Posts by fs-support_KW

    Just for tolerance of the crystal there are 20ppm. For aging additional 3ppm/year. This are vaues at exact 25 degree. If the crystal temperature is rising just to 45 degree on you working system you must add 16ppm.
    So it is impossible to reduce offset without temperature compensation.
    This would need a always working board to measure the temperature near the crystal - impossible.


    So the best way for more exact values to use a temperature controlled oscillator (TXCO) on I2C. We have a good experience with Maxim DS3231 with just 3.5ppm between -40...85 degree. This means daily offset is max. 0.3 sec.

    From point of EMV noise and easy to handle we strictly recommend using a LVDS displays. There are cable with JAE FI-S20S and Hirose DF19G on display side available from F&S. This will match 95% of useful LVDS displays.


    Every RGB panel is different. So we can't ship a "standard adapter" or "standard cable" for exactly your RGB display.
    You must design your own adapter doing this steps:
    1) buy a B.MKAB.29 SINTF-LVDS-JAE cable from F&S with JAE FI-S25S on armStone side and a JAE FI-S20P-HFE for you display adapter board
    2) load datasheet for 75LVDS86A from http://www.ti.com
    3) Make a design similar to page 11 on this datasheet. Between SN75LVDS86A and your panel use a connector with the right pinout for easy using. The panel datasheet should refer the right fitting connector.
    Note: don't forget the 100 Ohm termination on LVDS lines
    4) design power decoupling for the SN75LVDS86A following application note SLLA300.pdf from http://www.ti.com
    5) connect your display and correct the timings in registry following the display datasheet using a oscilloscope or logic analyser on digital side.
    On this way we can give you a paid support depends our effort. Please contact our sales or your distribution channel.


    Another way (we can't support you doing this, but you are able to adapt without making a PCB):
    1) go to <!-- m --><a class="postlink" href="http://en.esskabel.de">http://en.esskabel.de</a><!-- m --> and order
    1x KAB-SET-FI-S20S-DF14-XXS-0500RK
    1x JAE FI-S25S housing for using this cable with armStoneA8
    1x ADA-LVDS-RX-TWIN LVDS to TTL adapter
    1x cable from ADA-LVDS-RX-TWIN to your panel (ESS supports you to adapt your panel)
    2) connect your display and correct the timings in registry following the display datasheet using a oscilloscope or logic analyser.
    On this way please contact ESS for the support.


    The LVDS channel did restrict display clock to 31..75MHz. Working with lower clock is not supported.

    Der zulässige und zuverlässige Spannungsbereich der PicoCOM2 beträgt 3.3V +/-5%. Sollte daher die Spannung von 3.135V unterschritten werden, können wir aufgrund der Herstellervorgaben der Chiphersteller nicht mehr für die Stabilität der Baugruppe garantieren.
    Solange der Resetgenerator nur dazu dient, beim Powerup die nötige Verzögerung des Reset bis zum Erreichen der minimalen Betriebsspannung sicherzustellen, spielt seine Schaltschwelle keine Rolle.
    Sinkt jedoch während des Betriebes die Spannung unter die oben erwähnten 3.135V ab, empfehlen wir aufgrund der Herstellervorgaben, einen Reset auszulösen.


    The recommended power supply range is 3.3V +/-5%. On voltage dropdowns below 3.135V we can't give guarantee for stabel working board because the datasheet values from the IC manufacturer.
    The reset generator on the baseboard is just for a delayed reset on power up to have a stabel voltage on start. Thats independed from voltage trigger value of the reset generator.
    On a voltage dropdown below 3.135V we strictly recommend to generate a Reset.


    fs-support_KW

    We will remove the onbard pullups in mass production. For customers using I2C we will update the manual to add pullups 1... 2.2 kOhm to 3.3V on baseboard. This additional pullups for I2C must be used, because internal pullups on chip are >20kOhm.
    For modification of your PicoCOM4 sample we will send you a personal message.


    KW

    Dear Dirk,
    sorry for the long delay. The maximum power consumtion from VLCD @ NetDCU8 is 180mA@3.3V and 2.7A@5V. Thats because the limitations of the onboard 5Vto3.3V regulator. So you need en external power supply for your 3.3V display.

    The Atmel Controller will control the core voltage regulator enable signal by themself with the SHDN signal. But this signal is not working correctly, if VDDBU (the battery voltage) is connected. Atmel is publishing this error in the newest version of the AT91SAM9260 datasheet (pg 776, Shutdown Controller "SHDN Signal may be Driven to Low Level Voltage During Device Power-on"). In systems without battery this problem doesn't occur.
    On our hardware fix we will enable the core voltage regulator all time without using the SHDN signal. This fix will be included on all delivered boards starting today. Older boards can be corrected by our repairing crew. Please contact your sales partner to get an RMA number.
    Experienced customer can make this fix by themself.


    KW
    [attachment=0]<!-- ia0 -->Rework instruction PicoCOM1 to fix Atmel SHDWC bug.pdf<!-- ia0 -->[/attachment]

    The Atmel Controller will control the core voltage regulator enable signal by themself with the SHDN signal. But this signal is not working correctly, if VDDBU (the battery voltage) is connected. Atmel is publishing this error in the newest version of the AT91SAM9260 datasheet (pg 776, Shutdown Controller "SHDN Signal may be Driven to Low Level Voltage During Device Power-on"). In systems without battery this problem doesn't occur.
    On our hardware fix we will enable the core voltage regulator all time without using the SHDN signal. This fix will be included on all delivered boards starting today. Older boards can be corrected by our repairing crew. Please contact your sales partner to get an RMA number.
    Experienced customer can make this fix by themself.


    KW
    [attachment=0]<!-- ia0 -->Rework instruction PicoCOM1 to fix Atmel SHDWC bug.pdf<!-- ia0 -->[/attachment]

    It's a known problem. The reason is most Linux distributions are making a RESET via the historical PS/2 keyboard controller on port 60/64h. Because the nanoModX1 doesn't have a keyboard controller this RESET sequence is not working. It needs a recompilation of the kernel with the parameter reboot=pci instead the default reboot=kbd.