128 X 64 mystery LCD identify

[mkiijam]

Hello All,

Here is the back story. This LCD module is from an Roland TD-11 drum brain. You can google it if needed. The schematic is NOT available, nor really is the part (maybe I'm just to cheap, or maybe I just like a good mystery?)

The LCD was cracked, no longer functions, and I would like to find something that will work.

Known details:

1. The resolution is 128 x 64 (I counted the dots on a close picture of a working one)
2. There are 12 pins going into the LCD module from the main board of the drum brain.
3. The LCD module does NOT appear to be COG or OLED as I can't see any chips and there is a separate backlight (not part of the 12 pins mentioned on point "2"
4. The LCD module can have its contrast adjusted, through a user setting in a menu. More about this later.
5. LCD driving IC is ROLAND RBA02021ABG. This MCU is used in other Roland gear and appears to drive many other styles of LCDs.

LCD / MAIN BOARD PINOUT, known details and the MYSTERY...

When powered on and sitting in a nominal state.
PIN 1. is sitting at 3.3Vdc with a "negative" pulse every second or so. I think it's negative because it appears below the scope trace in AC mode.
PIN 2. is at 3.2Vdc with no pulses. I'm pretty sure this the LCD_RST line as it is 2.2k to ground and matches a known Roland RESET scheme.
PIN 3. 3.3Vdc, has shorter pulses than PIN 1 but with the same 1 second interval. It is also the PIN that has activity when the unit is asked to adjust CONTRAST in the user menu. ( I know this by going through a "blind" procedure to adjust contrast step by step without the need to see the LCD)
PIN 4. 3.3Vdc very similar to PIN 1. Longer pulses every second.
PIN 5. is held low at .03 Vdc and pulses "up" to 3.3v every second
PIN 6. is 3.3Vdc. No pulses.  It appears to be the VDC trace as it is thicker than the others AND connects to the D+3.3V line
PIN 7. GND
PIN 8.  Capacitor to PIN 9
PIN 9. see above
PIN 10. Capacitor to PIN 11
PIN 11 see above
PIN12. Capacitor and 4.7M in parallel to GND

So, any guess as to what this is?

SPI and charge pump capacitors??? If so, how to ID SO, CLK, CD, CS etc?

[DrG]

https://www.aliexpress.com/i/4000370015852.html



Maybe.

[mkiijam]

Yeah... that's dang close if not it! How did you find that so fast!?!?!?

PINs 6,7, 12 looks right for sure. The cap pins make sense.

PIN 2 looks right as another schematic lists pin 2 as RESEST and has the same 2.2k to ground as this main board does.

[DrG]

Yeah... that's dang close if not it! How did you find that so fast!?!?!?
/--/

I'd like to say that it comes from years and years of accumulated knowledge, skills and experience, but really all I did was search for "128 x 64 lcd 12 pin". Yes, I have used similar a few times.

Do yourself a favor though, show some close up pics of the LCD back (any numbers etc) and see if you can identify a controller. Maybe you can't do all that, but you can find plenty of interface examples, but there could be some timing issues.

[mkiijam]

Yeah.... There is no usable info on the LCD module.

okay, so here is some progress.

I hooked it up best I could to another SPI type OLED.

https://www.amazon.com/HiLetgo-128x64-SSH1106-Display-Arduino/dp/B01N1LZT8L

PIN 1 of my main board went to "CS" on the OLED
PIN 2 to "RES"
PIN 3 to "SDC"
PIN 4 to "CLK"
PIN 5 to "MOSI"
Then of course just VCC and GND

AND this is what I got when I powered it up.

The "STICK" and "TEMPO" make sense as this is a drum module LCD, but the other stuff???

[ataradov]

It looks like it could not handle the fast SPI clock or firmware tried to do some windowing commands that are not supported by this LCD.

There is a number of very similar, but subtly different protocols/controllers. You can see that in a later part the firmware tried to draw a frame, but every other pixel is blank.

Do you have any pictured of the broken display? Can you grab some SPI commands being sent using a logic analyzer? Especially the part after the initialization.

[mkiijam]

Here is a close up of the LCD working.

[ataradov]

So the dots may be normal, but at the very least the image is flipped. Which again, tracks well with differences in displays. They all have similar commands, but some configuration bits, like flipping are different.

Having captures of the communication may help figure out the controller type.

[mkiijam]

Ahhh...

Only a scope, no logic analyzer.

[ataradov]

It would be pretty hard to capture enough data with a scope.

Any way, most commonly used controllers are SSD1306 and SH1106, which is almost a clone of SSD1306, but with minor differences. Specifically with the display RAM size and layout.

So the easiest thing you can do is figure out the display controller in the display you bought and get the other one.

And buying things from https://www.buydisplay.com/ is a way to ensure that you actually get the display with a controller you need.

[DrG]

I'm am guessing (and it is only a guess) that it was an ST controller - something like an ST7567 or an earlier one consistent with the time that the Roland was made. I'm guessing because of the fit with the 12 pin interface. I know I could be wrong.

It was probably not an SH1106. Also note that he never connected the A0 line. Timing problems? Sure all that stuff. But look, this "needle in a haystack approach" is likely not going to end well.

@mkiijam I and now @ataradov have asked you about pics of the broken display, but it is as though you never heard us. Do you or do you not have the broken display - if you don't have it then at least we can stop asking about that part. If you do, please provide some pictures as that could, potentially provide important information. Also, clear pics of where the 12 pin connector attached to the main board - anything that might be helpful like that.

If you don't have the display anymore, you have to research as much as you can to get some information about that display. Repairs of the unit, repairs of similar units and so on.

[mkiijam]

PIN 3 "A0" went to SDC

[DrG]

PIN 3 "A0" went to SDC

Ok gotya SDC is like SD which is like A0. So, yes, you connected A0 and I was wrong to say that you had not connected A0.

So, what does the other side of the the broken LCD look like? If it is completely encased and there are no numbers on the case, can you peel it off so we can see the board under the case.  Edit: Stripping away the backlight if you can. Just looking for clues.

Also, what is around where the 12 pin connector attaches to the main board. Is it possible to also take a picture of that? Edit: IOW where are those components caps and rs that you mentioned in your first post.

[ataradov]

I still think the only viable approach here is to just try multiple displays with different controllers.

Or capture the initialization command sequence and match the used commands/registers to known controllers.

Those displays rarely have useful markings on them.

[DrG]

I still think the only viable approach here is to just try multiple displays with different controllers.

Or capture the initialization command sequence and match the used commands/registers to known controllers.

Those displays rarely have useful markings on them.

Yeah, well, I can't disagree with you, depending on how you want to define "viable" and if you have a number of different displays/controllers to try ESPECIALLY ones that have a 12-pin connector that seems to match his interface as he described in his first post.

So, yes, if he had that display with the ST controller that I linked to in my first post, I would definitely say, "try that".

Can't hurt to search for clues though. Especially since you damn well know that somewhere at Roland at some time was a service manual that identifies the LCD. What do they do, burn all the service manuals when they stop selling/supporting the product? I think it is only 8 years old - sheesh.

[DrG]

Those displays rarely have useful markings on them.

Just curious - is the "TSE8G0708FPC-A2-E" the identifier for the LCD the way it is on this one https://www.romasterl.com/index.php?main_page=product_info&products_id=204219

[ataradov]

Just curious - is the "TSE8G0708FPC-A2-E" the identifier for the LCD

Doubt it. It is the identifier of the cable by itself.

the way it is on this one

What makes you think it is a display identifier in this case? I mean it says 12864 as part of the name, at least. But is this the actual ordering number?

Well, it looks like LX-12864B11 is the generic name for this type of display. But I see it with all sorts of connector style.


I've looked at a few displays I have for which I know exact part number, and none of them have anything close to the part number printed on the cable or the display.

[amyk]

Ahhh...

Only a scope, no logic analyzer.

FX2LP (Saleae clone) or similar would be enough to capture the commands - and are cheap as chips.

The fact that you're even getting something remotely sensible with that OLED suggests you are already in the right ballpark for identifying the controller command set. My guess is that it is a descendant of the classic command set made famous by the S6B0724/0708 and the like from Samsung.

(I have done similar work in the past.)

[DrG]

OK, I'll leave it up to the experts.

[mkiijam]

Here is a portion of a schematic that uses the same Roland driver MCU.

It uses some 18 pin LCD...

[mkiijam]

Another picture with my jumper wires partially in the way.

The top left ORANGE wire is PIN 1, then the GRAY is PIN 2.

You can see the 4.7M and caps on PIN 12 etc..

[ataradov]

This looks like a typical set of components for SSD1306. But 4.7 M for reference current resistor is a bit high.

But realistically it van be any controller, again.

If you really  want to figure it out, get one of those cheapo logic analyzers and get the initialization commands https://www.ebay.com/itm/USB-Logic-Analyzer-Device-Set-Compatible-to-Saleae-24MHz-8CH-for-ARM-FPGA-M100/253841718379

[mkiijam]

Okay, I think I will...

Side question. Is the a COG LCD??? I can't see any "chip" on the glass... it looks completely transparent to me. I think I may not understand COG?

[ataradov]

The controller is mounted on the glass. It is not etched on the glass itself.

In your case the chip is sandwiched under the place where the cable is attached. You can actually see it on your picture, just a small part or bare silicon sticking out.

[mkiijam]

AH!

It's all becoming clear now.

So the MCU puts out some kind of language that the LCD controller then sends to the actual GLASS unit. So, if I can see the language then we can guess which imbedded chip(controller) we need?