Roland D-110 dim backlight

[Robber1956]

Changing led or whole backlight panel?

ALso just changing out the backlight behind the element is not so much work, the LCD still gets back correctly in its place with the rubber conductors to glass and no need to change the all unit altogether.

I changed a backlight panel once this way but the characters where gone, the actual LCD was finished forever. Couldn't get it right anymore. But anyway, I have enough LCD's to experiment so we'll see.
Buying? What does it cost nowadays? From China just a few dollars, some time (within 2 weeks) and often free shipping. Looked it up, I paid $8.30 total order $24.28 for 6 displays. Still a bargain.
Okay, changing color. Just change the led? What if it's moulded into the diffuser? Does the new diffuser fit in the old frame? I would like RGB. How to invert the appearance of the character? Black like now or white like with a blue backlight? Afaik you'll need to change the character code of the LCD but I vagualy remember that swapping two data lines will do the job. I'm not sure though.
"just increase the current for the LEDs" to make it brighter?
How? Resolder the MB, change capacitors? Put in different resistors? Another transistors? Are we talking Roland D110?

check that that it gets some voltage to its module edges. Mine had 4V on it and it was quite dim. You may reduce the R5, so it will get 4.5V so you get higher light output.

Okay, have to get the PCB out to resolder R5....no way. Anyway, my D110 gives of 5.05 volt, enough for the backlight*



pin one for the LCD is top left, pin 14 bottom right (not visible but obviously there ) of the LCD connector and pictue bottom right the little 2 pin white connector for the backlight* which you don't need anymore using the 1602 ERM 2.1.

The only problem here is the size of the display. I was hoping the 14 pin LCD with gaps instead of holes for mounting would just fit as the original. Should have read the specs a bit closer but then again, finding a similar lcd as the original DM029Z is almost impossible. I found a Winstar (called 1602D) but the MO was a 1000 pieces and I just checked the specs: It is like the ERM 2.1 30.0 mm and not 33.0 mm like the DM029Z.
As a matter of fact, none of the specs concerning size match with the old DM029Z display.

And also, can you tell me what you meant by this:

"the backlight "module" is a LED array of 36 LEDs. there goes 5V to the whole module."

I really thought you were kidding when I read this.

[jxalex]

Okay, have to get the PCB out to resolder R5....no way. Anyway, my D110 gives of 5.05 volt, enough for the backlight*

we arent afraid of screwdrivers, dont we?

Well, my solutions come from me as a result of decades of dealing with electronics, research, prototype engineering and reverse-engineering.
I soldered directly on top of the resistor. And I really do mean that, also the current I check what it consumes which is quite surprising.

And also, can you tell me what you meant by this:
"the backlight "module" is a LED array of 36 LEDs. there goes 5V to the whole module."
I really thought you were kidding when I read this.

No I am not. I have opened that backlight module. Just remove the LCD element and the backlight unit is covered with one PVC plastic-film. Once You remove it You see 36 bright green dots (in power-on state). Once satisfied the curiosity (or swapped the diodes, whatever), put back the plastic film cover, and LCD screen on its contact rubbers, see that the edges are firmly in contact, and mount the metal cover.
The 5V goes to the module array edges where the voltages go.

Thats what I did in person. Now perhaps I have given a impression as a very hilarious with these remarks?
Sorry, no youtube video.

IF someone has trouble with the contact rubbers, I am more than happy to help with it. Come on, its fun to get that new skill.

[logix]

Well, I gave it another go..
This time I soldered each connector pin to ensure a proper connection and re-checking with a continuity tester afterwards.
Same result though, which could be due to polarity reversal of the LCD and/or the fact that I carefully removed the black metal frame out of curiosity to what jxalex said about replacing the backlight etc. When I press the metal frame now some more boxes appear or go away, so it's obvious I've done something with the contacts. I also tried to desolder the new display from the cable earlier (so I could reattach the old display) but didn't succeed and might have damaged some circuit traces on the backside (underneath the IDC connector) of the new LCD. Lots of reasons to start fresh with a new display which I'll be getting in a few weeks time!

What we really need are some clear instructions and overview of the whole thing. So here's a diagram I've drawn with all the pinouts:


When I get the new IDC cables and connectors I think I'm going to make it the same way as the original cable, but solder the IDC header (the display end) on the other side as mentioned earlier, to reverse the pins. There should be room for the cable to go around on the side.
I also looked at the mounting holes on the display PCB sides and I'm thinking that for the new ERM1602xxx-2.1 ("xxx" is for the different backlight colour names) it will possibly fit if we just cut off the outer parts of the PCB:


No more time for experimenting today, but I'll give it a try later. Fortunately I have nothing to lose

Oh, another thing: I attached the separate backlight cable to the old display and checked the voltage which was just 3.9V (across the backlight pins) when attached. If I removed the backlight and checked the CN4 (2-pin) connector in the D-110 I measured 5V. So is this due to the internal resistance of the backlight LED and normal, or does it indicate a damaged LED backlight?

[Robber1956]

Why the tracer of the flat cable goes to pin 2 instead of pin 1 on the LCD board:



you may think encircled pin is pin 1 (and the blue tracer is actually going to pin 1 on the D110 board), but when you flip the module over in the longitudinal direction:



you will see that it's connected to pin 2

the PCB was designed for the connector to sit on the other side to match with the numbering of the pins on the D110 itself.

So why can't you just do so? On the mobo of the D110 pin 1 is VDD or Logic Power Supply and pin 2 is Ground. And this is how I connected the 1602 2.1 to it.
So where is the LPS for the DM029Z coming from? Is this why pin 3 is bridged to pin 1? And is the 5+ volt that is directed from pin 1 on the mobo to pin 2 of the LCD smothered with the 90k resistor?

Must admit, this is a bit too complicated for me but at least I've got the 1602 2.1 working.

Oh, another thing: I attached the separate backlight cable to the old display and checked the voltage which was just 3.9V (across the backlight pins) when attached. If I removed the backlight and checked the CN4 (2-pin) connector in the D-110 I measured 5V. So is this due to the internal resistance of the backlight LED and normal, or does it indicate a damaged LED backlight?

Afaik is this normal. It's called 'voltage drop' caused by the LED to start emitting light, I don't think the 90k resistor on the DM029Z LCD has anything to do with this.

More info:

https://www.baldengineer.com/led-basics.html

well, Google is your friend here.

[logix]

I suspected we were talking past each other, something those photos confirm as I recognize it as the old display (I've been referring to the new one).

Yes, I completely agree that pin 2 of the old display goes to pin 1 of the D-110 PCB and is VDD (+5V), but opposite for the new one.
Actually I realized that the display diagram in the D-110 service manual doesn't show the pinout for the display, but just the CN1 connector pin numbers/and signal names. And the pin numbering printed on the old display PCB is completely misleading as it doesn't correspond with the CN1 pin numbering (but flipped over)!
Now, the actual numbering markings on both displays next to its connector is the same, but the signals are different! So in conclusion, using the original cable -with the old display it should be soldered to the back of the display PCB while with the new display it should be soldered to the front its PCB. Do we agree on this?
My latest diagram comparing all three connections (CN1, old and new LCD) should be correct due to my findings (tracing signals and comparing with the datasheet/service manual).
UPDATE: I see what you mean about pin 1 becoming 2 and opposite with the connector, but that's due to soldering it on the rear of the display PCB. Looking at my latest diagram I do notice an irregularity: that although the pins on each row are flipped (when comparing the old and new displays), pins 1 and 2 go the opposite way on the CN1 connector! But this doesn't make sense -shouldn't all or none of the pin rows be flipped with the cable/connectors? I didn't see any twisted wires on the cable (i.e. wires for pins 1 and 2 flipped around) but just connected straight to each of the two IDC connectors, so how can this be?
This makes me wonder that although I have the power (VDD and GND) working on the new display, should I perhaps flip all the other rows around? But again, this doesn't make sense (the IDC connector pin pattern is the same all the way, and the pinout for the new display corresponds to that of its datasheet)..... frustrating

I made some more discoveries with the new display. First (since I believe I now have the correct wiring for the new display) I tried to wiggle the LCD a bit more and then removed the metal frame again to see if I could get a better connection. Of course this just made it worse (pressing down on different sides showed different things on the LCD), so with a new display I'm certainly not going to remove it as Jxalex suggested.



While having the frame removed (the display and contacts strips stuck to its sides) I also had a chance to look at the backlight. I'll desolder it and if possible take a look inside when I get a new, working display, but it appears only to be soldered by two pins and there's a foil or something underneath.

[logix]

Must admit, this is a bit too complicated for me but at least I've got the 1602 2.1 working

I just received my new 1602xxx-2.1 displays and also the IDC cables/pin-headers, so my first task was to make a new cable for the old (Roland) display to get it working again, just in case.....
I also find this very confusing, but despite being opposite of what I thought it worked the way you described it (I think). So that's good. At least I'm back to where I started now

So you managed to get a 1602xxx-2.1 display working in your second D-110?
If you don't mind, and to avoid any further confusion: if you have a multimeter, could you check which pin of the D-110 CN1 connector goes to which pin of the 1602xxx-2.1 and post the results here?

[Robber1956]

Just checked, CN1 pin 1 goes to 1602xxx-2.1 pin 1! No need for the anode and cathode connector.

About mounting the new display so it is positioned rightly behind the fron panel I made a bracket out of an aluminium profile that I had laying around to replace the original one:



which holds only the top left of the display and on the right and a another with 2 little nuts and bolts to keep it straight up:



It should be shifted a bit to the right but all in all the result is quite satisfying:



Can do better though. Anyway, I didn't have to alter anything on the D110 itself, it can be brought back in its original state like this.

[logix]

Thanks for checking.

I just received my replacement LCDs (also 1602xxx-2.1) and soldered a temporary pinhead connection (so I don't have to mess up the cables again) at the front of the PCB. It lights up as you can see (so I have the polarity right -just as you checked on your setup), but something is still wrong. Actually it appears very similar to my earlier replacement display -the black boxes at the top row which I showed in an earlier posting.


I suspect the data lines are somehow wrong, so if you have the opportunity, could you also check to see which where on the new LCD each pin from the D-110 connector goes to? Even the 1602a to the D-110 connector will help as I can look up the 1602a datasheet and compare it to the 1602xxx-2.1.
What I do know at this stage is that my D-110 and the original LCD (with a new cable) is working, and I have a brand new 1602xxx-2.1 which is working, so that helps not wasting more time figuring out that stuff

Your 1602a mounting in the D-110 looks nice! I totally agree that it's smart to make modifications which you can change back to the original if you need to later. I'm going to experiment and see how that damaged 1602xxx-2.1 will fit once I do some cutting etc.

[logix]

(preliminary) SUCCESS!!



Aside from the bad contact (after disassembling my first 1602xxx-2.1) I can definitely read something there now, so I think the replacement display will work as intended with the new replacement.
In hindsight I should have looked closer at the signal names instead of the pin numbers, just getting confused. Oh well

The way I got it working was to have the same cable pinout connection as the original Roland cable except for pin 1 (+5V) and 2 (GND) which were swapped. I think I got it all the other way round before, then flipped pins 3 to 14 around after switching the polarity to add to the confusion. The diagram below should now be correct:


Some other things of interest:
First, here's a comparison between the old (Roland) and new (1602xxx-2.1) display backlight (please disregard the strange backlight colour of the new display which is wrong because of the camera's colour balance). No doubt the new display's backlight level will be better on the eyes!


I also took a look at the original display's pin 1 (GND) and 3 (V0) connection. By removing it I got a very slight difference in the contrast as you can see here:



There's some variation on the information available online about how to set up the contrast:

1. A 1K trimpot with the centre pin connected to pin 3 and the two end-pins going to +5V and GND.
2. A 2K resistor between pin 3 and GND (alternatively a 5K trimpot which I suppose is meant to have the center and one of the end-pins connected together, going to pin 3 and the other trimpot pin going to GND).

Does anyone know if any of this matters? Either "balancing" the contrast level between +5V and GND or just using pin 3 and GND (leaving +5V out)?

[Rasputin]

There's some variation on the information available online about how to set up the contrast:

1. A 1K trimpot with the centre pin connected to pin 3 and the two end-pins going to +5V and GND.
2. A 2K resistor between pin 3 and GND (alternatively a 5K trimpot which I suppose is meant to have the center and one of the end-pins connected together, going to pin 3 and the other trimpot pin going to GND).

Does anyone know if any of this matters? Either "balancing" the contrast level between +5V and GND or just using pin 3 and GND (leaving +5V out)?

Connecting pin 3 directly to ground gives maximum contrast. Connecting pin 3 directly to 5V will give minimum contrast, probably too washed out to read. Using a variable resistor between 5V and ground will let you set a balance between 5V and ground therefore giving you an adjustable contrast knob.

If you want a fixed contrast that isn't either fully washed out or fully contrasted then you need to put a resistor between a 5V line and pin 3. This works just like leaving the variable resistor in one spot because, well, it's a fixed resistor instead of a variable one. In general, something a little over 4V is what to aim for, so pick an appropriate value.

[Robber1956]

(preliminary) SUCCESS!!

Well done! Congrats

Now if you connect the new display with all lines visible exatctly the same you're there.
I don't think the display in the pic where the rubber contact was disconnected from the PCB can be mended.
jxalex said he could do it, I really wan't to know how.

The inquiry of putting a trimpot or resistor to adjust the brightnes of the backlight and the contrast (two different things) also looks very promising. I shouldn't put 5.0v directly on the led, you always need a resistor to prevent the led from getting overloaded and it also will be far too bright.

[jxalex]

Now if you connect the new display with all lines visible exatctly the same you're there.
I don't think the display in the pic where the rubber contact was disconnected from the PCB can be mended.
jxalex said he could do it, I really wan't to know how.

How? Carefully.

I took that from its rubber contacts and put back, with careful positioning.
The same way I have done with several other displays with rubber contacts. However I have not done this if these contacts are glued to the plastic ribbon cable stripes.

[Robber1956]

Thanx for replying promptly.

Okay, that's very plausible but it doesn't give me enough confidence to sacrify a good working LCD. However, I must have a garbled LCD laying around somewhere that was detached from the PCB. Let's see if I can achieve some improvement on that one by carefully realigning the rubber contacts.

[jxalex]

Well... there is not so much think about it. It is just to be done. It is simple, almost nothing can go wrong and also it is ok as the display itself (which is in abundance and just $20).

The roland D110 display has the same backlight technology like the Korg Triton Rack has (LED array), but the LCD itself is attached in a way like it is in Korg Wavestation A/D. Which means - simple and reliable. The otherwise around would be h**l.

HOwever I need some brains for much tougher delicate experimenting.

The worst scenario so far I have seen is in Korg Triton Rack. Its backlight wont fade away, but its LCD is nuisance which has glued ribbon-cables. Their connections depend on temperature, they break down, loose contact from LCD. Many complain these dead stripes coming to LCD. I have old LCD still but have not found no way of restoring these dead stripes and neither there are alternative replacements than scarce and expensive originals (imagine $150 USD cost for a display).
Compared to these experiments what someone have done in this thread for 3 pages, the $20 spent on another LCD unit can be considered as a "school tax" if not done right.

[Robber1956]

HOwever I need some brains for much tougher delicate experimenting.

The worst scenario so far I have seen is in Korg Triton Rack. Its backlight wont fade away, but its LCD is nuisance which has glued ribbon-cables. Their connections depend on temperature, they break down, loose contact from LCD. Many complain these dead stripes coming to LCD. I have old LCD still but have not found no way of restoring these dead stripes and neither there are alternative replacements than scarce and expensive originals (imagine $150 USD cost for a display).
Compared to these experiments what someone have done in this thread for 3 pages, the $20 spent on another LCD unit can be considered as a "school tax" if not done right.

This sounds exactly like the problem with a CASIO FZ1 sampler, its graphical 96x64 display is also connected the way You discribed.

The Korg:

https://http2.mlstatic.com/display-tecl ... 2015-F.jpg

The Casio FZ:

http://www.midi-rakete.de/images/casio11.jpg

This is how they made things in the 80's.

If these ribbon cables break, or in my case became brittle and just fell apart when I touched them, it's the end of the story. No replacement available only second hand stuff. Some are sold cheap as "only for parts" with hopefully a working display. The latter is preferable than no display at all because editing can only be done with a working display. Backlight isn't really necessary to work with an old synth or sampler.
Old synths like Yamaha DX100/21/27 and the TX7 module didn't have backlit displays but they can be easily build in.