For the flickering problem, try setting the refresh rate.
matrix.setRefreshRate (120); // Modify the refresh value
The other problem seems to be more related to the management of the input data, and how it feeds with this data the output matrix of Smartmatrix.
@pinballsp is correct, the ADDX pins all have to be in the same port. When you changed to the Teensy 3.6, did you use pin 37, or stick with pin 16?
SmartLED Shield V4 has a solution to the shortage of pins, actually not requiring any dedicated pins for ADDX, and using an external flip flop to store the ADDX signals instead. Itās out of stock right now but you can check the hardware design files in github and wire up the circuit yourself. 1/32 using the V4 circuit isnāt there yet, with @pinballspās help I hope to get this in the next release.
Thanks to both of you!
@pinballsp Ah hah, the PORTC makes sense, Iāll try going back to Teensy 3.2 and try pin 15 instead. Iāll also try the refresh rate, thanks!
@Louis Correct, I adjusted to pin 37 when using Teensy 3.6. Great, I assume the 74245 & 74374 are the only necessary components needed for that. Iāll order those and give it a shot.
Besides the ADDX signals on the 74374, there are 9 signals that need to be buffered, so youāll need two 74245s, or one 74245 and one single gate buffer.
So Iāve tried setRefreshRate(120) and still getting some flickering, but I think that may just be due to the jumper wires. Iāll try milling a board with short traces and see how that works.
However, when using Teensy 3.2, half the screen doesnāt work (two separate quarters), even though Iām using pin 15 now (PORTC). Switching to Teensy 3.6 and port 37, it does work properly. (I am changing ADDX_TEENSY_PIN_4 when testing between them).
Any thoughts?
You need to also change ADDX_PIN_4 in addition to ADDX_TEENSY_PIN_4. Teensy pin 15 is C0, ADDX_PIN_4 is 0.
There are other PORTC GPIO on the underside of Teensy 3.2 if you want to keep pin 15 free for something else:
30 C11
29 C10
28 C8
27 C9
Oh, I see the internal port numbers need specified. Got it, thank you!
I hooked up the supposedly 1/32 scan panel I have and started adding @pinballspās changes to the code, but quickly found that my panel is actually 1/16 scan. I placed an order for this panel on Aliexpress and hope to have it within the next month so I can make these changes and test.
@Louis ,
Have there been any updates on the new version of the library that would support 1/32 panels on the V4 shield? I plan to use a set of panels like @pinballsp, but Iām using your latest shield and Iām not sure if his code changes are compatible with the external flipflops (or did you use a counter?) that the V4 shield uses to drive the address pins on the panels. I just thought Iād ask before I start trying to implement 1/32 changes myself.
I just pushed changes to support 64-row /32 panels to GitHub. Itās not packaged up into a release yet, but you can download the code. I tested it on the 64x64/32 panel I received yesterday.
@samy commented on poor refresh rate when driving with the Teensy 3.2: thatās to be expected, itās a lot of work for the Teensy 3.2 to refresh a 64x64 panel, and itās even more work when itās /32 scan and not two 64x32/16 panels strung together. The Teensy 3.5/3.6 should do better.
I didnāt add @pinballspās changes to support V1-V3 shields as there are more changes needed. I donāt want to break support for any existing users by driving a new pin unexpectedly. I donāt have the time to wire up a 5th pin on my Teensy, so I wouldnāt be able to test the changes myself. If someone really wants /32 panel support added to the library, and you canāt make the changes yourself following @pinballspās code above, please ask. There are other SmartMatrix Library features I feel are more important, like adding APA102 support, and getting the library ported to a platform with WiFi capabilities.
Thanks Louis. I started having a look at the code and although Iām quite well versed in C, I know very little about the DMA on the Teensy, so it was going to be a big job for me to try to get it working.
@Louis one last question. Do any code or circuit precautions need to be taken to make sure that the SmartLED V4 doesnāt drive the E address line high on a panel that is /16 and may have that pin tied to ground in the HUB75 port?
To be safe, donāt use set panelType to MOD32SCAN when connected to a MOD16 panel. I should probably put a current limiting resistor between the level shifter and the panel on the next rev of the shield, to prevent accidentally burning out that driver if accidentally driving that pin high when connected to ground.
I hooked up a SmartMatrix Shield V4 and Teensy 3.2 to a /16 panel that had the E line grounded, and had it run a sketch with panelType=SMARTMATRIX_HUB75_64ROW_MOD32SCAN for 8 hours. I switched back to a /32 panel after, and the shield was able to drive all the rows, so no critical damage to the driver after 8 hours.
Still probably a good idea not to use MOD32 paneltype on a /16 panel
I just received my matrices from China this week and I couldnāt resist adding one more reply to this thread. I now have a panel that is 256x64 functioning on the teensy. Gif playback is a little slower and I could only do 24 bit color refresh, but it works and looks better than I thought it would.
Now to find a clean way to mount these. 
Looks great! Thanks for posting
Hi all! I am currently working on a project and trying to connect two 64x32 boards (64x32 RGB LED Matrix - 5mm pitch : ID 2277 : $49.95 : Adafruit Industries, Unique & fun DIY electronics and kits), to form a 64x64 board. Although it says that the scanrate is 1/16, I have tried the FeatureDemo example on one board with SMARTMATRIX_HUB75_64ROW_MOD32SCAN and that worked fine. In the images the boards are next to each other, but I would like to create a square. However, the cables that I am currently using are not long enough for that. 
I am using a Teensy 3.2, with the 5th ADDX connected to pin 15 (PTC0). The others are connected to PTC1, PTC2, PTC3 and PTC4. I do not have a SmartLED shield and I hope that I can make it work without it.
I have changed the code as described above, but when I plug in the second screen, it does some weird things (see image). It copies the first screen to the second and only half of the rows are working. Also the letters are not displayed correctly.
Does anyone know what could be wrong?
Thanks in advance!
Hi, did you modify the SmartMatrix Library code to drive the 5th address pin?
Did you change the width and/or height in featuredemo drive the second panel?
I have made some changes to the code, I will list them underneath.
FEATURES DEMO
First line is commented out, width, height and kPanelType are adjusted.
// #include <SmartLEDShieldV4.h> // comment out this line for if you're not using SmartLED Shield V4 hardware (this line needs to be before #include <SmartMatrix3.h>)
#include <SmartMatrix3.h>
#include "colorwheel.c"
#include "gimpbitmap.h"
#define COLOR_DEPTH 24 // known working: 24, 48 - If the sketch uses type `rgb24` directly, COLOR_DEPTH must be 24
const uint8_t kMatrixWidth = 64; // known working: 32, 64, 96, 128
const uint8_t kMatrixHeight = 64; // known working: 16, 32, 48, 64
const uint8_t kRefreshDepth = 36; // known working: 24, 36, 48
const uint8_t kDmaBufferRows = 4; // known working: 2-4, use 2 to save memory, more to keep from dropping frames and automatically lowering refresh rate
const uint8_t kPanelType = SMARTMATRIX_HUB75_64ROW_MOD32SCAN; // use SMARTMATRIX_HUB75_16ROW_MOD8SCAN for common 16x32 panels, or use SMARTMATRIX_HUB75_64ROW_MOD32SCAN for common 64x64 panels
const uint8_t kMatrixOptions = (SMARTMATRIX_OPTIONS_NONE); // see http://docs.pixelmatix.com/SmartMatrix for options
const uint8_t kBackgroundLayerOptions = (SM_BACKGROUND_OPTIONS_NONE);
const uint8_t kScrollingLayerOptions = (SM_SCROLLING_OPTIONS_NONE);
const uint8_t kIndexedLayerOptions = (SM_INDEXED_OPTIONS_NONE);
MatrixHardware_KitV1.h
Added line 4 to define the 5th pin (at PTC0 in Teensy), changed last part of last line: 1 << ADDX_PIN_3 into 1 << ADDX_PIN_4.
#define ADDX_PIN_0 3
#define ADDX_PIN_1 4
#define ADDX_PIN_2 1
#define ADDX_PIN_3 2
#define ADDX_PIN_4 0
#define ADDX_PIN_MASK ((1 << ADDX_PIN_0) | (1 << ADDX_PIN_1) | (1 << ADDX_PIN_2) | (1 << ADDX_PIN_4))
SmartMatrix_Impl.h
I added this part:
#ifdef ADDX_PIN_4
if (i & 0x10) addressLUT[i].bits_to_set |= (1 << ADDX_PIN_4);
#endif
#ifndef ADDX_UPDATE_ON_DATA_PINS
int i;
// fill addressLUT
for (i = 0; i < matrixRowsPerFrame; i++) {
// set all bits that are 1 in address
addressLUT[i].bits_to_set = 0x00;
if (i & 0x01)
addressLUT[i].bits_to_set |= (1 << ADDX_PIN_0);
if (i & 0x02)
addressLUT[i].bits_to_set |= (1 << ADDX_PIN_1);
if (i & 0x04)
addressLUT[i].bits_to_set |= (1 << ADDX_PIN_2);
#ifdef ADDX_PIN_3
if (i & 0x08)
addressLUT[i].bits_to_set |= (1 << ADDX_PIN_3);
#endif
#ifdef ADDX_PIN_4
if (i & 0x10) addressLUT[i].bits_to_set |= (1 << ADDX_PIN_4);
#endif
// set all bits that are clear in address
addressLUT[i].bits_to_clear = (~addressLUT[i].bits_to_set) & ADDX_PIN_MASK;
}
#endifIām having a hard time following everything you changed going from it working to not working.
When it āworked fineā is that after you āchanged the code as described aboveā?
If you remove the second screen, and keep the code the same, does it work better (at least filling the first panel)?