Whats stalling the P2 and P8X64A?

My gripes about the P1 being just a niche toy.. Video HW for example, the only inputs I have left on my 49" 4K PC " monitor" (Thanks Santa!) are composite YCRB, DVI, and HDMI. The prop 2 was supposed to be produced over 2 years ago. I recognise the issue with a great chip (like that of the latest ARMs) comes a hefty man hours price tag. But it urks me that the Arduino bunch came out later than Parallax and has had more success than Parallax in MCU's. I really liked the Prop 1 when it became available, I have about a dozen boards and a few DIP's, but keep finding myself leaning toward AVR's and ARM's for projects because they have dedicated peripherals like UART, ADC, SERIAL. Yes, all this can be done on the prop 1, but its all emulated. This makes the prop even more limited than the 32K of RAM and limited pin count! The OG's in the forums achieved some UNBELIEVABLE things with the prop 1 but there was a lot of clever coding and much PASM (also niche) to achieve these things. And whatever happened to the release of the P8x64A?! 64 pins are much more usable than using shift registers for fast things like FRAM for video storage, or being able to drive something else than a 24 bit DAC & maybe a keyboard...

The ATMEL's inside a Phillips hue bulbs have many peripherals including UART,ADC,SPI,I2C,WIFI(Xbee proto IIRC)

Is the Prop 2 ever going to be released with relevance to a modern MCU/MPU. Seems the prop 2 is a confusion of a Raspberry Pi ($5 now) and a Prop 1 ($8 now) with nearly ALL peripherals needing to be emulated taking cores away from your "parallel" processing with proprietary language and not much mainstream support for many libraries in C/C++ like AVR's for instance.

Sorry if this sounds more like a whiny post, but I think it has some relevancy, since I did my share of supporting parallax over the years with more than a dozen of these boards more or less just sitting around now, plus sensors and a $40 HYDRA book collecting dust...
«1

Comments

  • Re: why is p2 delayed?
    It's delayed and Chip is working every day on it. I think it's best left at that. He will provide an ETA as soon as he can.

    Re: Arduino did all this cool stuff in less time
    Remember that Arduino is not a micocontroller or chip, but rather just a bunch of software and dev boards. Arduino did not develop a new MCU. They took existing MCUs and put them on very simple dev boards and then wrote a bunch of software to go long with it. We actually have wonderful competition for that in the Prop 1 world: SimpleIDE and the Learn folder being the most common example, but there's also my own creation, PropWare, and even lib-propelleruino. We also have the Propeller ASC+ for an Arduino-compatible dev board, along with all the dev boards out there.
  • Within a few days, I will have a new FPGA image posted here which will feature the new smart pins, which can do UART, SPI, ADC, DAC, PWM, and many measurements functions. This will unburden the cogs quite a bit.

    After we've refined the smart pins and done any final tuning to the rest of the design, we will proceed to make the chip.

    It's true that things like Arduinos do impressive things for cheap, but they are not often useful for strict real-time control. This is what the Prop2 is all about. You code exactly what it does. It's quite a simple and direct system, in that sense.
  • cgracey wrote: »
    UART, SPI, ADC, DAC, PWM, and many measurements functions

    how cool! I read about the UART hardware but didn't know SPI was also going to be included! Does that include SPI slave?

  • cgracey wrote: »
    Within a few days, I will have a new FPGA image posted here which will feature the new smart pins, which can do UART, SPI, ADC, DAC, PWM, and many measurements functions. This will unburden the cogs quite a bit.
    What about I2C?

  • cgraceycgracey Posts: 13,125
    edited 2016-01-01 - 17:16:20
    DavidZemon wrote: »
    cgracey wrote: »
    UART, SPI, ADC, DAC, PWM, and many measurements functions

    how cool! I read about the UART hardware but didn't know SPI was also going to be included! Does that include SPI slave?

    When I say SPI, I mean just the bit shifting in and out with a sychronous clock. It uses the configured pin as the data I/O and listens to the adjacent pin for the clock. The adjacent pin can either be toggled manually, by another system, or configured for multiple timed-transition output.
  • If your Propeller boards are collecting dust, that's on you. As for I/O, the P1 has more available pins than any ARM I've designed with. They are also more flexible and more easily used. Yes, you do sound a tad whiny. ;) My prescription? More doing and less waiting.

  • A week ago, I was running around doing Christmas errands and I was talking to my phone a lot for Google searches to get directions, phone numbers, etc. I considered how all this effort was in pursuit of certain outcomes that would bring satisfaction. For fun, I asked Google to bring about desired outcomes without any effort on my part. It gave me links titled, "How to avoid purgatory."
  • That's hilarious!

    I can just see placing a call....

    "Hello? Is this Bob with Purgatory?"

    "Bob here, and let me tell you it just never ends, they keep changing the forms, and I got people to process, newbies coming in, no where to put 'em, no way to cycle the others out, it just goes on and on and on, and don't get me started on those clowns in legal...."

    [click]

    :)

  • David Betz wrote: »
    cgracey wrote: »
    Within a few days, I will have a new FPGA image posted here which will feature the new smart pins, which can do UART, SPI, ADC, DAC, PWM, and many measurements functions. This will unburden the cogs quite a bit.
    What about I2C?

    I think the synchronous serial circuit that facilitates SPI can also facilitate I2C.
  • jmgjmg Posts: 14,540
    cgracey wrote: »
    David Betz wrote: »
    What about I2C?

    I think the synchronous serial circuit that facilitates SPI can also facilitate I2C.

    i2c has some tricky details.
    Some vendors use SPI shifters to move BYTE info in master mode, and manage the START.ACK.STOP in software.
    Often they shift the SDA changes toward the middle of the SCL low time. (SysCLK/4 max can do that )

    That's simple, but lacks Clock stretching and does not work as a i2c Slave.

    Other MCUs had small Logic to capture Start or Stop, and fire an interrupt, to give some slave support.

  • I2C is very limited (compared to SPI) in bus speed and is easily handled in software
  • Mike Green wrote: »
    I2C is very limited (compared to SPI) in bus speed and is easily handled in software

    That's how I feel about it, too.
  • jmgjmg Posts: 14,540
    Mike Green wrote: »
    I2C is very limited (compared to SPI) in bus speed and is easily handled in software
    That's true for master, not quite so correct for slave and clock stretching.
    i2c Specs are 100/400/1MHz and 3.4MHz and 5MHz

  • I'll start by apologizing to Chip. The P1 was what it was, 8 identical cores with counters and video hardware with 32K shared RAM. It was great for the hobbyist community when it hit. Still has some relevance even today, but it did have some severe limitations.

    Up to this point all I could gather on info was pretty much speculation and endless feature requests. I applaud What Parallax is trying to accomplish with the P2 having to learn from what the P1 was and wasn't capable of. And forgive me if i'm impatient. I'm glad Parallax semi is doing something the rest of the world isn't, listening to their customers and even getting feedback on the latest design! That's unheard of.
    Tough time of the year for projects, but glad to see that the Prop 2 isn't dead in the water and will be bristling with peripherals on ALL pins?!

    I'm not an expert, i'm 37 and my profession is Autobody Collision for the past 21 years. I do not come from 6502, 8051, or Pascal.. My whole experience started from a Forrest M Mims mini books from radioshack and a pile of parts, my first MCU was a BS2. I just found that C was easier to pickup than SPIN. There was more tutorials and support for C when arduino came about. More "shields" and library support.
    No doubt the Prop 2 won't be criticised as the prop 1 was.
    -UserName: It would take 2 P8x32A's to do the job of one ATMEL mega 2560 from a shear pin count standpoint, ie: itead 3.2S touchscreen shield. Prop 1 could do this with shift registers i guess since the screen alone takes 20 I/O lines, touch another 5, then what? Faster? sure, but with limited room left for a program after the video stuff, how many pins left? How bout an SD card to fetch images? oops, out of pins, and probably RAM.

    Just sayin is all. Unrelated - Sparkfun at least since cyber monday has had for sale a FLIR camera for under $300!
  • I'm not an expert, i'm 37 and my profession is Autobody Collision for the past 21 years.

    Do you ever access settings of the management computers at all? I've recently purchased my first modern car and found the electric window controls just a pain. They stop working the moment the ignition is turned off! I've asked an auto electric shop and they said something about a chassis management computer and that no they weren't able to change the setting.

    I'm thinking of selling the car again because of this.
  • I'll start by apologizing to Chip. The P1 was what it was, 8 identical cores with counters and video hardware with 32K shared RAM. It was great for the hobbyist community when it hit. Still has some relevance even today, but it did have some severe limitations.

    Up to this point all I could gather on info was pretty much speculation and endless feature requests. I applaud What Parallax is trying to accomplish with the P2 having to learn from what the P1 was and wasn't capable of. And forgive me if i'm impatient. I'm glad Parallax semi is doing something the rest of the world isn't, listening to their customers and even getting feedback on the latest design! That's unheard of.
    Tough time of the year for projects, but glad to see that the Prop 2 isn't dead in the water and will be bristling with peripherals on ALL pins?!

    I'm not an expert, i'm 37 and my profession is Autobody Collision for the past 21 years. I do not come from 6502, 8051, or Pascal.. My whole experience started from a Forrest M Mims mini books from radioshack and a pile of parts, my first MCU was a BS2. I just found that C was easier to pickup than SPIN. There was more tutorials and support for C when arduino came about. More "shields" and library support.
    No doubt the Prop 2 won't be criticised as the prop 1 was.
    -UserName: It would take 2 P8x32A's to do the job of one ATMEL mega 2560 from a shear pin count standpoint, ie: itead 3.2S touchscreen shield. Prop 1 could do this with shift registers i guess since the screen alone takes 20 I/O lines, touch another 5, then what? Faster? sure, but with limited room left for a program after the video stuff, how many pins left? How bout an SD card to fetch images? oops, out of pins, and probably RAM.

    Just sayin is all. Unrelated - Sparkfun at least since cyber monday has had for sale a FLIR camera for under $300!

    RinksCustoms, don't worry about a thing. I understand people's flabbergasted reaction to how long this has taken. I never would have imagined, myself. I thought this was going to be done in 2008, now eight years ago! A lot of special (I think so, anyway) work has gone into this, with a lot of formative input from people on the forums. What this is shaping up to be is not a "me, too" chip, but something that is real-time capable and excellent for addressing problems from a first-principles perspective. I don't know how well it will do in the market, but I hope it does all right. If we can return our investment and then some, that would be great.
  • PublisonPublison Posts: 11,787
    edited 2016-01-01 - 23:19:06
    No doubt the Prop 2 won't be criticised as the prop 1 was.

    I'm sorry, where did that come from? Who criticised it?

    I think the Prop 1 was fully embraced by the community as being a forward thinking MPU. Many people went on to produce successful products with it.

  • It would take 2 P8x32A's to do the job of one ATMEL mega 2560 from a sheer pin count standpoint, ie: itead 3.2S touchscreen shield. Prop 1 could do this with shift registers i guess...
    Sure! A clock I made years ago has 210 individually-addressable LEDs yet ties up only five Prop pins. It could have been done with fewer pins and it could have been done with more. Five seemed about right. More would have been extravagant.

    In commercial designs every pin is strategic and many do double-duty. 20 pins just for a screen sounds like the wrong screen.

    Also, why refer to the P1 in the past tense? People are using it to do cool things every day. If Phil Pilgrim hasn't fully exploited its possibilities, I doubt you or I have. :)

    Finally, who wants all microcontrollers to be the same? There are a wide variety of chips for a wide variety of situations. If you have a 50¢ task use a 50¢ PIC. If you're all about running canned software, perhaps Arduino is your cup o' tea.

    Arduino doesn't work for me simply because I have no faith that it can be made to precisely and dependably drive the MOSFETS of six power converters while it is also generating composite video and demodulating DPSK. And if it *could* be made to juggle all those balls at the same time, the coding would be a nightmare. The Prop makes such things a breeze. In fact I'm always pirating chunks of old code for for new projects. They generally fold-in without a hiccup, no matter how demanding or punctilious their nature.
  • potatoheadpotatohead Posts: 10,121
    edited 2016-01-02 - 01:53:47
    Yes, the P1 will continue to be sold, supported, etc... for a long time yet.

    Past tense is tricky. When we are working on, testing, etc... for P2, it slips out easy. Ken probably feels a little tickle behind his ear each time one of us does it!

    @Chip, you are a gentlemen. That's humble, honest and a considerate assessment of where it is at today. Well said. Nobody had any idea what had to be learned, nor how long it would take.

    I'm pretty stoked about this one though. It's gonna happen. The right trade-offs are being made.

    @All: You know, in the near future, we may be able to put together complete, all Propeller dev systems. P2 chips programming themselves and P1 chips too. For some use cases, where say a long time stable or accessible dev / maintenance environment is needed, people will pay nicely for that. It's a hard thing to secure these days, and who here hasn't seen "that system" running on DOS, or something, being carefully maintained via e-bay, etc...?



  • cgracey wrote: »
    David Betz wrote: »
    cgracey wrote: »
    Within a few days, I will have a new FPGA image posted here which will feature the new smart pins, which can do UART, SPI, ADC, DAC, PWM, and many measurements functions. This will unburden the cogs quite a bit.
    What about I2C?

    I think the synchronous serial circuit that facilitates SPI can also facilitate I2C.

    I'm giddy :D
  • MJBMJB Posts: 1,198
    more than a dozen of these boards more or less just sitting around now, plus sensors and a $40 HYDRA book collecting dust...
    there are still many great things that can be done with the Propeller 1

  • I have made a nontrivial pile of money for my company selling P1 based designs. It's a great platform when you need 8 serial ports or 2 vga or some other weird combination nobody's thought of. The things it can't do can often be done by something else; I'm finding out right now that P1 + ESP8266 is a way powerful combination. The one propless ESP design I am working on needs a transistor, because the ESP UART levels can't be inverted in software. It only takes a few things like that to show the genius of the Propeller design.

    Also, P1 is capable of very low power operation which the 180nm P2 won't be. That will always be a thing in a lot of applications.
  • Well, I just love the P1. I agree that there are a lot of things that cant be done with it, mainly in measurements due to the lack of ADCs and DACs, but you start being a fan is when you realize ALL that can be done with it and not with the others, unless you pile up a few extra (and may times not cheap) hardware. I have sold quite good portable RFID readers with 40X20 NTSC video display for portable ID, access and loging devices using only the P1 and a few passive parts (and a lot of ideas from OBEX).

    Now, Chip: when will I put my hands over the P2? Remember that I am old and, if I die before having the P2, I (or my gost) will give you a very hard time.

    Last question: any video upgrade? new LED TV are not "VGA friedly devices"
  • Does your TV do component video?

    I've finally got one setup right now to begin testing S-video and Component. Right now, we've got code running for composite and VGA, with both working pretty well.

    Component works out nice. It's basically the VGA of consumer TV's.

    Oh, and Rayman has done an LCD that worked out really nice too.

    For digital, we are going to need to source a dead simple, analog to HDMI type chip, or a little converter box for the same. I'm kind of hoping a chip like that can be dropped onto, or provided for as an optional, "solder it in right here" type solution.

    ...or we find out who is selling realllly cheap HDMI converters. :)

  • potatohead wrote: »
    Does your TV do component video?

    potatohead wrote: »
    ...or we find out who is selling realllly cheap HDMI converters. :)

    Thanks For your advice.

    Its not "My TV", A few years ago I did a very simple and cheap display for cheap TV sets to be used as info booths. Now the client wants a few more, but the new cheap TV set comes with a couple of HDMI and a very bad composite, not able to render the VGA quality. Is just a matter of prices and the HDMI adapter, even being cheap will skyrocket the final price, so now I am diving in Alibaba to find some old sold out TVs. As you can see in the attached pic, the product is small but loaded and a redesign would not do any good at this time.

    For new designs will use the adapter until a cheap and nice chip appears on the marked. The "solder it in right here" is something I like a lot, and use it for the Li Ion batery cxharger chip or for the Blue Tooth!!

    have agreat year, glad to meet you :thumb:
    2396 x 1200 - 268K
  • potatoheadpotatohead Posts: 10,121
    edited 2016-01-03 - 04:57:58
    You used the Prop VGA for "information booth?" Interesting! What was the content?

    Nice and compact. Loaded as you say too. :)

    Yes, I've run mostly analog sets and am finding composite varies on the digital ones a whole lot, and it's mostly not as good.

    A neighbor has a digital "SD" LCD that actually performs pretty well, but those are actually kind of rare. They were not on the market long before it all went HD 720p and greater, even in cheap-o consumer sets. Not sure those SD digital ones can be sourced at all, so I'm ignoring them.

    My testing different displays basically results in lower overall expectations for digital composite. That's fine. It's been a very nice long run. And it's there should someone have a good display.

    VGA has dropped off more newer sets, but I'm seeing component enough to feel that's what will make the most sense the majority of the time. There is new HD video gear shipping with component and the coaxial surround as well as the optical one. Component should be good for a long time yet.

    You've reminded me to go and get an adapter that takes in the lot, composite, component, s-video, VGA. Maybe that is the best overall test / development target right now. I've got a nice HDMI / DVI monitor to use, and I've been ignoring it.

    Good luck with your TV search. Hopefully, you can find enough to get you over the hump.

    Yeah, you too, and likewise. :)
  • potatohead wrote: »
    You used the Prop VGA for "information booth?" Interesting! What was the content?

    If you have been at Parallax (or seen some Ken´s videos), there is a TV secreen at the main entrance showing the day program beinf feed by a Quickstart card. Mine is similar but shown multiple pages, the pages changes on a timer or by IR control, and at the lower side of the screen there is a banner runing with hot spots or ads, all text. User chooses the colors and blabla....

    With the "S Video" you brought me back to my "Commodore 64" hacker times... Great video for those days.

    Will try on component, never thought about it, thanks. You are right, just need to deal with the sync and stuck it into the green, and that will be done in the cable from the VGA to YPbPr..

    Now lets find the time for doing it

  • jmgjmg Posts: 14,540
    JotaChip wrote: »
    ...Now the client wants a few more, but the new cheap TV set comes with a couple of HDMI and a very bad composite, not able to render the VGA quality.

    Find a better TV brand ?
    The cheap TV's can come with Multiple HDMI, and Component and Composite and VGA and USB.



    Even small ones have VGA+HDMI
    These on eBay manage to fit a connector shelf into small screens :

    http://www.ebay.com/itm/7-Inch-800x480-TFT-LCD-Screen-AV-HDMI-VGA-Car-Rear-View-Monitor-Adpater-/171882647468?hash=item2805010fac:g:Sw0AAOSw~gRVwiy-


    http://www.ebay.com/itm/8-1080P-TFT-LCD-Car-Monitor-Rear-View-MP5-Player-TV-Support-BNC-AV-VGA-HDMI-USB-/121775068519?hash=item1c5a5c1567:g:aU0AAOSw4HVWDkzo

  • Drooling!

    With smart pins, I'll dust off my DE2-115

    Any idea how many cogs & smart pins will find Nano and DE2?
    cgracey wrote: »
    Within a few days, I will have a new FPGA image posted here which will feature the new smart pins, which can do UART, SPI, ADC, DAC, PWM, and many measurements functions. This will unburden the cogs quite a bit.

    After we've refined the smart pins and done any final tuning to the rest of the design, we will proceed to make the chip.

    It's true that things like Arduinos do impressive things for cheap, but they are not often useful for strict real-time control. This is what the Prop2 is all about. You code exactly what it does. It's quite a simple and direct system, in that sense.

  • I think I2C would best be handled with an interrupt on the clock (due to clock stretching) - with the new cog speeds even 3.4Mbps I2C will be possible in an interrupt routine without chewing up a full cog (in silicon, the FPGA will probably handle 1Mbps I2C)
    cgracey wrote: »
    David Betz wrote: »
    cgracey wrote: »
    Within a few days, I will have a new FPGA image posted here which will feature the new smart pins, which can do UART, SPI, ADC, DAC, PWM, and many measurements functions. This will unburden the cogs quite a bit.
    What about I2C?

    I think the synchronous serial circuit that facilitates SPI can also facilitate I2C.

Sign In or Register to comment.