Why the Propeller works...
cgracey
Posts: 14,256
Kloss said...
Interesting that someone is brave enough to use a propeller chip in a
professional application.
This chip doesn't even have a datasheet, and it is very likely that
it has never been tested against any of the usual standards for integrated
circuits.
I would not even think about using it for anything other than hobby
projects.
DATASHEET NOW AVAILABLE (30 March 2007):
http://forums.parallax.com/showthread.php?p=641840
Kloss,
I understand what you're saying, and I'd have the same worries, myself.
In this case, though, I know that the Propeller is solid because I designed, debugged,·tuned, and tested·it myself. The only other person involved in the silicon·design was·one other layout engineer. This took eight years of my time, and two years of the layout engineer's time. An excruciating amount of attention went into every aspect of the Propeller's design and testing, and I allowed no compromises.
As for industry-standard tests, we recently hired a company called Nano Measurements·to perform their own environmental and ESD testing on the Propeller. Specifically, the following tests were conducted:
PPOT - Pressure pot, autoclave: 121ºC, 100% RH, 15 PSIG, 336 hours
TMCL·- Tempurature cycle, air, standard ramp: -65ºC to 150ºC, 500 cycles
PREC - Preconditioning (simulates soldering process): 30ºC, 60% RH, 192 hours
HTSL - High tempurature storage life: 150ºC, 1000 hours
ESD HBM - ESD human body model: All pins tested up to +/-8kV
Here are the results:
PPOT - 3 lots of 77 devices (PDIP, LQFP, QFN): 0 failures·(a lot of 77 ensures 3-sigma-quality with one failure)
TMCL - 2 lots of 77 devices and·1 lot of 76: 0 failures
PREC - 3 lots of 77 devices: 0 failures
HTSL - 1 lot of 230 devices (near-even mix of PDIP, LQFP, QFN): 0 failures
ESD HBM - 3 devices each of PDIP, LQFP, QFN:·all I/O pins survived +/-8kV (limit of tester) with VSS and VDD grounded, VSS-to-VDD·zap failed at·+/-3kV or greater
So, no environmental failures, and no testable ESD failures on I/O pins, but with the power supply pins failing at +/-3kV or greater·(which is quite acceptable). All these environmental tests say almost nothing about the quality of the chip, but only that the plastic packaging was good. The ESD tests start to give some metric of the silicon quality, but say nothing of what is far more likely to be a problem - design quality. There are no universal metrics for this, and it can only be understood by actually applying the device. This is the critical life test!
Here's why the Propeller is·high-quality:
The Propeller was an entirely full-custom effort. Every polygon of·the Propeller's·mask artwork was made here at Parallax - we designed our own logic, RAMs, ROMs, PLLs, bandgap references, oscillators,·and even·ESD-proof I/O pads.·All these structures were first fabricated on test chips and thoroughly tested before being applied to the final chip. This resulted in known-good blocks which could be confidently applied to the overall design. Then, the whole chip was fabricated and tested at·many·levels, in order to fix any problems resulting from integration and to fine-tune the clocking system and memory timings. The final chip, which is the only version we've ever sold, is the third iteration of this whole-chip process.
Towards the end of development, we invested in a Micrion FIB machine and a Schlumberger·e-beam prober to diagnose any problems and fine-tune the silicon. While the money to buy these used machines·was only about 0.5% of what they cost new, the time needed to get them running and to learn how to use them was a good six months. Now, we can do our own maintenance work·on them, which is not trivial. These machines·made it possible to see what was actually happening on the silicon (contactless, non-loading, 7Ghz oscilloscope function via an·electron beam)··and to perform modifications (via gallium beam and metalorganic gas)·before having another full chip made. All this was a huge adventure in itself, but invaluable in getting the silicon perfected.
I believe that the Propeller has received more attention to its quality than likely any other microcontroller in production. Big companies would never approach·such a project in the way we did. It wouldn't make any sense to them. They have a formulaic path·they follow which·minimizes design time and·ensures interchangeability of engineers, in exchange for a passable result. This involves RTL hardware descriptions, synthesis, place-and-route, IP blocks,·etc. It's·mainly managed chaos, not bottom-up design. The Propeller is correct by construction, not just given a stamp of approval by some $500k·software tool that managed to close timing on an inefficient·rat's nest of wiring and synthesized gates, that is bound to be big and power-hungry. The other fact is, big companies don't cultivate an environment in which any individual would have·the occasion, let alone reason, to know everything about a design, and then be able to ensure quality throughout. They rely on teams of people, none of whom knows everything about the chip. They also leverage IP developed by yet other people, in order to avoid reinventing the wheel. Even if the IP isn't buggy, it's rarely a perfect fit. If you've ever programmed Windows apps, you know how frustrating it is to be forced to rely on questionable black-box objects to get your application done. It's like trying to build a custom home, but being limited to shopping at·WalMart's Garden Center·for your building materials. The Propeller·comes from the antithesis·of this approach.
We should have a data sheet soon with quite a bit of characterization data in it. I hope it·will give people more confidence about using the chip. I think people's response so far on the forum validates what I've said here about quality, though. It is no accident or windfall that the Propeller is tough and reliable (not to mention low-power). It's very intentional.
I want people to·understand that big companies don't have a monopoly on quality. Many of the ways in which they do things actually undermine quality in the quest for expediency. We took the time to do everything right on the Propeller, and left no stone unturned.
We plan on a very long sales life for this chip, and have no intention of dilluting the concept with many slight variants, for which you'd inevitably be getting end-of-life notices for after a few years. This is good news for customers because they are the ones who·are going to be making investments in programming that will, in sum, dwarf the energy that we spent making the Propeller. We made a platform that is, hopefully, deserving of their coming efforts.
P.S. Kloss, I'm glad that you voiced your skepticism. I'm sure it's shared by others.·I needed this·impetus to explain a few things. Thanks.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chip Gracey
Parallax, Inc.
Post Edited (Chip Gracey (Parallax)) : 3/31/2007 8:16:21 AM GMT
Comments
gives me even more confidence using Parallax products - Chip - it is obvious you all hold this dear to your hearts .. well done - maybe the antithesis - should be the thesis !
Chip - can I use this in a parts and operators manual for my machines - this has Quality all over it....
I am going to make a poster of this also if thats ok with you ? and put it in the workshop .... would
look great for customers and maybe inspire staff .........
Its out of the park .....
Kloss - Like I said the mentioned industrial app was not a Parallax / Propeller issue - I was only stating that I had it out there and my only problem was with a 3rd party 3 phase motor - easily sorted by a filter...
Bleeding edge - gives the edge ! - It has already proved itself and orders for more machines - specifically requesting the same type of control and display are already on my desk - I have never been so busy and excited by a chip in my career.
*************************************************************************
We need this verbatim in PDF form - with a Chip Gracey signature at the end .(no - not signature - Autograph !)
for the moment its here as an attachment ...
Goosebumps anyone ?
Quattro
Post Edited (QuattroRS4) : 2/6/2007 12:54:16 AM GMT
Great Chip.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Regards.
Alberto.
Like anything new with engineers... it takes time .. why live outside the comfort zone.
The Propeller is like the Audi quattro of the 80's it came on the scene it took everybody by surprise
and blew the competitors away for a few years. It dominated the rally scene for years..
The Propeller has an unfair advantage... it is now well priced.. it will be succesfull no doubt.
Ronald Nollet Australia
Close to my heart - had a 'shorty' a few years back - have the new Quattro now RS4 B7 420BHP 4.2L - its new ,its cool and its fast ! - Just like the prop !!!
have attached an image of it for you....
Post Edited (QuattroRS4) : 2/6/2007 4:19:55 AM GMT
I sent a copy of this to Max Maxfield , he writes computer books & writes columns for various trade publications.
Brian
PS. I know most of us are not employees of parallax but I would like to think that we all·are helping carry out the dream !
Post Edited (Brian Beckius) : 2/6/2007 11:52:30 AM GMT
0-80MHZ in no time ( 0-100 KMH how fast Quattro ??)
The Prop accelerates nicely after the wait pin type (various types) instruction... awesome.
It is amaziong how it can sit asleep and wake up and run fast so quick.. .. totally cool.
To have that available in spin and in asm is great ...
Ronald Nollet Australia
The prop is truely outstanding - Like I said the first time in years I have been excited by with a chip !!!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
'Necessity is the mother of invention'
Those who can, do.Those who can’t, teach.
Post Edited (QuattroRS4) : 1/24/2009 2:32:19 AM GMT
Every one should read this post, not only does it say what a quality chip the prop is but what a quality company Parallax is. And frankly what a god damn genius Chip is.
Thank You Chip and the rest of Parallax for all your hard work. Your products are what got me into micro controllers and what keep me motivated.
Lyle
I can tell you that definitely the chip seems sound from a actual 24/7 operational status with full 100% silicon usage.
The last kind of tests really are MTBF under long term use studies.
But, of all the microcontrollers I have used, many of them have little problems and bugs, and thus far the propeller seems to have no problems thus far.
So I would have no reservations is saying use it for industrial applications, embedded systems, toys, and other products.
Andre'
·
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
NerdMaster
For
Life
Sometimes I forget I'm programming a single chip microcontroller...
Instead thinking I'm programming a full retro computer with RAM, VideoCard, SoundCard and so on.
I've never seen a single chip do so much and all from a single package.
If you are designing and building a safety critical system then you are concerned about understanding what could fail and what you are depending on to prevent or tolerate these failures. These are matters of engineering skill and professionalism and when you know you are doing this properly, you get a real sense of pride and achievement.
If you are selling safety critical systems then you are concerned about product liability and how the risks it presents can be off loaded to the providers of the components and tools you are using or to other third parties such as insurers or the publishers of standards. These are legal, commercial and actuarial matters and when you do them properly you tend to make money.
I have the impression that what Chip says belongs to the first set of “realities” and that what Kloss is on about belongs to the second. Both are necessary for progress.
I hope Chip (and Parallax) continue to make sufficient money because the results of his engineering represents a lot of stimulation, opportunity and interest for the rest of us. When making money becomes the primary objective you end up with stuff like Windows…
:-/
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
NerdMaster
For
Life
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Paul Baker
Propeller Applications Engineer
Parallax, Inc.
It was really nice to read that. It's nice to see a company so obviously concerned with things other than just making money. Thanks for not just "slapping something together" and then talking it up.
It's obvious that you have vision and passion. Long may they last.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
I stand on the shoulders of giants
I am getting ready to complete a commercial device that uses the Propeller.· The device is fantastic.· I have some limited environmental testing to complete and could use some more information from you.
My interpretation of the temperature testing "TMCL·- Tempurature cycle, air, standard ramp: -65ºC to 150ºC, 500 cycles" is that the device undergoes thermal cycling and then after that, the device is checked to make sure that it still works (at some given temperature).
I want to know the operational temperature range for the part.· Do you have this information?
Regards,
Joel-
·
http://forums.parallax.com/forums/default.aspx?f=25&m=169486
I have an industrial·application running faultlessly in a cabinet and have measured the cabinet ambient at @ 70 deg C - the prop is the is the 'coolest' I.C on the board .
So commercial Temperature range will not be a bother..
I believe a DataSheet is iminent !
Post Edited (QuattroRS4) : 2/7/2007 3:11:45 AM GMT
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Paul Baker
Propeller Applications Engineer
Parallax, Inc.
Like Paul said, it will work across quite a temperature range. I hope we'll be able to quantify the range soon.
As for your question about TMCL testing, you're right. The device is cooled and heated for 500 cycles·and then, presumably at room tempurature, it is re-tested.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chip Gracey
Parallax, Inc.
Post Edited (QuattroRS4) : 2/9/2007 6:25:19 AM GMT
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chip Gracey
Parallax, Inc.
Post Edited (Chip Gracey (Parallax)) : 2/8/2007 7:16:14 PM GMT
EDIT: Rereading old post and noticed I forgot to add a P.C.B cmpleted image after test .. so its there now ..
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
'Necessity is the mother of invention'
Post Edited (QuattroRS4) : 8/2/2007 10:48:12 PM GMT
Brian
Well 5 propeller based machines in and running with each having processed > 10 million products (1.4 to 2million/day each ) - and not as much as a glitch - have videos and pics to upload soon.
Also have BS2's in an even harsher environment - have been running 24/7 since build.
With regards to the prop based machines - all the datasheets in the world do not or could not point to absolute suitability for any given application - what generally happens is that they are a guide and implementation and references to actual 'real world' applications point towards suitability. Prior to installation the customer was aware that there was not any real reference sites or comprehensive data available for the prop - however based on previous projects I had successfully completed for them and prior to Chips 'Why the Propeller Works' document - they were more than willing to install - that position was further enhanced when they read the document and after the successful trial period had expired with Zero issues.- these people are the leaders worldwide in what they do and they like to be first ... I can't justify my stand point on this one any more but feel that time will prove this to be a good decision. I am also grateful to the customer to allow this to happen - I have since moved on to yet more Propeller based applications.