P2 Evaluation Board - Accessory Set 64006-ES

124»

Comments

  • msrobotsmsrobots Posts: 2,655
    edited 2019-06-27 - 03:48:43
    oh Henrique,

    I looked at the picture and it took a while until I had to smile wide. Yes. Mechanical connect them and have still pins for wires. Very cool idea and possible quite cheap and easy to produce when one has a working production line like Parallax has.

    I also think I need more then one eval board.

    @Ken Gracey you should better run 400 Eval-B boards not 200...

    And yes, @Yanomani I do intend to try 32 bit with my current FullDuplexSerial2 driver currently doing 8,n,1 with smart pins. But I need to fight down the streamer first, it is getting personal. I can transfer with the receiver waiting, but having the transmitter waiting fails timing-wise. for whatever %$%%@^!^** reason I can't find in 277 lines of code.

    Enjoy!

    Mike
    I am just another Code Monkey.
    A determined coder can write COBOL programs in any language. -- Author unknown.
    Press any key to continue, any other key to quit

    The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this post are to be interpreted as described in RFC 2119.
  • Hi Mike

    The four 2 X 6 male pin headers are expected to be used in place of the ones already present a the Evals, since the connection of the Interconn breakout (temptative designation; it will not be me who will define its name, if and when it'll become available as a product) would cover/use them.

    Beyond anableing the interconnection of up to four other Eval boards around a single one, they're also meant to serve as a support for any other breakout one intends to share between two P2.

    There will be also the possibility of "bridging" several P2 Evals, and, if its lenght is designed accordingly, perhaps some kind of "P2 ring" could be devised too.

    At this moment, as they are only innitially sketched, they are some kind of a fresh mass of pottery clay; each of us can use its own creativity, in order to help shape it as useful as it can be. :smile:

    Henrique
  • At the risk of going a little OT...

    Did you ever tried using one of the other streamers as a sniffing logic analyzer?

    Perhaps you can catch at exactly which point the handshaking between the sender and receiver Cogs goes nuts, or start to chew its own tail, in circles.
  • Like the idea, and the drawing too.

    My first reaction was that it might be better as single-header width. You can use 2 if you want, or only 1 for 8-bus access if that would be preferred.

    Then I was thinking about the accessory sharing the same header as the high-speed bus. Maybe not the best thing.

    And that got me back to the IDC cable. Probably pre-made from many distros, specify the length you need, plug and go !


    The PCB "group plate" would look might fierce though !
  • VonSzarvas wrote: »
    Like the idea, and the drawing too.

    My first reaction was that it might be better as single-header width. You can use 2 if you want, or only 1 for 8-bus access if that would be preferred.

    Then I was thinking about the accessory sharing the same header as the high-speed bus. Maybe not the best thing.

    And that got me back to the IDC cable. Probably pre-made from many distros, specify the length you need, plug and go !


    The PCB "group plate" would look might fierce though !

    But to make this work right, the connections need to be reversed so that LSBs and MSBs match. You'd have to twist the ribbon cables and cross them over each other. Maybe better to have a controlled-impedance PCB for the job.
  • msrobotsmsrobots Posts: 2,655
    edited 2019-06-27 - 08:40:05
    cgracey wrote: »
    VonSzarvas wrote: »
    Like the idea, and the drawing too.

    My first reaction was that it might be better as single-header width. You can use 2 if you want, or only 1 for 8-bus access if that would be preferred.

    Then I was thinking about the accessory sharing the same header as the high-speed bus. Maybe not the best thing.

    And that got me back to the IDC cable. Probably pre-made from many distros, specify the length you need, plug and go !


    The PCB "group plate" would look might fierce though !

    But to make this work right, the connections need to be reversed so that LSBs and MSBs match. You'd have to twist the ribbon cables and cross them over each other. Maybe better to have a controlled-impedance PCB for the job.

    I thought that @Yanomani just wanted the mechanic connection and let the user put wires between the headers where needed, not sure if it would be good to connect VSS/VIN.

    But yes, single wide, cross over and some buffered LEDs on it to show action would be nice to have.

    a optional 90 degree header version so we can use 3 dimensions and build P2 cube cluster...

    Mike
    I am just another Code Monkey.
    A determined coder can write COBOL programs in any language. -- Author unknown.
    Press any key to continue, any other key to quit

    The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this post are to be interpreted as described in RFC 2119.
  • msrobots wrote: »

    I thought that @Yanomani just wanted the mechanic connection and let the user put wires between the headers where needed, not sure if it would be good to connect VSS/VIN.

    But yes, single wide, cross over and some buffered LEDs on it to show action would be nice to have.

    a optional 90 degree header version so we can use 3 dimensions and build P2 cube cluster...

    Mike

    Agree on not connecting +ve's.

    Buffered LED's would be nice.

    90 degree could be installed- on all the accessories we've used a dual SMT/TH footprint, ready for all those possibilities.

    with all those points, and now realising I didn't consider cross-over before, the PCB version is starting to make more sense. Not sure how much demand it would have though.

    Would it be sensible to sell it without the SMT header installed, so that "cube guys" can make full use of the board too ?

    In fact, if we forget the buffered LEDs, then it becomes a PCB+header kit product. No manufacturing would reduce the cost somewhat, and the end-user can choose to fit straight or R/A headers.

    A kit might make more sense, as I suspect this could have a limited market, with diverse requirements?




  • msrobotsmsrobots Posts: 2,655
    edited 2019-06-27 - 09:26:27
    VonSzarvas wrote: »
    msrobots wrote: »

    I thought that @Yanomani just wanted the mechanic connection and let the user put wires between the headers where needed, not sure if it would be good to connect VSS/VIN.

    But yes, single wide, cross over and some buffered LEDs on it to show action would be nice to have.

    a optional 90 degree header version so we can use 3 dimensions and build P2 cube cluster...

    Mike

    Agree on not connecting +ve's.

    Buffered LED's would be nice.

    90 degree could be installed- on all the accessories we've used a dual SMT/TH footprint, ready for all those possibilities.

    with all those points, and now realising I didn't consider cross-over before, the PCB version is starting to make more sense. Not sure how much demand it would have though.

    Would it be sensible to sell it without the SMT header installed, so that "cube guys" can make full use of the board too ?

    In fact, if we forget the buffered LEDs, then it becomes a PCB+header kit product. No manufacturing would reduce the cost somewhat, and the end-user can choose to fit straight or R/A headers.

    A kit might make more sense, as I suspect this could have a limited market, with diverse requirements?




    Well a Kit with pcb and header is not really a product.

    We need PR and BLING here. A ready made Plug-In able thing to connect P2's to each other and have them stable on your desk. Nice Parallax Logo and - hmm - we need a name - sure - Parallax Hyperlink.

    It is one run of whatever number thru Parallax Production line. Simple PCB, header LEDs, should not cost more as the existing LED board.

    But one could plug it in, can see the status and can connect whatever one wants on the other end. Does not need to be another P2, one has the second header there and can plug in wires.

    But one could see the action on the pin. So it has to be buffered and as transparent as possible to not disturb whatever happens to the pin.

    I could imagine that they sell quite fast, each eval board has 8 Ports times eval boards, maybe run 500?

    the topping would be to use the LED board layout and show the analog voltage level of each pin, just dreaming.

    You asked for...

    Enjoy!

    Mike
    I am just another Code Monkey.
    A determined coder can write COBOL programs in any language. -- Author unknown.
    Press any key to continue, any other key to quit

    The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this post are to be interpreted as described in RFC 2119.
  • VonSzarvasVonSzarvas Posts: 1,543
    edited 2019-06-27 - 09:28:17
    Agreed on the product points.

    The reason for me testing the point, was more about the cube and other uses that may be desirable.

    I suppose instead of hacking off the SMT header, and fitting the RA option direct to the PCB, users could just as easily create a little R/A adapter that plugs between the P2 headers and HyperLink, to create a cube without any destruction and complication !


    Edit: Probably an "off the shelf" R/A TH socket would do. They must exist :)

  • yeah, even better just provide 90 degree pin header so one need 3 hyperlinks and two header to get to the next dimension!

    And you can run 1000 instead 500, we have even more BLING and cheap ones like me just plug in wires to go up and down.

    So do not care about 3D right now, just the simple thing, header both sides, cross connected, GND yes, VE no, and LED board 90 degree rotated that you have 8 lines of 7 LEDs showing the analog level of the pins. Or just 8 LEDs showing high/low.

    The main thing is that it not as less as possible disturbs the original pin output/input/drive level.

    And it is small, simple, cheap to produce, almost completely senseless but a must to have at least 6 of them for every eval user, because it will BLINK!

    Sorry, I am just more than happy right now because I finally go the streamer to what I asked for since days. This is just fun, this P2.

    Enjoy!

    Mike
    I am just another Code Monkey.
    A determined coder can write COBOL programs in any language. -- Author unknown.
    Press any key to continue, any other key to quit

    The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this post are to be interpreted as described in RFC 2119.
  • jmgjmg Posts: 13,609
    Yanomani wrote: »
    Also, to the extent of my knowledge, there are many ways to interconnect two or more P2, but they seem to be yet unknown. The SERDES are also mostly unnexplored; I've yet to heard something about synchronous comms.
    See the threads on triggered streamer connections.
    Yanomani wrote: »
    Well, perhaps some new means could be devised, to easy the interconnection path, while not loosing nothing that have been conquered so far.

    I hope the following image could explain what I've been planning, better that a lot of blah blah blah blah...
    Do you have dimensions on that ?

    The idea of being able to 'clip together' P2 Eval's is good, but the leads need to be kept very short.
    That suggests a compact staple-like U PCB, that links (almost) touching P2Evals.

    The next issues that show up, are the connectors are 45' in the corners, which rather conspire against compact adjacent-hops :(




  • jmg wrote: »
    See the threads on triggered streamer connections.

    It was my fault (the way I write in english almost never matches what I mean to say in english)...

    The intention of the text line written just before the one you'd commented about, was refering (whithout specifically mentioning them) to evanh's and msrobots efforts, whose evolution I had closelly followed, with great interest.
    Do you have dimensions on that ?

    It was draw proportionally to the female/male pin headers:

    about 4" long, 2.5" wide, IIRC
  • jmg wrote: »
    ...The next issues that show up, are the connectors are 45' in the corners, which rather conspire against compact adjacent-hops :(

    Not completely right. If 4 P2s are connected so that the corner with the reset button is in the center, the rest might fit well...

    Mike
    I am just another Code Monkey.
    A determined coder can write COBOL programs in any language. -- Author unknown.
    Press any key to continue, any other key to quit

    The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this post are to be interpreted as described in RFC 2119.
  • msrobots wrote: »
    jmg wrote: »
    ...The next issues that show up, are the connectors are 45' in the corners, which rather conspire against compact adjacent-hops :(

    Not completely right. If 4 P2s are connected so that the corner with the reset button is in the center, the rest might fit well...

    Mike

    We should have made a way to common P63's and RESn's, so that we could download to a whole group at once.
  • cgracey wrote: »
    We should have made a way to common P63's and RESn's, so that we could download to a whole group at once.

    Were you meaning a mass-production-assisting tool, as a general means of providing a delivery path of identical flash memory contents, to many Eval boards, at the same time?
  • Yanomani wrote: »
    cgracey wrote: »
    We should have made a way to common P63's and RESn's, so that we could download to a whole group at once.

    Were you meaning a mass-production-assisting tool, as a general means of providing a delivery path of identical flash memory contents, to many Eval boards, at the same time?

    That would be good, too. I was just thinking that because the download protocol supports multiple P2 chips, we could have had a way to talk to many P2's on eval boards at one time.

  • Wouldn't be enough to have some kind of micro test-clip, or single soldered wire, to reach P2_RESn, just at one of the legs of S100 tactile Reset switch?

    Perhaps two wires, twisted, with one being GND, and two solder joits to be made.

    P2_IO63 is readily available at J210, thus a very tinny breakout board could be enough, to bring all needed signals.

  • If you could have access to P2_RESn, you could also substitute the SPI Flash chip by any of those 8-pin QPI-enabled PSRAMs.

    Multiple boot-ready incarnations of software could be them distributed, before unleashing all them, to execute their individual tasks.
  • VonSzarvasVonSzarvas Posts: 1,543
    edited 2019-06-29 - 03:21:47
    RevB Eval has P2 RESn available on a header, top left, beside P63-56 header.

    An accessory with an extra wide socket would plug in there, or use a single jumper wire, etc..
  • YanomaniYanomani Posts: 850
    edited 2019-06-29 - 05:35:42
    Fantastic VonSzarvas!

    Due to poor eyesight, I'm still unable to find that header, both at the published 64000-ES_RevB_P2 EVAL Board Schematic_0617.pdf and also at 64000-ES_RevB_Top_3D.png.

    Are the files I have here, some outdated versions of the docs?

    P.S. Correction... I did found it, unpopulated, near another one, providing access to DTR.
  • VonSzarvasVonSzarvas Posts: 1,543
    edited 2019-06-29 - 05:49:36
    That's it.

    The DTR at that header is wired so you could cut the trace between the header pads to disconnect DTR from RESn.

    Optionally, could fit a header and shunt to quickly control the connection.


    ps. Next time I'm in the office, will check those docs/images are the final ones.
  • What about making P31's common? it would mean making all P31's undriven, except for the one which is connected to a computer via USB.
  • Yanomani wrote: »

    Are the files I have here, some outdated versions of the docs?

    I think the 3D image you found was the latest: http://forums.parallax.com/discussion/comment/1471956/#Comment_1471956

    Attached is the schematic in pdf format. Not sure if that was shared anywhere yet. These preview files will be cleaned up and published officially when the Eval starts shipping. Gosh, I can't wait !
Sign In or Register to comment.