Satistronic RT2002A2 LCDs behaving strangely

[EsotericLabs]

Hello Midiboxers,

 

I have strange behavior with the Satistronic RT2002A2 20x2 LCD: it keeps displaying a line of blocks on the upper lines.  The cabling is one-to-one and I can measure 5V at pins 1&2 and a variable voltage (backlight contrast!) on pins 15/16.  So this looks OK, right?

 

It is not in the LPC17 core nor the MIDIO128 software as my old and unpractical 16x2 LCD functions properly. I also double checked with a second RT2002A2, which also gives only a line of blocks.

 

Any clues to look into?

 

Thank you!

Edited February 24, 2013 by Martijn

[TK.]

Hi,

 

sounds like the LCD isn't initialized - either the RS/RW/E, or the data lines are not connected properly (or in the right order).

 

The latest MIDIO128 release provides a "testlcdpin" command which allows to troubleshoot such problems.

Just enter this command in the MIOS Terminal to get the testing instructions.

 

Best Regards, Thorsten.

[EsotericLabs]

Thank you! Great I'll give it a try tonight.

[EsotericLabs]

Hello TK,

 

might the problem also be related to this issue: ?? The datasheet of my LCD http://www.satistronics.com/myfiles/file/LCD/RT2002A2.pdf does not say anything about 5V tolerance mentioned in the other topics.

 

Anyway, i will do the testlcdpin tonight: I must measure the voltages at the LCD connector iiuc?

 

Best,

Edited February 25, 2013 by Martijn

[TK.]

might the problem also be related to this issue:  ??

 

this problem has been solved (as you can see at the end of the thread), and the latest MIDIO128 release contains the changed LCD driver.

 

 

 

I must measure the voltages at the LCD connector iiuc?

 

Yes - it makes sense to measure the voltages at the LCD.

And if you notice a failure, continue measuring at the core side to clarify the wrong (or missing) connection.

 

Best Regards, Thorsten.

[EsotericLabs]

hello TK,

 

I think I managed to track down the problem to the RW line: Without the display connected, measurements for RS, RW en E (when set to 1) are ca 3.3 and ca 0,2V when set to 0. D0-7 have ca 5V for 1 and 0V for 0.

 

But when I connect the display, the behavior of RW changes:  it has always 5V!  It is not a short at the pins soldered on the display (confirmed by multmeter beep test).  Could the pull-up issue play a role (in other thread) or do you still suspect my soldering ;-)  ?  (btw. the other pins act normally with the display connected).

 

If it is the pullup problem, where should I make what changes to the code?

 

Thanks a lot!

[TK.]

If it's related to the push-pull configuration, you would only need to compile the latest version in the repository.

 

Here is a prebuilt binary: http://www.ucapps.de/mios32/midio128_v3_014_pre1.zip

 

Please let me know if this solves the issue - in this case I will release v3.014 officially

 

Best Regards, Thorsten.

[EsotericLabs]

Thank you very very much. But alas, it does not help.  The old 16x2 happily says Midio128 V3.014 but the new 20x2 only shows an upper line of blocks. Should I retest with testlcdpin?

Edited February 25, 2013 by Martijn

[TK.]

It would be interesting if RW is still 5V, or 3.3V now

 

It also makes sense to measure the voltage directly on the core, including the pin on the LPCXPRESSO (see schematic).

Measure this with LCD cable connected and disconnected!

 

Best Regards, Thorsten.

[EsotericLabs]

hello TK,

 

now I measured the following:

with testlcdpin rw 1:  3V36 at j5a/b and p1.27on lpcxpresso without cable and display; and with cable only (no lcd); with testlcdpin rw 0; same spots and connections 0v24

 

with cable and LCD connected; however, the same measurements at j5b (cable is at j5a) and p1.27; but still always 4v96 (rw 0 and rw 1) on the lcd itself.  But measuring between Vss on the LCD and j5b.rw // p1.27 both with rw 0 and rw 1 gives proper voltages ie. 3v36 and 0v24.   It is either the cable or the LCD itself then, is that correct? There should be something with the cable, otherwise measurements at j5b.rw = p1.27= lcd.rw, right? But why only the RW line? that does not make sense to me. I'd expect an improperly connected cable to give more problems, not on one line.

 

Measuring resistance between Vd (+5V) and RW pins on the lcd gives about 830R with - on vd and + on rw. Other way around is about 750k.

 

Should I get more flatcable and try make another cable?

 

Thanks for helping me so much!

Edited February 25, 2013 by Martijn

[TK.]

Yes, this seems to be a problem with the cable.

 

Especially this measurement:

 

 

but still always 4v96 (rw 0 and rw 1) on the lcd itself.  But measuring between Vss on the LCD and j5b.rw // p1.27 both with rw 0 and rw 1 gives proper voltages ie. 3v36 and 0v24.

 

shows, that there is no direct connection between the LCD RW pin, and the core RW pin (btw.: you mean J15B, not J5B)

Otherwise you would measure exactly the same voltage.

Instead it looks like LCD RW is connected to J15A:Vd (5V)

 

Best Regards, Thorsten.

[EsotericLabs]

Confirmed and solved: cable issue! Thank you very much.