Convert 16x2 chars display to big TFT display

Do you have a question? Post it now! No Registration Necessary

Translate This Thread From English to

Threaded View
I have a CPU board that interfaces to a standard 16x2 chars display,  
based on HD44780 controller.

I'd like to replace this display with a full-color big TFT display,  
maybe 10", but I can't change the firmware of the CPU (because I don't  
have the source code and I can't rewrite it, because of cost and time).

The simple idea is to create an interface between the current CPU board  
and the TFT display. This interface should be able to detect the signals  
from the CPU and that was directed to HD44780, interpret the  
instructions (write CGRAM/DDRAM and so on) and write the same things to  
the big TFT display.

Of course, exactly the same texts will be written on TFT display (of  
course with a bigger and visible font) and this is acceptable.

So the MCU on the interface should be able to read 8 data bits, RS and  
R/W signals syncronous to signal E.

I'm thinking to configure E signal as an external interrupt on the  
falling edge. In the interrupt, I read data bits, R/W and RS and decode  
the instruction or data.

Any suggestions? Maybe a similar code is already online.


Re: Convert 16x2 chars display to big TFT display
On Tue, 21 Sep 2021 12:31:54 +0200, pozz wrote:

Quoted text here. Click to load it

I guess you don't need an interrupt. The signals to the 44780 are quite  
slow and well sequenced. So simple polling would do.  
Beware that the E signal is high when data is valid. So I would look for  
the rising edge.


--  
Reinhardt

Re: Convert 16x2 chars display to big TFT display
Il 21/09/2021 15:44, Reinhardt Behm ha scritto:
Quoted text here. Click to load it

Are you sure? Enable cycle time could be 500ns at 5V, look at page 52 of  
datasheet[1].

Quoted text here. Click to load it

Are you sure? Look at Figure 25[1], data are valid at the falling edge  
of E signal.

[1] https://www.sparkfun.com/datasheets/LCD/HD44780.pdf


Re: Convert 16x2 chars display to big TFT display
On Tue, 21 Sep 2021 16:01:42 +0200, pozz wrote:

Quoted text here. Click to load it

An eternity.
  
Quoted text here. Click to load it

Data should be valid for the whole E time. Look for the rising edge and  

time of your signal source.

--  
Reinhardt

Re: Convert 16x2 chars display to big TFT display
On 2021-09-21 pozz wrote in comp.arch.embedded:
Quoted text here. Click to load it

Datasheet indeed says 500ns minimum. It ofcourse depends on the mcu what
the actual timing is. 500ns is 'slow' for todays controllers, so I think
you could do polling as long as you don't do too much else (depending on
the speed of your mcu). But an interrupt may be a safer choice and makes
other processing (outside irq) easier. But even with using an interrupt
you have to make sure latency is small enough, especially if there are
other interrupts.


Quoted text here. Click to load it

Both is true: data is valid when E is high, but with a small delay (Tddr
and Tdhr), so data is still valid on the falling edge. But you must be
very fast in your response, minimum Tdhr is only 5 ns. This would
require an awfully fast interrupt response. Setting an interrupt on the
rising edge gives you another challenge: Tddr is 160ns max, so you may
not actually read the data within 160ns of the rising edge of E. If you
have a fast interrupt response, you may need to wait a little before
reading.

Maybe an FPGA solution is easier for this problem. ;-)

have you already thought about the other end of the interface? How do
you plan to get data to the TFT panel? And don't forget about the part
in between, the frame buffer etc.


--  
Stef

Yow!  Those people look exactly like Donnie and Marie Osmond!!

Re: Convert 16x2 chars display to big TFT display
Il 21/09/2021 17:09, Stef ha scritto:
Quoted text here. Click to load it

Immediately before the falling edge data must be valid, but they aren't  
valid immediately after the rising edge. So I don't think the sentence  
"data are valid when E is high" is correct.


Quoted text here. Click to load it

So I think it's much more robust to read on the falling edge.


Quoted text here. Click to load it

I got the point and indeed is critical. Maybe rising edge is better,  
because most probably MCU set data bits and after make a pulse on E.


 >Setting an interrupt on the
Quoted text here. Click to load it

Consider that the read operation for HD44780 will be a write operation  
for my interface. In other words, when a read operation is send from  
MCU, my interface should write on data bits.

Howevere your point is valid: my interface should write on data bits  
before Tddr, 160ns maximum, and this could be difficult.

Anyway read operations aren't usually used on this kind of applications.


Quoted text here. Click to load it

I will send data to a serial port.


Re: Convert 16x2 chars display to big TFT display
On 9/21/2021 3:31 AM, pozz wrote:
Quoted text here. Click to load it

You will likely find there are issues that you've not even considered
that the original software relies on (directly or indirectly).

Can your emulation handle a *steady* stream of reads/writes?  Or,
are you expecting the host to be gentle with the display interface?
Keep in mind, that you have to run a fair bit of code to paint
pels onto the display from "character codes".

Does the host rely on the timing of the display's "busy" indication
in any way?  Are there any bug in the display controller that
the host code works around -- that you'd have to emulate (but
aren't disclosed in the datasheet because they *are* bugs)?
Are the visual characteristics of your display similar to those
of the original display?  Cursor/character blink and other
attributes?

Put a pair of back-to-back latches on the data bus and some
steering logic and don't sweat the fine-grained timing issues.


Re: Convert 16x2 chars display to big TFT display


 >You will likely find there are issues that you've not even considered
 >that the original software relies on (directly or indirectly).

Of course, but in these day much is possible.

https://hackaday.com/2021/09/02/pi-pico-emulates-rom-for-speedy-retro-hacking/

But it feels strange to do this. :-)


Olaf


Re: Convert 16x2 chars display to big TFT display
On 9/21/2021 9:04 PM, olaf wrote:
Quoted text here. Click to load it

It used to feel strange to only partially decode address spaces for
devices -- so, a single 8b latch would occupy 128KB of address space.

<shrug>  Times change.

By comparison, a ROM is highly predictable.  You don't, for example,
expect a read of address X to take longer when it follows a read of
address A vs. B.  Or, when the datum returned is Q vs. R.

A display controller, OTOH, is a serial machine and the time it
takes to perform an action is typically related to the amount of
"effort" that must be expended.

It's likely that the original code took this into consideration
(implicitly or explicitly).  and, without delving into that code,
it's hard to decide how you can reasonably expect to emulate it
without unexpected consequences in the host.

[E.g., the host may expect to have to wait when it does X.
But, will expect X to have completed before some particular
future time.  If you fail to do so, there's no easy way
of knowing what might "misfire".]

Re: Convert 16x2 chars display to big TFT display
Il 22/09/2021 04:34, Don Y ha scritto:
Quoted text here. Click to load it

Yes, it could be possible.



Quoted text here. Click to load it

The host is a 16-bits MCU from Fujitsu. I will measure timings by an  
oscilloscope, but I don't think it writes so fast.
And I can implement a busy flag trick: after any instruction, I can  
simulate a busy state... if the host really use busy flag check (I  
suspect yes, because I saw R/W signal goes high and low often).


Quoted text here. Click to load it

I will send sniffed character codes by a serial line to a "smart TFT".


Quoted text here. Click to load it

Most probably the host polls for busy flag and I think I can replicate this.


Quoted text here. Click to load it

I wrote some code for HD44780 in the past and I don't remember of any  
workaround. Of course it depends on how many features of controller the  
host uses.


Quoted text here. Click to load it

Quoted text here. Click to load it

If I can sniff these things, I think I can replicate on the TFT.


Quoted text here. Click to load it

Yes, it could be done in this way.

And I was thinking to leave the original display connected and put the  
interface as a passive sniffer. In this way I don't need to replicate  
the exact behaviour of the controller, but only detect the stream of  
data and commands. Of course, in this case I have to be fast enough as  
the controller (I can't stop communication by busy flag).

Re: Convert 16x2 chars display to big TFT display
On 9/21/2021 11:49 PM, pozz wrote:
Quoted text here. Click to load it

But you will only be able to observe the activities that *happen*
while you are watching.  Without access to the sources, you won't know
if this is representative of every possible set of interactions with
the display (and, thus, YOUR display).

Quoted text here. Click to load it

Again, you are assuming that the host software always checks for this
and honors it.  Are you sure the developer didn't "forget" in some
cases?  (in which case, you have to emulate the "forgotten" case)?
Or, didn't implicitly rely on knowledge of how the (old) display
device would react in a given situation?  (e.g., "I know that if I
issue an XYZ command, the display controller will finish processing
it in 1.3 microseconds... and, the time it takes for my code to
advance to the next display access happens to be 1.8 microseconds
so no need to check the busy flag")

I can't know that the code does this.  Nor can you -- without
a look at the source code.

Quoted text here. Click to load it

See above.


Which you won't know without a look through the code...

Quoted text here. Click to load it

I'm not sure I see what that buys you.  It certainly doesn't help
the cost.  And, the sniffed signals *should* be identical to your
emulation of them.  What will you do if there is a discrepancy,
at run time, 3 months from now?  Light a big "error" indicator?

If I was tackling something like this, I'd try to build a model of
the controller's internal state in the "emulator".  Then, design
a hardware interface that handled the fine-grained timing issues
(to decouple the emulator as much as possible).  Then, see how
fast I could get the interface between that hardware and the
*model* to run.

Then, in the background, run a task that examines the state of
the model and "paints" the new display, accordingly.  Hopefully,
the slippage won't cost you any features or result in display
artifacts that aren't already part of the original design
(e.g., updating the contents of the display asynchronously
wrt it's being painted into the hardware)

Re: Convert 16x2 chars display to big TFT display
On Wednesday, September 22, 2021 at 2:49:47 AM UTC-4, pozz wrote:
Quoted text here. Click to load it
  
Quoted text here. Click to load it
  
Quoted text here. Click to load it
.  
Quoted text here. Click to load it
  
Quoted text here. Click to load it
  
Quoted text here. Click to load it
  
Quoted text here. Click to load it
  
Quoted text here. Click to load it
  
Quoted text here. Click to load it
  
Quoted text here. Click to load it
is.
  
Quoted text here. Click to load it

The timing of the HD44780U and the various emulations is specified assuming
 a given internal clock speed set by an RC.  If you look at the timing deta
ils of the instructions as the clock speed varies you can get timings that  
can let software work without reading the READY flag.  The data sheet timin
gs are based on the ideal timing of the internal clock and are not accurate
 for production variations as well as temperature and voltage.  But since t
here is never any harm in the controller working faster than the given timi
ngs, you don't really need to worry with this as long as your MCU runs fast
er than the HD44780U with a 270 kHz clock.  I can't think of a reason why y
our emulator device would not work much faster than the HD44780U.  

--  

Rick C.

+ Get 1,000 miles of free Supercharging
We've slightly trimmed the long signature. Click to see the full one.
Re: Convert 16x2 chars display to big TFT display
On Tuesday, September 21, 2021 at 10:34:52 PM UTC-4, Don Y wrote:
Quoted text here. Click to load it
sed on  
Quoted text here. Click to load it
be 10",  
Quoted text here. Click to load it
urce  
Quoted text here. Click to load it
 and the  
Quoted text here. Click to load it
he CPU  
Quoted text here. Click to load it
AM/DDRAM  
Quoted text here. Click to load it
urse  
Quoted text here. Click to load it
R/W  
Quoted text here. Click to load it
ing  
Quoted text here. Click to load it
truction  
Quoted text here. Click to load it

No need for adding hardware.  Read the data lines and the enable signal (an
d R/W, RS if you want) in a single read operation.  Save the previous value
 in a memory location and watch the new value for E to drop from 1 to 0.  T
hen read the previous sample where E was high.  If your reads are fast enou
gh it will contain valid write data.  The tDSW is only 80 ns with 5V power,
 so the interface reads will need to be faster than 12.5 MHz to assure vali
d data.  With a 3.3V unit the timing is 195 ns, so a bit faster than 5 MHz  
will do the job.  If you are emulating a 4 bit data bus the entire set of s
ignals fits in a single byte port.  

Otherwise a rather simple FPGA can do the job.  If you want to emulate the  
full set of features of the HD44780 it might take a bit of work since it do
es some fancy cursor operations and has an internal buffer for user defined
 characters.  

The HD44780 has a few bugs that require software work arounds in the driver
s.  The big one is the reset issue that requires an arcane sequence to init
ialize the device.  This does not need to be emulated since the work around
 sequence will work even if the chip doesn't have the bug.  The operation o
f the cursor and character writing is a bit clumsy, but should not be hard  
to emulate.  I don't recall any timing you will be obligated to emulate oth
er than meeting the bus speeds which is obvious.  The various delays in the
 chip are handled in the interface software either by reading a flag or sim
ply by waiting some minimum time.  If your emulation runs faster it won't m
atter.  I'm pretty sure there is no "too fast", but it's been a year or so  
since I worked with it and don't recall every detail.  

--  

Rick C.

- Get 1,000 miles of free Supercharging
We've slightly trimmed the long signature. Click to see the full one.

Site Timeline