PIC32MX AVR or ARM Hobbyist's Platform
microcontrolleruser
Posts: 1,194
Could Parallax please produce a system for hobbyist's that
already have learned a fair amount about MCU's?
already have learned a fair amount about MCU's?
Comments
Once you have a custom-language module, like Stamp, the actual core used matters not one jot to the user.
It might make commercial sense to choose a lower cost, easily available part, or one that you already know.
Users do care about things like 5V IO, and nice quick programming.
The ideal part would be HS-USB, 5V IO and small package.
You can get close to that sweet spot with NUC442RG8AE, gives HS-USB, 2.5V ~ 5.5V, TQFP64, 84 MHz $3.31/1k
or, if you relax HS-USB to FS-USB, perhaps NUC126 FS-USB, 2.5V ~ 5.5V LQFP48/54/100 72MHz $?
Notice that spec does not include the Core, because once you have your 'Stamp' language, that matters less.
Parallax could also look carefully at a Prop1 and Level Translators, and I think someone has already done that.
If you are OK with 3V IO only, the FLiP is available right now...
'Err, that's illogical.'
'Stamp' is an industry term. Probably more in use in the 90's.
Think now most outfits call it a 'Stick'.
It means MCU on a PCB. Probably other uses too.
How come I get picked to go out in the real world and find out what's going on?:)
I think there is a 'market opening' for this for Parallax version.
Digilent fumbled the ball and made there compiler/IDE a plug in for Arduino.
So. Parallax can be 'the only kid on the block' with an 'in house' 32MX compiler/IDE.
Another note.
Digilent PI was a 32MX but they have moved on to it looks like ARM.
So Parallax might just go with ARM.
And another note.
Think Arduino has been around since 2000.
No Propellor Arduino board. That means something. Don't know what.
Then you should logically also call it a 'Stick', otherwise readers will not be able to extract what you are trying to say.
If I google Stamp Module every single hit on the first page, bar 1, is Parallax Stamp related, and this is a Parallax Forum.
Now you switch to talking about some imagined Compiler/IDE for the 32MX, which is quite a leap from the #1 ?
Clearly, Parallax can never be 'the only kid on the block' with an 'in house' 32MX compiler/IDE !
There's been a lot of discussion over the years about using other microprocessors or IP like ARM. Parallax owns the IP for the Propeller. If they use someone else's, someone else gets the income. If that someone else feels like changing the design, Parallax is at their mercy and is stuck. These chips are cheap enough to be a commodity item and there's very little to distinguish Parallax's design from their competitors'. Witness what's going on with the Arduino. Parallax's Propeller is unique, has compilers and libraries that are specific to the Propeller, and their teaching materials, in addition to being good, are unique to the Propeller. The Propeller is easy to program and teach about programming in assembler as well as in C or Spin. I wouldn't want to teach ARM or Intel assembly to a beginner's class.
Indeed, there is one school of thought that says 'best to avoid 1000+ page manuals for beginners classes', and start with a simpler MCU.
The 8 bit MCU space is still very active, and if you are worried about being at any one suppliers mercy, the 8051 MCU is enjoying something of a resurgence.
The 8051 has many flavours of BASIC, C, Pascal, and is multi-sourced.
Mr Green
'secondary market of hobbyists'
Guess where the rub is. Leave's serious hobbyist out in the cold.
It would be rude to go into why.
''best to avoid 1000+ page manuals for beginners classes'
Great quote!
I think this is a bit of a distraction. The BASIC Stamp based on a PIC16C57 which has a 500 page data sheet and user manual. Seems to have been a good platform for Parallax back then.
There are lots of ARM chips with built in USB, code protection, 5V tolerance and less cost than the PIC16C57, and their user manuals are about the same size. It would be easy enough to put Blocky, or BASIC or JS on these and make it ready for the education market. And they are not vaporware.
JMG
There. I fixed the title.
Okay. The Propellor is just fine for teaching.
The SX is more complicated and leads into PIC's and AVR's.
That is because the Ubicom chip's were PIC clones.
So. We're all squared away here.
A quick sampling of sensors at Parallax website shows everything is 5V or 5V capable. How much would it be a feature if P2 could operate at 5V?
Just curious to see some feedback of how attractive this feature actually would be.
You have totally changed the question in you opening post. Which renders peoples people's replies rather confusing.
To answer your new question: What's wrong with the kit that Parallax offers already?
As someone that has used all kinds of MCU's since 1980 or so I find the Propeller to be a wonderful thing. It has the nicest instruction set, what with being 32 bit and very regular and simple. It has multiple cores which means it's
very easy to get deterministic timing without messing with the complexity of interrupts or an operating system. Not to mention a nice turn of performance. It has a drop dead easy higher level language, Spin, for beginners. Or C for the MCU old guard.
I presume, with your claimed MCU experience, the Propeller should also be a joy for you. As is for me.
To answer your original question about using a different MCU:
Well, what would you suggest?
My take on this is that there is a huge spectrum of computing modules available today. From the 8 bit AVR Arduino up to the 4 core, 32 bit, 1GHz, Linux running, Raspberry Pi. With all kinds of things in between like the STM32 based Esprunio and Micro Python boards, or the ESP devices with dual core 32 bit processors with WIFI and Bluetooth thrown in. They are amazingly cheap today. It would be pointless for Parallax to try and compete with just another ARM board or whatever. Very hard to make any money on that. Or even get any attention from the purchasers of such things.
I am very happy there is the Propeller. A lot more potent than an Arduino but a lot less complex than ARM based solutions. The instruction set manual for the ARM runs to nearly a thousand pages. Then you need to get to grips with the thousands of registers those devices tend to have.
The Propeller is a unique architecture, so it has a niche market. If Parallax boards used ARM they would just be, well, just another ARM board.
I think getting kids used to the multi-processing way of the Propeller is a very good thing.
NXP and Atmel have recently added 5V ARMs (actually Wide Vcc) to their Indistrial/Automotive Microcontroller range, joining Infineon, Nuvoton, Cypress & Zilog who all have Wide Vcc ARM Microcontrollers.
Yes, 5V / Wide Vcc ability on P2 would be nice, but the custom analog pins rather excludes that....
If there is an equivalent choice between Wide Vcc and 3v only parts, the Wide Vcc will win every time.
There have been 3V ARM Ardunios for a while, but they have not taken over the 8b segment, as some expected.
Part of that will be 3v issues, and part will be 'complexity overload'.
Atmel now offer 5V ARMs, but they also invest heavily in the Wide Vcc 8b AVRs, with the new Tiny817 series.
There have been a lot of new 8051 variants released in the last 2 years, and 8051 are now the 8b price leaders.
For ARM, the oft-claimed '8-bit replacement' are becoming less frequent, as it becomes apparent 8b are not going away.
NXP released an 8 pin ARM LPC810 with much hoopla back in 2012, and in 2017, I see it has quietly vanished from their line card.
Seems that was more marketing gimmick, than serious commercial demand driven product
Expanding on this from an IP side, and still looking at modules, as time moves forward there is another option, that Parallax are increasingly better placed to use.
This table shows QFN48 FPGAs listed by Lattice :
iCE40 Ultra / UltraLite / UltraPLus Device Selection Guide Parameter iCE40 Ultra iCE40 UltraPlus iCE5LP1K iCE5LP2K iCE5LP4K UP3K UP5K Density LUTs 1100 2048 3520 2800 5280 NVCM Yes Yes Yes Yes Yes Static Power 71 uA 71 uA 71 uA 75 uA 75 uA EBR RAM (kbits) 64 80 80 80 120 SPRAM (kbits) - - - 1024 1024 PLL 1 1 1 1 1 I2C Core 1 2 2 2 2 SPI Core 1 2 2 2 2 Oscillator (10 kHz) 1 1 1 1 1 Oscillator (48 MHz) 1 1 1 1 1 24 mA Drive 3 3 3 3 3 500 mA Drive 1 1 1 - - 16 x 16 MAC 2 4 4 4 8 PWM Yes Yes No Yes Yes iCE5LP1K iCE5LP2K iCE5LP4K UP3K UP5K QFN48 0.5mm (7 x 7 mm) 39 39 39 NA? 39 50:$3.13 50:$4.06 500:$5.00 - ~$6, nys Compatible Companion parts ? - QuadSPI and i2c Master/Slave FT4222H $1.48/1k or possibly NUC505DLA $1.74/1k TXB0108PWR 8-Bit Bidirectional Voltage-Level Shifter with Auto Direction Sensing and +/-15-kV ESD $0.58/1kParallax already have P1V, and the 128K Bytes of RAM on these larger parts, can support a lot of code.
Once P2 is over the hump, a module based on some choice from above (or something not released yet..), could fit between Venerable Stamps & FLiP.
or, they could choose a larger FPGA, and slot between FLiP and future P2 module, or even both, as a lot of the infrastructure is common.
You still haven't answered the question WHY Parallax should do all those things you are asking for.
WHY should Parallax, as Chip Producer build something with the parts of the competition? Would you also advice AMD to 'get over it' and use Intel-Chips instead?
And to your current post #1, Parallax does actual have a system that is "a system for hobbyist's that already have learned a fair amount about MCU's" and give them a some way enhanced and advanced MCU, food for thoughts and so, with a 8 core multiprocessor version of a MCU.
Far advanced compared to a Basic-Stamp. And far more easy as ARM or PICs with hundreds of variations and registers and pages of Data-Sheets just to figure out what chip you need to have the combinations of protocols on the pins you need.
Same goes for your quest about them building a complete development-system for PIC and all the needed documentation.
WHY the hell they should do that?
Maybe its you who needs to 'get over it' and program ARMs if you want ARMs, program PICs if you want PICs or dive into the multi-core parallel programming adventure what the P1 and hopefully someday the P2 provide?
As you have said, the documentation and course-material Parallax offers is brilliant, compared to other stuff of other manufacturers. But part of this simple and understandable stuff IS simple and understandable BECAUSE it uses the multiple-core aspect of the P1.
It is just a different way of structuring stuff and programming a solution, inherently build for multiple cores. And this is very needed for the future, since the classical single core processor can't go faster anymore because of general process limitations. Shrinking simply stops at atom-level and current processes are not far away from that, at a 10nm process a layer is just a couple atoms wide.
That is why all other manufacturers now produce multiple-core Systems. But the problem lies in the software, standard languages and standard approaches usually run sequential code, interrupted by other sequential code. Same goes for threads and handling of processes in the OS. All kludges, not solving the real problem.
Manufactures of graphic-cards did understand that (years? after Parallax), thus we have GPUs with hundreds of processing units.
But the Art and Science to program inherently multiple-core bound system has not really advanced. Sure we have various tries to solve that, extension for C, languages like Occam or Oberon and others, but the main thing is that it needs to be done to get around Moore's law and the physically ending shrinking process.
So even the 15+ years old P1 was and is way more advanced as any PIC you can find.
But it needs a different way of thinking and programming, and there you seem to struggle with.
Like the move from 'classical' to 'object-orientated' programming needed a new way of thinking for a programmer, the inherently multiple-core programming needs a different way of thinking from the programmer also.
And that is where Parallax shines in the MCU world.
Enjoy!
Mike
What does 'line card' mean? Mouser has 1400 in stock. You think it is EOL?
The web PDF called LPC Microcontrollers Line Card 2017, has no mention at all of LPC810, but does have LPC811+
The LPC810 parts that are stocked, have "go away" prices.
Microcontrolleruser, you seem like a smart person, I wish you would spell Propeller correctly...
Dave
Propeller.
Bob
Didn't the original Arduino folks use the BASIC Stamp first and then decided to spin their own board based fairly close to the Parallax design?
If you are suggesting that Parallax produce an ARM based board, then which vendor and which ARM processor do you suggest they go with; STM, TI, Qualcomm, Cortex-M0/M3/M4, A7, A9 and so on or should they plop down a billion or so to create their own fab just for hobbyist? Do they go with Thumb or Thumb-2 or both? What IDE should they use; IAR, Keil, TrueSTUDIO, gasp, Arduino IDE? All of these are different and since ARM is licensed IP, each vendor has the freedom to implement the design to suit their own product.
Personally, I would prefer to see a combo Propeller and FPGA based board, but that is just me.
Actually I was off the mark.
Parallax could have a C and Assembler learning course.
PIC 32MX. Chip is not that important. As long as it's a PIC.:)
Parallax SX is Basic and Assembler. Close. C has gotten very big for MCU's.
Parallax made it's name with PIC's.
First one to have debugger that loaded into memory.
Probably still getting royalties from Microchip.
That's the onboard debugger you see in PIC's and AVR's now.
That had products for just about anything you wanted to do with a PIC.
The cost of the development board and software is hundreds of times the
the cost of the chip.
It's called 'value added'.
Parallax's value is writing software and documentation.
Do these not already exist?
What, like since the 1990's or earlier?
So, if I understand this correctly, you want Parallax to create a PIC32MX based board and the tools for it?
I'm still confused as to what you are looking for.
Jon
'create a PIC32MX based board and the tools for it?'
Pull out there old PIC material and dust it off.
Create a Second System. Busy hands are happy hands.
Parallax's PIC skills are just going to waste.
The way Parallax mastered PIC's is no small feat.
There are 500-900 page manuals about aspects of PIC's.
Parallax already knows that stuff. They have written quite a few themselves.
Just to 'pick up you marbles and go home' for whatever reason is ridiculous.
Time for Parallax to rejoin the PIC world. They could contribute alot.
Wow, some diverse and entertaining claims there
Says who ? and which 'PIC' ? Microchip now owns AVR, AVR32 and numerous ARMs, and PIC already covers a great many cores....
... and yet for all the chanting, you have yet to make a sound commercial case for WHY they should do this ?
This does raise an interesting point - did you notice the recent news about Intel pulling the Plug on a number of their IoT 'experimental' parts ?
One big mistake Intel made there, was to forget who their embedded customers were. They literally lost the skills.
Intel used to be an Embedded MCU leader, and one of their cores still is a very big player in 8b MCU space, but they behaved like a total novice.
Skills only remain skills if they are used, and very little (none?) of what Parallax does in 2017, involves ANY PIC opcodes at all.
However...
Are PICs still being used, at least in a mass amount? Seems to me that they are starting to fade away especially since Microchip bought Atmel. Is that your issue? Maybe you need to direct that energy toward Microchip.
I suspect that Parallax could outlast the PIC line and that the PIC may end up in the board pile with the Edison and Galileo.
Not just you....
I think this is an area Parallax should look closely at.
It is an area where they do have active and growing skills and a growing infrastructure, and FPGA prices continue to deliver more.
There can be multiple possible 'sweet spots', paired with P1 :
* Smallest FPGA/CPLD as Pin-expanders on P1 - get those extra IO everyone claims they need
* Medium FPGA as IQ and COG expanders, - at least one P1 COG, and perhaps custom HW like Quad Counters, Capture, PWM
* Larger FPGA, that can have at least one P2 COG, as P2 market seeding and support.
You mean like this ?
It seems the D2000 is still active, but the real test of such parts, is do they have others coming, and have critical mass ?
eg The core of the D2000 has appeal, and the price is tolerable, but what if someone wanted a 100~200MHz version ?
JMG
'Wow, some diverse and entertaining claims there'
Whatever you say.
'Back to erase-burn-plug-pray-repeat. The Parallax Downloader used a special PIC bond-out chip and was the first tool available that allowed code to be run in RAM'
That is the last thing I'm posting about Parallax from 20 years ago.
That was pre internet. It only exists in old timer's memory.
Manuals and hardware have been tossed a long time ago.
http://forums.parallax.com/discussion/164923/new-boards-in-for-my-project-p1-max10m25#latest
Here is a P1 + FPGA combo test board. On the to do list to assemble. All pins broken out on the FPGA and most on the P1.