Very unusual problem with dot-matrix LCD display (Sharp PC-1600)

[Fliz]

Well, I thought that one pixel at dot matrix LCD display can't be split further, either is ON or OFF, no half ON half OFF state!
but, look at this (I adjusted the contrast a bit higher so it would be more pronounced):



How is this even possible???



It's an old "pocket"  computer Sharp PC-1600 that actually works 
calculates correctly, executes commands written in Basic... it's just not visible because of the screen problem.

What would be your best guess, what (where) could be the problem?
Never saw something like this before!!!

[poptones]

I've had a couple of the EL5100s that have a similar problem. It's actually well known, has something to do with the LCD material degrading over time due to UV exposure. If you look carefully in the upper left of this display you can see some of the "clouds." Those that don't suffer from this command a good deal of money (about five times the original 1970's selling price!)

http://www.ebay.com/itm/Vintage-SHARP-EL-5100-Scientific-Calculator-in-Original-Hard-Case-Works-Fine-/150956412191?pt=Vintage_Electronics_R2&hash=item2325b40d1f

[Fliz]

thanks for a reply,
so i guess it's either LCD screen replacement or recycle bin
i was hoping it has something to do with lcd's controller, that would be fixable... 

[EEVblog]

Wow, I wonder how it ends up with those angled lines half way through the pixels 

Dave.

[amyk]

My guess would be something like stress cracks in the ITO.

Some materials prefer to crack along one direction, having to do with their crystal structure.

[6502nop]

This may be a bit off-topic, but I notice that it has a "Romaji -> Kana" in the menu. Is this an Input Method Editor (IME) to use Japanese characters (Kana) from the English (Eigo/Romaji) keyboard?

It'd be a treat to see how they got an IME on such a small device.

nop
?? ?
^ {edit} That was supposed to be a Nihon-go joke, but this board apparently doesn't do Unicode.

[Fliz]

Wow, I wonder how it ends up with those angled lines half way through the pixels 
Dave.

That's what puzzles me to. If it were just horizontal or vertical lines that would potentially be
easy fix but across the pixels. Weird...

My guess would be something like stress cracks in the ITO.

ITO? Indium tin oxide? Sounds dangerous  I should probably crack it open now 


Guess I could buy another one and try to do LCD transplantation 
6 or 7 years ago when I was on college I had the same model and it worked excellent but just
when I get to use it they forbid the use of Sharp-s calculators  so I sold it 

Interesting thing is that the top row of functional letter (BUSY SHIFT... etc) is not affected by this "effect"!!!

[PA0PBZ]

Interesting thing is that the top row of functional letter (BUSY SHIFT... etc) is not affected by this "effect"!!!

That's because it is a separate part of the display, it is not the dot matrix stuff but dedicated words.

[Fliz]

Yes, but if it's a consequence of liquid crystals material degradation due to UV exposure or ITO (?) wouldn't that affect entire crystal material?

[poptones]

How smart is the pc-1600? One of the most advanced features of the el5100 was its ability to factor algebraic functions. You could enter AxB=C and run the formula, it would prompt for A and B, but if you didn't enter B (for example) it would prompt you for C and then solve the problem!

[Fliz]

if i am not mistaken PC-1600 was the most advanced from pocket series,
whatever you could program with MS Basic he can/could interpret...

[RAYZOR-SHARP]

Hello Fliz,

I only just came across your post. Having decades of experience with SHARP pocket computers and LCD displays,
this is indeed the weirdest thing I've ever seen.
Forget about most of the 'explanations' given: they're either beside the point (not unusual on Forums), or bull....
The only thing I can think of - though far-fetched, I must admit - is a camera shutter time, so small, that it actually
registered the 'building-up' of the triangular-shaped dots in the process from OFF to ON state or vice versa.

I'm still curious: have or someone else succeeded in repairing it, or found the cause ?
If you're still interested in fixing the PC-1600, I may very well be able to help, having factory-new spare parts,
and/or PC-1600 itself.
In that case, please mail me.

Kind Regards
Ray

[SeanB]

I have one with a broken display ( it was dropped) and it was a nice unit.

[RAYZOR-SHARP]

Hi again Fliz,

I realize now that my own 'explanation' was also "beside the point", as it would apply to the situation, where someone
tried to produce an 'impossible' picture, rather than you showing an actual anomaly/malfunction of your equipment.

So, back to reality, we're still confronted with your wondering, how an LCD dot can assume a state that seems to be between
ON and OFF ? From the bizarre pattern shown in your picture, it is clear to any electronics engineer, that there is something
seriously wrong with the circuits that control the LCD. (This has however, nothing at all to do with the LCD 'plague' in earlier
calculators and pocket computers. Which looks very different and is caused by imperfect sealing of the display, resulting in a
deterioration of the highly hygroscopic liquid crystal material itself - shown by dark stains, over time slowly extending, from
the sealed edges to the middle of the display.)

From LCD technology theory we know, that controlling an LCD is a rather complicated process, involving precise pulse
sequences on the electronics side, along with (twisted) electrostatic fields and more on the physics side. Keeping these
facts in mind, it will be less of a problem to imagine circumstances in which triangular-shaped dots can exist ...
Still, it would be very interesting and enlightening to have an expert's opinion on this. Shouldn't be too difficult on an EEV-
forum, should it ... ?

Kind Regards
Ray

P.S.: the complete 'innards' of a PC-1600 are available here, as a factory-new sub-assembly. Can I help-out anybody ?

[SeanB]

Quite possible the upper glass is cracked from either stress or impact, but the cracks are being held together by the top polariser film, and just show up as broken ITO traces on it. Only going to show the classic black marking when one of the cracks reaches the end of the panel and lets in oxygen to degrade the fatty acids in the LCD matrix.

[amyk]

I'm still going by the cracked ITO layer theory - on the other side of the crack, the pixels appear to be completely dead, and the crack happens to go through a pixel, hence the diagonal effect. The consistency in angle is probably an artifact of the manufacturing process.

Edit: SeanB beat me to it.

[miguelvp]

Only time I've seen LCDs missbehave is when I left my mulitmeter outside at below 0 (Fahrenheit) to see if it it will go negative. It was way under it's operating parameters and a fractal pattern formed on the LCD. The majority went away but it has some permanent "sup-pixels" in the liquid crystal on.

[SL4P]

If you're pulling it apart - my sight-unseen thought would be to look at the squishy foam conductor strips (if it uses them) which connect the PCB to the LCD panel.

I'm thinking the angled pixels may in fact be normal to support the 'other' character sets.
Just 2c worth.

[Fliz]

I already had one PC1600 long time ago, that was fully functional and it did not have horizontal split pixels.
Theory that sounds most plausible for me is that the screen was subjected to very low temperature, under 0C because this horisontal patern is eqally on entire screen surface.
Yes, conductive strips are used inside but they are not squishy, just oposite they are quite firm, you can see here how it looks like inside.