[2020/12/29]
PC110 No.3 WVGA-TFT front frame renewal
Until now, " No.3:no named " (with the WVGA-TFT-LCD and shaving off of almost of " IBM logo-like") was replaced with the "front frame" that was removed from another machine with some processing and with adjustment. ( Reference )
Machining and adjustment are the numerical settings of the "shaving off" on the right side of the front frame and the (my home made) "screen right justification" program.
"Shaving off" is "about 9mm", "Screen right justification" uses two types of programs, with "m8"(Actually, 16 = 2*8) on the DOS screen, and with "15" on the Win95 screen". (See below)
Now, I can see a beautiful display without significantly damaging the original front configuration of " PC110 " to the left. (Fig.1)
If I wish to see it on the original screen, "4.7 inch VGA-FSTN-LCD" is the most desirable, but unfortunately, my technical ability has not been able to realize a low contrast display. (Fig.2 [click] right side)
As a second best measure, this "6 inch WVGA-TFT-LCD" was used aiming for a normal screen display by compensate for the "difference in screen size" by scraping the front frame.
Since the frame top and bottom are left as they are, the screen is hidden by about 2 to 3 mm, but I don't care so much, so I say "good !".
However, when only the right side of the front frame was cut off in order to reduce the procedure, the balance of screen left and right was not very good.
If the left side is also scraped a little and the left side is 4mm and the right side is 5mm, the balance may have been a little better.
－－－－－－－－－－－－－－－－－－－－－－－－
+++ Grinding / widening of front frame +++
The front frame of " PC110 " has some experience of "cutting off", but it is difficult to cut straight because the material is hard and there are some irregularities.
In the left-right direction , I decided to cut off only the right side in order to reduce the number of cut points if possible.
I addressed the "metal ruler", made a line with a "cutter knife", and made a deep cut while tracing it.
However, in an unexpected place, the "groove course" bends.
Even if there is a deviation in the course, it is cut so that it is on the inside, so I can correct it by scraping it later.
But if the "metal ruler" shifts and the course bends toward that, it will become difficult to correct later.
So, the scraping work did not go as expected and I did not get very good results, but it can not be helped. (Fig. 3)
Since " frame top and bottom " are not cut, so the flatness of frame window was kept.
The top and bottom alignment of the "LCD module" was done quite finely using a part of the "flat wooden skewer". (Fig.4 [click])
After that, about 6 places around the "LCD module" were fixed with " hot melt bond ".
+++ Screen shift program +++
When the "VGA screen" of " PC110 " is displayed on the "WVGA-TFT-LCD screen", " only displayed at left-justified ".
I made a program that right shift with the number of characters (x), and shifted the screen to the right.
+ For DOS: ct65535e.exe /my (y=x/2 (Fig.6 [click])
+ For Win: ct65535w.exe x (Fig.5)
These x values had to be adjusted slightly depending on the actual position where the "LCD module" was placed.
Switching between LCD screen and CRT screen (external display) does not work with just "Fn+F7 key", so on DOS , I wrote the "ct65535eexe" in "autoexec.bat", make it resident at startup, and use "Alt+F7 key" to return to the LCD screen.
On Win , start it with the "shortcut" of " ct65535w.exe " (with options), and switch between the LCD screen and CRT screen using the menu that appears.
Due to these, the screen is displayed normally only after DOS and Win are started. Although there is dissatisfaction , there is no particular problem in running other apps, so let's say "good !".
PC110 No.3 WVGA-TFT brightness adjustment, successful
One of the things left unfinished this year was "verification of the brightness adjustment circuit".
At the time of creating the "relay board", brightness adjustment circuit was included, but this was not used, and checked only with "simulation" and "experimental circuit".
I have all the chip parts, and a good chance, so I put the chip on the board and checked it.
At the originally planned R2 = 5.1KΩ, no display appears.
So, when I changed to R2 = 5.6KΩ , the result was good ♪
Since the "automatic brightness adjustment function" of the digital camera used for shooting has an effect, I took pictures with the surrounding scenery as much as possible, but the difference is smaller than it looks in the picture. (Fig.7 and Fig.8 [click])
In reality, it can be changed from a fairly "bright display" to a fairly "dark display", but it is difficult to see in the photograph.
As for "operability", both hands are still required, but when I hold down the "Fn+Ins key" or "Fn+Del key", I can change the brightness significantly in about 3 seconds.
Well, I don't think it's usually used in the dark.
The "current" is about 20mA when the brightness is moderate (it seems that a maximum of about 35mA can flow), and about 10mA when it is dark (can't be lower less than that currently).
I forgot the "upper limit of the allowable current value" that Mr. Moonlight investigated before, so I don't know that this value of 20mA is a reasonable value.
However, with that current, I feel that I can continue to use " PC110 " without degrading it.
－－－－－－－－－－－－－－－－－－－－－－－－
+++ Brightness adjustment circuit and key group +++
Initially, I used the resistance values set in simulation , R1 = 56KΩ, R2 = 5.1KΩ, R3 = 10Ω, but the backlight did not turn on. It was.
Next, when I chose from the purchased items and set R2 = 5.6KΩ, the value (although the chip size is too small) seems to be good. (Fig.9)
For "brightness adjustment", "Fn+Ins key" (brightness increase) and "Fn+Del key" (brightness decrease) are used. (Fig.10 [click})
Actually, these are diversions of what was originally "for contrast adjustment" of "FSTN-LCD".
On the other hand, the "Fn+Home key" and "Fn+End key" are useless.
By the way, with "FSTN-LCD", I must always pay attention to "contrast adjustment" in order to obtain "optimal display", but with this "WVGA-TFT-LCD", the brightness I like once. If I set it to, there is almost no need to make any adjustments.
Regarding the display of " PC110 " (except for the poor balance), I was almost satisfied ♪
(... Unfortunately, " DOOM " (VGA screen) etc. may lack the right side of the screen. After that, you need to execute "ct65535w" again.)
+++ Difference in atmosphere +++
The same brightness setting , a photo taken in the daytime (Fig.11) and a photo taken at night (Fig.12 [click]).
(By Panasonic DMC-FX9 )
Looking at the picture, the screen seems to be too bright and halation.
This may be because my eyes are weak and brighter than necessary.
Later, let's lower the brightness a little more !
[2021/01/05] PC110 No.3 WVGA-TFT B/L current confirmation
Currently, ("PC110") " No.3 " is displayed with the brightness of the backlight (B/L) lowered to the extent that it is easy to see (Fig.13).
But by chance, I suddenly became anxious when I saw the notation that the (LEDs) "standard current value" of the backlight in the data sheet of another (old) "WVGA-TFT module" (TM060RDH01) was 140mA . (Fig.14 [click]) Isn't the current value (about 20mA) mentioned earlier really " a big mistake " ?
Looking at the B/L (LEDs) current value in the recent "VGA-TFT module" ( AT050TN23V.1 ) material, it is 20mA . So, the current value of " No.3 " seems to be reasonable, so I'm a little relieved !
+++++ Leisure story ++
This may be because the "TM060RDH01" is a fairly old module, so the luminous efficiency of the LEDs are poor and those consume a lot of current.
Actually, I have two " TM060RDH02 "(can't using to " PC110 ") close to this. So, I'd like to make this a "6-inch display", but I couldn't find the datasheet, and I was looking at the "RDH01 version" that I got by chance.
... Hmm ? Wait ... I think the reason why these two cards could not be used on the " PC110 " is that the B/L (LEDs) current was "extremely large" like this. (I can't confirm the value because I can't find the datasheet)
When I use this, I have to be careful there !
However, it was a problem that I couldn't find the pin assign !
Also, I have to look it up with an oscilloscope.
(Later, when I actually measured it, it was quite bright with about Volt:20V, Current:20~25mA. It seems that there is no need to pass 100mA.)
++ closed +++++
I couldn't accept the value of the datasheet or the verification with an experimental circuit, so I actually measured " No.3 ".
The actual circuit (Fig.16 [click]) has a current limiting resistor of "10Ω", so measured the terminal voltage there and measured the current. When calculated, it was about 12.5mA. (Fig.15)
130mV / 10.4Ω ≒ 12.5mA
With this, I'm finally relieved !
Well, the "luminous efficiency" of " LED " has improved considerably compared to the past, so maybe there is a difference of about an order of magnitude in the past.
When I use an old module with LED backlight, I need to consider it as well.
PC110 No.3 WVGA-TFT Jitter Suppression
So far, this " No.3 " has only been operated with " Win95 ", and the state of "screen" was stable, and no "jitter" or "flicker" was seen. I just assumed that the screen was stable.
However, the other day, Mr. Moonlight taught me "the meaning and effect of XR6C", and I tried " PC-DOS ". I decided to change the value of " XR6C ".
So, the first thing I found out was that there was a lot of "jitter" left on the "DOS mode". (Fig.17)
One can clearly see that the letters on the upper right are broken. (Oh my God !?)
Even so, when you touch or push the "LCD cover", the degree of jitter changes.
This is poor! I thought, but before removing the "LCD cover" and checking the state of the "relay board", I had to change the " XR6C value " first.
when "Bit2" of this register XR6C is set to "1", the drive capacity of " output buffers " can be doubled .
Until pointed out by Mr. Moonlight , I hadn't noticed the capability of this XR register.
When I looked at the datasheet again, I found that it was written, and I start this experiment.
When I start any of " ct65535c, d, e " in "DOS mode", I can manually change the value of the XR register, so if I check the value of XR6C, it was XR6C = 18h . (At this time, there was "jitter")
I rewrote it to XR6C = 1Ch and return to the normal screen, then " Tamegoro said "Att!" and suprised extreamly" "（japanese old slap-stick joke）
"jitter" has disappeared any more !!! XR operation and screen status: " YouTube Video "
"jitter" is similar to the state in which a corresponding noise is added to a weak signal received by a radio receiver or the like.
My common workaround is to keep the noise source as far away as possible to increase the S/N ratio, but this time XR6C = 1Ch increases the Signal strength and S/N. It is equivalent to a method of increasing the ratio.
Even so, Mr. Moonlight seems to have read the manual carefully, and I admire the fact that it is being used.
Thanks for this information !
