Minimum TX2 system requirements
Hello I'm designing a bare bones TX2 system. It only has Ethernet and a debug serial port. I am strapping CHARGER_PRSNT# to ground for auto-start. I am specifically interested in the minimum connections necessary to get a TX2 to boot. I know that Table 88 of the design guide has suggested external connections, but don't put much stock in it as I checked the Carrier Board schematic and not a single one of the recommended terminations were followed. Also, I have strapped my Carrier Board for auto-start and measured the key signals as there is no timing diagram for the auto-start case. Interestingly, CARRIER_PWR_ON goes true 200 mS after Vin is applied and VIN_PWR_BAD# goes false (high) ~300 mS after power is applied. This is the opposite of Figure 3 in the DG. question 1: What is the absolute minimum external parts count (assuming Vin is external) required to boot a TX2? Question 2: Given that the Carrier Board de-asserts VIN_PWR_BAD 100 mS after CARRIER_PER_ON what is the proper sequence for this bare bones system? thanks Fred
Hello

I'm designing a bare bones TX2 system. It only has Ethernet and a debug serial port. I am strapping CHARGER_PRSNT# to ground for auto-start. I am specifically interested in the minimum connections necessary to get a TX2 to boot. I know that Table 88 of the design guide has suggested external connections, but don't put much stock in it as I checked the Carrier Board schematic and not a single one of the recommended terminations were followed.

Also, I have strapped my Carrier Board for auto-start and measured the key signals as there is no timing diagram for the auto-start case. Interestingly, CARRIER_PWR_ON goes true 200 mS after Vin is applied and VIN_PWR_BAD# goes false (high) ~300 mS after power is applied. This is the opposite of Figure 3 in the DG.

question 1: What is the absolute minimum external parts count (assuming Vin is external) required to boot a TX2?

Question 2: Given that the Carrier Board de-asserts VIN_PWR_BAD 100 mS after CARRIER_PER_ON what is the proper sequence for this bare bones system?

thanks
Fred

#1
Posted 02/09/2018 03:33 AM   
Hi, assume you are testing on your own board. The timing of VIN_PWR_BAD# is decided by dedicated circuit on carrier board (page 27 in schematic), do you make any changes to it?
Hi, assume you are testing on your own board. The timing of VIN_PWR_BAD# is decided by dedicated circuit on carrier board (page 27 in schematic), do you make any changes to it?

#2
Posted 02/12/2018 05:51 AM   
Correct, I'm using my own board. The TX2 is powered by 12 volts instead of 19 so the timing will be different. Is the timing from application of input power until VIN_PWR_BAD# goes false at all critical? If so, what is the requirement? The timing diagram (Fig 3) is for a user power button case. I can change the capacitor (equiv to C176) to change the timing as required if you can tell me the specification. thanks Fred
Correct, I'm using my own board. The TX2 is powered by 12 volts instead of 19 so the timing will be different. Is the timing from application of input power until VIN_PWR_BAD# goes false at all critical? If so, what is the requirement? The timing diagram (Fig 3) is for a user power button case.

I can change the capacitor (equiv to C176) to change the timing as required if you can tell me the specification.

thanks
Fred

#3
Posted 02/12/2018 01:43 PM   
Do you measure the timing with dev kit? Is it same as with your custom board? It's weird that CARRIER_PWR_ON is active before VIN_PWR_ON#, in fact, the CHARGER_PRSNT# is controlled by VIN_PWR_ON# in module. Do you make any changes on these signals related circuit design?
Do you measure the timing with dev kit? Is it same as with your custom board?

It's weird that CARRIER_PWR_ON is active before VIN_PWR_ON#, in fact, the CHARGER_PRSNT# is controlled by VIN_PWR_ON# in module. Do you make any changes on these signals related circuit design?

#4
Posted 02/13/2018 04:08 AM   
I mis-spoke about the timing being reversed. It made no sense to me so I re-measured and discovered I had probed the wrong pin (easy to do with 400 pins on the back side). I have attached two scope captures from both the Nvidia carrier board and my own board. I have tried changing the timing capacitor on the VIN_PWR_BAD# line to make it take both more or less time than the Nvidia board. Neither had any effect. I suspect that I am missing something obvious. One clue is that while my board does not boot it is only consuming about 1.3W while the Nvidia board consumes 2.8 to 3.7W during boot up. Note attachment is on original post as it was the only option
I mis-spoke about the timing being reversed. It made no sense to me so I re-measured and discovered I had probed the wrong pin (easy to do with 400 pins on the back side). I have attached two scope captures from both the Nvidia carrier board and my own board. I have tried changing the timing capacitor on the VIN_PWR_BAD# line to make it take both more or less time than the Nvidia board. Neither had any effect. I suspect that I am missing something obvious. One clue is that while my board does not boot it is only consuming about 1.3W while the Nvidia board consumes 2.8 to 3.7W during boot up.

Note attachment is on original post as it was the only option

#5
Posted 02/13/2018 06:09 PM   
Trumany I checked in several times last night to see if we could get a more real time response between us. Did the attachment come thru OK? Since it looks like the power timing is correct I'm back to my original question. What is the minimum number of signals necessary to get a TX2 to boot. thanks Fred
Trumany

I checked in several times last night to see if we could get a more real time response between us. Did the attachment come thru OK? Since it looks like the power timing is correct I'm back to my original question. What is the minimum number of signals necessary to get a TX2 to boot.

thanks
Fred

#6
Posted 02/14/2018 01:49 PM   
problem resolved thanks Fred
Answer Accepted by Original Poster
problem resolved

thanks
Fred

#7
Posted 20 hours ago   
Scroll To Top

Add Reply