colstart is not correct

[lesamouraipourpre]

With the Adafruit 2.13" HD Tri-Color eInk / ePaper Display FeatherWing the default colstart value passed to the displayio.EPaperDisplay constructor is wrong

Adafruit_CircuitPython_SSD1680/adafruit_ssd1680.py

Lines 82 to 100 in eefe74c

	super().__init__(
	bus,
	start_sequence,
	stop_sequence,
	**kwargs,
	ram_width=250,
	ram_height=296,
	busy_state=True,
	write_black_ram_command=0x24,
	write_color_ram_command=0x26,
	black_bits_inverted=False,
	set_column_window_command=0x44,
	set_row_window_command=0x45,
	set_current_column_command=0x4E,
	set_current_row_command=0x4F,
	refresh_display_command=0x20,
	colstart=1,
	always_toggle_chip_select=True,
	)

Using the simpletest example with the included display-ruler.bmp I get the following with lines of noise at the bottom and lines missing from the top:

If I change the colstart in the library to 8 it lines up correctly:

Do the other Adafruit products that uses the SSD1680 library work with default colstart value or do they all suffer from this noise?

I can either do a PR to:

• change the default value to 8
• expose colstart as a SSD1680 constructor parameter and add constants and documentation

Could someone with access to the other displays report back on their behaviour please?

[Neradoc]

Somebody on discord ran into that issue too, setting colstart to 8 worked. With this wing:
Adafruit 2.13" Monochrome eInk / ePaper Display FeatherWing - 250x122 Monochrome with SSD1680
I don't have the hardware to properly test this. So, to whoever knows about the hardware:

• is it common to all such displays ?
• or is it a case of different models needing a different setup ?
• if so, can the hardware be detected in the code or should there be variants in the library ?

[jposada202020]

I have tested this in the following products

• `Adafruit 2.13" Tri-Color eInk Display FeatherWing
• `Adafruit 2.13" Mono eInk Display FeatherWing

In both cases problem is solved when using colstart=8 as per example in #9 . However, did not test on

• `Adafruit 2.13" Tri-Color eInk Display Breakout

So maybe for this particular case colstart=8 would not work

@dave-ct just pinging you in case you own an Adafruit #4947 Display

[dave-ct]

Just have the same two featherwings also.

[jposada202020]

Thanks dave-ct. Hopefully I can test over the weeeknd. Thanks again for your help dave

[tannewt]

Note, I don't expect the default to always be right because multiple displays use the same driver IC in different configurations. I'd expect to have a simpletest for each product that would include the correct colstart.

[makermelissa]

I can either do a PR to:

• change the default value to 8
• expose colstart as a SSD1680 constructor parameter and add constants and documentation

This may have been caused by #9, which was an attempt to fix #10 by doing the second item on the list.

I have tested this in the following products

• `Adafruit 2.13" Tri-Color eInk Display FeatherWing
• `Adafruit 2.13" Mono eInk Display FeatherWing

In both cases problem is solved when using colstart=8 as per example in #9 . However, did not test on

• `Adafruit 2.13" Tri-Color eInk Display Breakout

So maybe for this particular case colstart=8 would not work

@dave-ct just pinging you in case you own an Adafruit #4947 Display

Keep in mind that there are older and newer versions of both the tri-color and monochrome 2.13" Displays. Newer displays use the SSD1680, the older monochrome uses the SSD1675 driver, and the older tri-color uses the ILI0373. Sometimes the older displays almost work using the wrong driver, which may have been the case with #9. So we just want to be sure the display that is being tested is in fact the newer one. If unlabeled, I usually tell these apart by comparing the tabs on the left side of the display and comparing to product photos.

Apparently it's been a non-settable 1 since I first wrote the driver. I'm not sure why it's suddenly shifted. I don't think there was any change to the displays themselves. Maybe an internal change in CircuitPython did something. It may be worth testing against an older build as well.

Note, I don't expect the default to always be right because multiple displays use the same driver IC in different configurations. I'd expect to have a simpletest for each product that would include the correct colstart.

Good point. We should probably just leave it 1 for now and just update the examples as needed.

[jposada202020]

Got it, I do have some SSD1680 and some SSD1675, ill take a look over the weekend and make the Pr for the examples. Thanks :)

[makermelissa]

Thanks. SSD1675 should use https://github.com/adafruit/Adafruit_CircuitPython_SSD1675

[jposada202020]

Some news, I tested on the Adafruit 2.13" Tri-Color eInk Display Breakout and the colstart=8 is indeed the correct values, showing that all three screens for the SSD1680 Adafruit family are good to go. Will submit a PR for a new example and with that close this issue. Next step the SSD1675