Shop OBEX P1 Docs P2 Docs Learn Events
P2X8C4M64P Approved for Production - Page 3 — Parallax Forums

P2X8C4M64P Approved for Production

13»

Comments

  • jmgjmg Posts: 15,140
    kwinn wrote: »
    Ken Gracey wrote: »
    potatohead wrote: »
    Me too. That may end up being a pretty hot item. I think of it like the DIP-40 version of P1. I have a couple of those, and they are still my favorite. Drop onto breadboard wire up a basic circuit and go!

    IMHO, something like that is needed to maximize P2 potential adoption.

    Market feedback from the potatohead. Is it important to you to "drop a P2 module into a breadboard" even if it becomes inordinately long due to the number of I/Os? This is the first question to be answered.

    How would you feel about featureless, low-cost "carrier" boards that can hold the module and have the USB circuitry on them and are cheap enough to be allocated to projects and messing around with soldering irons?

    Ken

    There were a few 64 pin dips before the switch to smd packaging so I think a 64 pin dip carrier board is a good idea if you take advantage of the extra width for additional inputs at the ends of the board.

    We looked briefly at those, and DIP 64 sockets are very much dinosaurs - they are very large, expensive, and cannot bring out all P2 pins, whilst the dual 40p header is more compact and can access all IO.
    There is a shrink DIP64, (1.778mm?) but that fails to connect to breadboards, which cancels the whole point...
  • cgracey wrote: »
    I have been thinking that for a low-pin-count module, we could double up the port A and port B pins, tying them together. That would allow us to have scope input for every single pin, making it kind of a handy experimenter module with reduced pins.
    I don't understand how this would be useful. You mean pin 0 and pin 32 would be tied together and connected to a single pin on the module?

  • cgraceycgracey Posts: 14,133
    David Betz wrote: »
    cgracey wrote: »
    I have been thinking that for a low-pin-count module, we could double up the port A and port B pins, tying them together. That would allow us to have scope input for every single pin, making it kind of a handy experimenter module with reduced pins.
    I don't understand how this would be useful. You mean pin 0 and pin 32 would be tied together and connected to a single pin on the module?

    Yes. We'd probably want to put a resistor in between the two, though.

    If you can't get that many pins out of the module, anyway, you might as well put them to some such use.
  • cgracey wrote: »
    David Betz wrote: »
    cgracey wrote: »
    I have been thinking that for a low-pin-count module, we could double up the port A and port B pins, tying them together. That would allow us to have scope input for every single pin, making it kind of a handy experimenter module with reduced pins.
    I don't understand how this would be useful. You mean pin 0 and pin 32 would be tied together and connected to a single pin on the module?

    Yes. We'd probably want to put a resistor in between the two, though.

    If you can't get that many pins out of the module, anyway, you might as well put them to some such use.
    But what use would that be? Communications between COGs?

  • Peter JakackiPeter Jakacki Posts: 10,193
    edited 2019-11-24 04:02
    This is indeed a major 13 year milestone, congratulations "Parallax" :)

    I've been very very busy getting these P2D2s developed and assembling the current batch and getting a pcb revision through, but I will talk about that more on the P2D2 threads.

    Initially the P2D2 was developed as a low-cost "breakout" module with a single switchmode regulator for 1.8V and linear regs for 3.3V. Of course it had an option for microSD but no microUSB with just a simple Prop Plug compatible header etc. Because there were 64 I/O all begging and screaming to be let out, even those which already had internal boot use etc, I decided upon the simple 40-pin dual-row header on each side to not only keep it compact but also to make the pins usable by keeping tracks short and noise free.

    But that was not good enough according to practically everybody, it had to have umpteen amp regulators and USB etc. So now I have P2D2r3 and r4 coming.

    What I would really like to do is not assemble these at all because it really needs Parallax behind it and I do what I do, just develop and prototype. I tried to do this with the compact WIZnet W5500 module years ago in getting WIZnet to manufacture them since they had some clumsy oversized module that wasn't suitable for integration and I need something similar to their earlier generation W5200 module. Well I went to very great expense to design and manufacture these modules but what ended up happening was that my design files were effectively stolen by their German rep who promised all kinds of stuff until the phone went silent. Now WIZnet have a compact W5500 module that I really wanted in the first place anyway. Surely now the P2D2 module, optioned up, or not, is the module that Parallax and customers want, what I really wanted in the first place, but I had to do it myself, at very great expense.

    Because Parallax didn't have anything at all, I've been giving myself lots of grief in doing up a datasheet that seems to be forgotten, in doing up reference schematics, and reference pcb (the P2D2), and of course in all the bits and pieces associated with it including TAQOZ of course which never really gets mentioned. What am I knocking myself out for?
  • cgraceycgracey Posts: 14,133
    edited 2019-11-24 04:04
    David Betz wrote: »
    cgracey wrote: »
    David Betz wrote: »
    cgracey wrote: »
    I have been thinking that for a low-pin-count module, we could double up the port A and port B pins, tying them together. That would allow us to have scope input for every single pin, making it kind of a handy experimenter module with reduced pins.
    I don't understand how this would be useful. You mean pin 0 and pin 32 would be tied together and connected to a single pin on the module?

    Yes. We'd probably want to put a resistor in between the two, though.

    If you can't get that many pins out of the module, anyway, you might as well put them to some such use.
    But what use would that be? Communications between COGs?

    Those phantom pins could serve as background scope inputs for each of the hard-connected pins. That way, no matter what else was going on, you could always observe those hard-connected pins as if from an oscilloscope, all from within the chip.
  • cgracey wrote: »
    David Betz wrote: »
    cgracey wrote: »
    David Betz wrote: »
    cgracey wrote: »
    I have been thinking that for a low-pin-count module, we could double up the port A and port B pins, tying them together. That would allow us to have scope input for every single pin, making it kind of a handy experimenter module with reduced pins.
    I don't understand how this would be useful. You mean pin 0 and pin 32 would be tied together and connected to a single pin on the module?

    Yes. We'd probably want to put a resistor in between the two, though.

    If you can't get that many pins out of the module, anyway, you might as well put them to some such use.
    But what use would that be? Communications between COGs?

    Those phantom pins could serve as background scope inputs for each of the hard connected pins. That way, no matter what else was going on, you could observe those pins as if from an oscilloscope, all from within the chip.
    I see. Yes, that would be useful. Thanks for explaining.

  • Because Parallax didn't have anything at all, I've been giving myself lots of grief in doing up a datasheet that seems to be forgotten, in doing up reference schematics, and reference pcb (the P2D2), and of course in all the bits and pieces associated with it including TAQOZ of course which never really gets mentioned. What am I knocking myself out for?

    Peter, we're going to give P2D2 a really good look when we get closer to considering a module. We'll need to define our goals, hear from customers, and get this totally right. We've also got requirements of our own to consider relating to cost, manufacturability, etc. Educational customers will have some special requirements that we've not even heard yet.

    I'd rather not make a module at all and simply see our revenue come from P2 Eval Boards and chips, but I know this won't sustain us in the first year. We will need a module to generate some income.

    P2D2 is the go-to small P2 solution at this time. I mentioned it in a recent newsletter to 28K people and will likely feature the frozen P2D2 at some stage, too.

    Ken Gracey

  • potatoheadpotatohead Posts: 10,253
    edited 2019-11-24 05:27
    The P2D2 may be just the thing Peter. I had to hold off getting one until after end of year. Xmas, and an almost 4 year old I am raising for the near term for now, got in the way. It is on my short list!

    Ken, this was going to be my basic answer.

    For some people, couple that with a bigger system like the PPDB, and it would definitely hit a sweet spot, IMHO.

  • NumPyNumPy Posts: 27
    edited 2019-11-24 05:44
    This is extremely exciting. Congratulations. Just received the P2-ES Eval Board in the mail!
    I had a blinker running in 10 minutes with PNUT v33k and some PASM code.
    Can't wait to dig in, trying to find the best IDE setup now.
    I'm not a programmer/developer so I feel kinda guilty. I've been watching this since the beginning.
    So many people worked so hard on this.

    I noticed that the new Google doc said that 1,2,4,8, and 16 cogs would be available.
    Is there any published info on the possibility of 16 cogs and will any more I/Os be available.
    (just curious not greedy :-) )
    I'm assuming that the new silicon is the P2X8C4M64P 8 cog unit.

    The P1 is pretty awesome, I still have 2 being used and 3 boards in the box.
    I remember an industrial company built a 4 propeller drone,
    it will be amazing to see what P2 can do.
  • cgraceycgracey Posts: 14,133
    NumPy wrote: »
    This is extremely exciting. Congratulations. Just received the P2-ES Eval Board in the mail!
    I had a blinker running in 10 minutes with PNUT v33k and some PASM code.
    Can't wait to dig in, trying to find the best IDE setup now.
    I'm not a programmer/developer so I feel kinda guilty. I've been watching this since the beginning.
    So many people worked so hard on this.

    I noticed that the new Google doc said that 1,2,4,8, and 16 cogs would be available.
    Is there any published info on the possibility of 16 cogs and will any more I/Os be available.
    (just curious not greedy :-) )
    I'm assuming that the new silicon is the P2X8C4M64P 8 cog unit.

    The P1 is pretty awesome, I still have 2 being used and 3 boards in the box.
    I remember an industrial company built a 4 propeller drone,
    it will be amazing to see what P2 can do.

    To build a 16-cog version of the chip, we are going to need a smaller process than the 180nm process we are currently using. However, we can make many reduced versions of the chip, where we have 4 or 2 cogs and fewer IO pins, with smaller hub memory. In fact, I think we have five different reduced versions tentatively planned, already.
  • potatoheadpotatohead Posts: 10,253
    edited 2019-11-24 06:30
    I'm not a programmer/developer so I feel kinda guilty.

    Oh don't be. It was the engineers soap opera for a decade! LOL Good times.

    If I were you, get FastGUI. Eric has done a nice job supporting PASM, a look at what SPIN 2 can do, BASIC, and C. It works right now, and runs pretty solid. The whole thing is spartan, but I see that as a good thing. The stuff you need is there. Some stuff you didn't expect, is there too.

    When the choices are more limited, finding the right ones is easier. That's my take anyway.

    Lots of us really believed, and now we are here with a killer chip. Have some fun. That was a design goal. No joke.

    Yup. You've got the 8 COG model, 512kb RAM.
  • RE: simple breakout module

    How about breaking out all 64 pins in two 2*20 headers, such that when you plug it into a breadboard, Port A is connected to Port B, as Chip says, but when you plug it into a custom PCB, all pins are available.

    Or another option would be to make it smaller by tying up some pins with onboard hyperram or smth (I hear that needs impedance matching for high speeds)
  • jef_vtjef_vt Posts: 23
    edited 2019-11-24 14:54
    Indeed congratulations to parallax!
    I LOVE the P1. It just fits and is a dream to program. I have limited experience with P2, but I can't wait.
    I have designed a few products with P1.
    Total qty is now over 500pcs.
    I am now designing a new product with P2.
    Estimated qty needed is now at 170-200 pcs and will rise as time goes on.
    Not big numbers, but it can add up.
    I am just hoping the 22 weeks is over-estimated and will be shorter.
  • jef_vtjef_vt Posts: 23
    edited 2019-11-24 15:00
    cgracey wrote: »
    However, we can make many reduced versions of the chip, where we have 4 or 2 cogs and fewer IO pins, with smaller hub memory. In fact, I think we have five different reduced versions tentatively planned, already.

    This would be just perfect. A smaller chip like 20 pin TSSOP with 16 I/O and 4 cogs would be fantastic. Please put the 16K EEPROM IN the chip.
    Or a 8 pin SOIC with 2 cogs. perfect for smaller sensor-projects.
    for most of my products the 44 pin is a bit too big. especially with the external EEPROM.
  • Ken Gracey wrote: »
    JRoark wrote: »
    @"Ken Gracey" Any teasers as to how the above-mentioned “Parallax P2 Module” will be different than the 64000-ES?

    Here's a mix of goals and features, albeit briefly presented because I have to get back to webinar planning and preparing a promotional e-mail for Tuesday morning.

    - Moves layout-critical placement to a small footprint, alleviating challenges for designers
    - Designed for small-scale production purposes, fitting in a socket or into a header
    - Includes only the core parts needed for the P2 to operate (if the extra RAM is warranted, maybe we make a second module) - likely excluding USB parts
    - Accompanying carrier PCB for quick use and demo (maybe this includes USB stuff)
    - Includes reasonable margin for Parallax, low quantity cost for customers ($50 alone, maybe $20 in 100s?)
    - Marketing theme that makes people want to have it (we're all subject to some marketing lure, aren't we?), either through naming or specific support (i.e., the P2 MicroPython module or P2 Spin Module)
    - Design considerations factor in our educational program and possible use

    But it doesn't matter what I think at this point, as these ideas are the only ones I've envisioned. What matters is that our customers [and the many new ones we've yet to meet] want to buy what we make. This means we have some research to properly form this product.

    I also don't want it to trample on P2D2 with a competitive product but support P2D2 with visibility. It's scheduled to be featured in a future newsletter, for example. It has been our desire that others start making tools and PCBs, but the Parallax business success will necessitate one or two long-lasting Parallax designs. We must be able to support hobby/education/low-volume commercial while customers begin to plan their products. Don't expect a series of short-lived P2 PCB designs from us this time; we hope to get it right with each product (i.e., the FLiP came out many years after we introduced the P1).

    I welcome a thread on this subject, too. It would help us learn about what specific needs exist and how we might shape this product concept.

    Ken Gracey

    Congratulations to all the team for the success!!!

    50$ will be 50 Euros here and, sorry to say it once again, this is too expensive for a "naked" module! For such price, this products can only be used for very special projects that really need the special unique features. Like the project, that uses 12 serial connections, that was mentioned by LocalRoger.
    I understand, that Parallax cannot handle big volumes with too low prices, so I want to suggest to make a more "rich" module:
    - Power regulators on board
    - USB- Programming interface on board
    - SD card on board
    - reset + 4 switches on board (esc, OK, +, -)
    - a small display on board
    - at least one LED connected to a pin
    - WLAN on board (This is supposed to be the age of IoT, isn't it?)
    - perhaps additional ram on board?
    - perhaps some connectors for servo motors?
    With these features, the number of pins to be brought to the dil connectors will be reduced too....

    To have a small display and a few buttons will be a big advantage for a lot of projects.
    .....Looking forward for attractive boards with documentation, compiler and drivers for an attractive price.....
    Christof
  • VonSzarvasVonSzarvas Posts: 3,273
    edited 2019-11-24 16:03
    @"Christof Eb."

    Good points Christof.

    About the display... Do you think having that externally might be still useful for the user? The benefit should be better upgrade-ability and longevity of the product. But that doesn't consider usability.

    Let's assume customers could buy a kit that included the main board and all the features you mentioned provided either on-board OR as plug-on modules in the "starter pack", then would that be as appealing as one big PCB with everything included? What do you think?



    Edit: I forgot to ask... what type/size of display are you thinking ?

  • Congratulations to all the team for the success!!!

    50$ will be 50 Euros here and, sorry to say it once again, this is too expensive for a "naked" module! For such price, this products can only be used for very special projects that really need the special unique features. Like the project, that uses 12 serial connections, that was mentioned by LocalRoger.
    I understand, that Parallax cannot handle big volumes with too low prices, so I want to suggest to make a more "rich" module:
    - Power regulators on board
    - USB- Programming interface on board
    - SD card on board
    - reset + 4 switches on board (esc, OK, +, -)
    - a small display on board
    - at least one LED connected to a pin
    - WLAN on board (This is supposed to be the age of IoT, isn't it?)
    - perhaps additional ram on board?
    - perhaps some connectors for servo motors?
    With these features, the number of pins to be brought to the dil connectors will be reduced too....

    To have a small display and a few buttons will be a big advantage for a lot of projects.
    .....Looking forward for attractive boards with documentation, compiler and drivers for an attractive price.....
    Christof

    What you have described is almost a WX Badge, which hits the $50.00 mark:
    https://www.parallax.com/product/20300

  • Yes indeed.

    A little known feature of the BadgeWX is that it was designed so it can behave as the frontend to user projects.

    If you pull off the battery pack (gently; its held on with two sticky pads) you will find all the expansion and power through-hole pads accessible to standard 0.1" pins.

    Flying cables or an appropriately designed carrier pcb would plug right in. A super easy way to add buttons, RGBs, wifi and display to your own project.

    If anyone wanted to do this I'd be happy to provide a Diptrace and gerber file with pins in all the right places, ready to plug into the BadgeWX!

    (Oops. Well OT! Please start a new thread if anyone has questions about this!!)
  • Wow! ...

    Sorry, but I'm kinda lost for words at the significance of this announcement, and the total transparency of Ken's note on Parallax as a company being "all in" with the P2.

    I guess for the rest of us (who have been unable to assist in more technical ways), now is the time, more than ever, to ramp up the "word of mouth" contribution to further spread the P2 capability and launch awareness as wide as possible amongst the maker and industry communities.

    Here's to P2 uptake quickly achieving the necessary tipping-point to establish itself as a solid Parallax income earner for future years! :smile:
  • I think so too, and am moving into, "Have you heard about the Parallax P2 Chip?" mode. Will do that unabashedly too.

  • jmgjmg Posts: 15,140
    VonSzarvas wrote: »
    @"Christof Eb."

    Good points Christof.

    About the display... Do you think having that externally might be still useful for the user? The benefit should be better upgrade-ability and longevity of the product. But that doesn't consider usability.
    ...
    Edit: I forgot to ask... what type/size of display are you thinking ?
    This 'What display' angle, is one of the driving reasons I moved to Pi pin-out on the dual 2 x 20 headers.
    No need to include a display, you just select what fits your project best and plug it in.


  • jmg wrote: »
    VonSzarvas wrote: »
    @"Christof Eb."

    Good points Christof.

    About the display... Do you think having that externally might be still useful for the user? The benefit should be better upgrade-ability and longevity of the product. But that doesn't consider usability.
    ...
    Edit: I forgot to ask... what type/size of display are you thinking ?
    This 'What display' angle, is one of the driving reasons I moved to Pi pin-out on the dual 2 x 20 headers.
    No need to include a display, you just select what fits your project best and plug it in.


    But until you actually get pcbs made and have a board assembled and tested, you don't "have a board". My latest P2D2 keeps all the I/O intact and in order but still includes an RPi header in addition to the standard I/O. Best of both worlds and you don't cripple your P2 trying to "put on a Pi Hat" :) I will have to send you one next week so you can test the UB3 chip out a bit more.

    Re the badge - That price on that badge looks more BlocklyProp promotional than profitable.
  • jmgjmg Posts: 15,140
    edited 2019-11-25 00:27
    ... Best of both worlds and you don't cripple your P2 trying to "put on a Pi Hat" :) ...
    There is no crippling involved at all, because all 32 P2 io are available per 40pin header, to anyone who wants them :) - and even the 0.050" strips are intact too,
    It is only a connector, with a minor pin shuffle, the important detail for most users is "What can you connect it to ?"

    Addit: I'll drop a P2D2Pi FLiP universal pin-out into the other thread, as a picture is worth 1000 words.


  • Peter JakackiPeter Jakacki Posts: 10,193
    edited 2019-11-24 23:43
    RPi pinout has 8 grounds and 4 power pins and that leaves 28 pins including the I2C pins.
    This is the pinout I am using for this extra header (but I think we would need a new thread perhaps if we say anymore)
    btw - there is a little mistake on that schematic
    1201 x 1572 - 595K
  • You mean all 64, right?
  • VonSzarvas wrote: »
    @"Christof Eb."

    Good points Christof.

    About the display... Do you think having that externally might be still useful for the user? The benefit should be better upgrade-ability and longevity of the product. But that doesn't consider usability.

    Let's assume customers could buy a kit that included the main board and all the features you mentioned provided either on-board OR as plug-on modules in the "starter pack", then would that be as appealing as one big PCB with everything included? What do you think?



    Edit: I forgot to ask... what type/size of display are you thinking ?

    Still thinking of a single low cost, compact, bread board friendly module.
    So that display has to be smaller than the module and cheap for Parallax.
  • Just want to say Congratulations to Parallax and all the community that have worked for the last decade on the P2.

    I've been around since the BASIC Stamp II days and even more so since I first laid my hands on my first P1 back in like May 2006. I was so blown away by the power that Parallax had put into a uController. It was crazy that it could drive a composite display with only 3 resistors. I think I was up programming that first chip like 20 straight hours that first day...lol

    I remember going to the unofficial Expo in 2008 and there was some guy there talking about the P2 and what it was going to be able to do. I didn't know him, but at first I thought it was just same crazy dude. He had his badge turned around. The more and more he talked, I leaned over to my friend and said I thought that guy may be Chip. And it was. He decided to make a surprise visit.

    Over the next several years I managed to make it to all but one of the expos. Some of the best times meeting the Gracey family, the Parallax staff and all the great people that do amazing things with the propeller.

    I've been laying low the last year, but have been following the progress of the P2 closely. It's really turned into an awesome chip and hope to get my hands on the released chip soon.
  • cgraceycgracey Posts: 14,133
    Jim, thanks for being around all this time. P2 will be even more fun to program. What once took three pins for composite now takes only one, and it works way better. Also, Ken is planning another expo in April/May.
Sign In or Register to comment.