Savel brain dump in English!

It was told that it is impossible to connect standard controller-less LCD screen to the computer’s printer port. The main reason it is not working is the speed of the computer. In fact, the speed of printer port.

I decided to test this situation and attached Sharp LM64P10 (640 x 480 STN) screen to my computer. And here is the result:

The image is repeating because the first 4 bits of the data byte is for upper screen and other bit are for lower part. As I was missing spare pins on my LPT to “breadboard” adapter, I just connected the lower nibble with upper.

The refresh rate of the screen is horrible. The only reason you see this nice image, is that digital photo camera is using very long exposure (half of the second). The real refresh rate is only 2…3 fps. The manufacturer recommends to use at least 80 fps.

The data is going to the right place- pixel to pixel. Now I need to attach some SED or Hitachi controller or build some self-made controller based on FPGA.

Technical stuff below…
(more…)

To program CPLD programmable logic) devices you only need very simple isp download cable. The cable is simple if we use computer’s printer port.

For my experiments I removed Lattice isp2032VE chip from old wireless router and soldered it to test PCB. This chip is supported by free licensed Lattice Semiconductors software. You can download software for free from Lattice web pages.

Somehow cable circuit diagram is hidden in Lattice pages, but while searching world wide web I found it. The first description was from Intel (!) web page. It is description of the cable and small schematics.
And from some personal pages I found full circuit diagram of Lattice ispDownload cable. It is universal isp and JTAG cable, but it is supported only by Lattice software.

New CPLD chips are 3.3V or less logic, but cable chips are powered by 5V. So don’t forget to use proper power supply voltage. My chip is 3.3V, but 5V tolerant, so no problems with interfacing with computer.

No need to build PCB for such small device, so I made everything in the “air” and glued with thermal glue.

These two resistors I added only to protect my computer if I’ll decide to connect the powered cable to computer. One resistor is connected to Vcc sense- I don’t like to put full Vcc power to my computer LPT port pin. Other resistor is connected between shield ground and data ground pins. When you connect the connector, the shield is connected first and all voltage difference with go threw this resistor first reducing possible power surges.

From old device I removed controlles graphic LCD screen and part of the mainboard with Toshiba controller. As controller was with all hardware, I just connected everything with air wires to PC printer port using my breadboard.

There are two main chips on controller “board”- the controller itself (Toshiba T6963C) and small RAM to store data. All other stuff left on board is leftover from the other device components and negative power supply.

(more…)

LCD module test or how to connect the beast to PC printer port.

I received few requests how to test the LCD module using easy to obtain software. Maybe people are to lazy to search internet… So, here is simplified instructions how to connect standard LCD module to PC computer port (LPT) and how to configure WinAmp plugin.

The circuit diagram is very simple. Attention! Some old or nonstandard, “winter”, modules require additional power supply to be connected to Vee pin (contrast). But most of the modules are working from regular voltage, and you connect Vee pin to ground or using some trimmer connect it between Vcc (+5V) and ground (GND). All standard modules after power-up show one dar line to demonstrate that they are working.

The circuit itself is very simple. Take note to trimmer on Vee.

(more…)

After I smashed about 5 mobile phones and collected some white LEDs, I decided to build some some LED searchlight. There are two ways to solve this problem: first, Chinese version, just connect some white LED’s to battery and use some ballast resistor. And watch you spotlight fading while the battery is loosing it’s power. Ante the second, hi-tech, way. To build special LED driver.

almost all recent white LEDs are in fact blue LEDs with special coating to emit missing wavelengths and the result is white color. To use LED and full efficiency, you must supply constant working current to LED. It is possible to connect LED in daisy chain way using only one current ballast, but the voltage must be quite high (~3V per LED) or connect them in parallel. But in parallel mode you must use separate ballast for each LED.

Chip manufacturers make special chips- LED drivers. This chip lets you to connect LEDs in daisy chain and is working from single cell. I received sample from Linear Technologies. It is LT3486, dual channel white LED driver from single Li-ion cell.

It is generic step-up converter with feedback from current shunt resistor and over-voltage protection. In sample circuit there are 2×8 white LEDs connected to chip. But other example is with 2×16!

I thing the limit is only max power on the chip, max voltage on the output (35V) and max current (~25mA typical) per channel. And the number of diodes are not very significant value.

The only bad thing is that chip is “new style”, lead-less and very small…

The circuit is very simple. All component are easy to get. As I love using second hand components, I had some difficulties collecting very small 10μH inductors. Thanks to Samsung digital cameras, I removed inductor from camera’s PCBs. Other component was fast schottky diode- recommended one was ZHCS400 (40V, 400mA), but I placed IR1F4 (10BQ040), 40V, 1A. I don’t know if it is fast enough.

This is first run. Only one channel is working- 9 White LEDs from mobile phone back-light. Power supply is one LiIon cell (~4V). The voltage across LEDs are 29V, empty circuit voltage- 36V. White LEDs are getting quite hot- I made some mistake in calculations and LEDs are overloaded.

(more…)