Another point that has not been brought up... "Parallax Could not sell an SX Replacement", because we do not own the IP (Intellectual Property) of the SX. To sell a verbatim replacement we would need to own the Silicon.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ Beau Schwabe
IC Layout Engineer
Parallax, Inc.
I don't understand the insistence that Parallax should support what it doesn't have or sell. Why not hammer Microchip and others to provide better, more Parallax-like support for their small, cheap chips?....
Gee. Aren't we talking about the LONG TERM? Parallax has made it clear that the SX is going to be around for quite a while. Ten years has been mentioned. (I forget by who.) What if it is ONLY five years? That is still long term in this biz. Although I do not know Parallax's business model, is it beyond imagination that they may come up with a NEW chip that we will all love and with which they can make money?
I think they will almost have to offer a new mpu in order to ensure long-term survival.
JonnyMac said...
I don't understand the insistence that Parallax should support what it doesn't have or sell. Why not hammer Microchip and others to provide better, more Parallax-like support for their small, cheap chips?....
Because microchip makes their money from selling lots of chips to a few people who can succesfully use them.·Actually, there are lots of people using them, but microchip only·really cares about those people who buy lots of them.
I really think that at some point in the future, Parallax should look into building a small fast low cost micro. Not the SX, but SX like.· I think parallax could compete with a low end fast 32 bit micro (single cog).·Until then,·those of us who need the speed may need to look elsewhere (There are ~100Mips 8051's out there @ a few $ea).
The SX is not the greatest micro out there,·but it is a great all around processor.
-Dan
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"A saint-like quantity of patience is a help, if this is unavailable, a salty vocabulary works nearly as well." - A. S. Weaver
The problem with Parallax creating their own new low cost micro is that the field is highly competitive and filled with major league players like Atmel with their AVR series, all the ARM vendors, Freescale with their innumerable Coldfire MCU variants, Microchip with their multiple chip lines, and then there's Zilog with its Z80 line and Sil Labs with their 8051x chips.
Worse these folks have new product lines every 18-24 months and fat advertising budgets to get the word out.
Doesn't make sense for Parallax to play in that hotly competitive field.
As far as a single chip Cog goes, it's not a good choice. Alone it doesn't compare well with ARM and the other 32bit processors out there. The Cog's strength is that there are 8 of them in a Prop and you don't need a thousand page manual to understand them.
"Until then, those of us who need the speed may need to look elsewhere (There are ~100Mips 8051's out there @ a few $ea)."
A 100-pin TQFP? May as well use a TMS320series DSP. Or burn the functionality into a 144-pin FPGA. I mean there are all sorts of solutions for speed. But none of them come in a little plastic DIP for $3. And none of them use a fabulous in-circuit development tool like the SX-Key that costs less than $50. And none of them are designed to exploit the virtual peripheral paradigm.
Meanwhile, if all you need to do is generate a 1kHz squarewave, two eight-cent transistors in an AMV config will do the trick. Or if a soldering iron puts you off, use a PIC12F508... $0.56 single piece price.
My point is this: Until you've used an SX for something that only an SX can do, I don't think you can relate to the profound loss that the EOL represents. I've got a collection of IP tied to a specific hardware platform, programming paradigm, and physical size that is now worth squat. A several year supply of chips helps very little. I can't design new products with a discontinued chip. It isn't ethical to my clients.
I'm certainly not angry with Parallax.. Quite the contrary... They are true public benefactors. If it weren't for Parallax the SX would already be gone. And we wouldn't have the SX-Key. The SX-Key they floated me at the 1998 Embedded Systems conference was a work of benevolence rarely ever matched in my experience. And my designs have employed thousands of the chips, not out of quid pro quo but because of the absolute superiority of the product.
I liked the comparison someone made to the Cash for Clunkers program...crushing to death perfectly serviceable cars. This saves energy and resources how?
Either lawyers or the government must be involved in the Ubicom decision. No one else would be this stupid.
Post Edited (David Jensen) : 9/22/2009 4:25:52 PM GMT
16-bit PIC24 or dsPIC devices should outperform the SX chips in most applications. They start at just $2.09 (quantity price) for the 18 pin PIC24HJ12GP201.
Leon
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM
Suzuki SV1000S motorcycle
Nah . . if you've been successful with Parallax and enjoyed the support, applications and hardware then consider staying with our microcontrollers.
David, consider a port to the Propeller. We'd welcome you as a customer the same way we did in 1998. I was at that Embedded Systems Conference 11 years ago and remember the SX introduction and our display, demos, etc. We can give you tools again today to get you started - just drop me a message.
Maybe this has already been asked, but has Parallax made an attempt to purchase the SX-related IP from Ubicom? Something tells me it's not quite that simple...
waltc said...
Worse these folks have new product lines every 18-24 months and fat advertising budgets to get the word out.
Actually, you hit the nail on the head. I want a chip that will be around for 20 to 30 years, not something that is released, and then vanishes in a couple of years. All of the other guys you mentioned are into big sales, in fast evolving products. That means that the markets come and go with the tide, and lots of the chips do the same.
This is why the SX has been a good seller for Parallax. If Ubicom didn't want to walk away, Parallax could keep selling these things forever.
Now that parallax has the Prop Chip and they own the IP, they could rip one cog out and build a chip around that. The language tools could remain the same or very similar (No reinventing the wheel), and the chip·should be·easy to design (because they have a good template).
Again, the Prop looks like a really really nice chip, but it's overkill for everything I do.·Run one cog·at 100 mips, and they'd do everything I need them to do.
-Dan
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"A saint-like quantity of patience is a help, if this is unavailable, a salty vocabulary works nearly as well." - A. S. Weaver
I like that idea, but I gather that is not too plausible..... the flash program requirements apparently are not compatible with the Propeller silicon process, so we stilll would need an external EE to hold the program.
But your concept is great.......... IF we could get a single cog running at some decent speed, say 50-plus MIPs, then with the COG assembly multi-tasking operating system that is soon to be available, we could run a dozen or so independent tasks in that COG, and out perform the SX. And have multi-tasking SX/B to boot!
Man, I'd beat THAT drum!...... solve all my problems.
Ken, that really was a fabulous demo at Embedded Systems 11 years ago! I can still remember how awed I was. And the amazement didn't stop when I got home. I feel confident in saying that never has more magic been packaged in a smaller volume than with the SX-Key. And its synergy with the SX processor has no equal to this day.
Speaking of great demos, the Propeller VGA generator using only seven external resistors is VERY impressive. I was compelled to buy a Propeller Proto Board just to witness it first hand. (Don Lancaster, are you out there...?)
Perhaps the simultaneous mastery of Spin, a new 32-bit assembly language, Cog Think, Cog Sync, and a new (but fabulous-looking!) IDE has been a wall too tall up to now. Spin alone doesn't have the needed throughput for my aps. A single cog doesn't have sufficient code or data space for them. So I feel like I have to conquer everything at once. Or nothing at all. And there I sit.
But you are persuasive... Almost thou persuadest me to be a Propper!
@David - Have you looked at Bill Henning's LMM (Large Memory Model)? This is a special tiny interpreter that reads assembly instructions from HUB memory into a cog for execution one at a time. It runs slower than straight assembly, but much faster than Spin and has all 32K of the Hub memory to hold the program. It's used as the basis for the several C compilers (Catalina and ImageCraft's).
I like that idea, but I gather that is not too plausible..... the flash program requirements apparently are not compatible with the Propeller silicon process, so we stilll would need an external EE to hold the program.
But your concept is great.......... IF we could get a single cog running at some decent speed, say 50-plus MIPs, then with the COG assembly multi-tasking operating system that is soon to be available, we could run a dozen or so independent tasks in that COG, and out perform the SX. And have multi-tasking SX/B to boot!
Man, I'd beat THAT drum!...... solve all my problems.
Cheers,
Peter (pjv)
I'm just guessing here:·The first prop·was probably done·in·VHDL on an FPGA, and the design migrated to needing a EEPROM from that...
I'm guessing that a good portion of the logic could be ported to any process.
Just think of it,·A low power screaming fast·32 bit processor,·with long term supply,·and all the great technical support and learning tools parallax has to offer. A processor designed for Assembly, not C!·· AND a whole library of Objects (from the prop) that could be utilized as examples.
Also, just think of the basic stamp Parallax could build with that thing! [noparse];)[/noparse]·
I'm drooling just thinking about it!
-Dan
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"A saint-like quantity of patience is a help, if this is unavailable, a salty vocabulary works nearly as well." - A. S. Weaver
Mike Green said...
@David - Have you looked at Bill Henning's LMM (Large Memory Model)? This is a special tiny interpreter that reads assembly instructions from HUB memory into a cog for execution one at a time. It runs slower than straight assembly, but much faster than Spin and has all 32K of the Hub memory to hold the program. It's used as the basis for the several C compilers (Catalina and ImageCraft's).
Mike,
I looked at the information on the Large Memory Model, and it was very interesting.· However, it seems like there is a significant loss in efficiency due to executing one instruction at a time.· The cycle overhead is from 4x to 8x depending on how much the loop is unrolled.· Also, all of the jump instructions have to be fully implemented by the interpreter.
Has anybody looked into an overlay scheme?· This method works by having a static portion of the code and several dynamically-loaded portions.· A new dynamic portion is loaded into memory whenever a function is called in that portion.· If done properly, this can efficiently execute large programs.· Of course, if it is not done correctly it can be very inefficient.· I recall an overlay program I once wrote on a PDP-15 that used DECtape as the mass storage device.· It was very entertaining watching the DECtape spin back and forth as the program needed new overlays.
Yes, in fact ImageCraft was looking into an automatic overlay scheme where the compiler would identify loops that could fit into a 64 or 128 instruction overlay area in the cog and generate code for the loop to execute there. A single LMM instruction would load a small block of code following the instruction and execute it. I don't know whether they actually implemented it. One problem is that it really has to be used only for small loops that are executed several times to make up for the overhead in copying the code into the cog ... not a big deal for a good optimizer.
Several suggested LMM interpreters that have been posted have included provisions for this sort of overlay mechanism. Look at some of the threads on the LMM idea from Bill Henning and others.
Properly used, the overhead is <10% - the key is to use FCACHE to load loops.
When I came up with LMM I thought people would use FCACHE extensively to cache loops - it makes loops of up to 128 instructions execute at full cog speed (after loading it).
Take a look at the pre-release version of my LAS assembler; it hides most of the differences of LMM. Fair warnings: it may still have bugs; but I do know that it generates correct code for the test file (I worked on it until the output was the same, bit for bit, with the Propeller Tool)
I will also have an LMM Basic compiler, but that won't be available until sometime next year.
You may wish to look at the Basic compiler Bean (Terry Hitt) and Jonny Mac (Jon Williams) are working on - it is VERY similar to SX/B.
Hope this helps,
Bill
Dave Hein said...
Mike Green said...
@David - Have you looked at Bill Henning's LMM (Large Memory Model)? This is a special tiny interpreter that reads assembly instructions from HUB memory into a cog for execution one at a time. It runs slower than straight assembly, but much faster than Spin and has all 32K of the Hub memory to hold the program. It's used as the basis for the several C compilers (Catalina and ImageCraft's).
Mike,
I looked at the information on the Large Memory Model, and it was very interesting. However, it seems like there is a significant loss in efficiency due to executing one instruction at a time. The cycle overhead is from 4x to 8x depending on how much the loop is unrolled. Also, all of the jump instructions have to be fully implemented by the interpreter.
Has anybody looked into an overlay scheme? This method works by having a static portion of the code and several dynamically-loaded portions. A new dynamic portion is loaded into memory whenever a function is called in that portion. If done properly, this can efficiently execute large programs. Of course, if it is not done correctly it can be very inefficient. I recall an overlay program I once wrote on a PDP-15 that used DECtape as the mass storage device. It was very entertaining watching the DECtape spin back and forth as the program needed new overlays.
Dave
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Please use mikronauts _at_ gmail _dot_ com to contact me off-forum, my PM is almost totally full Morpheus & Mem+dual Prop SBC w/ 512KB kit $119.95, 2MB memory IO board kit $89.95, both kits $189.95 www.mikronauts.com - my site 6.250MHz custom Crystals for running Propellers at 100MHz Las - Large model assembler for the Propeller Largos - a feature full nano operating system for the Propeller
Should Parallax sell an SX replacement?· It seems that the answer is yes and no.· As far as low-cost small-size microcontrollers are concerned, the answer is no.· There are a lot of solutions out there, and it is difficult to compete in that arena.
However, the SX is a very interesting solution for the higher end microcontroller applications.· With its high clock rate and its single-cycle exectuion, the SX is/was very competitive against other solutions.· One of the drawbacks·to the SX is its 8-bit architecture.· Other processors have moved on to 16 bits and 32 bits.· I do believe that Parallax should compete in the higher end microcontoller market· -- and the Propeller is the SX replacement!
The Propeller has a lot going for it.· 160 MIPs and 32-bit architecture.· The library of applications continues to increase over time, and more tools are being developed for it.· I've had the Propeller Education Kit for several months, but I haven't done much with it.· However, the more I read about it the more interesting it looks.· I'll definitely use it in one of my next projects.
For those of us who were around back in the PBasic days, or who have worked with the Basic Stamp in non-FIRST related projects, I have a quick question....has anyone been able to find a drop in replacement for the stamp which offers similar functionality at a lower price? Basically, I've developed a small product using the Stamp, and now that we have a working prototype, we're looking for any way to reduce costs. Specifically, I'm using the following features of the stamp:
-Serial I/O
-X-10 Output
-Frequency Output
Any recommendations? I'd even consider maybe splitting the load between two chips(for instance, a separate X-10 control module), if the final cost would be cheaper than the $49 that Digikey wants for each Stamp 2 module...
Post Edited By Moderator (Bean (Hitt Consulting)) : 1/4/2010 10:27:03 PM GMT
People don't use the BASIC Stamp because it is cheap, so don't compete on price alone. To make a similar product viable you'll need to develop an educational program and accessory hardware with lots of free documentation and code examples. That's your best advice! Took us about fifteen years to do that. Save people time and offer excellent support.
As for the technical aspect, how about using the Propeller? You could skip the whole single-threaded design entirely. The language interpreter is already done and the software modules you mentioned exist. All you need is a form factor of a BS2.
This question would be best asked in the BASIC Stamp or Sandbox forums, too, as this thread has strayed beyond the poster's original question.
If you are doing a product with a Stamp why don't you use an OEM design -- the Stamp firmware pre-burned runs about $11 and then you'd need minimal external components (in a finished board), *especially* if you preburn EEPROMs with your Stamp program and don't need the PC/Stamp serial interface.
Depending on volume, it would seem you could have a Stamp 2 on each board for around $13.00 (esp. if you order EEPROMs in bulk from a regular wholesaler).
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ When the going gets weird, the weird turn pro. -- HST
I have 2 SX chips but never bought an SX-key. Time used to go a lot slower so I downloaded the programming spec
and since I had already made a PIC16C54 assembler and burner for all those that I used to use ,I thought one day I'd
get around to upgrading to SX (which I think is a 12x faster compatible version of that very old PIC). But then I used 8051s
and after that I use Propellers. Sad to not have used SX but for me a 2009 day = a 1999 hour so I'll never get
around to using SX.
Zoot is right: use OEM parts if you want to cut your cost and keep the code you have. You're not going to find a work-alike, Stamp-type product that is cheaper AND will run the code you've already tested. You're not saving a lot of money if you have to redevelop your code.
I'm new to Parallax and I was watching one of the Youtube videos on the SX and I started to catch the vision behind it but I want to invest in something that will be here for a long time. I think there needs to be a lot of variety which is why I would like to see an SX replacement.
Chuckz -- well, there are tens of thousands of DIP SX28s available, and I'm presuming lots more will be available when the final orders arrive late this year (but Ken Gracey would have to speak to that). It seems that the surface mount SXes disappeared immediately, most likely because folks with production items stocked up in order to stay in business. If you are more of a hobbyist or one-off developer, that's a different story.
Still it's a shame. I would not look to Parallax for a similar replacement -- there are PICs and AVRs and the like (though not exact replacements, certainly there are oodles of micros in <$4 range that are suitable, and certainly places like AVRfreaks.net offer great support) -- clearly Parallax is migrating to the Propeller, not just for a micro, but as the backbone for many modules. E.g., the new servo controller, which is Prop based.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ When the going gets weird, the weird turn pro. -- HST
Chuckz,
In some ways, the Propeller is the SX replacement. Remember that Parallax has no control over the situation with the SX. It's all tied up in an intellectual property dispute between Microchip and the owner of the SX design. The Propeller is an attempt to make a unique microcontroller based on Parallax's experience in the business of marketing microcontrollers for years. Parallax owns the Propeller. They can develop the market for it secure that no one can take it away in the future.
As Ken and others have mentioned, there's no future in Parallax marketing someone else's low end microcontroller except as a "commodity" component of a larger product like in the BS2pe MoBo where there are two AVRs that act as peripheral processors for the Stamp. It's too expensive for Parallax to support some other microcontroller that has very low margins and is pretty well supported elsewhere.
Although I did skip the SX product from Parallax, and go straight to the Propeller....I do not think Parallax should sell or try to sell a SX replacement.
Here are the reasons. First, the development time would cause most users to move to a new chip any how. Second, the Propeller may be over kill for your project, but if you put some thought into it......it can become a better project, which requires the silicon your are supplying (in this example the Propeller).
If you built a Load Cell for motors, why not go ahead and make it stand alone. It could not only do the load calculations, but also show a real time graph on a display (TV) and also log the information on an SD card. And do most of this on a simple battery configuration.
I know most want a simple project.....make it simple with the possibilities of expansion.
I use the propeller, almost exclusively. I'm trying to get my wife to learn the Basic Stamp, but have a hard time getting back to a linear process. The Propeller will spoil you very fast, and can do just about everything I ever wanted to do.
There are new programming languages being written for it, and Parallax owns the design. I see that as a win/win situation for anyone who helps support the supply line of the Propeller.
If you want to go and learn another language to support a limited processor, I understand. But at a little over 3X the price I would rather have the 454 under the hood, especially because I do not have to put petrol in it.
I'm not knocking other limited processors.....some times a simple solution is just that. But since we know the SX is on the Green Mile, lets leave it there....... As for Parallax, they have done very impressive things for the micro-controller world....and seem to have a pretty good game plan. I figure what ever decision they make, they do for good reason.
People need to remember, even if Parallax got permission to proceed with the SX (or designed their own like model), it's price would change drastically. They would definitely need to increase it's price to offset the added cost of taking the SX on (or development cost....which ever is the case).
In that case, the Propeller would probably be cheaper and do more. Ok.....so a 454 in a moped is a little overkill. There is nothing wrong with that if your brave enough to drive it.
This is truly exciting! I have spent time lobbying Parallax for BASIC on the Prop. (As have a lot of others.) I hope you are well rewarded for your efforts. If it wasn't for you I would never have gained exposure to the great world of the SX! Now, it looks like I will have a whole 'nother world to explore.
I know you realize the need for good documentation. I hope that is in whatever agreement you may have with Parallax. They stand to sell a LOT of Prop-related stuff due to the availability of BASIC. If good BASIC documentation is available, I plan to buy a bunch of it myself.
Comments
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe
IC Layout Engineer
Parallax, Inc.
Gee. Aren't we talking about the LONG TERM? Parallax has made it clear that the SX is going to be around for quite a while. Ten years has been mentioned. (I forget by who.) What if it is ONLY five years? That is still long term in this biz. Although I do not know Parallax's business model, is it beyond imagination that they may come up with a NEW chip that we will all love and with which they can make money?
I think they will almost have to offer a new mpu in order to ensure long-term survival.
--Bill
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
You are what you write.
I really think that at some point in the future, Parallax should look into building a small fast low cost micro. Not the SX, but SX like.· I think parallax could compete with a low end fast 32 bit micro (single cog).·Until then,·those of us who need the speed may need to look elsewhere (There are ~100Mips 8051's out there @ a few $ea).
The SX is not the greatest micro out there,·but it is a great all around processor.
-Dan
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"A saint-like quantity of patience is a help, if this is unavailable, a salty vocabulary works nearly as well." - A. S. Weaver
Worse these folks have new product lines every 18-24 months and fat advertising budgets to get the word out.
Doesn't make sense for Parallax to play in that hotly competitive field.
As far as a single chip Cog goes, it's not a good choice. Alone it doesn't compare well with ARM and the other 32bit processors out there. The Cog's strength is that there are 8 of them in a Prop and you don't need a thousand page manual to understand them.
A 100-pin TQFP? May as well use a TMS320series DSP. Or burn the functionality into a 144-pin FPGA. I mean there are all sorts of solutions for speed. But none of them come in a little plastic DIP for $3. And none of them use a fabulous in-circuit development tool like the SX-Key that costs less than $50. And none of them are designed to exploit the virtual peripheral paradigm.
Meanwhile, if all you need to do is generate a 1kHz squarewave, two eight-cent transistors in an AMV config will do the trick. Or if a soldering iron puts you off, use a PIC12F508... $0.56 single piece price.
My point is this: Until you've used an SX for something that only an SX can do, I don't think you can relate to the profound loss that the EOL represents. I've got a collection of IP tied to a specific hardware platform, programming paradigm, and physical size that is now worth squat. A several year supply of chips helps very little. I can't design new products with a discontinued chip. It isn't ethical to my clients.
I'm certainly not angry with Parallax.. Quite the contrary... They are true public benefactors. If it weren't for Parallax the SX would already be gone. And we wouldn't have the SX-Key. The SX-Key they floated me at the 1998 Embedded Systems conference was a work of benevolence rarely ever matched in my experience. And my designs have employed thousands of the chips, not out of quid pro quo but because of the absolute superiority of the product.
I liked the comparison someone made to the Cash for Clunkers program...crushing to death perfectly serviceable cars. This saves energy and resources how?
Either lawyers or the government must be involved in the Ubicom decision. No one else would be this stupid.
Post Edited (David Jensen) : 9/22/2009 4:25:52 PM GMT
Leon
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM
Suzuki SV1000S motorcycle
Post Edited (Leon) : 9/21/2009 11:53:38 PM GMT
David, consider a port to the Propeller. We'd welcome you as a customer the same way we did in 1998. I was at that Embedded Systems Conference 11 years ago and remember the SX introduction and our display, demos, etc. We can give you tools again today to get you started - just drop me a message.
Sincerely,
Ken Gracey
Parallax Inc.
This is why the SX has been a good seller for Parallax. If Ubicom didn't want to walk away, Parallax could keep selling these things forever.
Now that parallax has the Prop Chip and they own the IP, they could rip one cog out and build a chip around that. The language tools could remain the same or very similar (No reinventing the wheel), and the chip·should be·easy to design (because they have a good template).
Again, the Prop looks like a really really nice chip, but it's overkill for everything I do.·Run one cog·at 100 mips, and they'd do everything I need them to do.
-Dan
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"A saint-like quantity of patience is a help, if this is unavailable, a salty vocabulary works nearly as well." - A. S. Weaver
I like that idea, but I gather that is not too plausible..... the flash program requirements apparently are not compatible with the Propeller silicon process, so we stilll would need an external EE to hold the program.
But your concept is great.......... IF we could get a single cog running at some decent speed, say 50-plus MIPs, then with the COG assembly multi-tasking operating system that is soon to be available, we could run a dozen or so independent tasks in that COG, and out perform the SX. And have multi-tasking SX/B to boot!
Man, I'd beat THAT drum!...... solve all my problems.
Cheers,
Peter (pjv)
Speaking of great demos, the Propeller VGA generator using only seven external resistors is VERY impressive. I was compelled to buy a Propeller Proto Board just to witness it first hand. (Don Lancaster, are you out there...?)
Perhaps the simultaneous mastery of Spin, a new 32-bit assembly language, Cog Think, Cog Sync, and a new (but fabulous-looking!) IDE has been a wall too tall up to now. Spin alone doesn't have the needed throughput for my aps. A single cog doesn't have sufficient code or data space for them. So I feel like I have to conquer everything at once. Or nothing at all. And there I sit.
But you are persuasive... Almost thou persuadest me to be a Propper!
I'm guessing that a good portion of the logic could be ported to any process.
Just think of it,·A low power screaming fast·32 bit processor,·with long term supply,·and all the great technical support and learning tools parallax has to offer. A processor designed for Assembly, not C!·· AND a whole library of Objects (from the prop) that could be utilized as examples.
Also, just think of the basic stamp Parallax could build with that thing! [noparse];)[/noparse]·
I'm drooling just thinking about it!
-Dan
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"A saint-like quantity of patience is a help, if this is unavailable, a salty vocabulary works nearly as well." - A. S. Weaver
I looked at the information on the Large Memory Model, and it was very interesting.· However, it seems like there is a significant loss in efficiency due to executing one instruction at a time.· The cycle overhead is from 4x to 8x depending on how much the loop is unrolled.· Also, all of the jump instructions have to be fully implemented by the interpreter.
Has anybody looked into an overlay scheme?· This method works by having a static portion of the code and several dynamically-loaded portions.· A new dynamic portion is loaded into memory whenever a function is called in that portion.· If done properly, this can efficiently execute large programs.· Of course, if it is not done correctly it can be very inefficient.· I recall an overlay program I once wrote on a PDP-15 that used DECtape as the mass storage device.· It was very entertaining watching the DECtape spin back and forth as the program needed new overlays.
Dave
Several suggested LMM interpreters that have been posted have included provisions for this sort of overlay mechanism. Look at some of the threads on the LMM idea from Bill Henning and others.
Properly used, the overhead is <10% - the key is to use FCACHE to load loops.
When I came up with LMM I thought people would use FCACHE extensively to cache loops - it makes loops of up to 128 instructions execute at full cog speed (after loading it).
Take a look at the pre-release version of my LAS assembler; it hides most of the differences of LMM. Fair warnings: it may still have bugs; but I do know that it generates correct code for the test file (I worked on it until the output was the same, bit for bit, with the Propeller Tool)
I will also have an LMM Basic compiler, but that won't be available until sometime next year.
You may wish to look at the Basic compiler Bean (Terry Hitt) and Jonny Mac (Jon Williams) are working on - it is VERY similar to SX/B.
Hope this helps,
Bill
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Please use mikronauts _at_ gmail _dot_ com to contact me off-forum, my PM is almost totally full
Morpheus & Mem+dual Prop SBC w/ 512KB kit $119.95, 2MB memory IO board kit $89.95, both kits $189.95
www.mikronauts.com - my site 6.250MHz custom Crystals for running Propellers at 100MHz
Las - Large model assembler for the Propeller Largos - a feature full nano operating system for the Propeller
However, the SX is a very interesting solution for the higher end microcontroller applications.· With its high clock rate and its single-cycle exectuion, the SX is/was very competitive against other solutions.· One of the drawbacks·to the SX is its 8-bit architecture.· Other processors have moved on to 16 bits and 32 bits.· I do believe that Parallax should compete in the higher end microcontoller market· -- and the Propeller is the SX replacement!
The Propeller has a lot going for it.· 160 MIPs and 32-bit architecture.· The library of applications continues to increase over time, and more tools are being developed for it.· I've had the Propeller Education Kit for several months, but I haven't done much with it.· However, the more I read about it the more interesting it looks.· I'll definitely use it in one of my next projects.
Dave
-Serial I/O
-X-10 Output
-Frequency Output
Any recommendations? I'd even consider maybe splitting the load between two chips(for instance, a separate X-10 control module), if the final cost would be cheaper than the $49 that Digikey wants for each Stamp 2 module...
Post Edited By Moderator (Bean (Hitt Consulting)) : 1/4/2010 10:27:03 PM GMT
People don't use the BASIC Stamp because it is cheap, so don't compete on price alone. To make a similar product viable you'll need to develop an educational program and accessory hardware with lots of free documentation and code examples. That's your best advice! Took us about fifteen years to do that. Save people time and offer excellent support.
As for the technical aspect, how about using the Propeller? You could skip the whole single-threaded design entirely. The language interpreter is already done and the software modules you mentioned exist. All you need is a form factor of a BS2.
This question would be best asked in the BASIC Stamp or Sandbox forums, too, as this thread has strayed beyond the poster's original question.
Ken Gracey
See this thread for several schematics of a "minimum" Stamp (at bottom right of page): www.parallax.com/Store/Microcontrollers/BASICStampOEM/tabid/135/CategoryID/10/List/0/SortField/0/Level/a/ProductID/21/Default.aspx and this thread for some useful discussion about the same setups: http://forums.parallax.com/showthread.php?p=567989
And this page for OEM Stamp interpreter chips (you certainly would want to build the Stamp circuitry into your board rather then purchasing an OEM board or building DIP versions of Stamps): www.parallax.com/Store/Microcontrollers/BASICStampOEM/tabid/135/List/0/CategoryID/10/Level/a/SortField/0/Default.aspx
Depending on volume, it would seem you could have a Stamp 2 on each board for around $13.00 (esp. if you order EEPROMs in bulk from a regular wholesaler).
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
When the going gets weird, the weird turn pro. -- HST
1uffakind.com/robots/povBitMapBuilder.php
1uffakind.com/robots/resistorLadder.php
Post Edited (Zoot) : 9/28/2009 8:14:16 PM GMT
and since I had already made a PIC16C54 assembler and burner for all those that I used to use ,I thought one day I'd
get around to upgrading to SX (which I think is a 12x faster compatible version of that very old PIC). But then I used 8051s
and after that I use Propellers. Sad to not have used SX but for me a 2009 day = a 1999 hour so I'll never get
around to using SX.
Still it's a shame. I would not look to Parallax for a similar replacement -- there are PICs and AVRs and the like (though not exact replacements, certainly there are oodles of micros in <$4 range that are suitable, and certainly places like AVRfreaks.net offer great support) -- clearly Parallax is migrating to the Propeller, not just for a micro, but as the backbone for many modules. E.g., the new servo controller, which is Prop based.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
When the going gets weird, the weird turn pro. -- HST
1uffakind.com/robots/povBitMapBuilder.php
1uffakind.com/robots/resistorLadder.php
In some ways, the Propeller is the SX replacement. Remember that Parallax has no control over the situation with the SX. It's all tied up in an intellectual property dispute between Microchip and the owner of the SX design. The Propeller is an attempt to make a unique microcontroller based on Parallax's experience in the business of marketing microcontrollers for years. Parallax owns the Propeller. They can develop the market for it secure that no one can take it away in the future.
As Ken and others have mentioned, there's no future in Parallax marketing someone else's low end microcontroller except as a "commodity" component of a larger product like in the BS2pe MoBo where there are two AVRs that act as peripheral processors for the Stamp. It's too expensive for Parallax to support some other microcontroller that has very low margins and is pretty well supported elsewhere.
Although I did skip the SX product from Parallax, and go straight to the Propeller....I do not think Parallax should sell or try to sell a SX replacement.
Here are the reasons. First, the development time would cause most users to move to a new chip any how. Second, the Propeller may be over kill for your project, but if you put some thought into it......it can become a better project, which requires the silicon your are supplying (in this example the Propeller).
If you built a Load Cell for motors, why not go ahead and make it stand alone. It could not only do the load calculations, but also show a real time graph on a display (TV) and also log the information on an SD card. And do most of this on a simple battery configuration.
I know most want a simple project.....make it simple with the possibilities of expansion.
I use the propeller, almost exclusively. I'm trying to get my wife to learn the Basic Stamp, but have a hard time getting back to a linear process. The Propeller will spoil you very fast, and can do just about everything I ever wanted to do.
There are new programming languages being written for it, and Parallax owns the design. I see that as a win/win situation for anyone who helps support the supply line of the Propeller.
If you want to go and learn another language to support a limited processor, I understand. But at a little over 3X the price I would rather have the 454 under the hood, especially because I do not have to put petrol in it.
I'm not knocking other limited processors.....some times a simple solution is just that. But since we know the SX is on the Green Mile, lets leave it there....... As for Parallax, they have done very impressive things for the micro-controller world....and seem to have a pretty good game plan. I figure what ever decision they make, they do for good reason.
People need to remember, even if Parallax got permission to proceed with the SX (or designed their own like model), it's price would change drastically. They would definitely need to increase it's price to offset the added cost of taking the SX on (or development cost....which ever is the case).
In that case, the Propeller would probably be cheaper and do more. Ok.....so a 454 in a moped is a little overkill. There is nothing wrong with that if your brave enough to drive it.
James L
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
James L
Partner/Designer
Lil Brother SMT Assembly Services
Please note: Due to economic conditions the light at the end of the tunnel will be turned off until further notice. Thanks for your understanding.
Well, just click on the link in my signature.
PropBASIC will compile BASIC code directly into PASM code. Multiprocessing is handled by using TASKs.
Bean.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Use BASIC on the Propeller with the speed of assembly language.
PropBASIC thread http://forums.parallax.com/showthread.php?p=867134·
This is truly exciting! I have spent time lobbying Parallax for BASIC on the Prop. (As have a lot of others.) I hope you are well rewarded for your efforts. If it wasn't for you I would never have gained exposure to the great world of the SX! Now, it looks like I will have a whole 'nother world to explore.
I know you realize the need for good documentation. I hope that is in whatever agreement you may have with Parallax. They stand to sell a LOT of Prop-related stuff due to the availability of BASIC. If good BASIC documentation is available, I plan to buy a bunch of it myself.
This is a great New Year's present!
--Bill
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
You are what you write.