Connecting TUMPA V2 to Raspberry PI2 with Mac OSX as UI
Just got my new shiny TUMPA V2 for debugging baremetal programs on RPI 2.
I decided to get one after I graduated from using USB-TTL console cable. With TUMPA I hope to use both comm. channels to run console as well as JTAG/OpenOCD/GDB debugger.
Here are some details and a few questions in the end.
Development Environment
Terminal: Mac OSX Sierra with minicom 2.2 emulating VT102, 115200 8N1
Target: Raspberry Pi 2 with ArmV7 quad core 1 GB RAM running Raspbian Jesse
JTAG/Console Programmer: TUMPA V2
Task 1: Get TUMPA channel B (TTL port) functional with Mac Driver with RPI2 has target
Followed the recommendation by “TUMPA Mac Driver” posting by mitchell486
to the letter. See
http://www.tiaowiki.com/forums/index.php/topic,4334.0.html
I got it working through loopback testing without target host attached.
To reiterate some important tips along the way are:
1. Before installing Mac VCP driver FTDIUSBSerialDriver_v2_3.dmg, disable Apple FDDI driver located at /System/Library/Extensions/AppleUSBFTDI.kext. by renaming it *.disable. In order to do that for Sierra, you have to disable SIP - System Integrity Protection. This requires booting mac in recovery mode (hold Command R while power-up): invoke terminal program from menu and type “csrutil disable”. You can verify later by “csrutil status”
2. Install FTDIUSBSerialDriver_v2_3, code signed by Apple. The installation should proceed without a hitch!! Edit info.plist per mitchell486 posting.
3. After mac reboot, and loading the FTDIkernel extension, you should see TUMPA device in system info (about Mac, System Report) under USB tab
4. Enable TUMPA TTL port by relocating 8 jumpers from default RS232 setting.
5. Setup TTL com port for loopback by shorting pins 1-2, 3-4, 5-6 ( note I did not connect DCD as I am short of extra cables !!)
6. Power up TUMPA from mac USB port and check for two new devices: /dev/cu.usbserial-?????A/B as channels A and B.
The loopback test was successful but connecting target RPI2 did not work.
For RPI2, I tried two scenarios – powering RPI from Mac and Powering RPI from another source as described below.
Case1 : Powering RPI from TUMPA TTL port
Connected TUMPA TTL pin 1,2 to RPI GPIO pin 8 & 10 for TX & RX signals
Connected TUMPA TTL pin 10 & 12 to RPI GPIO pin 4 & 6 for +5V and GND
I see /dev/cu.usbserial-?????A & B but the terminal did not respond
Then I tried by shorting TUMPA pin 3-4 and 5-6 to hardware disable flow control, no luck
Case2: Powering RPI from separate power source
Same connection as above except disconnected 5V and GND cables in previous case. This time the minicom terminal responded with two ??. Every time I type I get ? Apparently the hardware is working but some comm. issues to be resolved. Looks like garbledup message. However, RPI booted as expected
Need help. Any guidance on how to proceed?
Note:
I did not try JTAG channel-A yet as still struggling with RPI GPIO pins to be reprogrammed for ALT4/5 function as a prerequisite to starting JTAG session.
I have not spent much time except I cannot use the gpio utility which is not available in baremetal programming. Checking out DWELCH raspberry pi github posting on how to.