Editorial
rjo__
Posts: 2,114
in Propeller 2
You might have the impression that the Engineering Phase of the P2 is just for Engineers... not true.
If you listen to the engineers, you may become forever and hopelessly confused.
Rest assured. I am a generic idiot... the kind they write books.
I'm not an idiot about everything (just the things that currently interest me).
You might also get the impression that the P2 is incredibly complicated and will have a much
steeper learning curve than P1....
NO!!!!!!
Well... yes and NO!!!!!
If you really care about how the P2 does what it does... YES.... it is incredibly complicated.
BUT!!! If you simply want to do what you want to do... there will not be a simpler super-controller to use than the P2.
(Assuming you want to do a lot of things you can't do right now.)
If you want to hit the ground running, THIS is precisely the time that you should be tinkering.
Get a P2v... they are cheap... Or treat yourself and get the whole hog, a P123a9.
There is no time like the present...
You can't do it without hardware in your hands.
Get going and let us know how you are doing.
Rich
If you listen to the engineers, you may become forever and hopelessly confused.
Rest assured. I am a generic idiot... the kind they write books.
I'm not an idiot about everything (just the things that currently interest me).
You might also get the impression that the P2 is incredibly complicated and will have a much
steeper learning curve than P1....
NO!!!!!!
Well... yes and NO!!!!!
If you really care about how the P2 does what it does... YES.... it is incredibly complicated.
BUT!!! If you simply want to do what you want to do... there will not be a simpler super-controller to use than the P2.
(Assuming you want to do a lot of things you can't do right now.)
If you want to hit the ground running, THIS is precisely the time that you should be tinkering.
Get a P2v... they are cheap... Or treat yourself and get the whole hog, a P123a9.
There is no time like the present...
You can't do it without hardware in your hands.
Get going and let us know how you are doing.
Rich
Comments
First is the power requirements they are crazy.
Second the unit is too big.
Third the unit is too expensive.
Fourth is that I don't think it fits Parallax's mission statement.
I love the P1 and what I can do with it. It fits. I think it may need a refresh, say more memory and faster processing speed but that's it. If I want to run an operating system or Python I would just use an STM product that has been doing that for year now.
Mike
I agree with Mike's first three items.
9 years ago I chose the P1 for several projects because:
* It was breadboard friendly (the DIP version, of course)
* Single power source - 2 AA batteries work fine in a pinch.
* Minimum of support hardware - a couple of caps and a crystal for development.
* Free development software that works.
* Low cost (both initial and ongoing).
On the fourth, I understand that Parallax needs to move forward and have a competitive product.
But it's overshot my needs.
I would be very happy with a P2 available as a P1-compatible 40 pin DIP module, if available for < $20. Why?
* For development purposes where there's a risk of damaging the chip due to bad wiring or design on my part. Yes, only 32 I/O. And less memory and/or cogs and/or performance is ok.
* As an upgrade to my existing designs where I've used the DIP version. I understand I'll still have to redo the software. But not having to redo the hardware (other than maybe the crystal?) is priceless.
Again, just my opinion.
Walter
If you are going to "breadboard" just solder it all up properly to double-sided plated through matrix board since you are not tied to limited strips of pins. Not only do the components "plug-in" just where you want them, but the connections are also reliable but still changeable.
This matrix board is cheap, the P2D2 module is small and you can plug it in. My unit is connected to a VGA monitor and running as a standalone PC playing continuous background music and videos but I can still talk to it and program it while it is doing this.
For others:
1. Power is power. Either you want it or you don't:) IF you are going mobile with your P2, you are going to need more than a AAA battery... but if you are going mobile, you aren't going to be using a AAA battery anyway!
2. I don't know. The P2 is a pretty small package. I personally wish it were bigger. I expect pretty small boards to be available. The primary reason they will be a tad larger is that you have twice the amount of i/o available.
If Mike is talking about the P2v variants there are some pretty small units out there. Granted you need a wall-wart to run them.
3. Not if you charge by the MHz, Hub Space, or ease of use. Then it is either exactly the same price as a P1, a little cheaper or a huge value, depending upon the metric.
4. Mission statement. I didn't realize that Parallax had a mission statement. I thought that the reason Parallax existed was to give me what I want, when I want it, at an affordable cost.
How about a new mission statement: "More, better, faster, easier." To be more woke you could leave out the first comma.
"I love the P1 and what I can do with it. It fits. I think it may need a refresh, say more memory and faster processing speed but that's it."
This is a secret. You can use the P2 almost exactly as if it were a P1. Nothing has been lost. You can view it as a P1 refresh with more memory and much faster processing. If you don't want to bother with the new stuff, you don't have to.
"If I want to run an operating system or Python I would just use an STM product that has been doing that for year now."
Exactly. All you need to run the P2 is PNut. Everything else is just gravy for some and confusion to others.
I personally use Python to automate the development cycle a little. I like NotePad++ for editing big code.. With Python I can edit my script in NotePad++ and when I am ready to go I hit a magic key and it automatically saves the changes opens PNut and runs my script. That's it... a convenience nothing more. I get code folding and super find functions... but without a Python helper... it is just too many keystrokes between my last thought and a result on the screen.
MicroPython is a different story for me. I can imagine the confines of a class that lasts for a semester. Trying to introduce the P2 to a class of students, all of whom already know Python, and having MicroPython right there to ease the task. That to me sounds incredibly cool... not absolutely necessary, but very, very useful.
Operating systems. There is a reason they exist, but you don't need or even necessarily want one on the P2. It is possible. It will happen, but it won't be for everyone. Python to me is the bridge between the P2 and the rest of the world. Serial is really all we need, and there are plenty of ways to do it, but for my money and time spent, Python is just about perfect.
Rich
I welcome the P2's speed and memory. For me to pursue using the P2 as a P1 I would expect documentation om a par with that for the P1's PASM . I realize that to program in PASM on the P2 I need to learn some new instructions. I found sufficient I/O pins on the P1. I would also expect a reasonably sized and priced basic board akin to the P1's Micro board to allow me the flexibility to use it in way that suits me.
Let's say that you want a 10MHz numerically controlled signal coming from one of your pins...
With the P1 you do something like this:
To a novice, this code is incomprehensible. How to change this to a 1MHz signal is not at all obvious.
you have to understand both the operators and the counters. Pretty steep learning.
Now let's look at the P2 equivalent:
WRPIN #%1_00110_0,PIN WXPIN #1,PIN qfrac ##10_000_000,##sys_clk 'set the frequency to 10MHz ' qfrac ##5,##sys_clk 'or hook up an led to PIN and set to 5Hz getqx pa WYPIN pa,#_PROPCAM_MCLK dirh #_PROPCAM_MCLKThis looks more complicated... but look more carefully. If you were a complete novice, the only line you need to worry about immediately is the qfrac. And you don't have to understand how it works, you just need to understand that to change the frequency... you make a single substation in the qfraq line and you are off to the races.
The rest of it you can learn at your own pace. You don't have to suck it all in at the beginning.
Until there is a large set of simple examples, this stuff will look fairly opaque. But remember that at a similar stage in development of the P1, there was a datasheet that was about 2 pages long and almost no-one understood any of it.
I am not an engineer who has cause to get into all the nitty gritty. I just want to use it and move on.
Not saying we shouldn't look forward to P2 or that we should judge P2 on the current documentation, tools or examples. Just saying don't sugar-coat the current learning process for P2. It's difficult.
dgately
I find that I often have to make several revisions before I get a circuit working properly. That, coupled with difficulty soldering (hand shaking) makes this a poor choice for my purposes.
I'm not trying to do high-speed activities such as memory access. For my purposes SPI (as an example) works fine. So breadboard circuits work fine for my prototyping.
Regarding the P2D2 module, what's the estimated purchase price for a completed module?
-Phil
It has been so long since I actually used a counter that I have forgotten how they work:) So, I don't know that much.
I would have to go back and look. But what I seem to remember is that for each cog we got 2 counters. This ate up cogs in a hurry. If you needed lots of counters... you needed a lot of cogs. Not any more.
Counter functions have been moved out of the cog and into the pin structure, so there is no limit to these functionalities.
In many cases they are "set and forget" until needed.
If you want a single cog using 64 different counter-type functions for 64 different pins... you can do it.
No doubt there is going to be a need for outstanding educational materials, including a ton of very simple examples.
Fortunately, this is one of Parallax's many strengths.
I didn't have any controller experience before coming to the P1. I didn't want to read anything, I just wanted to find a bunch of good examples and get moving. I did end up reading a lot of stuff shortly thereafter:)
Day 1 the P2 will be easier to approach than the P1...for an absolute novice. To actually master the P2 will take longer, but the journey is the destination.
About the "substation" … I can type real fast but I can't actually see what I am typing very well;) Too much sun for my aging macula.
I see it more as evolution (or directed development, depending upon your preferences.)
Since the PropCam is hot on my mind these days let me give you another example. Your i2c code for the PropCam.
When I look at your code and then I look at the manufacturer's datasheet. I simply can't see why your code should work... but it does.
PRI _config(register, data) {{ Write configuration information to sensor chip register. '' '' `Parameters: '' '' `register: sensor chip's register address. '' `data: byte to write in `register. '' '' `Return: none }} outa[Scl]~~ dira[Sda]~ outa[Sda]~ dira[Sda]~~ outa[Scl]~ _write($88) _write(register) _write(data) dira[Sda]~~ outa[Scl]~~ dira[Sda]~When I directly transliterate this to the P2.... it doesn't work. But if I code from the Datasheet directly into P2 assembly, my code does work.
According to the Datasheet... the start condition is supposed to be a transition from high to low on sda while the clock is held high... and then the clock goes low... Yours does work on the P1. I just don't see how.
Smart pins are way better than counters. Here's one cog controlling up to all pins on the P2. click here
I don’t pretend to understand everything about P1 yet, and I’ve been using it for over 10 years now.
It’s only necessary to understand what you need to to do what jobs you want to get done. Don’t over complicate things.
So, there is no requirement to understand smartpins unless you actually want to use those features. You can use code objects that others have/will write without the need to understand how they work.
There is no need to understand how the streamer works. Or the Cordic functions. Get the idea. Even the complex instructions are not required for base use. Just the equivalent P1 instructions and a few of the CALL/JMP instructions will do many jobs.
So, in summary, stop stressing about the P2’s complexity. And if you still are too afraid to jump in, stay with P1 - it will be around for many years to come - it’s not going anywhere just because the P2 is here.
P.S. There will not be a DIP package for P2. Time to get over it!!! Most micros are ditching the DIP packages, and almost no company is making any new ones in DIP packages. The volumes are too low to be viable. The hobbyist market is way too small - and here i mean those who actually make their own circuits/boards. Add to this that many complex chips no longer fit into the DIP packages either. So either try SMT or buy modules.
I politely disagree that this is simple, this early in the game. I would much rather wait for a less expensive non-eval board, bare bones, hopefully proto board style P2 to be released.
Is there a release date for any smaller form factor P2 boards?
But later on I did just bought assembled units, I loved those 5-pack propeller Proto boards without USB. Perfect for one offs.
Same will go for the P2. I will not be able to solder that thing. But Peters P2D2 is small and can plug into anything I can build without me needing the knowledge to design a board around the P2.
So in some form it is a DIP chip, just uncommonly sized. I hope that Parallax will support Peters work and use this as a common mass produced module to get the price down.
On the software side I was eager to try the P2 and have a Eval-Board. The documentation is somewhere sparse and tense, but like with the P1 all of it fits together like it does on the P1.
Chip has some mindset trying to avoid 'usual' things and has his own view of what should be done. And like on the P1 I really like what he did with the P2.
Sure I never managed to really understand the P1 video circuit, but I was able to use it. Same goes for them Smart Pins. I have not even scratched the surface of what they do, but the jump from P1 to P2 is not that complicated.
I do not regret paying for 2 eval-boards, it is honestly really fun to program the P2.
And the tools are evolving fast. P2gcc, fastspin and now support for VS-code, all of this already working with just 120 chips alive.
Sure Parallax should get work organized for redoing gcc or switching to clang, but my guess is that they are currently very stretched.
I am lucky, I can play with P2 already and have a lot of fun with it.
Enjoy!
Mike
An example is worth 1000 words.
Reduces time and frustration, too.
I am glad that I was able to get a P2 Eval board and when the respin makes it obsolete I am seriously thinking of framing it. I wasn't there in the earliest days of P1 and I suppose it must have been a lot like this, with documentation sparse and a bit inscrutable when it existed, and a world of possibility not yet really explored. I don't know everything P2 can do and I'm not one of the people who will be pushing its boundaries, but I have learned this: It solves all the problems I had with P1 for certain important commercial applications.
Those aren't "high volume" applications, so I know my needs are not a priority to Parallax, but I also know there are a fair number of people like me out there so if I'm happy they will be happy too. And I am very happy with what I've seen. I wish we had the clarity for development environments that we had with P1, but I know a lot of people saw Spin-only as a problem rather than a feature. I've focused on FastSpin and have been pleased with the results so far.
I got my start with the Propeller 1, after reading about the YBox2, via SpinStudio and the DemoBoard. The P2 Eval board nicely splits the difference and makes a great intro package, and I'm glad Parallax seem to be planning a similar board for the respun "real" P2 chip. It's not a board any of us would put in a product or permanent project, but it's what we need to learn and experiment, and aren't those things what Parallax is all about?
I remember what some of the topics of discussion were at the last UPEW in 2011, when I was privileged to meet the Parallax crew and a lot of our forum regulars. It's amazing how far things have come since then. This is going to be a great product and I've already got one design engineer at a real manufacturer sniffing at the data sheets. Make it real, and they will come.
What I'm doing can only be done with a Prop1 / Prop2 type processor.
Looking forward very much to the Prop2 going into production later this year. I think it will make Rev 2 of my product much easier to write and will take it to the next level.
What function is the PIC18F handling and is the reason because of hardware or software or memory limitations of the P1? It's just that I have been able to cram a mighty lot into the humble P1 with Tachyon.
There is the nice and compact P2D2 module, which is 0.1" pitch, and brings out all P2 IO, on 2 x 40 pin headers. It includes a USB Bridge, and SD card and Flash.
I have a PCB layout variant fork called P2D2Pi, that has pin-swaps applied to give RaspPi pinout compatible headers, same PCB size, to allow access to all the RaspPi boards eco system. (std P2D2 lacks RaspPi pinout compatible)
If you want a DIP40, clearly that has to lose some P2 pins, and then you have some mechanical issues - see image of P2D2 with DIP40 dropped onto it.
You might be able to use the P1 xtal, but since this is a module with regulator & caps, you would probably include a higher MHz xtal
The DIP+QFP is quite a challenge to route and package, you might choose to do a middle board that is DIP40<->P2D2, but that's physically even larger...
If you decide to map the lower 32 pins, that's going to be at least 4 layers, maybe 6, depending how squeezed you want width.
~0.9" looks smallest possible.
Next issue is vanilla 0.1" headers are cheap, but not really DIP socket friendly, whilst finer pin headers cost more, and are more fragile.
It may overshoot your needs, but mind that the P2 is excellent for analog and digital real-time applications. This chip would be great for DDS signal generation, when paired with a DAC. It may have potential to be used in an oscilloscope as well.
Kins regards, Samuel Lourenço
All this stuff is ridiculously inexpensive.
Mike
10MB washing machine sized HDDs were $16,000 ea in 1976 when they were discontinued. Field Engineers who repaired the minicomputers and mainframes of the day were paid about $6,500 pa.
Now, those 64GB SD cards are worth a couple of hours pay.
I am still looking for my free car in the weeties box
I am eager to start using the P2, but I am a busy person and I can't deal with the constant changes being made.
Once the design and commands are "set in stone" then I will get involved.
From the brief understanding I have of the P2 it appears to be a true quantum leap from the P1, but I would have been happy with a P1 with 64 I/O and more RAM.
It's a shame that the P2 is only 8 cogs. I was really hoping for a 16 cog P2.
I once commented that the P2 would be "The greatest microcontroller never made." I hope I was wrong about that, but that was several years ago...
[/soap box]
Bean
Are you basing that on the relatively power-hungry ES version of the chip on the evaluation boards? The production chip is going to be much less thirsty: about 790mW max.
Also, regarding cooling, most applications aren't likely to push the chip as hard as the evaluation boards have.
Barring a complete disaster with the re-spin it is "set in stone".