TCAN1167-Q1: tPOWER_UP

Hi Hirata-san,

I see you posted a few related questions around the same time, so I'll address them all here and remove the other posts. 

Q1. When if this CVCCOUT is changed, what has become of tpower_up?

tPOWER_UP defines the delay between when Vsup rises past 4.5V to when Vccout reaches 90% of its final value. Because the rise-time of the Vccout supply is dependant on the capacitance of the supply rail, we define this behavior at a given output capacitance. If this output capacitance is increased, the rise-time of the Vccout voltage will increase and therefore the effective tPOWER_UP will be greater. The amount that this time increases will depend on many factors including the Vsup voltage level and rise time. We do not provide an equation to calculate such effects as the specification is based on device observations under certain conditions. If a customer plans to have significantly different external conditions, the most accurate results for expected behavior would be to observe the rise-time in an experimental test. Evaluations modules are available for TCAN1167 if the customer wishes to conduct such a test. 

Q2. How long does it take for the TCAN1167-Q1 to output the VCCOUT after the WAKE pin rising/falling edge in the Sleep mode?(min, typ, max)?

The delay between a registered transition on the WAKE pin and the device transition into Standby mode (INH=high) is defined by tWAKE as 40us. At this time, the Vccout supply will also be enabled and start to rise. The rise-time of this supply is dependant on the output capacitance, but will likely take on the order of a couple of milliseconds. 

Q3. Do you have the response time (min, typ, max) of the blow state transition by SPI communication?
Standby Mode　→　Normal Mode
Normal Mode →　Standby Mode
Standby Mode　→　Sleep Mode

This timing is not characterised for TCAN1167. However, from my experience they are on the order of around 100us - 200us maximum. 

Datasheet timing specifications are created based on observational data from device characterisation. If a test was not conducted during this characterisation, we cannot specify a min, typ, max timing for the behavior. When we receive inquiries like these on certain behaviors that were not included during characterisation, we can provide estimates and approximations based on individual measurements or experience. However, because these values do not come from characterisation data we cannot guarantee such values. 

Let me know if you have any more questions.

Regards,
Eric Schott

Hi Hirata-san,

Your idea here is right, but I think the tmode_slp_reset time would be better applied here than twake. 

The time between a WUP or LWU event and when INH asserts high (tmode_slp_reset) is specified to be 50us maximum. This also represent the delay between such a wake event and when Vccout becomes active (with some margin for INH rise time). Taking into account the rise-time of the Vccout supply at the measured capactive loads, we can add these times together to get an estimate of the combined delays:

10uF load:   800us + 50us ~= 850us
100uF load: 2000us + 50us ~= 2050us

Is this the condition that the customer is looking for? Do you know if there is a requirement for Vccout response time from a wake event? Or are these questions to estimate the expected response time of the system?

Regards,
Eric Schott

Hi HIrata-san,

I was interested by this difference as well. I used the same Vsup (12V), but did not include a bus load on CANH and CANL. This may be where some of the difference comes from as my setup did not have the extra load on the Vccout supply which is also used by the CAN transceiver. However, I think it's quite likely there were other differences in the setup as I was testing this on an EVM. When such testing is done for characterisation, it is done on high-end automated test equipment that is a much more controlled environment than an evaluation tool. In addition to this, datasheet values typically build in some healthy margin to ensure all corners of operational conditions are covered well. 

I think the variance seen in my measurement values are representative of the expected difference given a larger capacitive load in an application. However, these one-off measurements are very specific. The datasheet values need to guarantee device behavior, so the maximum expectation must be different. 

Regards,
Eric Schott