TJA1021 has by default on power on, that the transmitter is off.
It needs the SLP_N line asserted for >Tgotonorm . The transmitter will not work only with bus traffic until it is woken up by signal SLP_N. So there is a scenario where the transmitter reboots due to brownout and our processor does not, leaving the transmitter offline.
TJA1027, TJA1029 appear drop in replacements do not have this feature. They work when they power on, and don’t require this initialization according to the state diagram. So this appears it will solve my problem.
The TJA1029 is specified differently than the TJA1021 mostly with respect to the operating voltage. Operating voltage is up to 18V on The TJA1029 (absolute max = 42) but on TJA1021 both operating and maximum is specified up to 40VDC. We have spent a lot of qualification time using TJA1021 and found this problem.
Question: 1) Is there anybody could answer if TJA1021 and TJA1029 is the same or similar IC design with just different state diagram and specifications? 2) Looking to select which is best TJA1027 vs TJA1029, I also cannot tell by the specification any difference between TJA1027 and TJA1029. | When there is bus traffic an interrupt is passed on the Rx line which wakes up the processor the processor can then set the transceiver enable high to receive the message. |
Lukas, thanks so much for that table. That helped the customer a lot. The final question the customer is asking is internally, what is the difference between the 1021 and the 1029? My customer would love nothing better than to not have to change any hardware due to all of the hours they have spent testing the system which included the 1021. So now moving to the 1029 adds new risk.
I can see that a conference call is planned with customer. This is a little bit out of my scope, so I hope that responsible engineers will help them.
Regards,
Lukas
update from customer:
Our customer cannot accept the software fix of during sleep periodically enabling the TJA1021 in case the transmitter has reset. Unless we do this every <100ms we will not meet LIN specification, and if we enabled it every 100ms we would not meet our power specification. We had proposed that we could do this every fifteen seconds during sleep.
They also are refusing to accept the proposed hardware change to TJA1027, so we need to convince them despite the inconvenience of HW changes this late in the program this is the path we need to take. Using TJA1027 or TJA1029 is cleanest solution since is drop in replacement and solves the problem. Their concern is that we change hardware to fix one problem and this somehow creates another.
I need NXP support to help convince our customer that TJA1027 is same functionality, same design as TJA1021, except it does not need initial enable to work – so this failure mode of the transceiver not being enabled after reset would not exist. TJA1027 Our customer insist to have all the answers within today’s time regarding the difference between TJA1021 design and design. Our specified voltage range is 9-14VDC with the worst case voltage being a clamped 27VDC on a load dump test we are not operating near maximum ratings.
Hi,
I have received this response from app engineer:
You can find the difference between TJA1021 and TJA1027/29 as table below, I have highlighted the main difference as the color blue character.
For the replacement from TJA1021 to TJA1027/29, pls refer to the attached document. ( page 16-19 )
TJA1020 | TJA1021 | TJA1027 | TJA1029 | TJA1022 | |
LIN physical layer | LIN 1.3 | LIN 2.x / SAE J2602 | LIN 2.x / SAE J2602 | LIN 2.x / SAE J2602 | LIN 2.x / SAE J2602 |
LIN channel | one | one | one | one | two |
EMC | - good emission | approved at automotive manufacturers (e.g. EU, US, etc) | further improved EMC performance (even better than TJA1021) | further improved EMC performance (even better than TJA1021) | further improved EMC performance (even better than TJA1021) |
ESD |
| ||||
Normal and low slope | programmable normal and low slope mode | - normal slope variant (/20) | - normal slope variant | - normal slope variant | - normal slope variant |
1st mode after power-on | sleep mode | power-on mode | sleep mode | sleep mode | sleep mode |
Operation during cranking pulse (for e.g. Start-Stop, RKE, …) | LIN operation down to 7.3V | LIN operation down to 5.5V (fully specified in data sheet) | LIN operation down to 5.0V (fully specified in data sheet) | LIN operation down to 5.0V (fully specified in data sheet) | LIN operation down to 5.0V (fully specified in data sheet) |
Defined undervoltage behavior | POR not specified in data sheet |
| - UV detection | - UV detection | |
Load dump robustness | 40V | 40V | 42V | 42V | 42V |
Sleep current | <10 µA | <10 µA | <10 µA | <10 µA | <10 µA |
LIN speed |
| 1 … 20 kBd | 0 … 20 kBd |
| |
TXD dominant time-out | yes | yes | no | yes | yes |
INH (High-side switch for VReg, master termination, battery monitoring) | yes | yes | no | no | no |
Local wake-up input | yes | yes | no | no | no |
Package | SO8 | - SO8 | - SO8 | - SO8 | - SO14 |
Regards,
Lukas
Thank for the answer Lukas.
This as I understand shows that 1029 has same or better performance as 1021 without power off mode and external wakeup function (using wake up input).
Could this be confirmed?
Thanks in advance.
Garabo