Rabbit 3400 CPU module

Yes, providing there's not too much data moving around. Stick to low baud rates or fewer com ports running simultaneously because the Rabbit doesn't include much in the way of FIFOs on the UARTs.

Dynamic C has cooperative multitasking features built in which work fine for many people. If you're looking for POSIX threads you're out of luck.

Kelly

Thank you for your reply.

Does Dynamic C has a simulator so that I can try writing code for Rabbit

3400 before I purchase the board? (Does Dynamic C avalible for free trial download?) Does the Rabbit 3400 CPU module is the only one with analog input?

Not that I know of - you might drop an email to the sales folks at Rabbit and find out for sure.

Not that I know of.

Yes.

There is a second C compiler for the Rabbit chip from Softools. It's widely considered to generate faster/smaller code than Dynamic C and also supports ANSI C, separate compilation of files, and the other things one generally expects of a contemporary software IDE. There's even a free trial version.

Kelly

You may wish to look at

where you will find a very good deal for a development system, complete with Dynamic C, and a free copy of the new book on embedded design with the Rabbit 3000, for a total of US$99.00

Bill

Great deal - I always like buying dev kits when the company loses money on the sale!

Kelly

I wonder why they aren't providing the 3700 rather than the 3720 in a development kit. Usually, the dev kits include the option with the most RAM, in my experience.

Jon

Oh, and Kelly... they sell the 3720 at $55 in singles. Why would you imagine they are losing money?

Jon

Software, cables, wall wart(s), and the book. The best part of the deal is that I'll now have an up-to-date copy of Dynamic C. (Whoohoo - new bugs to discover and work around!)

The most egregious dev kit I ever bought was from Zilog - 3 wall warts, a 5-port 10/100 hub, at least two Cat-5 cables, and the usual RS-232 cables. I've received a ton of use from the various pack-ins although I can't I played with their board for more than a few hours.

Kelly

The dev kit includes the wall wart, the Dynamic C compiler, a development board into which to plug the RCM3720 card, a wall chart of info on the CPU...

I believe the module price is just that -- the price for the bare module.

Bill

Yes, that was my understanding. I also note that the $99 price appears to be gone. But my logic for asking the question follows this line of reasoning:

(1) Products, including development kits, sell in qty-1 pricing with a factor of about 4X the actual BOM costs.

(2) If the 3720 is selling (not on special, but normally) in qty-1 for $55, I tend to imagine that the cost to put it into the shipping box is in the $14 range. Certainly, no more than $20.

(3) I wasn't able to find the prototyping board separately, but lets grant it the same cost -- somewhere from $14 to $20.

(4) The software's manufacturing cost is nil. There may be some supposed lost opportunity in including it, but it's not a hard cost item. Further, if a developer decides to use this system in a product, then the profits will arrive from the sales of the modules when more than a few of them are involved (which, I'm sure, is Rabbit Semi's hope -- no one likes the idea of having to support lots and lots of one-off customers.) Let's call the CD cost as $2.

(5) The AC adapter is close to nil. Being very gracious, call it $4.

(6) The serial cable is also close to nil. Neck and neck with the AC adapter. Frankly, I'd be surprised if it is over $2 to Rabbit. But let's call it $4, being the gracious folks we are.

Total? $20*2+4*2+2 = $50 cost. Even if I am off by a factor of two on all these (and I'm sure I'm not that far off) then we aren't talking about losing money.

The short of this is that I don't believe they lose money every time they sell the kit, even at $99. Of course, I completely understand their desire for $199 -- as that is about 5X what I believe is their actual cost and that is about what one tends to price at, ignoring knowing more about the business (which I cannot pretend to know.) And I am willing to accept that $99 is probably good pricing worth serious consideration, if you need their development kit and their software, too.

By the way, I didn't notice the chart being mentioned (though I can now see something that looks like a fold-over in the zoomed up picture of the kit) and I haven't talked about the reference book they are tossing in or other printed docs. I'm sure these account for a few dollars, as well. But nothing to cause this to become a loss.

I remain unconvinced that they are losing money selling these at $99. But I do suspect it is a nice deal for those needing exactly these features and intending to buy, anyway.

Jon

Whoops! I was looking NOT at the promotion page. My bad.

Jon

As may be, and Kelly's assumption of a loss was likely erroneous. But so what? Why would we wish anyone to sustain a loss?

More relevant is that their development bundles normally list for anywhere from $130 on up, with the lowest being for older Rabbit 2000 products. The product at the URL I specified has a normal list of $199, which appears to be consistent with their other pricing. It has been offered on that URL at $99, for some time now, and more recently, they have offered the book as a free bonus.

I sincerely hope they do not lose money. I'm happy to get a good deal, but if the vendor bleeds for my benefit, it is not, in the long run, likely to benefit me -- I need a supplier whose long-term health is not uncertain.

It is indeed a nice deal, and if it makes some modest profit for them, so much the better. I'm a great fan of profitability, for developers, suppliers, and others. ;)

Bill

And there, we can agree. I just didn't see why anyone should imagine they were cutting their nose off on this one. Some manufacturers really do, you know, lose money with the idea of making something very nice and cheap enough so that an engineer might just buy it or get it 'on spec.' Like an impulse item, really -- speculating that it might be useful in the not-too-distant future. This is how I picked up a system for the Mot. '908's, for example -- it was exactly free. No question here that Motorola lost money. And the world is confusing enough that it is helpful to avoid going around adding to it by conflating the cases where some manufacturers really DO lose money with those cases where they don't.

Jon

Oh, absolutely! If not for profit, why are we doing all this work? ;)

I'm in favor of low-margin introductory items, but as I said before, I want the vendor to remain healthy, else why bother to learn to use his stuff?

Bill

Hmm? I do it because I love it. I'd do it if I weren't paid for it, though probably less of it.

What can I say? I want more choices than less -- at least, that is, if they are offering something I may need in my short lifetime, anyway. But I want them getting rich off of my customers, not off of me!

More to the point, I suppose, is this: They are the ones who have the greater benefit if I help sell their product by designing it into a product. Since they reap the upside of their business, I think it is more their capital that should be risked in the first place. And less, mine. (If they want to cut me into a percentage of their profits, I'll take up more risk.) Or, in other words, I understand why companies do give away excellent development tools at zero or very reduced cost. It can make good business sense.

But focusing back on why I said anything at all, I just prefer not to conflate what *is* clearly a real loss-leader with what isn't.

Jon

The way I look at it, I paid them $99 for a compiler upgrade and couple of software modules. They are shipping me some additional hardware that I don't care much about.

I don't worry too much about the why's of product pricing anymore. The marketeers and business wonks are supposed to be professionals. If they underprice their product and go out of business, it'll just make them smarter the next time around. Trust Darwin.

Kelly

There is that! ;)

Jon