TMP451-Q1: TMP451 detects OPEN FAULT after initialization

Hi Meredith,

We do not have the waveforms at the fault site yet . The waveform during normal operation can be attached, and we ensure that the signal of D+ is acceptable and within the design range.

Hello Xiaochen,

The signal that you provided looks like it would function but it is a little noisy. Is this the signal right after power up or after restart? Can you provide the signal right after power up?

Best Regards,

Meredith McKean

Hi Meredith,

we physically disconnected the resistor between SOC and tmp451-Q1, the 0x02 register comes 0x0C, (open and Rlow), but actually the issue we met the 0x02 register comes 0x84( only open fault), So I think it not a physical bug as it recovered in next power cycle, right? so M y question is that what will cause the tmp451-Q1 get ge OPEN fault?  thanks.

Hi Meredith,

The signal provided is the signal just after the power on. The green one is the power supply of the TMP451, and the yellow one is the D+ input signal of the TMP451.

hi TI expert, we catch the failed waveform as below, when we write register 0x09 to 0x40 ,which means let tmp451 enter SDmode, and the alert pin of tmp451 drive low and report OPEN fail. and the yellow channel is D+ , I think there is no obvious  abnormal. could you help to explain why entering SD mode and report OPEN ocassionally.thanks,

Hello Ziyang,

There should not be a connection between SD mode and the alert line pulling low. An open condition is when the D+ exceeds (V+) – 0.3 V. The D+ may be idle between conversions. A short condition is when D+ is 0V. What was in the configuration register before writing 0x40? Can you provide a register dump before and after this change? What are your alert and therm limits set to? What are your status bits?

Best Regards,

Meredith McKean

hi TI expert

Meredith McKean said:

What was in the configuration register before writing 0x40?

A：0x00

add aonther re-produce waveform,only a samall glitch when alert pin of tmp451 drive to LOW.  and the register dump is shown as below.

add normal waveform which alert pin of TMP451 stay high after writing 0x09 register to 0x40

Hello Ziyang,

It sounds like the device was with default values and your wrote 0x40 to place the device in shutdown mode. I'm still confused to what you are writing to the device before the 0x40 command to the 0x09 register. The glitch makes me think that there is something else in your system causing the open status issue. Does the device recover with no additional changes or writes? 

Best Regards,

Meredith McKean

hi TI expert

1、no writing process to tmp451 before 0x40 commands to the 0x09 register. we only read the register as below picture green  area show

2、the glitch is so small that could not cause "OPEN " which should meet with D+ exceed V+ -0.3v, right? maybe it is caused by something wrong inside the tmp451.what I mean is that the glitch maybe a phenomenon/outcome of the "OPEN " issue;

3、yes, the device recover (what it mean the alert pin drive high) after reading the 0x02 register (0x84) of tmp451

Hello Ziyang,

This is not normal behavior for TMP451. I believe that something else is causing this issue. Is the remote junction inside another IC? If so, what IC. What are you measuring temperature of?

Best Regards,

Meredith McKean

yes, the remote junction inside another soc chip. ambient temperature 85C and water coolling 60C. the issue only happen in these condition. 

firstly, D+ exceed V+ -0.3v, report "OPEN",exactly there is no such waveform. 

secondly, if there is something wrong inside the SOC, what happen will cause "OPEN", the BJT inside the SOC short or open or any other issue?

Hello Ziyang,

I'm currently not seeing any issues presented in this thread. I want to investigate in how the SOC device is being powered on in comparison to the local TMP451 die. Is the supply stable? If not, are there glitches? Does the SOC and TMP461 turn on at the same time? 

Best Regards,

Meredith McKean

sure ， firstly tmp451 powerup , then SOC powerup and then I2C access the tmp451 and begin to configure the 0x09 register to 0x40 command.

and I found there is also other customers meet same issue, but also do not have RC:TMP451-Q1: TMP451 random OPEN false fault after device initialization - Sensors forum - Sensors - TI E2E support forums, Ithink it maybe something wrong with tmp451?

Hi Ziyang,

The diode can sometimes be hidden when the SOC is turned off. Can you confirm through waveforms that the power lines are stable for both the SOC and TMP451?

Best Regards,

Meredith McKean

Hi Ziyang,

The diode can sometimes be hidden when the SOC is turned off. Can you confirm through waveforms that the power lines are stable for both the SOC and TMP451?

Best Regards,

Meredith McKean

sure we can test， but if stable,then ? 

even it is unstable, can we see D+ exceed V+ -0.3V? is there any possible reasons that make tmp451 report "OPEN" except the D+ -0.3V?  you mean the supply for tmp451 will unstable to 0 V? 

Hello Ziyang,

Lets confirm it the signal is unstable. Can you provide a schematic? I want to see what bypass caps and resistances that you are using. 

Best Regards,

Meredith McKean