Talk-Polywell.org

I thought this was an interesting leader for a discussion here.

http://www.bbc.co.uk/news/education-14949538

ladajo wrote:"In China, you see children going into school at 6.30am and being there until 8 or 9pm, concentrating on science, technology and maths. And you have to ask yourself, would European children do that?

1. I wasn't aware that China is known for innovation. It seems more likely to me that this sort of grim study system kills the joy of learning and suppresses unstructured time that really allows the generation of innovative ideas.

2. If the system is so hyper-competitive that people have to work like this from cradle to grave (and can't even really afford to have children) in order for nations to survive, we need a new system.

I agree with your take on the Chinese bit. I also thought similar things. Also to be considered though is that the Chinese export and subsequent re-patriation of under-grad and grad students has been on a steady climb for a while. They are no fools to that end and seek continual expansion of the education base they function from. They have even expanded embedding and recovery of technical and engineering graduates to foreign companies to even further their experience and knowledge before bringing them home to China. Systematic and somewhat scary.

Liberty is the greatest promoter of innovation.

It sounds like they have the same problem the former USSR had. They have to follow the plan.

Here is something being done in the US by a bunch of old guys with time on their hands who just don't know when they have been beaten. Their first chip design was about 20 or 25 years ago (a gate array) and never caught on.

http://www.ecnmag.com/Blogs/2011/09/M-S ... rimenters/

In fact nothing they have done has really caught on.

Maybe this time.

China hard landing?

http://www.zerohedge.com/news/here-come ... ion-slides

http://www.zerohedge.com/news/prominent ... and-800-sp

The Chinese government has a well founded fear on managing the Yuan. Whatever happens, it will not be easy or probably go well for them.

MSimon, if you haven't, you should check out the Propellor from Parallax.

TDPerk wrote:MSimon, if you haven't, you should check out the Propellor from Parallax.

I have. It is a kludge. A funky kludge. It has been around for a while. It is useful. It is in no way elegant or fast.

Because you think it is a kludge, I suspect you mistake it's purpose. It is carefully designed to provide perfectly determinate behavior and real time processing without classic interrupts, in the context of inexpensive uP use by the hobbyist clientele they grew their business by serving. It does that very well with 160MIPs for $8. The GA144 is $20, and the eval board for it is more than twice the cost of the equiv parallax product. An array of programming languages is avail for the propeller,where I have only seen Forth for the GA144.
I have heard and see the point that every person's Forth is a kludge.
The effective buyin for the GA144, since it has a min buy of 10 is about $700. The propeller buy in can start @ $20.
I kind of have the feeling you are accusing an apple of not being a banana.

TDPerk wrote:Because you think it is a kludge, I suspect you mistake it's purpose. It is carefully designed to provide perfectly determinate behavior and real time processing without classic interrupts, in the context of inexpensive uP use by the hobbyist clientele they grew their business by serving. It does that very well with 160MIPs for $8. The GA144 is $20, and the eval board for it is more than twice the cost of the equiv parallax product. An array of programming languages is avail for the propeller,where I have only seen Forth for the GA144.
I have heard and see the point that every person's Forth is a kludge.
The effective buyin for the GA144, since it has a min buy of 10 is about $700. The propeller buy in can start @ $20.
I kind of have the feeling you are accusing an apple of not being a banana.

You are correct about the buy in at this time (actually it is closer to $400 but that is a mere detail - get a gang together and divide up the cost). The GA144 uses no interrupts and is totally deterministic. Why use interrupts when response time to a pin level change is on the order of 20 ns or better? If you want to do a volume buy the cost is rather good. Plus they will do a custom chip for you if you like.

If the Propeller suits your needs fine.

The intended market is cell phones and high end controllers. No one cares about programming at that level. What is desired is low cost and low power (esp for cell phones).

If you want to do some work with it I can probably get you an experimenters set up for about $30 or $40. Let me know. This offer is for you personally. What they are looking for at this time are people who can do something relatively quickly so they can get publicity or orders.

I expect pricing will improve greatly once they go into volume production.

They are sending me an evaluation board so I can write up my experiences I will know the chip better in a month or so.

MSimon wrote:What they are looking for at this time are people who can do something relatively quickly so they can get publicity or orders.

Right now I'm sure I'm not that guy. I've got enough irons in the fire and on the back burner, and a few hung on the wall.

(Before I croak, I'm gonna try that axial opposed chamber pulsejet. Lockwood, Hiller, eat your heart out.)

My point was the board plus the chip is north of $650.00. I'd happily pay $50.00 for a chip on a bare bones carrier board, just to have it if I got a chance to work on it sometime.

But not 20 apiece for a minimum of ten, with w/a $450.00 board. Pfft!

I will keep an eye on them, and downloaded their Forth manual today.

TDPerk wrote:Because you think it is a kludge, I suspect you mistake it's purpose. It is carefully designed to provide perfectly determinate behavior and real time processing without classic interrupts, in the context of inexpensive uP use by the hobbyist clientele they grew their business by serving. It does that very well with 160MIPs for $8. The GA144 is $20, and the eval board for it is more than twice the cost of the equiv parallax product. An array of programming languages is avail for the propeller,where I have only seen Forth for the GA144.
I have heard and see the point that every person's Forth is a kludge.
The effective buyin for the GA144, since it has a min buy of 10 is about $700. The propeller buy in can start @ $20.
I kind of have the feeling you are accusing an apple of not being a banana.

I use pic chips. They are cheap. Hate their opcode and architecture, but you can't beat the price. I've been looking at TI's stuff, (especially their zigbee rf/processor combos) but I'm getting tired of having to learn yet another processor's opcode and architecture. Seems like everyone wants to do everything differently from everyone else. Yuck.

MSimon wrote:

TDPerk wrote:Because you think it is a kludge, I suspect you mistake it's purpose. It is carefully designed to provide perfectly determinate behavior and real time processing without classic interrupts, in the context of inexpensive uP use by the hobbyist clientele they grew their business by serving. It does that very well with 160MIPs for $8. The GA144 is $20, and the eval board for it is more than twice the cost of the equiv parallax product. An array of programming languages is avail for the propeller,where I have only seen Forth for the GA144.
I have heard and see the point that every person's Forth is a kludge.
The effective buyin for the GA144, since it has a min buy of 10 is about $700. The propeller buy in can start @ $20.
I kind of have the feeling you are accusing an apple of not being a banana.

You are correct about the buy in at this time (actually it is closer to $400 but that is a mere detail - get a gang together and divide up the cost). The GA144 uses no interrupts and is totally deterministic. Why use interrupts when response time to a pin level change is on the order of 20 ns or better? If you want to do a volume buy the cost is rather good. Plus they will do a custom chip for you if you like.

If the Propeller suits your needs fine.

The intended market is cell phones and high end controllers. No one cares about programming at that level. What is desired is low cost and low power (esp for cell phones).

If you want to do some work with it I can probably get you an experimenters set up for about $30 or $40. Let me know. This offer is for you personally. What they are looking for at this time are people who can do something relatively quickly so they can get publicity or orders.

That lets me out. My only idea is going to require a long period of development, and that is WHEN I can get a chance to mess with it.

MSimon wrote: I expect pricing will improve greatly once they go into volume production.

They are sending me an evaluation board so I can write up my experiences I will know the chip better in a month or so.

Will be following your efforts with interest.

TDPerk wrote:

MSimon wrote:What they are looking for at this time are people who can do something relatively quickly so they can get publicity or orders.

Right now I'm sure I'm not that guy. I've got enough irons in the fire and on the back burner, and a few hung on the wall.

(Before I croak, I'm gonna try that axial opposed chamber pulsejet. Lockwood, Hiller, eat your heart out.)

Did you happen to see that Boeing concept video? Their design looks like your pulse jet rotated about it's axis.

http://youtu.be/chmmUF9fPSE

If the numbers work out, this ought to be a pretty good system.

They have have the flow pattern all wrong in a Kadenacy effect engine (the red arrows), and their augmenters are drastically of the wrong cross section and are stupendously oversized. I suspect the video is the result of an artist's impression had on a first description without further engineering input.

Here's the big idea I may or may not get to before I croak.

To improve internal airflow to make the best use of available ram pressure, a Kadenacy effect engine so far will have around twice to twenty times the cross sectional area of the cross section of it's intake, and its operating cycle will be long compared to what I think is possible.

By having two chambers fire alternately into each other (with the entire flow momentum of the forward facing rear chamber compressing the charge in the forward chamber and a small fraction of the flow of the larger forward but rear facing chamber compressing the rear chamber charge, and the rest of the forward chamber's flow used for thrust) I hope both to create more continuous flow in the intake, have a larger proportion of the shrouded volume and cross sectional area employed for creating thrust, have a higher operating pressure ratio at any given operating condition than the usual valveless pulsejet, and have a more fixed and higher operating frequency.

The thrust specific fuel consumption of the Lockwood's suck at at best 1.1 or so pounds of fuel per pound of thrust per hour, and they are better than most. Even with inexpensive construction and very low maintenance and high reliability, that's quite a penalty to design around.

Kadenacy Effect

Valveless Pulsejet