Page 4 of 6 FirstFirst 123456 LastLast
Results 61 to 80 of 112

Thread: msay MKII interface project (gw1 hacks)

  1. #61
    Junior Member
    Join Date
    Mar 2009
    Age
    47
    Posts
    46
    Thanks
    3
    Thanked 4 Times in 4 Posts
    Rep Power
    185
    Reputation
    24

    Default

    Sorry for late replying.
    after 10 times using light & magnifier , 1 of the rail had a damage (disconnected) and i think it should be new.
    anyway , right now the LEDs and 13 volt are ok for me but i still cannot read my cards with it.
    Also 555 pin3 still shows 3.8 volt !!
    my components:
    NE555N
    HCF4053BE (couldn't find BP and N )
    CD74HCU04E
    74HC04N
    MAX232N , MAX232CPE
    C547B
    C558C , BC557
    1N4007 INSTEAD OF 1N4004

    Modifications to Rev. 5:
    R21: 10K
    R5, R15, R16 : 820R

    Power: Wall adapter that shows 17.5 Volt , after connecting to programmer it shows around 12.5 Volt.

    Today i will spent many hours on it and will inform you again.
    Last edited by bobzombie; 21-03-09 at 01:30 AM.



  • #62
    Junior Member
    Join Date
    Feb 2009
    Posts
    55
    Thanks
    0
    Thanked 0 Times in 0 Posts
    Rep Power
    186
    Reputation
    12

    Default

    Quote Originally Posted by gw1 View Post
    Yes, of course, but it does mean the user needs to come up with a 15V supply which is less convenient than 12V. That's what I meant, I wasn't really clear about it.
    Is hard to get a 15V PSU??
    I'm getting more trouble to get a VISHAY 2N7000KL than a 15V adapter.

    Quote Originally Posted by gw1 View Post
    I have a feeling we haven't seen the last of the switching issue. And I encourage you to explore the 2N7000 option
    I will do some tests but i don't have a lot of options here, i can't have access to a empty card to play with. Also I'm working blind because i don't understand the details about SC.

    Quote Originally Posted by gw1 View Post
    I have a feeling the I/O pulldown resistor will cause trouble in some cases if it hasn't already.
    How much trouble this can cause?
    The original design was using the I/O line to power a led, at least 20ma was "stolen" from that line. Using the switch will draw much less than that, why not increase the R10 to about 120K?

    I was reading some info about the RS232 interface from this site:


    The standard voltage range on RS-232 pins is _15V to +15V. This voltage range applies to all RS-232 signal pins. The total voltage swing during signal transmission can be as large as 30V. In many cases, RS-232 ports will operate with voltages as low as _5V to +5V. This wide range of voltages allows for better compatibility between different types of equipment and allows greater noise margin to avoid interference.

    I think you have access to many type of cards than I, if you could build the interface for yourself it could be easy to debug this erros, if any.

    Quote Originally Posted by gw1 View Post
    I'm pleased to hear so many others have had success with your work
    Thanks but is your hacks

    Many of the people that have build this interface are cellphone repair man, they use the interface to dump the phone book (so they say). Others to program ID cards.

    So far, i did not receive any complain by this people, that makes me think what is wrong with the one that bobzombie have build.

  • #63
    Senior Member gw1's Avatar
    Join Date
    Jan 2008
    Location
    Hobart
    Posts
    957
    Thanks
    49
    Thanked 608 Times in 213 Posts
    Rep Power
    267
    Reputation
    1901

    Default

    Quote Originally Posted by bobzombie View Post
    Sorry for late replying.
    after 10 times using light & magnifier, 1 of the rail had a damage (disconnected) and i think it should be new.
    That makes sense. I think msay is probably correct, your problems may be because of construction issues rather than design since others have it working correctly.

    555 pin3 still shows 3.8 volt !!
    This is normal. I was wrong earlier when I thought it should be closer to 5V. The 555 is unable to drive all the way up to Vcc - its maximum output voltage will always be a volt or so below.

    my components:
    NE555N
    HCF4053BE (couldn't find BP and N )
    CD74HCU04E
    74HC04N
    MAX232N , MAX232CPE
    C547B
    C558C , BC557
    1N4007 INSTEAD OF 1N4004
    All those components will be fine. Apart from the 'B' suffix on 4053, none of the other suffixes are important.


    Modifications to Rev. 5:
    R21: 10K
    R5, R15, R16 : 820R
    Yes, that's what I recommend. If problems persist you could try taking R5,15,16 back to 470R in case my suggested value is too high, but I think 820R should be right.

    Power: Wall adapter that shows 17.5 Volt , after connecting to programmer it shows around 12.5 Volt.
    That's normal. Your adapter isn't regulated, which means its output voltage without load will be higher than when loaded.

  • #64
    Senior Member gw1's Avatar
    Join Date
    Jan 2008
    Location
    Hobart
    Posts
    957
    Thanks
    49
    Thanked 608 Times in 213 Posts
    Rep Power
    267
    Reputation
    1901

    Default

    Quote Originally Posted by msay View Post
    Is hard to get a 15V PSU??
    I'm getting more trouble to get a VISHAY 2N7000KL than a 15V adapter.
    It doesn't need to be a Vishay, any brand will do. In fact most n-channel FETs/MOSFETs will be ok (provided Vgs threshold is around logic level). They differ from NPN bipolar transistors in that they are voltage controlled, not current controlled. On FETs the gate presents very high input impedance, while on bipolars the base-emitter junction is very low impedance once conduction starts. Changing T3 to a FET means the only resistors biasing the I/O line are pulling high, not low.

    How much trouble this can cause? The original design was using the I/O line to power a led, at least 20ma was "stolen" from that line. Using the switch will draw much less than that, why not increase the R10 to about 120K?
    Certainly there's a limit to how much current a card can sink when it pulls the I/O line low (20mA is excessive). But my concern isn't about current overloading the card, it's about the voltage level on I/O when neither the card nor the PC are transmitting. Smartcard I/O is an open collector bus which needs a pullup to card VCC and ought to have impedance around 5-20K, not too much less or more. My concern was that the T3 configuration might cause the quiescent bias on card I/O pin to be significantly less than the 5V it ought to be. My concern may be unfounded, it's just a possibility I'm raising.

    if you could build the interface for yourself it could be easy to debug this error, if any
    I might make one at some stage but don't have time to etch a board at the moment, I threw out my ferric chloride years ago :-)

  • #65
    Junior Member
    Join Date
    Mar 2009
    Age
    47
    Posts
    46
    Thanks
    3
    Thanked 4 Times in 4 Posts
    Rep Power
    185
    Reputation
    24

    Default

    i'm ****ed up.
    Ordered a new board to start from the beginning .

  • #66
    Junior Member
    Join Date
    Feb 2009
    Posts
    55
    Thanks
    0
    Thanked 0 Times in 0 Posts
    Rep Power
    186
    Reputation
    12

    Default

    Quote Originally Posted by gw1 View Post
    It doesn't need to be a Vishay, any brand will do. In fact most n-channel FETs/MOSFETs will be ok (provided Vgs threshold is around logic level). They differ from NPN bipolar transistors in that they are voltage controlled, not current controlled. On FETs the gate presents very high input impedance, while on bipolars the base-emitter junction is very low impedance once conduction starts. Changing T3 to a FET means the only resistors biasing the I/O line are pulling high, not low.
    Not that i like vishay, I have to depend on farnell to buy parts and vishay is the only options, is very cheap though.

    Quote Originally Posted by gw1 View Post
    Certainly there's a limit to how much current a card can sink when it pulls the I/O line low (20mA is excessive). But my concern isn't about current overloading the card, it's about the voltage level on I/O when neither the card nor the PC are transmitting. Smartcard I/O is an open collector bus which needs a pullup to card VCC and ought to have impedance around 5-20K, not too much less or more. My concern was that the T3 configuration might cause the quiescent bias on card I/O pin to be significantly less than the 5V it ought to be. My concern may be unfounded, it's just a possibility I'm raising.
    I will apply to your suggestions since I'm blind here, I don't understand this stuff.

    Quote Originally Posted by gw1 View Post
    I might make one at some stage but don't have time to etch a board at the moment, I threw out my ferric chloride years ago :-)
    Come on, are you trowing the towel already?
    Is very easy to transfer the design to a board this days, I use emulsion on the PCB, very easy, it gives millimetric details as professional ones.

    Like my last board:
    (big image)




    Quote Originally Posted by bobzombie View Post
    i'm ****ed up.
    Ordered a new board to start from the beginning .
    What did you do?

  • #67
    Junior Member
    Join Date
    Mar 2009
    Age
    47
    Posts
    46
    Thanks
    3
    Thanked 4 Times in 4 Posts
    Rep Power
    185
    Reputation
    24

    Default

    I ordered a new PCB board and is waiting for it to come.

    But i think my first board is working too. here is the problem:
    most of the time, I connect the power, the LEDs are not ok.
    I disconnect the power , Remove IC 5 , connect the power , disconnect the power put back IC 5 , and the LEDs are ok now and i can get ATR.

    But my mistake was the i was trying my 4 viaccess cards , and none of them is working in your card reader , but they work in my infinity & Cas2 .

    But today i tried my Gsm sim card and a card called Abracadabra , and it's working too.

    I couldn't find out yet , what's the problem with IC 5 ?
    And why my Viaccess cards are not working in your card reader ?

    My next step will be testing the response time of your card reader !!
    Viaccess card in my Infinity : ~110 ms
    Viaccess card in my Cas2 : ~90ms

  • #68
    Junior Member
    Join Date
    Feb 2009
    Posts
    55
    Thanks
    0
    Thanked 0 Times in 0 Posts
    Rep Power
    186
    Reputation
    12

    Default

    viaccess?
    This is a satellite card.

    Don't know if my interface will work with it, gw1 could assist you more with this.

  • #69
    Junior Member
    Join Date
    Mar 2009
    Age
    47
    Posts
    46
    Thanks
    3
    Thanked 4 Times in 4 Posts
    Rep Power
    185
    Reputation
    24

    Default

    Quote Originally Posted by msay View Post
    viaccess?
    This is a satellite card.

    Don't know if my interface will work with it, gw1 could assist you more with this.
    I thought u guys are making this for satellite cards.
    Cause most of pplz i know that are using phoenix readers , they use them for satellite cards.
    And i want to use it to connect to my Dreambox receiver as external card reader, cause your cards reader doesn't need driver , then it's the best solution for linux pc & dreambox receiver.
    For example , my infinity usb card reader , need driver for linux , and running it under linux is a real pain in the ass.
    So what do u recommend for making it more compatible with thoes cards ?

  • #70
    Senior Member gw1's Avatar
    Join Date
    Jan 2008
    Location
    Hobart
    Posts
    957
    Thanks
    49
    Thanked 608 Times in 213 Posts
    Rep Power
    267
    Reputation
    1901

    Default

    Quote Originally Posted by bobzombie View Post
    So what do u recommend for making it more compatible with thoes cards ?
    Remove R10 temporarily. The Data LED won't work but it may fix your card compatibility problem. Give it a try. That's my best guess at this stage as to why you're seeing ATR for some cards but not others.

    If removing R10 solves your problem then put R10 back in and replace T3 with a 2N7000 (it needs to be inserted backwards as described above).

    A reminder, make sure the 'COMMA' link is present - you'll usually need it, and you shouldn't ever need to remove it.

    Be aware that Viaccess doesn't work at 6MHz - you should use 3.58MHz for those cards.

    And make sure the serial cable you're using with Dreambox is fully wired (all 9 pins).

    The interface is designed for all 5V ISO7816 cards (lower voltage cards will also work if they're 5V tolerant). Since DVB encryption is a major application for smartcards, that's where many electronics enthusiasts find smartcards to experiment with. But there's nothing about the ISO7816 standard or this interface which is specific for satellite or TV applications. If you find the interface works with some cards but not others it may be because the interface is not compliant with the ISO specification in some regard that the card is fussy about (eg the card wants 1.8V or 3V, or the I/O impedance is wrong, or the card doesn't like the interface's power up timing sequence). But the most common cause is the card is communicating at a different bit rate to what your PC/software is configured for.
    Last edited by gw1; 23-03-09 at 05:31 PM.

  • #71
    Junior Member
    Join Date
    Mar 2009
    Posts
    11
    Thanks
    0
    Thanked 0 Times in 0 Posts
    Rep Power
    0
    Reputation
    10

    Default

    Hi GW1

    I can't download the schematic, the page gives ''The file link that you requested is not valid. ''

    can u send it to me to my mail if possible? :$

    fox2006eng@googlemail.com

    thanks in advance

    Best Regards

    Fox2006Eng

  • #72
    Junior Member
    Join Date
    Feb 2009
    Posts
    55
    Thanks
    0
    Thanked 0 Times in 0 Posts
    Rep Power
    186
    Reputation
    12

    Default

    Enable javascript fox, the link is working just fine.

  • #73
    Junior Member
    Join Date
    Mar 2009
    Age
    47
    Posts
    46
    Thanks
    3
    Thanked 4 Times in 4 Posts
    Rep Power
    185
    Reputation
    24

    Default

    Thank you very much gw1 & msay.
    you rocks

    The trick "Removing R10" was great, now my viaccess cards are working too.
    The response time is 90ms~100ms the same as Duolabs Cas2 , and faster than Infinity usb

    Tomorrow i'll go for 2N7000.

  • #74
    Senior Member gw1's Avatar
    Join Date
    Jan 2008
    Location
    Hobart
    Posts
    957
    Thanks
    49
    Thanked 608 Times in 213 Posts
    Rep Power
    267
    Reputation
    1901

    Default

    A note about response times. Nearly all of the response time you're seeing is the time taken by the smartcard itself to process the commands. That's determined mainly by the type of command and to a lesser extent the card clock frequency. Also significant to a small degree is the baud rate, as some lengthier commands take time to pass through the serial port - 9600 baud was fast in its day but it's extremely slow by modern standards.

    Another factor is latency - the time between when an event happens in hardware and when your application knows about it. True phoenix interfaces, like this one, that use simple logic and no microprocessor buffers will always have better latency than the expensive programmable interface products. USB has inherent latency, small and variable but insignificant for most applications (except for timing analysis applications of course). Latency may be less noticable on Duolabs because they use High Speed USB rather than Full Speed. But High Speed really only matters when programming multi-megabyte devices, not so much when transferring a few dozen bytes as is the case for Phoenix. With Phoenix the most significant contributor to latancy is the device driver.

    For systems like this the hardware delay is insignificant compared to the time taken for enough characters to accumulate before the UART generates an interrupt, the driver to collects the new data, merges it with other data sitting in the buffer, then sends a notification to the application's message queue etc. Windows is nowhere near a real time operating system. Some companies that choose Windows for embedded systems usually do so because they're familiar with Windows programming, because of Windows' universal hardware support, and because many applications such as kiosks don't need predictable event processing or high reliability. But mostly IMO those companies choose Windows because they don't know any better. In the world of engineering where reliability matters almost nobody uses embedded Windows because predictability and reliability (not to mention licensing cost) are usually far more important than high level library availability and development time. Applications that need beefy library support usually prefer Linux; those that need true realtime use QNX, VxWorks or similar; those which are simpler or which are extremely time or resource critical tend to use no operating system at all (which I imagine accounts for 80% or more of embedded systems). Realtime programming in a busy multitask environment is quite tricky, which is why PVRs for example are so buggy. Controlling and predicting latency is only easy if you can keep things very simple.

    With virtually all smartcard applications what you're interested in is the software that's driving it. The hardware is irrelevant provided it supports your type of card (if you're programming a blank device rather than just sending ISO7816 commands) and has a driver that suits your software (if not a standard serial port). Response times are relevant for things like card sharing but they're almost entirely due to card processing speed, which is why people overclock their cards. Latency is almost never an issue, unless you're doing very specific timing analysis (in which case you're probably using custom hardware that also does power analysis).

  • #75
    Junior Member
    Join Date
    Mar 2009
    Age
    47
    Posts
    46
    Thanks
    3
    Thanked 4 Times in 4 Posts
    Rep Power
    185
    Reputation
    24

    Default

    After so much time i put on this , i still have problem.
    I have 8 4053 IC & 2 Max232 ( Max232N and Max232CPE ).
    ( 2 different brands)
    And 2 different wall adapter.

    Only 2 out of 8 4053 and 1 of the Max232 are working with this design.
    But the others work too BUT : i connect the power, the LEDs are not ok.
    disconnect the power, remove IC 5 , connect the power , disconnect the power put back IC5 in place , connect the power , and everything is working.
    If i disconnect the power and connect it again , in about 30 seconds , everything is working, but if i disconnect the power for about 2 minutes and connect it again, then it's not working unless i do the trick of removing IC5 !!

    Any suggestion please ?

    Also Max232N is the working one with 2 of 4053 .
    Max232CPE is not working with that 2 special 4053.

    I found only 4053BE, and 4053 SMD type. Do you think SMD is better ?
    Also i cann't find 2N700 , i only found BS170 , i think they are the same , only different pinouts !?

  • #76
    Junior Member
    Join Date
    Mar 2009
    Age
    47
    Posts
    46
    Thanks
    3
    Thanked 4 Times in 4 Posts
    Rep Power
    185
    Reputation
    24

    Default

    Quote Originally Posted by gw1 View Post
    I suggest you instead replace T3 with a FET. You'll need to insert it backwards so that its source is on GND, its gate is on R10 and its drain is on R11. I haven't tried it myself but it should work. (Circuits typically have additional pulldowns to ground on the FET gate but I don't think it's necessary here.)
    I used BS170 , Source on GND, Gate on R10, Drain on R11.
    It's not working.Also Data LED is on all the time.

  • #77
    Junior Member
    Join Date
    Feb 2009
    Posts
    55
    Thanks
    0
    Thanked 0 Times in 0 Posts
    Rep Power
    186
    Reputation
    12

    Default

    I was able to reproduce your problem bobzombie, using a power supply without the bridge diode rectification at the PSU, this means that you are feeding the circuit with AC instead of DC.

    Using 12VAC at my board i got the exact same problem with IC5, the leds getting crazy, flashing and such. I recommend that you check your PSU, the circuit was not developed to handle AC:


    D5 is not there to act as a half-wave rectification, but to prevent some one to invert the polarity and damage the on-board circuits.

    Use a well rectified and filtered 12VDC (or more) adapter.

    PS: My both IC4/IC5 (4053BP) got damaged after this test.

  • #78
    Junior Member
    Join Date
    Mar 2009
    Age
    47
    Posts
    46
    Thanks
    3
    Thanked 4 Times in 4 Posts
    Rep Power
    185
    Reputation
    24

    Default

    i use two different DC wall adapter, not AC.
    Both are DC and have capacitor and rectifier.
    My LEDs are not flashing.

  • #79
    Senior Member gw1's Avatar
    Join Date
    Jan 2008
    Location
    Hobart
    Posts
    957
    Thanks
    49
    Thanked 608 Times in 213 Posts
    Rep Power
    267
    Reputation
    1901

    Default

    msay, this track on your V5 PCB appears to be incorrect. It links IC4 pins 12 & 14 which isn't right.



    I don't understand how an error like this can slip through.
    Doesn't Eagle verify the PCB against the schematic using a netlist?

  • #80
    Senior Member gw1's Avatar
    Join Date
    Jan 2008
    Location
    Hobart
    Posts
    957
    Thanks
    49
    Thanked 608 Times in 213 Posts
    Rep Power
    267
    Reputation
    1901

    Default

    Quote Originally Posted by bobzombie View Post
    I used BS170 ... it's not working ... Data LED is on all the time.
    Let's not waste any further time on a FET approach. Instead I suggest we utilise a spare IC2 gate.
    • Remove T3 completely
    • Desolder and lift the +5V leg of R11 (ie the one that's connected to R18)
    • Cut the track between IC2 pin 9 and GND
    • Connect IC4 pin 14 to IC2 pin 9 using hookup wire
    • Connect IC2 pin 8 to the floating leg of R11 using hookup wire


    Also, because you've had problems, I recommend:
    • Revert R5,R15,R16 back to 470R
    • Change R8 to 10K (if you haven't done so already)
    • Ensure COMMA link is present


    And cut the track I've highlighted in red above. There may be other errors on the PCB, I've always assumed it was correct and not looked at it closely. When I get time I'll give it some study.

    Regarding the mode not behaving consistently, please make the following temporary modification for test purposes.
    • Remove R21
    • Solder a wire link across C & E pads of T2
    • Replace R22 by 2K2 (or 1K8, 2K7 or 3K3)
    • Power up and measure the voltage across D6 for us.

    This bypasses the mode selection circuit and forces it to Phoenix mode. I've also dimmed the mode LEDs in case excess 4053 current is part of the problem. Keep it this way until we get it working properly with Phoenix mode at least.

  • Page 4 of 6 FirstFirst 123456 LastLast

    Bookmarks

    Posting Permissions

    • You may not post new threads
    • You may not post replies
    • You may not post attachments
    • You may not edit your posts
    •