any application failure would most likely be due to the non-recommended design. 

    4. Due to differences in die between devices you could see differences in typical operation (i.e. why results are different between devices). The non-ideal design may have been okay on legacy material but the typical differences may be too much - but considering that both devices are bounded by same values it makes me think its actually an application issue and that you guys got lucky with legacy material. 

    I want to verify the connections to see where the large IOL current is coming from - because it very much seems like that is the issue. So please let me know!

    Best,

    Parker Dodson

  • 0 in reply to Parker Dodson
    Prodigy 10 points
    Parker Dodson said:
    I want to verify the connections to see where the large IOL current is coming from - because it very much seems like that is the issue. So please let me know!

    That's where the problem is, the excess current is leaving the MAX3232 via pin 12. Eliminating all the peripherals connected to the line, there is still a voltage of 3.3V, VOH voltage.

  • 0 in reply to jeferson paz
    Guru 317480 points

    The CN8 connector in the schematic has three pins. What exactly is CN8-2 connected to? RO?

  • 0 in reply to Clemens Ladisch
    Prodigy 10 points

    to the RS485 converter, with the exception of the power pin, this voltage comes from another point in the circuit.

  • 0 in reply to jeferson paz
    Guru 317480 points

    And to which pin of the RS-485 converter is CN8-2 connected to?

  • 0 in reply to jeferson paz
    TI__Guru 52552 points

    Hi Jefferson,

    Pin 12 does not source current on a low output - it sinks current on low. The excessive current is not coming from the device - its coming from, most likely, a pull-up that you are not showing. 

    That is the cause of the problem - you have too strong a pull-up on that bus - for both devices. If it is an integrated pull-up or not I can't tell - there are at least 1 10k pull-up (which shouldn't be an issue by itself - but if there is another pull-up operating in parallel it will matter)  - but the only way to fix that is to weaken the pull-up bias on the bus. There is nothing on the 232 device that would cause pin 12 to source current at a low output unless the bus was negatively biased - which doesn't seem to be the case. 

    To confirm ROUT pin doesn't have issues  - disconnect it from load if possible and measure voltage - it should be near 0V when outputting a low value. 

    Connect loads to ROUT 1 at a time and measure the decrease in VOL - you shouldn't have much current going into these pins.  This will help isolate where most of the current is coming from. 

    If you see issues on ROUT without loading then you can submit for a customer return and potential Failure analysis - link here: https://www.ti.com/support-quality/additional-information/customer-returns.html . What I will say about that is this however - I really don't think there is damage and it looks like there is too much current being pushed into pin 12 when that pin drives low - this pin will sink current when driving low - without a current source (pull-up resistor + source act as a current source) on the bus it should output near 0V. If we do end up doing an FA on the device the most likely outcome is part is okay and issues is apps related - more unlikely there is damage but most likely would be caused by EOS - and the most unlikely event is a quality issue. I don't think there damage and the mode of failure matches up pretty much perfectly with too much current being injected into pin 12 from the system when driving low. Driving high doesn't have a problem because VOH current is going to be much lower in these configurations. 

    These types of failures are almost never the part and always due to too much current above our tested values - weakening the pull-up strength of the bus is the only real way to fix this issue. 

    Best,

    Parker Dodson

  • 0 in reply to Parker Dodson
    Prodigy 10 points

    In fact, pin 12 is keeping the circuit in HI and the other device (rs485) is unable to consume the current and take the voltage to LOW. They are connected in parallel, MAX3232 output with RS485 output, and communication is carried out only with the Raspberry Pi.

  • +1 in reply to jeferson paz
    TI__Guru 52552 points

    Jefferson,

    Since you have repeatedly not responded to my replies to this post and just keep repeating yourself - there is nothing else I can do - this is most likely an apps issue caused by poor design because someone wasn't bothered to read a very short datasheet  as it is very clear the design does not follow instructions that we set out in the datasheet - it seemed like this error was okay with legacy material but not new - however since your VOL is large on both images and outside of what we test - the design is not good - the R pin is a push pull driver - when its low its pulling current from the system - which is not the IC; this isn't a complicated circuit and there is no more hand holding I can do for you. I gave you a path forward yesterday - there is absolutely nothing else that be accomplished in this thread especially since you have ignored almost everything I have already said and have only giving incomplete information 

    -Parker Dodson