TL16C2552: About TXB output level during RESET = H.

Hi Kazuya,

I can double check this for you but can you tell me how the customer is biasing some of the other pins? (Chsel. IOR, IOW, CS)

Thanks,

-Bobby

 Hi Bobby,

 Thank you very much for your prompt reply.

 The customer checked each signal level as the follows.

 CHSEL = "L" level,

 /IOR = "H" level,

 /IOW = "H" level,

 /CS = "H" level.

 If you have any requirment for other signal, please let me know.

 Thank you and again and best regards,

 Kazuya.

 Hello Bobby,

 I know you are very busy but could you please give me your reply?

 Thank you and best regards,

 Kazuya.

Hey Kazuya,

I've checked to see if I had any boards set up with this device (both in office and at home) but it looks like I only have the 2550 and 550 but not the 2552. I need to order samples and boards to verify this for you.

I'll aim for the end of next week.

-Bobby

 Hello Bobby,

 Thank you very much for your reply and strong supports.

 I'm looking forward to receiving your reply this week.

 Thank you again and best regards,

 Kazuya.

 Hello Bobby,

 Thank you very much for your strong supports.

 Do you have any update for TXB output level when RESET = H?

 The customer really needs your feedback.

 Thank you again and best regards,

 Kazuya.

Hi Kazuya,

I received the breakout board and samples for the device last Wednesday and put in a soldering request on Thursday but it looked like they had a lot in the queue (wouldn't finish until late on Friday). I am at he office today and will pick up the units. I should be able to do the testing for you this week. Sorry for the delay.

-Bobby

 Hi Bobby,

 Thank you very much for your strong supports.

 I'm looking forward to receiving your measurement result.

 Thank you again and best regards,

 Kazuya.

Hi Kazuya,

Sorry for the delay. I just did the testing today and found that when RESET is HIGH with the configurations you gave me, TXB was found to be driving LOW. When I pulled RESET to GND, the TXB pin went back to HIGH.

-Bobby

 Hi Bobby,

 Thank you very much for your test result.

 Could I ask you a few additional questions as the follows?

 Q1, TL16C2552 datasheet (SLWS163A) says in Table 2 on page 22

       that TX(B) becomes "High" level under RESET STATE.

        Does this "RESET STATE" mean "Just after RESET goes Low from High"?

 Q2. The customer wants to keep TXB level high when RESET is High level.

        Is there any way to keep TXB high level when RESET is High level?

        If CHSEL, /IOR, /IOW, /CS are set to other setting, does TXB becomes High level?    

 Q3. The customer replaced PC16552 with TL16C2552 for their current products because PC16552 was EOL.

       PC16552 TXB was High level when RESET was High level as Table 2.

       Is this one of different point between TL16C2552 and PC16552?

  Your reply will be very helpful for the customer and me.

 Thank you again and best regards,

 Kazuya.

 Hi Bobby,

 Could you please give me your reply for the customer and me?

 Thank you very much and best regards,

 Kazuya.

Hi Kazuya,

I've been out of the office for a few days.

---------------------

Kazuya Nakai59 said:

 Hi Bobby,

 Thank you very much for your test result.

 Could I ask you a few additional questions as the follows?

 Q1, TL16C2552 datasheet (SLWS163A) says in Table 2 on page 22

       that TX(B) becomes "High" level under RESET STATE.

        Does this "RESET STATE" mean "Just after RESET goes Low from High"?

[Bobby] It doesn't look like it explicitly states whether this is during the reset condition or immediately after. If I were to interpret this, I would assume after (not during). I believe this table to be taken from a competitor device when this device was being designed many years ago. The engineer who spec'd the device isn't here for us to ask so I don't for certain if the table for aimed at 'a during reset' condition when it was created.

 Q2. The customer wants to keep TXB level high when RESET is High level.

        Is there any way to keep TXB high level when RESET is High level?

        If CHSEL, /IOR, /IOW, /CS are set to other setting, does TXB becomes High level?    

[Bobby] I can try to test this out, I'll get back to you tomorrow night (US time) and see if modifying some of the inputs will do anything. I suspect it won't affect the TX pin but may affect the D0-D7 pins.

 Q3. The customer replaced PC16552 with TL16C2552 for their current products because PC16552 was EOL.

       PC16552 TXB was High level when RESET was High level as Table 2.

       Is this one of different point between TL16C2552 and PC16552?

[Bobby] Just to verify, the PC16552 is the National Semiconductor device under the lit number:SNLS387B?

  Your reply will be very helpful for the customer and me.

 Thank you again and best regards,

 Kazuya.

 

 

-Bobby

 Hi Bobby,

 Thank you very much for your strong supports.

 I'm looking forward to hearing your check result.

 And yes for Q3. The PC16552 datasheet lit number is SNLS387B.

 Thank you again and best regards,

 Kazuya.

 Hi Bobby,

 Could I ask you a few additional questions as the below?

 Q1. Is it necessary to input H level of RESET signal after TL16C2552 is powered up?

 Q2. Does TL16C2552 have POR (Power On Reset)?

 Thank you and best regards,

 Kazuya.   

Hi Kazuya,

" Q1. Is it necessary to input H level of RESET signal after TL16C2552 is powered up?"

I would recommend this, my experience with some of the UART devices (not this one in particular) I've seen some weird bugs pop up before due to not toggling RESET at the beginning.

" Q2. Does TL16C2552 have POR (Power On Reset)?"

Normally for this, we would check the design files to look for a PoR circuit but I am not able to locate the design database for this particular device. I know some of our newer UART devices don't have this so I can only assume the older ones likely don't either. My suggestion is for the customer to toggle the RESET pin anyways because I've also seen devices power up incorrectly sometimes due to the Vcc ramp rates (not the UART devices specifically).

-Bobby

 Hi Bobby,

Thank you very much for your comments and suggestions.

 Those are very helpful the customer and me.

 I'm looking forward to hearing your check result.

 Thank you again and best regards,

 Kazuya.

Hi Kazuya,

I wasn't able to get the TX level HIGH during the RESET condition when playing around with the other inputs. If the customer is able to redesign the board, maybe they can place a switch or something infront of the TX line as a work around.

-Bobby

 Hi Bobby,

 Thank you very much for your strong supports.

 I understood there is no way to set TX high during the RESET = H condition.

 Could I ask you about Power On Reset again?

 The customer found your thread about the POR in E2E site as the follow.

 https://e2e.ti.com/support/interface/f/138/t/922597

  In this thread, you said that the device should reset itself upon power up.

 Does this mean TL16C2552 have Power On Reset circuit?

 Thank you again and best regards,

 Kazuya.

 Hi Bobby,

 I think you would answer for my question like the following.

 TL16C2552 probably have a POR though it is not 100% because the device database was not found.

But the function reliability of the POR is not so high if so.

Therefore TI recommends to input reset signal from externally after powered up to ensure the device reset.

 Is my understanding correct?

 Thank you very much and best regards,

 Kazuya.

"TL16C2552 probably have a POR though it is not 100% because the device database was not found.

[Bobby] Correct, a PoR is good design practice and I would expect this to be part of all digital based devices but I can't say 100% that the device has it without looking at the design database for this device.

But the function reliability of the POR is not so high if so.

[Bobby] I've seen this be dependent on slew rate of Vcc before. Usually there is only one PoR condition (common with older devices) but newer devices nowadays have two (rising and falling) to help reliably reset going below and above the falling and rising PoR thresholds. 

Therefore TI recommends to input reset signal from externally after powered up to ensure the device reset.

[Bobby] Yes, sometimes Vcc ramp does not always reset a device properly due to the slew rate or even off time. A reset pulse is the most reliable way to properly reset a device."

-Bobby

 Hi Bobby,

 Thank you very much for the detail explanations.

 The customer would like to ask you the following question. Could you please give me your answer for the customer?

 Q. If the previous your answers are correct, are your answers in the following thread incorrect?

 https://e2e.ti.com/support/interface/f/138/t/922597

 Thank you again and best regards.

 Kazuya.

 Hi Bobby,

 Thank you for your strong supports.

 Could you please give me your answer to the question?

 Thank you gain and best regards,

 Kazuya.

Hi Kazuya,

Your previous question is the same as the last question, isn't it?

The device should have a PoR though typically we would want to toggle reset as good practice. My answer in that link previously assumed we had the PoR (which I can't verify without a database). I suspect this since the reset trigger is not inverted. When a reset is inverted (ACTIVE LOW), typically the reset mechanism is the power up with the reset tied to Vcc.

-Bobby

 Hi Bobby,

 Thank you very much for your reply.

 Your reply in the URL says "You can tie the reset pin to GND if you don't have any need to control it with a GPO."

  I think that the reply means TL16C2552 has POR clearly.

  But at this moment, you assume TL16C2552 doesn't have POR because the reset trigger is not inverted (RESET is active "H").

 If so, .Is "You can tie the reset pin to GND if you don't have any need to control it with a GPO." incorrect at this moment?

 Thank you again and best regards,

 Kazuya.

Hi Kazuya,

Sorry for bringing this thread back up but I noticed that in the system i forgot to follow up with you.

" If so, .Is "You can tie the reset pin to GND if you don't have any need to control it with a GPO." incorrect at this moment?"

Yes, I think it is better to think this way as so that we can take care of the 'what if' case.

Thanks,

-Bobby