TCAN1145-Q1: TCAN1145-Q1 CANINT

Arabella,

Thanks for your patience, we'll get back to you by Wednesday next week with an update.

Regards,

Eric Hackett 

Hi Eric,

Could you please reply to me tomorrow? The customer is so urgent.

Many thanks!

Arabella Zhang

Hi Arabella,

Can they capture some waveforms? Just want to ensure there's no CAN activity on the bus but current can still reach that high.

Regards,

Sean

The customer has two additional questions.
1. When the tcan1145 transceiver configures the selective wake-up enable register through SPI, are there any requirements for the transceiver mode?

2. After testing, the customer found that after the tcan1145 sets the selective wake-up function from standby and switches to sleep mode, there is a 100ms+ process of the static current of about 180uA gradually decreasing. The customer did not find relevant description in the manual. The customer wants to ask if this is a chip feature and is this normal? If so, where is the relevant description? If not, what is the cause?

The customer has two additional questions.
1. When the tcan1145 transceiver configures the selective wake-up enable register through SPI, are there any requirements for the transceiver mode?

2. After testing, the customer found that after the tcan1145 sets the selective wake-up function from standby and switches to sleep mode, there is a 100ms+ process of the static current of about 180uA gradually decreasing. The customer did not find relevant description in the manual. The customer wants to ask if this is a chip feature and is this normal? If so, where is the relevant description? If not, what is the cause?

Hi Arabella,

1. There is no requirement for the transceiver, SPI interface will be available in sleep mode if VIO supplied. Typically the selective wakeup is enabled in standby mode, since the device will first enter standby mode after power on.

 2. In order to decode the CAN data to monitor for a wakeup frame (WUF) the device must draw additional current. Is there any communication on the bus when they saw that current?

Regards,

Sean

1.Thanks for reply.

2. When the bus is silent, the normal sleep current of the transceiver is about 20uA, which is the actual measured value of the customer.
The current situation is that after being configured as sleep, there is no communication on the bus, and the actual measurement shows that there is a current drop of more than 180uA for 100ms+. Why is this?

You can see the current values ​​of maker2 in the two pictures below are 175uA and 90uA respectively.

Hi Arabella,

Would you please ask them for the schematic? And the waveform plots of CAN bus. I would like to see how bus is changing when this current drop happens.

And the current measured is VSUP, not VCC right?

Regards,

Sean

First, let me answer your last question:the current measured is VSUP, not VCC right?
The current of the entire board of the customer's ECU from wake-up to sleep is measured here. As can be seen from the figure, the current drops to the end and lasts for a period of time with a current value greater than 20uA. In addition, we soldered off the 1145 and simulated INH pull-up and pull-down. We found that it only takes about 60ms for the current of the entire board without the 1145 to drop to 0, while the ECU with a complete transceiver keeps the current higher than 20uA for much longer than this time.

The following is a test of the entire board current drop after removing 1145.

Also, has anyone responded to the question about the 400uA current?

Many thanks!

Hi Arabella,

Which supply pin of transceiver is probed for this current measurement? Can you included the voltage on the waveform as well? Current dropping time of the entire board doesn't make sense to me, the decoupling capacitors on supply pins can also contribute to it. The INH signal is an open-source output, it takes time for the signal to decay. Can you plot the INH voltage on the same waveform? In that way we can see how INH was decaying.

For that 400uA current, that happens before entering sleep mode? Or after entering sleep mode? Which supply pin is measured on? Can you also include 400uA current waveform as well (The waveforms you uploaded above have not included the 400uA plot, right?)? 

Regards,

Sean

Hi Sean，

The following is the waveform of the ECU bus captured by the customer after the inh power is cut off.
The environment is that there is no message sent on the bus, the timeout period is reached, the spi configures the transceiver to perform a series of operations, and finally enters the sleep mode and pulls down the inh pin.

Thanks,

Arabella

Hi Sean,

关于您之前提到的问题客户的回复分别如下： 

Which supply pin of transceiver is probed for this current measurement?

--当前产生问题的400uA是测量了整个ECU端的整板电流，当INH关断后，只有收发器工作。

 Can you included the voltage on the waveform as well? Current dropping time of the entire board doesn't make sense to me, the decoupling capacitors on supply pins can also contribute to it.

--我们当前的设备无法同时在同一个时间轴上精确测量电平和电流变化，这个若您有解决办法可以提供一下，否则无法提供此图像。

 The INH signal is an open-source output, it takes time for the signal to decay.  Can you plot the INH voltage on the same waveform?  In that way we can see how INH was decaying.

--同上。我们无法同时提供电流下降和INH拉低的图像。若想要参考INH压降的波形，可以参考图1（见附件“图1_INH引脚拉低时CAN总线波形.jpg”）.这是我们采集的INH引脚下降同时CAN总线变化的图像。这个之前提供给张工了。

For that 400uA current, that happens before entering sleep mode?  Or after entering sleep mode?  Which supply pin is measured on? Can you also include 400uA current waveform as well (The waveforms you uploaded above have not included the 400uA plot, right?)?

--400uA产生在INH引脚拉低后，即SLEEP模式后，测量的是整板ECU电流。

至于为什么测量的是整板电流，并且我们能够认为是收发器导致的，我在下面对当前问题再做一次我们分析过程的描述以及当前情况和我们想要解决的问题。

 我们当前ECU需要支持选择性唤醒，超时未收到指定报文后会配置1145进入SLEEP模式，拉低INH引脚后，仅收发器工作在低功耗模式等待WUF。休眠时需要ECU整板电流不能超过100uA，且从开始下电至100uA以下要求的的最大时间为100ms。但按照我们当前配置收发器进入休眠的流程，产生了异常情况，不能满足我们的要求。在此过程中目前存在两个问题：

问题1：

我们配置收发器进入SLEEP模式的流程为：设置收发器NOMAL模式->0x51寄存器的CANINT写1（0x40）->关闭选择性唤醒->设置STANDBY模式->0x51写0x20、0x52写0x5F、0x53写0x80、0x51写0x4B、0x53写0x40（清除FLAG）->设置选择性唤醒报文、使能选择性唤醒->0x51写0x20、0x52写0x5F、0x53写0x80、0x51写0x4B、0x53写0x40（清除FLAG）->设置STANDBY模式->设置SLEEP模式

之后INH拉低，设备电流下降进入休眠。但我们经过测试发现，在此休眠流程后，设备持续一段时间的400uA电流，大概持续时间为1s左右（自电流下降开始（此时认为INH引脚拉低并关断DCDC电流开始下降）至电流下降至20uA）,在图像中可以看到明显的电流台阶，如图2图3 图4所示。

我们经过排查发现，将上述步骤中的步骤2（0x51寄存器的CANINT写1（0x40））删除后，此400uA持续电流现象消失。我们需要确认的是为什么清除CANINT会导致这一现象，其背后的原理是什么？

问题2：

在临时删除上述步骤2后，再次测量发现，设备在下降至100uA的时间仍超时，设备电流下降至180uA左右后，经过约100ms以上可以下降至100uA以下，最终经过一段时间缓降至20uA左右的正常静态电流值，如图6图7所示。为了验证是否为其余器件导致的问题，我们将收发器焊下来，手动模拟INH拉高拉低，并测量设备电流下降时间，如图5所示。此时整板电流的下降时间大概在60ms左右可下降至8uA。则我可以认为，除了DCDC关断后的电流，其余为收发器的工作导致的。同时软件上我们仅保留配置选择性唤醒、设置STANDBY、设置SLEEP模式后，电流现象一致。软件配置流程应该无问题。

我们需要解决这一部分电流缓降的时间。我们期望的是收发器配置为sleep模式后，电流能够在规定时间内下降至正常的静态电流值。另，收发器原理图如图附件中图8 所示

之后客户又补充测试了产生400uA程序下、产生180uA程序下，收发器VSup\Vcc\Vio三端分别在下电过程中的电流变化，分别对应问题1和问题2

所有的图片见附件

希望能得到回复，谢谢

 TRANSCEIVER_TEST.zip

Hi Arabella,

After talking to digital designer, here's what we are thinking to those phenomenon:

The current they were seeing here is the effect of the 1s tSilence timer. By entering Normal mode, they’re changing the CAN bus bias to 2.5V and starting this timer. Even after entering Sleep mode, this 1s timer must expire before the CAN bias changes back to GND and the selective wake logic and high speed oscillator turns off. As long as the bus bias is 2.5V, the selective wake circuits are active, hence the extra current draw. I would suggest modifying the procedure to not enter Normal mode, because tSilene timer will start in Normal mode  and there is no method to disable it. Just do all the configuration in Standby mode.

We can’t explain the behavior related to CANINT. It’s possible some other step is done or not done, and in the scenario where the current is low by not clearing CANINT, perhaps selective wake isn’t being enabled (can double check that SWE_EN=1, SWCFG=1, and SWERR=0). However, assuming the change suggested in above works, then there's no need to revisit.

Let me know if that solves your customer's question.

Regards,

Sean