Jetson Nano GPIO Drive Capability
Jetson Nano GPIO Drive Capability
Hi,
Its very hard to find this out but the Nano uses a TI level shifter TXB0108 to drive the Expansion GPIO pins from a 1.8V level.
And this device can only drive a low static current (via a 4K resistor). So you need to buffer it (to 50K it says with <70pF).
But worse, the device is bi-directional so if you pull its output high, it will latch high, assuming the uP Port driving it is not yet configured as an output.
https://www.ti.com/lit/ds/symlink/txb01 ... e.com%252F
I hope this helps someone else.
Do not use the GPIO unbuffered, as you would with a RPi. See the data sheet.
73, Mike
Its very hard to find this out but the Nano uses a TI level shifter TXB0108 to drive the Expansion GPIO pins from a 1.8V level.
And this device can only drive a low static current (via a 4K resistor). So you need to buffer it (to 50K it says with <70pF).
But worse, the device is bi-directional so if you pull its output high, it will latch high, assuming the uP Port driving it is not yet configured as an output.
https://www.ti.com/lit/ds/symlink/txb01 ... e.com%252F
I hope this helps someone else.
Do not use the GPIO unbuffered, as you would with a RPi. See the data sheet.
73, Mike
Re: Jetson Nano GPIO Drive Capability
Hi Mike
A timely reminder. There should be a brief article in the next CQ-TV describing the buffer that I use for PTT switching on my Jetson Nano.
Dave
A timely reminder. There should be a brief article in the next CQ-TV describing the buffer that I use for PTT switching on my Jetson Nano.
Dave
Re: Jetson Nano GPIO Drive Capability
Hi Dave,
I just 'formally' built my Jetson nanobox (its all in a box - that took ages). I used a Verobox 75-265740 with a 1.5mm thick ally sheet (Wickes) mid box height, double sided chassis. It is a very, very tight fit to get it all in. I have a 5V fan atop my Jetson (so added height there).
I put the Jetson and the +8V linear and +5V SMPSU on the top. And a PI Cam V2 to HDMI extender socket.
LKV373A and router are on the bottom. I used 2mm threaded metal stand-offs to mount the Router and 2.5mm threaded metal for the Jetson. All else on 3mm threaded metal stand-offs (Farnell). I used 2mm/2.5mm and 3mm screws nuts and washers (CPC/Farnell).
Anyhow, I put in a BSS138 (digital MOSFET) interface to the Jetson PTT on Pin 40 of the expansion header for near perfect electrical isolation.
This is the first time I have tried the brand new Jetson Install with the PTT function/circuit. This MOSFET circuit worked when I checked it with a 4K7 to +5V.
And I get no 3V3 on Pin 40 on Tx. I tried all the Expansion pins with my DMM and none change state on PTT. Some are always on 3V3. Even when no Tx.
Does Pin 40 actually work as a PTT, please, with this latest (September/October) Nanobox? I got everything else working.
73, Mike
I just 'formally' built my Jetson nanobox (its all in a box - that took ages). I used a Verobox 75-265740 with a 1.5mm thick ally sheet (Wickes) mid box height, double sided chassis. It is a very, very tight fit to get it all in. I have a 5V fan atop my Jetson (so added height there).
I put the Jetson and the +8V linear and +5V SMPSU on the top. And a PI Cam V2 to HDMI extender socket.
LKV373A and router are on the bottom. I used 2mm threaded metal stand-offs to mount the Router and 2.5mm threaded metal for the Jetson. All else on 3mm threaded metal stand-offs (Farnell). I used 2mm/2.5mm and 3mm screws nuts and washers (CPC/Farnell).
Anyhow, I put in a BSS138 (digital MOSFET) interface to the Jetson PTT on Pin 40 of the expansion header for near perfect electrical isolation.
This is the first time I have tried the brand new Jetson Install with the PTT function/circuit. This MOSFET circuit worked when I checked it with a 4K7 to +5V.
And I get no 3V3 on Pin 40 on Tx. I tried all the Expansion pins with my DMM and none change state on PTT. Some are always on 3V3. Even when no Tx.
Does Pin 40 actually work as a PTT, please, with this latest (September/October) Nanobox? I got everything else working.
73, Mike
Re: Jetson Nano GPIO Drive Capability
Hi Mike
As part of writing the article, I rebuilt all the Jetson software from a blank card, and I can confirm that the PTT does work on Pin 40. There is a 15 second delay after selecting TX on the Portsdown to allow for Lime calibration.
You can check on the Jetson there should be a file /home/nano/tx_on.sh. You can use the commands in it to check the GPIO functionality:
Is the LimeSDR Mini that you have connected to your Jetson transmitting (even without a PTT indication)?
Dave
As part of writing the article, I rebuilt all the Jetson software from a blank card, and I can confirm that the PTT does work on Pin 40. There is a 15 second delay after selecting TX on the Portsdown to allow for Lime calibration.
You can check on the Jetson there should be a file /home/nano/tx_on.sh. You can use the commands in it to check the GPIO functionality:
Code: Select all
#! /bin/bash
cd /sys/class/gpio
echo 78 > export
cd /sys/class/gpio/gpio78
echo "out" > direction
echo 0 > value
sleep 15
echo 1 > value
exit
Dave
Re: Jetson Nano GPIO Drive Capability
Hi Dave,
Thank your for the reassurance that the Jetson PTT/GPIO operation has been tested with the latest nanobox build.
The only odd thing I did was to not use the default Jetson user name and password (I now know NOT to do this). I had to edit the install script.
This is the only thing I did that was in any way a deviation from the official BATC install and build procedure.
Yes, the Lime Mini IS providing valid H265 DVBS2 output on 2.40925 GHz directly into my Minitioune (via -20dB attenuator).
I can see the video and hear the sound for BOTH the Pi CAM (V2) and LKV373A inputs (again, using my GoProHero3+ as a reliable HDMI source).
So the only thing that looks duff is the PTT operation.
I will try your suggestion to manually run the script you kindly highlighted and see what happens then.
That was going to be my next test - just to make sure something is really happing with the GPIO, given a simple test program to drive it.
I will let you know what happens...
73, Mike
Thank your for the reassurance that the Jetson PTT/GPIO operation has been tested with the latest nanobox build.
The only odd thing I did was to not use the default Jetson user name and password (I now know NOT to do this). I had to edit the install script.
This is the only thing I did that was in any way a deviation from the official BATC install and build procedure.
Yes, the Lime Mini IS providing valid H265 DVBS2 output on 2.40925 GHz directly into my Minitioune (via -20dB attenuator).
I can see the video and hear the sound for BOTH the Pi CAM (V2) and LKV373A inputs (again, using my GoProHero3+ as a reliable HDMI source).
So the only thing that looks duff is the PTT operation.
I will try your suggestion to manually run the script you kindly highlighted and see what happens then.
That was going to be my next test - just to make sure something is really happing with the GPIO, given a simple test program to drive it.
I will let you know what happens...
73, Mike
Re: Jetson Nano GPIO Drive Capability
Hi Dave,
I tried manually executing the tx_on and tx_off scripts and they did not work with "Permission denied" and "Operation not permitted" errors reported.
I ran the individual script commands independently and the permissions then seemed to be magically granted - until a reboot (always check after a reboot).
So there was a script permissions problem, which I reasoned to being due to my use of Jetson user and passwords other than those mandated.
And so, to make totally sure, I rebuilt my Jetson completely with the "standard" Jetson user name and password.
Apart from one EXTRA SSH key authentication step not described (where I had to delete the initial Pi->Jetson SSH key THEN re-login) all is working with the DATV output AND the PTT operation.
So the problem was, permissions to execute tx_on.sh and tx_off.sh were not added to my Jetson user group, or so I read.
Solved: Never deviate from the specified user and passwords is my advice.
Thanks for your help,
73, Mike
I tried manually executing the tx_on and tx_off scripts and they did not work with "Permission denied" and "Operation not permitted" errors reported.
I ran the individual script commands independently and the permissions then seemed to be magically granted - until a reboot (always check after a reboot).
So there was a script permissions problem, which I reasoned to being due to my use of Jetson user and passwords other than those mandated.
And so, to make totally sure, I rebuilt my Jetson completely with the "standard" Jetson user name and password.
Apart from one EXTRA SSH key authentication step not described (where I had to delete the initial Pi->Jetson SSH key THEN re-login) all is working with the DATV output AND the PTT operation.
So the problem was, permissions to execute tx_on.sh and tx_off.sh were not added to my Jetson user group, or so I read.
Solved: Never deviate from the specified user and passwords is my advice.
Thanks for your help,
73, Mike