Newhaven Display Forum

Newhaven Products => OLEDs => Topic started by: erlopez on May 08, 2015, 08:26:46 AM

Title: OLED dark horizontal bands
Post by: erlopez on May 08, 2015, 08:26:46 AM
Hello,

I’m evaluating the NHD-1.69-160128UGC3 OLED display. I got it working and can draw pictures and graphics on it. I’m using the SPI interface.

I’m experiencing a series of dark horizontal bands that appear in the screen when something is moving (i.e. when animating graphics). These dark horizontal regions span left to right through the entire screen width.

I did not find anything like this in the forums - Is my unit defective? Is this normal for these OLED modules? Have you seen this before?

The picture below shows a vector balls animation; dark bands appear in regions where the animation (This is not a interlace effect due to screen refresh rate and the camera taking the picture; these dark bands can be seen with the naked eye); the background behind the animated balls is supposed to be solid blue.

(http://www.lopezworks.info/darkbands.jpg)


Best Regards,
-Edwin
Title: Re: OLED dark horizontal bands
Post by: Paul_B on May 08, 2015, 02:14:30 PM
Hello Edwin,

Thank you for reaching out, a quick way to verify is to move the location of the 'balls'. If the dark horizontal bands appear in the new location there is an issue with your code.

Please post the updated pictures once the changes have been made, we can then debug further  ;)
Title: Re: OLED dark horizontal bands
Post by: erlopez on May 08, 2015, 07:29:04 PM
Thanks for your reply Paul_B. I have identified the problem with the “dark bands” and I’m posting my findings here for documentation purposes; should anyone run into this problem in the future.

I narrowed down the issue to the values set in the SEPS525’s 0Bh, 0Ch, and 0Dh pre-charge current registers. I was using the values given in the Arduino example code from NHD. I played around until I found the values that worked for me (the SEPS525 datasheet is very vague about describing these registers). After tampering with these values, I found better color accuracy by using lower values that the ones given in the Arduino example. Lowering pre-charge current created better-contrasted images; however lowering them too much start making the colors in a photo look a bit off, for example “cyans” to start to look “greenish”. Bottom line: one has to play with these registers until one finds the right color temperature that works for the application.

The picture below describes the problem and solution for the “dark bands”, or better termed, contrast issues:

(http://www.lopezworks.info/darkbands-doc1.png)

The following values did the trick for me:
 
Code: [Select]

  oledWriteCommand( 0x0B ); // Set Pre-Charge Current of Red
  oledWriteData( 0x0A );
  oledWriteCommand( 0x07 ); // Set Pre-Charge Current of Green
  oledWriteData( 0x17 );
  oledWriteCommand( 0x0D ); // Set Pre-Charge Current of Blue
  oledWriteData( 0x02 );
 

Best Regards,
-Edwin


UPDATE:

After much tinkering; the "dark bands" are still appearing around bright objects moving over medium-to-dark colored backgrounds; adjusting the the pre-charge currents help reducing the "dark band" effect but the artifact is still visible.

Here is a video showing the problem - vector balls animation over a medium-dark blue background; observe the ghostly dark bands dancing over the solid background color (this is not a flicker/interlace artifact between the screen refresh rate and the camcorder - you can see these bands clearly with the naked eye)

http://www.lopezworks.info/darkbands-oled.wmv (http://www.lopezworks.info/darkbands-oled.wmv)

I'm frustrated. Any comments are welcome.

-Edwin
Title: Re: OLED dark horizontal bands
Post by: Paul_B on May 14, 2015, 08:15:23 AM
Hello Edwin,

Since these OLEDs are passive matrix displays unfortunately there will always be slight banding. You can reduce the effect by adjusting the registers like you previously stated.