Dual In-line Adaptor Board?

XCore Project reviews, ideas, videos and proposals.
nagmier
Member++
Posts: 21
Joined: Fri Dec 11, 2009 3:42 pm
Contact:

Post by nagmier »

I'll throw my vote in for the XTAG2 connection, for people that have the xk-1 kits could then reuse the xtag2 there...


nisma
Active Member
Posts: 53
Joined: Sun Dec 13, 2009 5:39 pm

Post by nisma »

Just a correction to some post read here, the 4pin ports (xlink) need 10 pins, not just 8.
nisma
Active Member
Posts: 53
Joined: Sun Dec 13, 2009 5:39 pm

Post by nisma »

Hi Leon, do you plan to make the 40 dip board or not ? If not, i do it.
Until you´r respond, i´m on hold for going furhter.
User avatar
The_RB
Member
Posts: 15
Joined: Tue Dec 15, 2009 7:03 pm

Post by The_RB »

I think it is critical to get the "cost per consumable unit" (yes I just coined the term) down as low as possible as soon as possible. In my opinion that would really put XMOS in the hands of users who will start buying and keep buying.

In it's present form XMOS seems like a nice package with neat development boards, and it's obvious people can program them to do really neat stuff. But this is firmly in the "experimenting" market. Good for University use and maybe product development.

But I'm coming from the other direction, in that I like to MAKE things. If I want to make some nice pieces of test equipment, or some VGA output cards, or a small run of commercial devices etc then at the moment the best CPCU is the Sparkfun board, but it is still going to cost $50 for every doodad I make. This is a serious limitation for what is really a great value for money $7 IC because to make anything with it, it then becomes a large $50 IC.

I know I'm new here to the XMOS scene, and this might sound presumptuous but I think what would REALLY put XMOS on the map would be a minimal "stamp" type PCB, a CONSUMABLE product, that keeps the CPCU at the lowest possible price.

Something like a $20 stamp (even $15 is possible) that has the SPI ROM, XS1-L1, 1v reg and some programming connector and the whole thing runs from 3.3v. Just the very bare minimum stuff, because they are going to get used up.

Maybe it's just because I come from a PIC background where I use up a LOT of PICs every year :D but I'm sure a product like that would sell in the big quantities and more importantly bring a ton of new users who will eventually buy the larger XMOS chips etc etc. It would be worth XMOS company themselves producing these, and even selling at cost (or subsidising) and picking up all those new users.

Now as a designer if I design a PCB (ie commercial product etc) using the XS1-L1 then it still looks like 5 to 7 chips needed, SMD placement costs, a number of fiddly SMD discretes etc which really lowers the attractiveness of designing with this IC. It maybe easier to stick a dsPIC in there and get a simple easy solution albeit with reduced performance. So likewise a low CPCU stamp type PCB would suit small product manufacturing runs, and be used as a "module" in the same way you would buy RF modules or GPS modules etc and incorporate them in small run commercial products.

I hope this hasn't offended anyone, especially Leon Heller who has already started a "stamp" type project in the projects section. But IMO the best solution for the XMOS company (and the users!) would be to make a XS1-L1 based low cost module, in significant quantity to get the price right down very quickly.

Then XMOS would go from "fantastic to play with" to "fantastic to use".
nisma
Active Member
Posts: 53
Joined: Sun Dec 13, 2009 5:39 pm

Post by nisma »

Please check out (and others) the attachement on the last post on page 3
Heater
Respected Member
Posts: 296
Joined: Thu Dec 10, 2009 10:33 pm

Post by Heater »

Yes indeed.

It really irks me to have to buy a 100 dollar board to use a seven dollar chip. I do so much miss the old days of DIP packages.
User avatar
jonathan
Respected Member
Posts: 377
Joined: Thu Dec 10, 2009 6:07 pm
Contact:

Post by jonathan »

The cost is pretty much just determined by quantity, right?

What volume do you think this low-cost STAMP board will sell in? It doesn't matter who makes it as far as I can see... It just matters that someone makes it, and makes it cheaply.
Image
User avatar
The_RB
Member
Posts: 15
Joined: Tue Dec 15, 2009 7:03 pm

Post by The_RB »

I think in $20 for singles or $15 for 3 (say) it's going to sell real well. That puts it squarely in the consumable range where a $30 module would be out of that range.

That price is quite do-able in 500 quantity or even a smaller quantity but won't leave much room for profit and/or retail markup. There are probably lots of us who could design and produce this in 500 type quantities, Leon could, Sparkfun, myself etc but everyone who is not the XMOS company will want room for wholesale and retail margins.

So my suggestion was mainly aimed at the mother company. They may be in the best position to design and produce a low cost consumable module in decent quantities, and they also have the most to gain from selling a TON of these things at minimal profit or even at cost or subsidised.
User avatar
Omer
Member++
Posts: 16
Joined: Thu Dec 10, 2009 6:11 pm

Post by Omer »

Thought I'd chime in with my recently uploaded 'stamp' design ;)

Image

The schematic and the BOM can be found at the project page

I will probably have some time to work on it in the next few weeks so I'll update the project page then.

If you spot any errors please let me know!


Cheers,
Omer.
User avatar
TonyD
XCore Addict
Posts: 234
Joined: Thu Dec 10, 2009 11:11 pm
Location: Newcastle, UK
Contact:

Post by TonyD »

Omer wrote:Thought I'd chime in with my recently uploaded 'stamp' design ;)
....
I will probably have some time to work on it in the next few weeks so I'll update the project page then.

If you spot any errors please let me know!
@Omer, Thanks for sharing.

Should U4 LP3879 pin 8 SHDN be tied to the INPUT pin 4 and instead of the OUTPUT pin 5 ?

also have you considered taking the RESET signal to one of the 40-pins?
Post Reply