Siemens ChipCards SLE4432 / SLE4442

[Wilhelm]

Hi,

I am a software developer (Linux GCC mainly) but have been tasked by a company in our group for a PCW Driver for the SLE 4442 Chipcard.

When done I will post the code in the Code Library for everyone to benefit from (Open source is part of my coding practices)

For those who are not familiar with the SLE4442 Chip, it is a 256-Byte EEPROM.

Datasheet : http://www.scdeveloper.com/datasheet/sle4442.pdf

Assume the following :
1) PIC16F84
2) RA0 -> RST on SLE
RA1 -> CLK on SLE
RA2 -> I/O on SLE
3) include lcd.c (so this will be on RB0-7) for output.

Page 10 of the SLE datasheet shows the ATR proceedure.

The SLE requires 31 clock pulses after a reset.

so I would set RA0 - High (Reset), then Low again to complete the reset command.

then to generate the 31 clock pulses must I do the following ?

[libor]

Just a few comments:
- If I get it right there must be a clock pulse during the reset also according to the datasheet.
- The datasheet specifies both min. and max. timing intervals and frequencies for all signals, you should also consider the timing issues depending on the frequency your PIC is running, the output_high() followed by the output_low() might result a pulse too short, the printf() may take too much time before the next loop iteration...
- Dont forget, that I/O line is an open drain, must be pulled up by an external resistor.
so much came to my mind so far, is it reading something at all?

[Guest]

Thanks for the reply.

Note Taken of the fist clock pulse on the reset signal.

You mention the Min and Max clock pulses. I see the clock freq is min 7 and max 50kHz.

(in my First post I said will be using 16F84 just to add this is at 4Mhz)

Then what are the implications of using a different PIC with say a 20Mhz Xtal ?

How can I ensure that this is maintained ? irrespective of the PIC being used ? Say keeping a value of 25kHz ?

If the Printf() may take to long, would you suggest that I do all the read/write operations to the SLE first then send output to the LCD ?

I/O Line is an open drain ? So RA2 can't connect directly to the I/O. Must there then be a resistor between VCC and I/O, With I/O to RA2 ?

I understand a pullup resistor to be between VCC and Pin X and a Pulldown as between GND and Pin X. Is this right ?

Thanks for all the help ...

[libor]

The simplest way of making pulses of certain length is to use the delay_us() or delay_ms() instructions between port manipulation.
The #use_delay(clock=4000000) in the beginning will inform the compiler how fast is your PIC running (4Mhz in this case), so it will generate the delay routines accordingly. If you move to another PIC running at different speed, you only have to change this directive.
Eg. to make a clock pulse of ca. 25 kHz: (length of high pulse is ca. 20us = half period at 25khz)

[Wilhelm]

Hi again,

After a bit of experimenting ...

The Datasheet states while sending the ATR you get the first four addresses on the eeprom read back during the ATR pulses.

This works 100% (with the code below) and the fist 2 addresses are dispalyed on line 1 of the LCD and the next 2 on line 2.

[libor]

- the bits on the data line are LSB first according to the data sheet diagrams: to send the command 00110000, you have to send the bits in reverse order: 00001100
- i would suggest to use #use fast_io(port) directive to allow setting each port's direction manually with set_tris_X() command, so you have to set each port's direction explicitly before using it either for output or for input, it is less confusing when you use a port that function in both directions (DATA line) you can be always aware of what direction a port is set to.
- I have not checked thoroughly if you deal with the start-condition and stop-condition mentioned on page 10. have a look at it.

[Wilhelm]

Hi,