Shop OBEX P1 Docs P2 Docs Learn Events
security - Page 2 — Parallax Forums

security

2»

Comments

  • mctriviamctrivia Posts: 3,772
    edited 2009-02-17 03:29
    edit only shows up if no one read it between time of writing it and editing it(i think)
  • Cluso99Cluso99 Posts: 18,069
    edited 2009-02-17 04:06
    If you can find an smt fuse of about 25mA you could use the prop to blow the fuse by outputting a high on the prop transmit pin and grounding the transmit pin at the propplug connector on your pcb (or visa versa). For normal use, limit the current by a series resistor in the propplug (or cable between).

    Another idea. Place a zener diode between the prop pin and ground to limit the voltage to the prop to 3V3. Place an smt fuse in series with the prop pin. Then to blow, place a higher voltage (you will need to calculate this so that the zener can handle the current, but the fuse cannot) on the prop connection and ground.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Links to other interesting threads:

    · Home of the MultiBladeProps (SixBladeProp)
    · Prop Tools under Development or Completed (Index)
    · Emulators (Micros eg Altair, and Terminals eg VT100) - index
    · Search the Propeller forums (via Google)

    My cruising website is: ·www.bluemagic.biz
  • PropabilityPropability Posts: 142
    edited 2009-02-17 04:17
    ·Applying a negative voltage to the RX pin was what I was trying to do (the reason for the diode). I've just played around with it (not on the propboard but just a breadboard) with another diode from the RX to the positive supply.

    With the negative supply of the battery applied to the positive of the breadboard and the positive applied to the RX I got a 10 ohm resister pretty hot.

    Added another way to just use the propplug connection.


    Post Edited (Propability) : 2/17/2009 5:13:36 AM GMT
    528 x 266 - 15K
    2432 x 848 - 348K
  • mctriviamctrivia Posts: 3,772
    edited 2009-02-17 04:49
    this probably would work but may hurt prop. negative voltage at tx pin will conduct to ground breaking fuse
    358 x 129 - 3K
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2009-02-17 05:53
    mctrivia,

    You could use a 6-pin (SOT23) PIC ($0.83/C) or 8-pin (SOIC) AVR ($0.84/C) to make a password-protected reset line — two pins (DTR, RxD) in, single pin (RxD) out. You'd unlock it by wiggling DTR with the proper code, after which the RxD line would simply pass thru, and you could reprogram the Propeller. Power cycle the board, and the RxD line would be locked again, i.e. no pass-thru.

    I know you said, "I would like it that not even I can bypass the security measures", but what if you could, securely?

    -Phil
  • mctriviamctrivia Posts: 3,772
    edited 2009-02-17 05:57
    could not power cycling the board at the right time act as the reset needed to program?
  • Cluso99Cluso99 Posts: 18,069
    edited 2009-02-17 09:19
    Blowit2 could damage the prop. Blowit3 should work.

    I was thinking about a circuit like opt3, but using a 3V zener to protect the prop pin. The zener will need to be able to handle the current to blow the fuse.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Links to other interesting threads:

    · Home of the MultiBladeProps (SixBladeProp)
    · Prop Tools under Development or Completed (Index)
    · Emulators (Micros eg Altair, and Terminals eg VT100) - index
    · Search the Propeller forums (via Google)

    My cruising website is: ·www.bluemagic.biz
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2009-02-17 10:41
    mctrivia said...
    could not power cycling the board at the right time act as the reset needed to program?
    If you're switching only /RST, yes; otherwise, no. (You must've seen the post I deleted. smile.gif )

    -Phil

    Post Edited (Phil Pilgrim (PhiPi)) : 2/17/2009 10:51:58 AM GMT
  • Beau SchwabeBeau Schwabe Posts: 6,568
    edited 2009-02-17 15:45
    It bugs me that there is a potential for damaging current so close to the I/O's,TX, and·RX, it just sends up a red flag.
    Here is a minimal parts design excluding the fuse that uses·a single transistor, diode, and 3 resistors to keep the high current away from the I/O's.· The idea is, that in normal operation, the transistor is "off" allowing data to pass in either direction.· When the fuse is blown, the transistor conducts via the 2.2K bias resistor which acts to snub the data on the·receiving end down to about 60mV.
    ·




    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Beau Schwabe

    IC Layout Engineer
    Parallax, Inc.
    751 x 481 - 113K
    OTP4.JPG 112.6K
  • mctriviamctrivia Posts: 3,772
    edited 2009-02-17 15:51
    with otp3 high current would never be at pin just through the diode and fuse and only briefly till fuse blows. the big question is will the prop mind the -0.2V that would be applied to the pin until the fuse blows.
  • Beau SchwabeBeau Schwabe Posts: 6,568
    edited 2009-02-17 15:57
    mctrivia,

    "the big question is will the prop mind the -0.2V that would be applied to the pin until the fuse blows." - most likely... you run the risk of latchup in the IC when you do that.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Beau Schwabe

    IC Layout Engineer
    Parallax, Inc.
  • mctriviamctrivia Posts: 3,772
    edited 2009-02-17 16:02
    latchup?
  • Beau SchwabeBeau Schwabe Posts: 6,568
    edited 2009-02-17 16:19
    mctrivia,
    ·
    In any CMOS circuit design, there are parasitic bi-polar transistors that are formed.· With a simple CMOS inverter circuit, which are practically everywhere inside of an IC, it just so happens that the parasitic bi-polar transistor arrangement forms an SCR.· Latch up is a term referred to when this parasitic SCR becomes active and turns "ON" which causes a direct short to the power rails of the IC.· The end result to the IC is usually irreversible.


    Reference:
    http://en.wikipedia.org/wiki/Latchup

    "A spike of positive or negative voltage on an input or output pin of a digital chip, exceeding the rail voltage by more than a diode drop, is a common cause of latchup."

    In the case of the Propeller, the·effective diode drop is about 0.3V to 0.4V under a load... much less without a load·... 0.2V is way to close·for comfort.


    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Beau Schwabe

    IC Layout Engineer
    Parallax, Inc.

    Post Edited (Beau Schwabe (Parallax)) : 2/17/2009 4:33:38 PM GMT
    963 x 505 - 98K
  • mctriviamctrivia Posts: 3,772
    edited 2009-02-17 16:41
    if there is no power to the prop just gnd will that be a problem? no short because no positive rail?
  • Beau SchwabeBeau Schwabe Posts: 6,568
    edited 2009-02-17 16:47
    mctrivia,

    You need to·be absolutely certain that power can not enter parasitically through another I/O pin for it not to be a problem.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Beau Schwabe

    IC Layout Engineer
    Parallax, Inc.
  • mctriviamctrivia Posts: 3,772
    edited 2009-02-17 16:57
    so short vcc and vss during burn out for safety
  • Jimmy W.Jimmy W. Posts: 112
    edited 2009-02-17 17:23
    I have nothing to contribute, but what a great thread! Lots of good ideas!
    Question though, is it more important to burn out the tx/rx that could later be used for troubleshooting, or is it more important that the eeprom doesn't get overwritten/altered?
  • dMajodMajo Posts: 855
    edited 2009-02-17 17:27
    Have a look at this ICs

    They have non volatile IOs (they preserve the configuration·at power cycling) managed through I2C. You can communicate through the serial port (P30/31): give a passw protected command to lock/unlock the DTR<->RST link (and/or Prop TX·leaving RX enabled to change the config again) also togling the eeprom WP pin could be a solution

    http://www.standardics.nxp.com/products/pca/datasheet/pca8550.pdf


    Post Edit: Or use a ramtron FM31L278 device in place of the boot eeprom.
    I have suggested this IC so many times because I really love it (no incomes from ramtron). In one IC you get the eeprom substitute (software four levels write porotection), RTC, WDT, Counters (operational also under Vbat), Power fail comparator, User definable 64bit SN, reset debouncer ...

    You can use its software write protection control through the i2c, conditioned upon a pws serial comand. (If the eeprom is wp no need to take care on tx/rx/rst since it can not be rewrited)

    Post Edited (dMajo) : 2/17/2009 5:53:40 PM GMT
Sign In or Register to comment.