Searching \ for '[PIC] 24F and USB' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/microchip/devices.htm?key=pic
Search entire site for: '24F and USB'.

Exact match. Not showing close matches.
PICList Thread
'[PIC] 24F and USB'
2008\11\05@111434 by Alan B. Pearce

face picon face
I am looking to take a quantum leap (and may need some solace in the
process, sorry James Bond) in my understanding, and do a USB based project -
a 'USB Interface Brick' that will be used to test and calibrate a bolometer
instrument to go on a space craft.

The ideal chip seems to be the 24FJ256GB110 family, with USB, a decent chunk
of RAM, the possibility of using external RAM through the PMP, 3 SPI
interfaces, Output Compare modules, and various other bits that look like
they will be useful. Chip cost is not a bother (at most about 6 'bricks'
will be built) so I expect to get the 100 pin chip with largest memory. Size
isn't a problem, and the extra couple of quid a time for the large chip is
way down in the noise of the scheme.


However it would be nice to get a development board to do some knock up
hardware for those in the early stages of the instrument hardware, and the
interface PC software. Looking at the various options that Microchip make
available, there does not seem to be anything that really suits this
purpose. I may have to just go straight to doing my own PCB, but it would be
nice to have something that could be knocked up for experimenting first.

The Explorer 16 has a 100 pin adapter that has an earlier version of the 24F
family on it, without the USB capability (and I cannot see how one would
connect to it if it did have it). The PIC24 Starter kit seems to be oriented
at playing with user interface things (display, capacitive switches, LEDs
etc) and doesn't appear to have a user definable area for building up ones
own interface circuits, but does have the USB interface available.

What have people used for these devices in the past? I appreciate they are
relatively new in the 16 bit family, but figured there might be a board that
would allow me to do something like this.

And as an aside, has anyone used the USB Framework that Microchip provide,
specifically the 2.3 release currently available?
What 'gotchas' etc does one need to be aware of?
Has anyone tried converting it to Linux, or written a Linux driver from
scratch?
Apart from buying a block from Wouter, what does one do about a USB VID/PID?

Any pointers to relevant info appreciated. I have the 'USB in a Nutshell'
pages, and am intending to get Jan Axelsons book all as part of the project,
but other pointers appreciated.

TIA

PS - am away 'till Tuesday, so if I don't reply, please don't think its a
one way communication.

2008\11\05@184703 by Xiaofan Chen

face picon face
On Thu, Nov 6, 2008 at 12:14 AM, Alan B. Pearce
<spam_OUTAlan.B.PearceTakeThisOuTspamstfc.ac.uk> wrote:
> The Explorer 16 has a 100 pin adapter that has an earlier version of the 24F
> family on it, without the USB capability (and I cannot see how one would
> connect to it if it did have it). The PIC24 Starter kit seems to be oriented
> at playing with user interface things (display, capacitive switches, LEDs
> etc) and doesn't appear to have a user definable area for building up ones
> own interface circuits, but does have the USB interface available.

Both are fine choice.
I have the Explorer 16 board, then I have to bought the PICtail+ USB
daughter board and USB PIC24 PIM. I have actually bought USB PIC32
PIM. So Explorer 16 is the most flexible system. But it is also the most
expensive setup.

> And as an aside, has anyone used the USB Framework that Microchip provide,
> specifically the 2.3 release currently available?

It works nicely.

> What 'gotchas' etc does one need to be aware of?
So far it is fine. There are more enhancement needed.

> Has anyone tried converting it to Linux, or written a Linux driver from
> scratch?

What do you mean by converting it to Linux? If you use CDC,
it works under Linux. If you use HID or Custom driver or WinUSB,
it is not difficult to use it under Linux with libusb. If you use the
host function, then it has nothing to do with Linux or Windows.

For developing, Windows is still easier.

> Apart from buying a block from Wouter, what does one do about
> a USB VID/PID?

You can get one free PID from Microchip. I think Wouter's offer is
very reasonably priced.

Xiaofan

2008\11\06@035440 by Jan-Erik Soderholm

face picon face
Xiaofan Chen wrote:

> On Thu, Nov 6, 2008 at 12:14 AM, Alan B. Pearce
> <.....Alan.B.PearceKILLspamspam@spam@stfc.ac.uk> wrote:
>
>> Apart from buying a block from Wouter, what does one do about
>> a USB VID/PID?
>
> ... I think Wouter's offer is very reasonably priced.

Maybe, but this have been on his page for something like
half a year now :

> Pending a discussion with the USB organisation I have
> suspended the sale of USB PIDs


Jan-Erik.

2008\11\06@040042 by Dumitru Stama

picon face
>> Apart from buying a block from Wouter, what does one do about
>> a USB VID/PID?
>
> You can get one free PID from Microchip. I think Wouter's offer is
> very reasonably priced.

No, you can't. He is not selling those anymore from what i saw on his webpage.

2008\11\06@044909 by Wouter van Ooijen

face picon face
>>> Apart from buying a block from Wouter, what does one do about
>>> a USB VID/PID?
>> ... I think Wouter's offer is very reasonably priced.
>
> Maybe, but this have been on his page for something like
> half a year now :
>
>  > Pending a discussion with the USB organisation I have
>  > suspended the sale of USB PIDs

And I must confess that it is not likely to change to 'available',
the usb.org simply ignores me, except for threatening letters when I
was selling PID blocks. AFIAK this is the same for the few other places
that sold PID blocks.

If it is for an open-source project feel free to contact me, otherwise I
am afraid the only option is to buy a VID. When I did so this option was
hidden somewhere on the usb.org website. IIRC the price was $1500 (one
time, no yearly fee).

And you can of course ask usb.org why they don't either
- sell small PID blocks themselves, or
- let someone else do so.

If it is for a product that uses PICs as USB endpoints there is the
option to ask Microchip for a PID. There is some established procedure
for this, I don't know how easy it is. Let us know if this works.

--

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2008\11\06@050414 by Xiaofan Chen

face picon face
On Thu, Nov 6, 2008 at 5:48 PM, Wouter van Ooijen <wouterspamKILLspamvoti.nl> wrote:
> And I must confess that it is not likely to change to 'available',
> the usb.org simply ignores me, except for threatening letters when I
> was selling PID blocks. AFIAK this is the same for the few other places
> that sold PID blocks.
>
> If it is for an open-source project feel free to contact me, otherwise I
> am afraid the only option is to buy a VID. When I did so this option was
> hidden somewhere on the usb.org website. IIRC the price was $1500 (one
> time, no yearly fee).
>
> And you can of course ask usb.org why they don't either
> - sell small PID blocks themselves, or
> - let someone else do so.

I agree with you that it is kind of strange.

Jan Axelson has this article and she talks about
"The Vendor ID Dilemma".
http://www.lvr.com/usb_on_a_budget.htm

Xiaofan

2008\11\06@054949 by Wouter van Ooijen

face picon face
> I agree with you that it is kind of strange.
>
> Jan Axelson has this article and she talks about
> "The Vendor ID Dilemma".
> http://www.lvr.com/usb_on_a_budget.htm

I suggest everyone keeps bugging them about this problem! Maybe in the
end they will get tired and offer a solution.

--

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2008\11\06@060703 by Xiaofan Chen

face picon face
On Thu, Nov 6, 2008 at 6:49 PM, Wouter van Ooijen <.....wouterKILLspamspam.....voti.nl> wrote:
> I suggest everyone keeps bugging them about this problem! Maybe in the
> end they will get tired and offer a solution.

Your site URL is still listed in Jan Axelson's web site.
http://www.lvr.com/usb.htm#DevelopmentTools

The other link from Jan's site is now saying
"Coming Soon".
http://www.ezprototypes.com/EZModulesMain.php

Xiaofan

2008\11\06@063202 by Wouter van Ooijen

face picon face
> Your site URL is still listed in Jan Axelson's web site.
> http://www.lvr.com/usb.htm#DevelopmentTools
>
> The other link from Jan's site is now saying
> "Coming Soon".
> http://www.ezprototypes.com/EZModulesMain.php

Funny! I contacted Jan years ago, but did not know I was mentioned. I
guess the other source did receive the same threatening legal mail I got :(

--

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2008\11\07@122545 by PicDude

flavicon
face

Microchip will provide a VID/PID at no charge (AFAIR) as long as you are
using their products and are producing no more than 10,000 units of the
product.

Cheers,
-Neil.




Wouter van Ooijen wrote:
{Quote hidden}

> --

2008\11\07@133639 by Wouter van Ooijen

face picon face
PicDude wrote:
> Microchip will provide a VID/PID at no charge (AFAIR) as long as you are
> using their products and are producing no more than 10,000 units of the
> product.

But I would like to hear from someone who has actually walked that road.
Details like how long it took, how many forms hadc to be filled in, etc.

--

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2008\11\09@013036 by PICS

flavicon
face
Wouter van Ooijen ha scritto:
> PicDude wrote:
>  
>> Microchip will provide a VID/PID at no charge (AFAIR) as long as you are
>> using their products and are producing no more than 10,000 units of the
>> product.
>>    
>
> But I would like to hear from someone who has actually walked that road.
> Details like how long it took, how many forms hadc to be filled in, etc.
>
>  
Doe's anyone know "how-to" get a PID from Microchip?

2008\11\09@140332 by Tamas Rudnai

face picon face
USB PID should have done in the very same way as the IP and the URL in the
Internet, right? USB is an organisation, not a company so they should
provide it free or a small amount of handling fee only and let companies act
as registerar. I bet many small companies just pick a random PID now and use
that one to avid paying extra fee reducing the small amount of margin they
have on their small run product line. Same as many companies put the USB
logo on the USB connector without actually having the right for that.

Tamas


On Sun, Nov 9, 2008 at 6:41 PM, Wouter van Ooijen <EraseMEwouterspam_OUTspamTakeThisOuTvoti.nl> wrote:

{Quote hidden}

> -

2008\11\09@153023 by Christopher Head

picon face
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

AFAIK the problem is probably that the USB Forum only sells VIDs, not
PIDs - and there are only 2^16 VIDs.

I've always thought they must be either pretty incompetent or a bunch of
money-grubbing b***ards to have written the spec to call for a 16-bit
VID and a 16-bit PID, rather than just dropping a GUID or something in
the descriptor that anyone could generate for free. Or even just a
64-bit monotonic ID number that the USB Forum's website could generate,
instantly, on the push of a button, and probably never run out. It's not
like the descriptors aren't pretty big already; adding another few bytes
wouldn't really hurt.

Chris

Tamas Rudnai wrote:
> USB PID should have done in the very same way as the IP and the URL in the
> Internet, right? USB is an organisation, not a company so they should
> provide it free or a small amount of handling fee only and let companies act
> as registerar. I bet many small companies just pick a random PID now and use
> that one to avid paying extra fee reducing the small amount of margin they
> have on their small run product line. Same as many companies put the USB
> logo on the USB connector without actually having the right for that.
>
> Tamas
>
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.9 (GNU/Linux)
Comment: GnuPT 2.7.2
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iEYEARECAAYFAkkXSEIACgkQiD2svb/jCb7IvwCeIF0RhkXMi6b9lzFPllCg8jao
r0YAoKHFhkMbYwQPKYSH0gXDZaPLNe7w
=piFQ
-----END PGP SIGNATURE-----

2008\11\09@163246 by Wouter van Ooijen

face picon face
> I've always thought they must be either pretty incompetent or a bunch of
> money-grubbing b***ards to have written the spec to call for a 16-bit
> VID and a 16-bit PID, rather than just dropping a GUID or something in

I don't think an off-line program can work, but they could have offered
let's say 10 PIDs for < $100.

>> USB PID should have done in the very same way as the IP and the URL in the
>> Internet, right? USB is an organisation, not a company

They are an organization founded by a few (big) companies, so they have
all the properties of a big company plus the properties of a committee.

Note that the committee or whatever that hands out ethernet addresses is
equally unfriendly to small companies.

--

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2008\11\09@171536 by Christopher Head

picon face
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Wouter van Ooijen wrote:
>> I've always thought they must be either pretty incompetent or a bunch of
>> money-grubbing b***ards to have written the spec to call for a 16-bit
>> VID and a 16-bit PID, rather than just dropping a GUID or something in
>
> I don't think an off-line program can work, but they could have offered
> let's say 10 PIDs for < $100.
>

Well, GUIDs seem to work pretty well in the software world, and they're
generated offline based on the current time and the MAC address of the
generating computer's Ethernet card... so assuming MAC addresses are
unique (a reasonable though imperfect assumption), GUIDs are too.
They're used all over Windows already for things that are assumed to be
unique.

Chris
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.9 (GNU/Linux)
Comment: GnuPT 2.7.2
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iEYEARECAAYFAkkXYOwACgkQiD2svb/jCb4QNgCfSu6QRQ1x4giA1eJ2rMd9aDfp
CDIAoItJrFqy9Mta1kNt0zA3PnCrolHB
=TRdA
-----END PGP SIGNATURE-----

2008\11\09@184647 by mamouda

flavicon
face


> Message: 28
> Date: Sun, 09 Nov 2008 08:43:41 +0100
> From: Wouter van Ooijen <wouterspamspam_OUTvoti.nl>
> Subject: Re: [PIC]  24F and USB
> To: "Microcontroller discussion list - Public." <@spam@piclistKILLspamspammit.edu>
> Message-ID: <KILLspam491694AD.1020406KILLspamspamvoti.nl>
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>
> >> But I would like to hear from someone who has actually walked that road.
> >> Details like how long it took, how many forms hadc to be filled in, etc.
> >
> > It's not taking long to get a VID/PID from Microchip. You'll have to use Microchip products.
> > Just fill in one formular (application for a sublicense) and fax it to Microchip. After your application have been processed and approved you'll get a VID/PID  (VID,Microchip).
>
> How do you know (did you actually go through this process?), and could
> you quantify 'not long'?
>



You'll find the formular here:   ww1.microchip.com/downloads/en/AppNotes/Application%20for%20USB%20Vendor%20ID%20Sublicense.pdf
Please complete it and fax it to the indicated number. I went also through the process and it took less than three weeks (don't remember how long exactly).

_________________________________________________________________________
In 5 Schritten zur eigenen Homepage. Jetzt Domain sichern und gestalten!
Nur 3,99 EUR/Monat! http://www.maildomain.web.de/?mc=021114

2008\11\09@213958 by William \Chops\ Westfield

face picon face
>> I've always thought they must be either pretty incompetent or a  
>> bunch of
>> money-grubbing b***ards to have written the spec to call for a 16-bit
>> VID and a 16-bit PID, rather than just dropping a GUID or something  
>> in


If they'd been smart, they would have used the same vendor ID as  
ethernet, or SNMP, or one of those existing protocols with a vendor ID.

> Note that the committee or whatever that hands out ethernet  
> addresses is
> equally unfriendly to small companies.

I recall it being pretty easy, and relatively inexpensive, back in the  
day.  (I have the dubious distinction of having requested cisco's  
first such ID.)  Things may have changed since then.  Also, some 1/4  
of the vendor id space in IEEE802 is defined to be "locally defined",  
although these tended to be used for dynamic addresses rather than ad-
hoc vendor ids...

BillW


2008\11\09@224259 by Xiaofan Chen

face picon face
On Mon, Nov 10, 2008 at 10:39 AM, William Chops Westfield
<RemoveMEwestfwTakeThisOuTspammac.com> wrote:
> If they'd been smart, they would have used the same vendor ID as
> ethernet, or SNMP, or one of those existing protocols with a vendor ID.

They are at least smart in the business sense. ;-)
And a big player in USB may not be a big player in Ethernet.


Xiaofan

2008\11\10@013436 by Wouter van Ooijen

face picon face
Christopher Head wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Wouter van Ooijen wrote:
>>> I've always thought they must be either pretty incompetent or a bunch of
>>> money-grubbing b***ards to have written the spec to call for a 16-bit
>>> VID and a 16-bit PID, rather than just dropping a GUID or something in
>> I don't think an off-line program can work, but they could have offered
>> let's say 10 PIDs for < $100.
>>
>
> Well, GUIDs seem to work pretty well in the software world, and they're
> generated offline based on the current time and the MAC address of the
> generating computer's Ethernet card... so assuming MAC addresses are
> unique (a reasonable though imperfect assumption), GUIDs are too.
> They're used all over Windows already for things that are assumed to be
> unique.

Are GUIDs unique over time or only at a specific moment in time?

Also, an ethernet address is 48 bits. An USB VID/PID is only 32 bits...

--

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2008\11\10@113519 by Philip Pemberton

face
flavicon
face
Wouter van Ooijen wrote:
> Note that the committee or whatever that hands out ethernet addresses is
> equally unfriendly to small companies.

There's a workaround for that though -- Dallas Semiconductor make a variant of
the DS2502 1-Wire EPROM that's pre-programmed with a MAC address. Technically
the MAC address applies to the 2401 chip, but it's still a valid MAC address
with the same "globally unique" properties...

"DS2502-E48: 48-bit Node Address Chip" --
<http://www.maxim-ic.com/quick_view2.cfm/qv_pk/3748>

$2.69 each in 1-off, $2.11 each in 25+, $1.74 each in 100+, either TSOC-6 or
TO92-3 package. Relatively expensive, but it might be a better option than
paying the IEEE OUI King's Ransom (tm)... :)

--
Phil.
spamBeGonepiclistspamBeGonespamphilpem.me.uk
http://www.philpem.me.uk/

2008\11\10@120242 by Wouter van Ooijen

face picon face
Philip Pemberton wrote:
> Wouter van Ooijen wrote:
>> Note that the committee or whatever that hands out ethernet addresses is
>> equally unfriendly to small companies.
>
> There's a workaround for that though -- Dallas Semiconductor make a variant of
> the DS2502 1-Wire EPROM that's pre-programmed with a MAC address.

I did not know that - nice, I'll make a note on my website. Or maybe
I'll stock some of these chips.

--

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2008\11\10@124744 by Christopher Head

picon face
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Wouter van Ooijen wrote:
[snip]
> Are GUIDs unique over time or only at a specific moment in time?

Yes, because a timestamp is part of a GUID. The MAC address gives
spatial uniqueness, and the timestamp gives temporal uniqueness.

>
> Also, an ethernet address is 48 bits. An USB VID/PID is only 32 bits...
>

The point I made in my first post was that USB descriptors are pretty
big already, so adding a few bits to the device ID wouldn't cause much harm.

Anyway, my premise is that very few small businesses/hobbyists want to
build Ethernet cards, but lots of us want to build USB devices - on the
flip side, almost everyone *owns* an Ethernet card or two, and thus has
the ability to generate GUIDs with MAC addresses in them.

Chris
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.9 (GNU/Linux)
Comment: GnuPT 2.7.2
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iEYEARECAAYFAkkYc6MACgkQiD2svb/jCb4FQwCeOLAhAAQ/8QBkDh56sGccC1Wi
UXkAnAjgtcmJu5MyXNLAAQS9VNth97Kt
=g4gs
-----END PGP SIGNATURE-----

2008\11\10@140634 by Wouter van Ooijen

face picon face
> Anyway, my premise is that very few small businesses/hobbyists want to
> build Ethernet cards

You'd be surprised how many ENC28J60's I sell...


But I agree that a large device ID (VID+PID) would have made much nicer
solutions possible.

--

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2008\11\11@065117 by Alan B. Pearce

face picon face
OK, thanks everyone for the pointers. It looks like I will need to apply to
Microchip for a couple of VIDs then, as I will be making two devices that
are functionally different (although I am hoping that the PCB will be
essentially the same for both).

Thanks Again.

2008\11\11@070913 by olin piclist

face picon face
Wouter van Ooijen wrote:
> But I agree that a large device ID (VID+PID) would have made much
> nicer solutions possible.

There is one big differenc between ethernet MAC addresses and USB VID/PID
combinations.  MAC addresses must be unique globally unique accross all
individual instances of all ethernet devices.  A VID/PID must only be unique
per device type.  For example, all my USBProg and USBProg2 programmers have
the same VID/PID.

Hmm.  I wonder if it's legal to sell each ReadyBoard-02 with a uniqe VID/PID
option?  Since it is a full USB device that actually works out of the box, I
can't see how they could make a rule against it.

For that matter, could I sell 10F200 with a preprogrammed uniqe VID/PID.
This would be like the Dallas part, but cheaper.  For that matter, I could
do the same thing with ethernet MAC addresses.  I guess I need to find and
carefully read the appropriate agreements and see how much wiggle room there
is.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2008\11\11@073036 by Wouter van Ooijen

face picon face
> There is one big differenc between ethernet MAC addresses and USB VID/PID
> combinations.  MAC addresses must be unique globally unique accross all
> individual instances of all ethernet devices.

true

> Hmm.  I wonder if it's legal to sell each ReadyBoard-02 with a uniqe VID/PID
> option?  Since it is a full USB device that actually works out of the box, I
> can't see how they could make a rule against it.

Interesting idea.

> For that matter, could I sell 10F200 with a preprogrammed uniqe VID/PID.
> This would be like the Dallas part, but cheaper.

In think the fine print will forbid that.

> For that matter, I could
> do the same thing with ethernet MAC addresses.

IIRC the request 'contract' for getting a MAC vendor prefix explicitly
disallow that.

> I guess I need to find and
> carefully read the appropriate agreements and see how much wiggle room there
> is.

When I applied for a VID I signed a request form, but it did not bind me
to any regulations. Yet the usb.org did order me to stop selling PID
blocks. That order might not be legally sound, but I don't want to take
the risk. So it is not just the visible rules and regulations, but the
invisible ones as well. Or you must be prepared to go to court.

--

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2008\11\11@075141 by Alan B. Pearce

face picon face
>For that matter, could I sell 10F200 with a preprogrammed uniqe VID/PID.

<VBG> Now there is a thought ...

2008\11\11@095239 by William Couture

face picon face
On Mon, Nov 10, 2008 at 11:35 AM, Philip Pemberton
<TakeThisOuTpiclistEraseMEspamspam_OUTphilpem.me.uk> wrote:
>
> There's a workaround for that though -- Dallas Semiconductor make a variant of
> the DS2502 1-Wire EPROM that's pre-programmed with a MAC address. Technically
> the MAC address applies to the 2401 chip, but it's still a valid MAC address
> with the same "globally unique" properties...
>
> "DS2502-E48: 48-bit Node Address Chip" --
> <http://www.maxim-ic.com/quick_view2.cfm/qv_pk/3748>
>
> $2.69 each in 1-off, $2.11 each in 25+, $1.74 each in 100+, either TSOC-6 or
> TO92-3 package. Relatively expensive, but it might be a better option than
> paying the IEEE OUI King's Ransom (tm)... :)

Of course, if you only need a couple of MAC addresses, just go to your
local electronics flea market / computer show, and pick up some ISA
ethernet boards for $1 each (or less).  Note the MAC addresses, and
throw away the boards.

Bill

--
Psst...  Hey, you... Buddy...  Want a kitten?  straycatblues.petfinder.org

2008\11\11@114741 by Peter

picon face
Olin Lathrop <olin_piclist <at> embedinc.com> writes:
> There is one big differenc between ethernet MAC addresses and USB VID/PID
> combinations.  MAC addresses must be unique globally unique accross all
> individual instances of all ethernet devices.  A VID/PID must only be unique

Small nitpick: there is no need for the mac to be globally unique, it is
sufficient for it to be unique per net. That means that certain products, such
as routers, of which type there MUST be only one per subnet, might use the same
mac address on the "inside" over and over again. The only reason one has to have
a unique MAC is to ensure that ARP works, and that is a local net issue only.
More importantly, ethernet devices with the same MAC are important for hardware
failover systems where another device must pick up traffic from one packet to
the other.

Peter


2008\11\11@135951 by William \Chops\ Westfield

face picon face

On Nov 11, 2008, at 8:47 AM, Peter wrote:

> Small nitpick: there is no need for the mac to be globally unique,  
> it is
> sufficient for it to be unique per net.

One per level-2 domain (bridged or switched set of nets.)

> That means that certain products, such as routers, of which type  
> there MUST be only one per subnet, might use the same mac address on  
> the "inside" over and over again.

Huh?  There is no reason to limit a configuration to one router per  
subnet.  Some home routers may want such a simplification, but there  
is no reason that an ethernet has to be a "stub" network (consider  
"metro ethernet."  For that matter, I think a good number of "wan"  
network protocols use Ethernet mac addresses for layer two, even for  
non-ethernet hardware.

>  The only reason one has to have a unique MAC is to ensure that ARP  
> works, and that is a local net issue only.

That's a very TCP/IP-centric view of the world.  It may be safe these  
days, but it's not RIGHT.

BillW

2008\11\12@063233 by Peter

picon face
William "Chops" Westfield <westfw <at> mac.com> writes:

> > Small nitpick: there is no need for the mac to be globally unique,  
> > it is
> > sufficient for it to be unique per net.
>
> One per level-2 domain (bridged or switched set of nets.)

True.

> > That means that certain products, such as routers, of which type  
> > there MUST be only one per subnet, might use the same mac address on  
> > the "inside" over and over again.
>
> Huh?  There is no reason to limit a configuration to one router per  
> subnet.  Some home routers may want such a simplification, but there  

There may be a cost and vendor policy reason to limit configuration to one
router of a certain kind per domain (those from the same manufacturer and type,
and habing the same mac). Also the shared mac would be on the "inside" interface
only, for obvious reasons. This would automatically enforce the use of just one
such router per domain (on the inside at least).

> is no reason that an ethernet has to be a "stub" network (consider  
> "metro ethernet."  For that matter, I think a good number of "wan"  
> network protocols use Ethernet mac addresses for layer two, even for  
> non-ethernet hardware.

That is a rarely stupid idea, and the main reason for such nets being massively
hacked into (set top boxes etc). It is also the reason for which most wan
routers that are worth the $25 they cost have a "clone mac" option that copies
the mac of an attached "inside" computer to its outside eth interface, thus
allowing the wan router of the provider to authenticate it. Which is a hole the
size of a hanger door in the wan provider's policy, by the way, and may provide
a way for illegal impersonation of that user. Mac cloning also plays a role in
wireless network cracking. I am not into an of that at all, but I do read what
needs to be read.

> >  The only reason one has to have a unique MAC is to ensure that ARP  
> > works, and that is a local net issue only.
>
> That's a very TCP/IP-centric view of the world.  It may be safe these  
> days, but it's not RIGHT.

Relying on the MAC being unique while the device numbers are such that people
are running out of 32 bit ipv4 addresses rhymes with "640k of ram should be
enough for anyone" imho. Especially for equipment makers who hope to be in
business for some time. Even without considering the ubiquitous mac cloning
feature described above.

$0.02
Peter

PS: I am so glad someone hijacked this thread. At least it shows it was read <G>


2008\11\13@191102 by Christopher Head

picon face
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Peter wrote:
[snip]
> Relying on the MAC being unique while the device numbers are such that people
> are running out of 32 bit ipv4 addresses rhymes with "640k of ram should be
> enough for anyone" imho. Especially for equipment makers who hope to be in
> business for some time. Even without considering the ubiquitous mac cloning
> feature described above.
>

MAC addresses are 48 bits. There are 65,536 times as many MAC addresses
as there are IPv4 addresses, so today there are around 32,768 MAC
addresses per human on the planet.

If you still don't like this... just have a button on usb.org that
returns a monotonic counter of, say, 64 bits. Now if 10 people get an ID
number every *second*, you'll run out in 58 *billion years*!

Oh, and you don't even have to pay thousands in software development
cost. Just go download a free website hit counter :) With so many
trivially easy ways to do this, I can't attribute their failure to
anything but greed. It's pretty sad when people pay thousands of dollars
for... a number.

Chris
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.9 (GNU/Linux)
Comment: GnuPT 2.7.2
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iEYEARECAAYFAkkcwfkACgkQiD2svb/jCb5BKACfT3ZNoun/OYexRH55azbzATyS
g0QAoKoUz+luAfgh5laSFLm+t8Y5CT59
=MBLE
-----END PGP SIGNATURE-----


'[PIC] 24F and USB'
2009\05\16@231414 by Jinx
face picon face
Xiaofan, I've been looking through some of these links

http://forum.microchip.com/tm.aspx?m=123533

and reading up on USB to prepare for my first project

Could I impose upon you to advise please. I'm bewildered by
all the information available for USB implementation, but not as
bewildered as I was yesterday

I've actually several projects in mind. One is new, others involve
updating older products from parallel (oh yes !!) and serial to USB,
mostly due to the disappearance of connectors on modern PCs.
Speed isn't a particular issue, as the quantities of data are small

The new one was originally sketched out using an 18F4550 as a
peripheral and serially via a Vinculum VNC1L as a host. However,
now I find that the VNC1L is around $20 and the programmer is
> $300, so it seems more economic to use a more capable PIC

It seems to me that the intended usage suits an OTG PIC. This
product is required to receive updating data (as emailed files) from
a PC and also datalog as a USB host into a thumb drive. The thumb
drive can be transferred to a PC for uploading, so it's not necessary
for the PIC to act as a conduit to do that

Does that fit with your understanding of OTG ? I'm on the right
track ?

I've downloaded the 24FJ256GB110 datasheet

http://ww1.microchip.com/downloads/en/DeviceDoc/39897b.pdf

and Section 27 of the family datasheet

ww1.microchip.com/downloads/en/DeviceDoc/39721a.pdf

2009\05\16@232516 by Xiaofan Chen

face picon face
On Sun, May 17, 2009 at 11:12 AM, Jinx <RemoveMEjoecolquittspamTakeThisOuTclear.net.nz> wrote:
> It seems to me that the intended usage suits an OTG PIC. This
> product is required to receive updating data (as emailed files) from
> a PC and also datalog as a USB host into a thumb drive. The thumb
> drive can be transferred to a PC for uploading, so it's not necessary
> for the PIC to act as a conduit to do that
>
> Does that fit with your understanding of OTG ? I'm on the right
> track ?

I think this is a good fit for USB PIC24F. You can download the
Microchip USB stack v2.4 and it has a few examples for USB host
of PIC24F. I think you do not need OTG here. There is only one
OTG example in the USB stack.

Look at this example and see if it fits your application.
C:\Microchip Solutions\USB Host - Mass Storage - Thumb Drive Data Logger>

Download link:
http://www.microchip.com/usb

--
Xiaofan http://mcuee.blogspot.com

2009\05\17@063439 by Jinx

face picon face
Xiaofan, thanks

> Look at this example and see if it fits your application.
> C:\Microchip Solutions\USB Host - Mass Storage - Thumb Drive Data Logger>
>
> Download link:
> http://www.microchip.com/usb

The C:\ above doesn't lead to a file unfortunately. I searched MC for
"Thumb Drive Data Logger" though and get a page with this link

http://www.microchip.com/stellent/idcplg?IdcService=SS_GET_PAGE&nodeId=2680&dDocName=en537044

and this one, which appears identical

http://www.microchip.com/stellent/idcplg?IdcService=SS_GET_PAGE&nodeId=1406&dDocName=en537044

Is one of these what you intended to link to ?

The problem I have at the moment is that many links and contents look
so very very similar. I've spent quite some time already downloading
files and packages that weren't what I expected and that I can't or
shouldn't use. I'm getting a little frustrated because when researching
and following links which have similar descriptions, I seem to end up at
a different USB page each time. When you're on dial-up and each
destination means a multi-meg download that turns out to be not what it
promised, it gets somewhat annoying. Either Microchip's site is a mess
or I'm really not understanding what I'm looking for

For example, my comments about the above page's text in [ ]

========

MCHPFSUSB Framework - MCHPFSUSB  is a distribution package
containing a variety of USB related PIC18 and PIC24F firmware projects,
along with other USB related drivers and resources intended for use on
the PC. The USB embedded host stack is API compatible with the USB
Device and Embedded Host Stack for PIC32. All release notes are
included in the .zip file bundle

[ The one for PIC32 is a zip, the one for 18/24 is an exe ]

Demos include Device CDC demo, CDC serial emulator, device composite
HID and mass storage, generic driver demo, HID mouse demo, HID
keyboard demo, SD card reader, SD data logger, thumb drive data logger
(host) and much more.

MCHPFSUSB Framework v1.3 (legacy version) - MCHPFSUSB is a
distribution package containing a variety of USB related PIC18
firmware projects, along with other USB related drivers and resources
for use on the PC.  This version is not recommended for new designs

[ 'not recommended for new designs' is less than helpful ]

2009\05\17@065336 by Xiaofan Chen

face picon face
On Sun, May 17, 2009 at 6:34 PM, Jinx <joecolquittEraseMEspam.....clear.net.nz> wrote:
> Xiaofan, thanks
>
>> Look at this example and see if it fits your application.
>> C:\Microchip Solutions\USB Host - Mass Storage - Thumb Drive Data Logger>
>>
>> Download link:
>> http://www.microchip.com/usb
>
> The C:\ above doesn't lead to a file unfortunately. I searched MC for
> "Thumb Drive Data Logger" though and get a page with this link

You need to go to http://www.microchip.com/usb and then
click "Software&Tools". Then you can download
MCHPFSUSB Framework v2.4.
http://www.microchip.com/stellent/idcplg?IdcService=SS_GET_PAGE&nodeId=2680&dDocName=en540668

After installation, you will get the Thumb Drive Data Logger projected.

--
Xiaofan http://mcuee.blogspot.com

2009\05\17@071319 by Jinx

face picon face
> You need to go to http://www.microchip.com/usb and then
> click "Software&Tools". Then you can download
> MCHPFSUSB Framework v2.4.
> http://www.microchip.com/stellent/idcplg?IdcService=SS_GET_PAGE&nodeId=2680&dDocName=en540668
>
> After installation, you will get the Thumb Drive Data Logger projected.

Thanks

2009\05\17@083959 by Jinx

face picon face
> MCHPFSUSB Framework v2.4.

BTW, the file is

http://ww1.microchip.com/downloads/en/DeviceDoc/Microchip%20Application%20Libraries%20v2009-03-24.zip

As it's such a big file (42MB) I want to use a download manager.
GetRight wasn't latching on to Microchip's link but it does to the
above

2009\05\17@090426 by Xiaofan Chen

face picon face
On Sun, May 17, 2009 at 8:39 PM, Jinx <EraseMEjoecolquittspamclear.net.nz> wrote:
>> MCHPFSUSB Framework v2.4.
>
> BTW, the file is
>
> http://ww1.microchip.com/downloads/en/DeviceDoc/Microchip%20Application%20Libraries%20v2009-03-24.zip
>
> As it's such a big file (42MB) I want to use a download manager.
> GetRight wasn't latching on to Microchip's link but it does to the
> above

Old version archive links are here:
http://www.microchip.com/forums/tm.aspx?m=422691

42MB is now considered small file by me. ;-) I need to routinely download
Linux distro updates about 50-100MBs. MPLAB is not small either.
So my 8Mbps broadband helps.


--
Xiaofan http://mcuee.blogspot.com

2009\05\17@092551 by Jinx

face picon face
> Old version archive links are here:
> http://www.microchip.com/forums/tm.aspx?m=422691

Thanks. I've a couple of quiet days ahead to get stuck into USB

> 42MB is now considered small file by me. ;-) I need to routinely
> download Linux distro updates about 50-100MBs. MPLAB is not
> small either. So my 8Mbps broadband helps.

I've got the patience for a big file ...... if it's the right one !!

One of the staff I get talking to at the local electronics store asked
me on Friday what I use micros for. I described the PIC timer I
built many years ago to disconnect (relays) the Internet at 6am. That
gives me about 9 hours, ~ 100+ MB, of overnight downloading,
leaving the phone working for other people in the morning. One of
the most useful gadgets I ever made

2009\05\17@115154 by Alan B. Pearce

face picon face
>The C:\ above doesn't lead to a file unfortunately. I searched MC
>for "Thumb Drive Data Logger" though and get a page with this link

Jinx, it does if you install the Microchip USB stack on your machine. The
'link' that Xiaofan gave is correct if you install the stack using the
defaults of the installer.

I have been having a little play with the Device CDC code, making a device
that looks like a COM port to the host. So far all my playing has been with
the v2.3 stack, and the new v2.4 stack is set up to use interrupts, which I
am hoping will help me. Also the whole stack is written in C, so you will
need to look at getting the C30 compiler.

As to the PIC 24 chip, if you want to use the input capture or Output
Compare (PWM) then be aware that the modules are very different to previous
ones, as they have their own timer. There is also no Timer0, but there are 5
16 bit timers, 4 of which are able to be cascaded to make 2 32 bit timers. I
also like the PPS feature, where certain of the digital peripherals can be
assigned to a selection of pins. There are some sneaky tricks with this as
well, in that one peripheral can be assigned to multiple pins. One example
of doing this is to assign two interrupts to one pin, with one interrupt
positive edge triggered, and one negative edge triggered, for doing fancy
pulse measuring, without changing the edge selection - but that is about the
only reason I can see to do it for any peripheral.

2009\05\17@193308 by Jinx

face picon face
> >The C:\ above doesn't lead to a file unfortunately. I searched MC
>>for "Thumb Drive Data Logger" though and get a page with this link
>
> Jinx, it does if you install the Microchip USB stack on your machine.
> The 'link' that Xiaofan gave is correct if you install the stack using the
> defaults of the installer

You mean The Internet * doesn't * have a great big C drive ?

> Also the whole stack is written in C, so you will need to look at getting
> the C30 compiler

Hmmm, guess I will. Had absolutely no practical exposure to C until now,
hopefully it's written well enough to be understandable

Downloaded the application but not yet installed anything. No 24F sample
to hand yet though, so there's time to figure it out beforehand

2009\05\17@202251 by Jinx

face picon face
>> product is required to receive updating data (as emailed files) from
>> a PC and also datalog as a USB host into a thumb drive. The thumb
>> drive can be transferred to a PC for uploading, so it's not necessary
>> for the PIC to act as a conduit to do that

> I think you do not need OTG here

Xiaofan, how would you see the USB cable being connected ? My
first thought was that the PIC product would have a Micro-AB for
OTG functionality. As a host for interfacing with the thumb drive and
as a peripheral for interfacing with a PC. If OTG isn't used, do you
think the product would have a USB-A socket for the thumb drive
and a leaded USB-A plug to the PC ?

2009\05\17@214117 by Xiaofan Chen

face picon face
On Mon, May 18, 2009 at 8:22 AM, Jinx <RemoveMEjoecolquittEraseMEspamEraseMEclear.net.nz> wrote:
>>> product is required to receive updating data (as emailed files) from
>>> a PC and also datalog as a USB host into a thumb drive. The thumb
>>> drive can be transferred to a PC for uploading, so it's not necessary
>>> for the PIC to act as a conduit to do that
>
>> I think you do not need OTG here
>
> Xiaofan, how would you see the USB cable being connected ? My
> first thought was that the PIC product would have a Micro-AB for
> OTG functionality. As a host for interfacing with the thumb drive and
> as a peripheral for interfacing with a PC. If OTG isn't used, do you
> think the product would have a USB-A socket for the thumb drive
> and a leaded USB-A plug to the PC ?

http://www.usbmadesimple.co.uk/ums_2.htm
As you mentioned, the thumb drive can be transferred to a PC
for uploading, so it's not necessary for the PIC to act as a
conduit to do that. So I assume you can take out the thumb
drive and transfer it to the PC through the PC USB port.

In that case, it is easier to do the following.
PIC24 will act as the host to the USB thumb drive.
Another PIC (PIC18F4550 or similar) will act as the
USB device. You can use SPI between the two.

If the speed requirement is not high between PC and
your device, then you can save the 2nd PIC and use
RS232 between the PIC24 and the PC. That is the case
with the Microchip example.

OTG might be a possibility. But it is a recent addon
to the stack and there is only one example which
is the generic driver demo. So I think you need significant
efforts to get it working for you.

--
Xiaofan http://mcuee.blogspot.com

2009\05\17@214419 by Xiaofan Chen

face picon face
By the way, it is highly recommended that you buy the Explorer 16 demo
board and the associated PICtail+ USB and USB PIM for PIC24. Then
you can play with the demos. I am still more on the device side and
have not played too much with the host/otg side.

PIC24 USB starter kit is cheaper but more limited.

--
Xiaofan http://mcuee.blogspot.com

2009\05\17@214728 by Marcel Duchamp

picon face
Jinx wrote:
>
> You mean The Internet * doesn't * have a great big C drive ?

Dude! You need to totally get the interweb on CD! Dialup? Waszat?

(this message from a nerd who contemplates rural living - with only
dialup...OMG!)

2009\05\17@220226 by Jinx

face picon face
> By the way, it is highly recommended that you buy the Explorer 16
> demo board and the associated PICtail+ USB and USB PIM for
> PIC24

Yes, I see now that this will be a significant undertaking. s/w will
need to be upgraded and new h/w bought. I'm looking into that now

I still tend to think, perhaps somewhat naively given my total lack of
practical experience, that I can write application s/w which will do the
simple and particular jobs I have in mind, even though it would fall
short of full compliance

However, it will be helpful to have those modules and then see what
can be pared back from a full-blown implementation

2009\05\17@221113 by Jinx

face picon face

> (this message from a nerd who contemplates rural living - with only
> dialup...OMG!)

Marcel Du Champs eh ?

2009\05\17@222147 by Jinx

face picon face
> So I assume you can take out the thumb drive and transfer it to
> the PC through the PC USB port

That would be desirable, as the thumb drive will be out in the field
with the product during logging. Periodically data will need to be
looked at back at the office, with the product still out in the field

One other possibility is to email the data through a GSM modem,
in which case the thumb drive will not be moved. Given that thumb
drive prices are through the floor it makes some sense to use one
rather than install on-board flash, although using on-board memory
in such a situation does rather mean that USB is no longer needed

But this is as much a learning exercise to make modern interfaces

> In that case, it is easier to do the following : PIC24 will act as the
> host to the USB thumb drive. Another PIC (PIC18F4550 or similar)
> will act as the USB device. You can use SPI between the two

Hmmm, OK, that's a thought

> If the speed requirement is not high between PC and
> your device, then you can save the 2nd PIC and use
> RS232 between the PIC24 and the PC. That is the case
> with the Microchip example.

I wouldn't want to get caught out trying to upload to a rep's laptop
for example that had no serial port. Although even this PC I thought
had no serial port until I spotted a COM1 header hidden away under
the modem card. A serial-USB adapter could be used if necessary

> OTG might be a possibility. But it is a recent addon
> to the stack and there is only one example which
> is the generic driver demo. So I think you need significant
> efforts to get it working for you.

Thanks for the advice. Once I get my head around what's actually
required to use the USB protocol the decisions should become easier

2009\05\18@074952 by Artem Zezyulinskiy

flavicon
face
Jinx a écrit :
>> (this message from a nerd who contemplates rural living - with only
>> dialup...OMG!)
>>    
>
> Marcel Du Champs eh ?
>  
LOL :)

very good wordplay (pun?) !

2009\05\18@200328 by Eoin Ross

flavicon
face
Jinx, you're on the wrong version of the internets. You need Web 2.0

>>> On 17 May 09 at 21:47, in message <RemoveME4A10BE40.5070907spam_OUTspamKILLspamsbcglobal.net>, Marcel
Duchamp <RemoveMEmarcel.duchampTakeThisOuTspamspamsbcglobal.net> wrote:
> Jinx wrote:
>>
>> You mean The Internet * doesn't * have a great big C drive ?
>
> Dude! You need to totally get the interweb on CD! Dialup? Waszat?
>
> (this message from a nerd who contemplates rural living - with only
> dialup...OMG!)


2009\05\18@210530 by Jinx

face picon face
> Jinx, you're on the wrong version of the internets. You need Web 2.0

Aw, rats. Had only a year to go to finish my download of Web 1.0

Web 2.0 is so much bigger isn't it ? I see that MPLAB is now a
Rubenesque 90MB download. Still have the original which comes
on a handful of floppies

More... (looser matching)
- Last day of these posts
- In 2009 , 2010 only
- Today
- New search...