miercuri, 19 mai 2010

Using a 5V CG LCD in a 3.3V design

Some application are requiring large viewing area character display LCD, powered at 3.3V. However this version can be hard to find, while the 5V version of the same LCD is widely available. Nowadays, the stress in our daily life is often reflected in some datasheet content, design engineers being forced to read datasheet information through the lines, trying to elucidate some hidden characteristics or omissions of the incriminated documentation. This make us to search for an unusual solution.

For example, the JHD762Y/YG, manufactured by E’go China, which is an excellent 40x4 characters-type LCD [1] with backlight and the largest size of dot matrix (9.54 mm x 4.84 mm character size or 118.8 mm x 38.5 mm effective viewing area), appears to be a 5V device, according to the electrical characteristics specified in its datasheet. However, being a HD44780 CMOS compatible device (as revealed after carefully reading the initialization sequence and timing diagrams and comparing with the datasheet of HD44780, which is the industrial standard device), an experienced user may wonder if the JHD762Y/YG LCD could also run at a lower voltage. Inspecting its back side, one can easily notice an unstuffed area containing an SO8 package footprint marked U4 and two capacitors C1 and C2, which compose a negative power supply.

On the opposite side of our first example is a very small and inexpensive 8x2 characters-type LCD named AC-082AYI, also HD44780 compatible, manufactured by Ampire [2], which appears to be, from the electrical characteristics provided in its datasheet, a 2.7V…5V device; however, after reading the initialization sequence (15 ms delay after VDD rises to 4.5V, where VDD is the positive logic voltage) it looks more like a 5V device (or else we have discovered an error in its datasheet). The bottom side of the AC-082AYI LCD reveals the same unpopulated footprints, marked as U2, C3 and C4 (fig. 1).

Fig 1. The AC082AYI LCD - bottom side

What are the conclusions? Both LCDs are almost for sure designed to run properly between VDD-VSS=3.3V…5V (where VSS is the reference voltage connected to ground). The AC082AYI LCD becomes a fully 3.3V device after the board is populated with missing components. The message on such a display becomes visible only when |VDD-V0| ≥ 5V (where V0 is the contrast voltage of the LCD). A simple math computation indicates a value of V0 ≈ -2V for a properly operated LCD at VDD = 3.3V.

Fig. 2 Embedded design using the 5V JHD762Y/YG LCD powered at 3.3V

Testing the above idea is a simple task, as long as we can turn to our advantage the use of the free resources provided by great enthusiastic people. First we need a microcontroller [3] and a programmer [4], then we need to download a compiler [5], an Integrated Development Environment [6] and some libraries and examples [7] for an easy learning curve. Based on these tools, writing the firmware [8] is quite easy. Some simple modifications on the JHD762Y/YG are required to make it run at 3.3V (fig. 2), that is mounting U4, C1 and C2 (which is a negative charge pump converter) and supplying the negative contrast voltage V0 around 2.5V. If there is an IO pin available, negative voltage may be generated directly from the microcontroller. At VDD = 3.3V, the LCD’s operating temperature (0 to 50C, according to datasheet) and viewing direction angle (6 o’clock) remains unchanged. Fortunately we have solved our problem (fig. 3) by using a single 3.3V power supply, since the backlight works as well with a slightly lower brightness.

Fig. 3 The JHD762Y/YG LCD powered at 3.3V

Note: This article has been rejected by the EDN Design Ideea Editor because:
“At this time, we have other Design Ideas that we feel will draw a broader audience.”

Did you find it useful?


[1] JHD762Y/YG datasheet http://www.lcdmodkit.com/specification/LKC-2004G-A2.pdf
[2] AC-082 http://pdf1.alldatasheet.co.kr/datasheet-pdf/view1/277539/ZETTLER/AC-082AYILH.html
[3] Microchip PIC18F25J10 datasheet: http://ww1.microchip.com/downloads/en/DeviceDoc/39682E.pdf
[4] Microchip Pickit2 programmer,
[5] Kyle York, JalV2.4m compiler: http://www.casadeyork.com/jalv2/
[6] Sunish Issac, Jaledit 07.20: http://code.google.com/p/jaledit/downloads/list
[7] Rob Hammerling, Joep Suijs, Stef Mientky, Sebastien Lelong, Matthew Schinkel, Albert Fabber, Eur van Andel, Javier Martinez, William Welch, Wouter van Ooijen, libraries for JalV2: http://code.google.com/p/jallib/
[8] http://www.itim-cj.ro/~vasile/18F25J10_lcd_EDN.zip

3 comentarii:

  1. Hi,

    And what are the elements of an article which can attract a large audience in their opinion?

    Or what are the elements which determined them to reject the article? JAL? Open source tools and not the commercial ones? Poor fellows...

    Anyway if their rejection was the reason of creating this blog, then it has a positive result so, Welcome to Blogspot!


  2. Hi Vasi, it seems you've understood very quickly, it seems there is a positive result after all.

    BTW, I like your daughter windmill...


  3. Hi Vasile, thanks for the pointer from jallist to your blogspot. Your article will greatly assist in my education in things PIC and JAL.