SC18IM704 Baud Rate Limits

[andrtnnr]

Hello,

I was able to successfully change my baud rate to 19200 on this board using the following code on my Particle Boron micro while connected to the SC18IM704 evaluation board using protowires and with the Jp4 jumper open. I was also able to read i2c from i2c devices quite well. But its still too slow and I can notice the difference.

      Serial.begin(9600)      

      Serial1.write(0x52);  // Start i2c message

      Serial1.write(0x00);  // BR0

      Serial1.write(0x01);  // BR1

      Serial1.write(0x50);  // End i2c message

      // Attempt Higher Baud Rate

      if (!once) {

         Serial1.write(0x57);  // Start i2c message

         Serial1.write(0x00);  // BR0

         Serial1.write(0x70);  // Data

         Serial1.write(0x01);  // BR1

         Serial1.write(0x01);  // Data

         Serial1.write(0x50);  // End i2c message

         delay(500);

         Serial1.end();

         delay(1000);

         Serial1.begin(19200);

         once = true;

      }

But the second I started changing the baud rate to 384000 (BRG0 = 0xB0, BRG1 = 0x00)  . Things start getting wonky. To be specific, the SC18IM704 begins to effect the TX signal coming from the Boron micro. On my logic analyzer I can see framing errors and it appears to prevents me from doing internal register or i2c operations. Based on the speed spec (460.8 kbit/s), this should be achievable but the chip seems to be limiting itself. When I removed the SC18IM704 from the uart and just probed the signal from the Boron, the messages are totally fine. Which makes me think something is wrong here.

Would love some help.

Thanks!

[andrtnnr]

I also tested this on my PCB which houses the direct SC18IM704 and have the same issue. So I wouldn't chop this up to lengthy fly wire interference either...

[diazmarin09]

Hello Andrew ,

Please accept my apologies for the delayed response.

This is definitely an unexpected behavior. Could you please share your complete configuration besides the BRG0 and BRG1 registers?

Regards,

David

[andrtnnr]

Hey David,

Im not altering any other registers. A reset is issued every run.

Thanks,

Andrew

[diazmarin09]

Hello Andrew ,

Please accept my apologies for the delayed response.

All the baud rates are generated by the same oscillator via a digital divider. So, if the oscillator frequency is accurate all the baud rates should be accurate.

I assume the internal oscillator is not accurate enough. Have you tried to replace the device?

Have you tried to give the device some delay after the reset that you mentioned so that the oscillator start up properly? I mean before the baud rate configuration.

Regards,

David

[diazmarin09]

Hello,

I hope all is great with you. Thank you for using the NXP communities.

Your source code seems correct to me. I am glad to know that your device is working as expected with a 19200 baud rate.

Do you mean that the device is not working with 38.4 kbit/s? Could you please share further details about the device behavior with such baud rate? I mean, a detailed description of the signals. Could you please confirm that you are using our SC18IM704-EVB? 

I hope this information helps.

Regards,

David

[andrtnnr]

Hey David,

Here is a screen shot of how the logic starts getting degraded at 38400

https://i.postimg.cc/WzB79wLL/Bad-38400.pnghttps://i.postimg.cc/WzB79wLL/Bad-38400.png

and the same message is totally fine at 19200

https://i.postimg.cc/y8HNK37V/Good-19200.png

I think you might be mistaken regarding 38.4 kbit/s. Because baud is symbols/second. But all definitions seem to say its (5x38400) kbit/s, which is still within the datasheets specs. Maybe Im wrong, could be the root of the problem.

Yes I am using the SC18IM704-EVB

[andrtnnr]

If there is no useful support here, I will probably just go with a different IC. This is a major blocker...