MiSTer FPGA Forum

jtsdram48 and jtsdram96 are done.

You can find here https://github.com/venice1200/MiSTer_tt ... s/XBM_Menu
some alternative MENU Pictures for your /media/fat/tty2oledpics/pri Folder.

//Edit
New Picture for Choplifter by RealLarry

I've been observing some strange interactions between my tty2oled and other hardware in my MiSTer setup recently and after chatting about it on the discord it was suggested I post about it here.

MiSTer Configuration:
128MB RAM
Digital IO board v1.2
USB Hub v2.1
- K400+ keyboard dongle
- bluetooth dongle
- 2.4GHz M30 controller dongle
- tty2oled (TTGO-T8 with SSD1322 on 6" USB cable)
MT32-pi (on 6" USB cable)
Wired ethernet connection
HDMI via CEC blocker to Sony sounder and LG OLED55B8

The MiSTer, MT32-pi and tty2oled are all running the most up to date software releases as of this morning 27/8/21

First problem:
With the tty2oled connected and running, at boot the M30 controller is unresponsive, but if I disconnect and reconnect the M30 dongle it works. Rebooting the MiSTer displays the same behaviour, controller is unresponsive until the dongle is replugged. With the tty2oled unit disconnected the behaviour is not observed.

Second problem:
With the tty2oled connected and running, spurious level meter activity is sometimes observed on the MT32-pi. This can be a single meter rising and falling or all meters showing lots of activity, eventually followed by a "sysex overload" message and the MT32-pi appearing to reboot. d0pefish stated this was down to something "wiggling" the serial port pin that the MT32-pi uses to communicate with the MiSTer. Others suggested it may be down to RF interference from a noisy OLED board on the tty2oled, so I connected it up via a 3m USB extender cable, moved the tty2oled to the opposite side of the room, and still observed the strange meters activity.

As a further experiment, while the meters were moving without any MT32-pi audio activity I disconnected the tty2oled from the USB cable and the behaviour persisted. I took this further and reset the MiSTer using the button on the Digital IO board and the meter activity persisted until the MT32-pi was initialised.

At one point while cycling through cores to precipitate the behaviour the MT32-pi display became completely corrupted and shifted. Again, d0pefish attributed this to corruption of the i2c data connection between the pi and the display and suggested RF interference.

I'm not looking for a fix, as I hardly use the tty2oled to be honest, but wanted to present my experience in case someone else has similar issues or there's an obvious solution fix. MT32-pi and tty2oled both use serial communications, although the MT32-pi connects directly rather than via a USB serial device, so I'm really not sure.

Thanks for reporting this.
I have two systems with external powerred USB Hub's v2.1 and never had problems while my tty2oled is connected.
One system has an MT32pi connected as well. Never saw "Self Meter Moving".

Which "2.4GHz M30 controller dongle" do you use?

WiteWulf wrote: ↑Fri Aug 27, 2021 2:32 pmSecond problem:
With the tty2oled connected and running, spurious level meter activity is sometimes observed on the MT32-pi. This can be a single meter rising and falling or all meters showing lots of activity, eventually followed by a "sysex overload" message and the MT32-pi appearing to reboot. d0pefish stated this was down to something "wiggling" the serial port pin that the MT32-pi uses to communicate with the MiSTer. Others suggested it may be down to RF interference from a noisy OLED board on the tty2oled, so I connected it up via a 3m USB extender cable, moved the tty2oled to the opposite side of the room, and still observed the strange meters activity.

I'd a similar issue some days ago after finishing and testing my MiSTer-AjSTer setup. 'till finishing up the whole setup (including mt32-pi by USB cable via user port and tty2oled) I've used a MeanWell GST40A05-P1J power supply and this was working fine. But after completing the setup with the official USB hub (Y-cable from power supply to MiSTer and hub) I could measure voltage drops down to 4.7V on one line (4.8V went fine, depending on moonphase and used drugs). Swapped the power supply against a MeanWell RS-25-5 set to 5.1V (and it's doing this and nothing other) all problems are gone and whole setup is running stable since then.

USB hub supplies BT, WiFi, Keyboard, Mouse, Joystick and tty2oled - MiSTer supplies mt32-pi

After I updated one of my MiSTer's with "Update All" today (Release 20210825) the tty2oled stops working.

I had to run the Updater again to get the Scripts back onto the MiSTer and I had to rename
/etc/init.d/S60tty2oled to /etc/init.d/S99tty2oled to get it working again.

Are there others with this problem?

venice wrote: ↑Fri Aug 27, 2021 3:32 pm Which "2.4GHz M30 controller dongle" do you use?

Whichever one comes with the 8bitdo 2.4GHz M30 controller, one of these:

RealLarry wrote: ↑Fri Aug 27, 2021 3:35 pm
I'd a similar issue some days ago after finishing and testing my MiSTer-AjSTer setup. 'till finishing up the whole setup (including mt32-pi by USB cable via user port and tty2oled) I've used a MeanWell GST40A05-P1J power supply and this was working fine. But after completing the setup with the official USB hub (Y-cable from power supply to MiSTer and hub) I could measure voltage drops down to 4.7V on one line (4.8V went fine, depending on moonphase and used drugs). Swapped the power supply against a MeanWell RS-25-5 set to 5.1V (and it's doing this and nothing other) all problems are gone and whole setup is running stable since then.

USB hub supplies BT, WiFi, Keyboard, Mouse, Joystick and tty2oled - MiSTer supplies mt32-pi

That's interesting, my PSU is pretty good, although I've not measured voltage on rails within the system. I've put my inline USB meter on the tty2oled, though, and it's only pulling 0.08A, so I'd have to be *very* unlucky for that to be causing the problems I think.

WiteWulf wrote: ↑Fri Aug 27, 2021 3:51 pm Whichever one comes with the 8bitdo 2.4GHz M30 controller, one of these:

Have you tried to swap/change the order of the USB Plugs?

Do you have MT32pi Problems with or without the tty2oled because you wrote

while the meters were moving without any MT32-pi audio activity I disconnected the tty2oled from the USB cable and the behaviour persisted.

@WiteWulf:

What happens if you do the following:

- disconnect the MT32-pi
- power up your setup
- connect the MT32-pi

I also think that your problems are related to your power supply. Because there is no power sequencing on the 5V rail the inrush current during startup can be quite high. This can lead to a momentary voltage drop. Maybe this causes an unexpected behaviour during startup of Raspberry Pi and 8bitdo 2.4 GHz dongle.

Double check this as well

Incompatibilites with 8-bitdo m30 controllers / adapters and USB hub board v2.1 when other devices are plugged in

https://github.com/MiSTer-devel/Main_MiSTer/issues/297

venice wrote: ↑Fri Aug 27, 2021 3:36 pm After I updated one of my MiSTer's with "Update All" today (Release 20210825) the tty2oled stops working.

I had to run the Updater again to get the Scripts back onto the MiSTer and I had to rename
/etc/init.d/S60tty2oled to /etc/init.d/S99tty2oled to get it working again.

Are there others with this problem?

With MiSTer Release 20210829 everything works normal again for me.

Sorry for the confusion my but my MiSTer running the 20210825 Release still need /etc/init.d/S99tty2oled as Init Script.

venice wrote: ↑Mon Aug 30, 2021 8:56 am

venice wrote: ↑Fri Aug 27, 2021 3:36 pm After I updated one of my MiSTer's with "Update All" today (Release 20210825) the tty2oled stops working.

I had to run the Updater again to get the Scripts back onto the MiSTer and I had to rename
/etc/init.d/S60tty2oled to /etc/init.d/S99tty2oled to get it working again.

Are there others with this problem?

With MiSTer Release 20210829 everything works normal again for me.

Sorry for the confusion my but my MiSTer running the 20210825 Release still need /etc/init.d/S99tty2oled as Init Script.

For the 2nd time since I have the Mister, a update_all update broke tty2oled and Xon Driver, both times, I had to run the update tty2oled script to install it back again, AND run the Xon Driver option in the pre-installed scripts to enable xbox controller support again.

stefer09 wrote: ↑Mon Aug 30, 2021 12:40 pm For the 2nd time since I have the Mister, a update_all update broke tty2oled and Xon Driver, both times, I had to run the update tty2oled script to install it back again, AND run the Xon Driver option in the pre-installed scripts to enable xbox controller support again.

This is a known issue and we are working on something more persistent.

venice wrote: ↑Mon Aug 30, 2021 12:42 pm This is a known issue and we are working on something more persistent.

The renaming part, i didn't have to do though.

stefer09 wrote: ↑Mon Aug 30, 2021 1:05 pm The renaming part, i didn't have to do though.

Which Release did you install?

After my manual update to the current Release 20210829 tty2oled works again with "/etc/init.d/S60tty2oled".

Well didn't really "install" since my initial install months ago... i ran mister's update_all, that broke tty2oled, like last time i ran the tty2oled update script and when I check now, I see it's /etc/init.d/S60tty2oled

RealLarry has asked to include the tty2oled Updater into update_all which fix this issue automatically

That would work yes!

I just ran the update script and was told I needed the newest version of the .ini so I went and grabbed that but am still getting the message.

Worth noting that the newest version still says 1.4 at the top:

# Version 1.4
#
# Copyright (c) 2021 ojaksch, venice
#
# v1.5 New Options CCSCRIPT
# Add "-echo" to TTYPARAM
# Add PICTURE_REPOSITORY

Also, looking at the default settings I just went and changed those to what I wanted, but might be worth setting some of these to yes as default for the average user. Average person probably does want the script to be kept updated and to overwrite pictures with the new ones, and also grab the text images for when there isn't a graphical one.

# Userdata
CONTRAST="200" # Set/send your Contrast from 0..255, default by Arduino Sketch = 200
ROTATE="no" # Set to "yes" if you want the Display rotated by 180 degrees
SCRIPT_UPDATE="yes" # Set to "yes" if you want to update the tty2oled Scripts by the "Updater"
USBMODE="yes" # Set to "yes" if you use the USB Mode Sketch, set to "no" for SD or Standard Version
OVERWRITE_PICTURE="yes" # Set to "yes" if you want to overwrite existing Pictures if they're updated upstream
USE_TEXT_PICTURE="yes" # Set to "yes" if you want to download the Text-Pictures-Version, default is no
USE_US_PICTURE="no" # Set to "yes" if you want the US Version of the Pictures, default is "no"

Moondandy wrote: ↑Thu Sep 02, 2021 6:28 pm Worth noting that the newest version still says 1.4 at the top:

# Version 1.4
#
# Copyright (c) 2021 ojaksch, venice
#
# v1.5 New Options CCSCRIPT

v1.4 is the actual stable.
Everything I add to the next stable version is already in the INI file. That's the reason why you see v1.5.

Do you still get "update" messages after the update?

Moondandy wrote: ↑Thu Sep 02, 2021 6:30 pm Also, looking at the default settings I just went and changed those to what I wanted, but might be worth setting some of these to yes as default for the average user.

The next version will get more active defaults.

I had swapped the .ini file for the new one and it still gave me the message I couldn't update as it wouldn't be safe, so deleted the ini, ran the script to get a new one, then edited it again with my settings and is working fine again. All good this end now.

I had a thought for a possible script, I am not sure how doable it would be, but it could be useful to have a cleanup script. Running it would check the files in the folder against what is in the repo and flag up which ones are in the drive but not in the repo (i.e. not used). Maybe an optional command to delete the unused ones if that is doable.

Hi everyone! I have issues with my tty2oled. I updated Mister and the tty2oled is not working. I deleted and installed again but i don't know what's happening. It seems load the core and i can see the core name in txt but after a few seconds the screen goes black.
Any clues?

jawler wrote: ↑Thu Sep 09, 2021 8:28 am Hi everyone! I have issues with my tty2oled. I updated Mister and the tty2oled is not working. I deleted and installed again but i don't know what's happening. It seems load the core and i can see the core name in txt but after a few seconds the screen goes black.
Any clues?

According to the photo it looks like that you have updated tty2oled as well, but are using an older sketch on your ESP. I'd update that, too...

Yes, you need to update your ESP.

That's it! i updated my ESP and all works like charm!
Thank you.

I figured out what was going on with me getting the message that I had to delete my .ini file when running the update script. It was because I wasn't connected to the internet. For some reason (maybe related to the linux updates, may be something my end) in the last week my Mister is being slower to connect over my wired internet connection that is was previously. So if I boot up and go run a script straight away then it is generally failing to connect the first time.

I am not sure if it is possible to have an "unable to connect" message instead in this scenario, as being told to delete the .ini file is a weird message to get in this situation.

Ok so since the linux update from the other day, i did have to re-run the tty2oled update script to get the service running again (it happened once before after a big update so no big deal...) but I noticed that the service doesn't start by itself anymore, I have to run the update script everytime or else it doesn't start...

stefer09 wrote: ↑Fri Sep 10, 2021 1:21 am Ok so since the linux update from the other day, i did have to re-run the tty2oled update script to get the service running again (it happened once before after a big update so no big deal...) but I noticed that the service doesn't start by itself anymore, I have to run the update script everytime or else it doesn't start...

This is caused by the new Kernel and Linux setup: The Kernel/System is now initialising the Init scripts faster before the corresponding USB device is setup.
Please be patient. This is fixed in our testing branch and will be released "the next days".