Searching \ for '[EE] 'C' PIC programmer from USB Port power' 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/devprogs.htm?key=programmer
Search entire site for: ''C' PIC programmer from USB Port power'.

Exact match. Not showing close matches.
PICList Thread
'[EE] 'C' PIC programmer from USB Port power'
2004\10\29@223330 by Peter Crowcroft

flavicon
face
Tony Nixon has been designing a PIC programmer to do 'C' DIP PICs from the
USB port. Getting the MAX761 chips, low ESR ecaps and choke for the
stepping up of the 5V to 15V has been a bit of a hassle. But all this seems
to have been solved during my visit to the Shanghai Electronics Show two
weeks back. There I found a small potted DC to DC converter, 22mm x 10mm x
12mm, 5V in, 15V out, 66mA or 132 mA versions available. No low ESR ecap
needed. Nice price of $5.20 in bulk.

http://www.mornsun-power.com


When we get prototypes of the new programmer we will ask piclisters for
help in testing.

But at this stage can anyone see why this DC to DC converter might not work?

Thanks,



regards,
                DIY Electronics (HK) Ltd
      PO Box 88458, Sham Shui Po, Hong Kong
     M/F, 97 Fuk Wa Street, Sham Shui Po, HK
Factory: voice 852-2304 2250    Fax: 852-2729 1400
Home: voice 852-2720 0255,      Mobile: 852-6273 2049
Web:  http://www.kitsrus.com    Email: spam_OUTpeterhkTakeThisOuTspamkitsrus.com

   Chinese/Thai language emails to  .....peter5998KILLspamspam@spam@netvigator.com
---------------------------------------------------------------

____________________________________________

2004\10\29@225127 by Ken Pergola

flavicon
face


Peter Crowcroft wrote:

> But at this stage can anyone see why this DC to DC converter
> might not work?

Hi Peter,

The USB port's output voltage might not meet the minimum input voltage
requirement for the DC-DC converter. If you have not read the USB
specification documents, take a look.

What is the input voltage tolerance on the DC-DC converter's 5 V input?

Best regards,

Ken Pergola


____________________________________________

2004\10\29@230156 by Ken Pergola

flavicon
face

Hi Peter,

Could you provide a DC-DC converter part number?

Best regards,

Ken Pergola

____________________________________________

2004\10\29@232607 by Bob Axtell

face picon face
Gee, Peter there are a LOT of low-cost solutions for 5V to 15V
up-switchers. Should be less than $3 for
50ma at 15V, with everything.

Maxim is a nice company, but their parts are pretty dear. Look at TI,
Linear Technology, National. Minimal
parts count, low noise, little $ in it.

Take another look at National, you'll be surprised. If I can get a
coupla programmers out of the deal, I'll breadboard
the up-switcher for you.

--Bob

Peter Crowcroft wrote:

{Quote hidden}

> ______________________________________________

2004\10\30@000525 by Matt Pobursky
flavicon
face
This is more for Peter, but I wanted to comment on Bob's reply also...
Here's a circuit I recently used in a medical design that will work well. Fits in an area about the size of a dime (or less), will do 15V @ ~60mA and will cost less than $3 in parts (closer to $2). All the passives are 0805 R's and C's, except C30 (1206) and L3 (1206). The C's are all mono ceramic caps (cheap). The coil is a standard Murata chip
inductor (cheap, I bought my proto parts from Digikey for ~$.40 single
piece price). The circuit is also easy on the USB power supply, which
has a limit of 10uF maximum input capacitance specification.

http://www.mps-design.com/misc-images/switcher.jpg

You can eliminate R54/C28 if you don't need soft start, although it
adds almost no cost. Q9/Q10 provide a zero current ON/OFF switch that
you may or may not need. If you are switching from a PIC output pin
that's at the same voltage as your input (VBAT in this case), you can
eliminate Q10.
Very efficient circuit, relatively cheap and very reliable. If you are
interested, I can give you an exact Bill of Materials with all
component part numbers and manufacturers.
Bob -- National has some OK parts, but they tend to be not very efficient. They ARE cheap and easy to use though. Every time I look at the National switchers I find them lacking in one way or another for my particular design (usually quiescent current or off-state leakage, etc.). But they are great parts where you can tolerate a little more slop. Don't get me wrong though, I use a lot of National parts for other things too. Lately I've been using a lot more TI and Linear Tech
parts, it seems. You are right about Maxim parts though -- even if you
can afford them, you probably won't be able to actually BUY them.

Matt Pobursky
Maximum Performance Systems


On Fri, 29 Oct 2004 20:26:03 -0700, Bob Axtell wrote:
{Quote hidden}

> > _______________________________________________

2004\10\30@002016 by Matt Pobursky

flavicon
face
On Sat, 30 Oct 2004 10:29:24 +0800, Peter Crowcroft wrote:
> Tony Nixon has been designing a PIC programmer to do 'C' DIP PICs from the
> USB port. Getting the MAX761 chips, low ESR ecaps and choke for the
> stepping up of the 5V to 15V has been a bit of a hassle. But all this seems
> to have been solved during my visit to the Shanghai Electronics Show two
> weeks back. There I found a small potted DC to DC converter, 22mm x 10mm x
> 12mm, 5V in, 15V out, 66mA or 132 mA versions available. No low ESR ecap
> needed. Nice price of $5.20 in bulk.

http://www.mornsun-power.com
..snip...

> But at this stage can anyone see why this DC to DC converter might not work?

You might have trouble with their input capacitance -- USB has a spec
of maximum 10uF load on the 5V supply. Too much input capacitance
causes problems for USB on hot plug-in (big cap, big short, bad news).
I'm not sure which model you are looking at, but the datasheets I
looked at all had ~.1W minimum load specs on the converters. That would
tell me they are probably using a simple magnetically coupled
oscillator with a post regulator (needs a minimum current to stay in
regulation). This may or may not be a problem for your design.

Matt Pobursky
Maximum Performance Systems 


___________________________________________

2004\10\30@002554 by Bob Axtell

face picon face
Matt Pobursky wrote:

{Quote hidden}

I know the efficiencies aren't high, but in Peter's case $ is important,
and National parts are always reasonable
and ALWAYS available. USB 5V supplies are rich in excess current,
efficiency isn't important. And Peter
can get REALLY GOOD, inexpensive inductors from China.

Maxim got into the habit of advertising stuff they didn't even have. I
vowed off of them for years, but they
HAVE improved in the last year.

I just LOVE the LM2653 TSSOP16 buck switcher. This thing stays within
50mV from 10mA to 900mA. And it will run right even
when the 8V LiIon pack is sputtering at 4V! It is fairly efficient at
7V, nominal operation. Makes you look like
a magician.

--Bob

{Quote hidden}

2004\10\30@014204 by Bob Barr

flavicon
face
On Fri, 29 Oct 2004 23:05:15 -0500, Matt Pobursky wrote:

>This is more for Peter, but I wanted to comment on Bob's reply also...
>
>Here's a circuit I recently used in a medical design that will work
>well. Fits in an area about the size of a dime (or less), will do 15V @
>~60mA and will cost less than $3 in parts (closer to $2). All the
>passives are 0805 R's and C's, except C30 (1206) and L3 (1206). The C's
>are all mono ceramic caps (cheap). The coil is a standard Murata chip
>inductor (cheap, I bought my proto parts from Digikey for ~$.40 single
>piece price). The circuit is also easy on the USB power supply, which
>has a limit of 10uF maximum input capacitance specification.
>
>http://www.mps-design.com/misc-images/switcher.jpg
>

Looks good, Matt. On my browser, though, both ends of the schematic
are cut off and I can't seem to get it to display the whole thing.

Any suggestions?


Regards, Bob

____________________________________________

2004\10\30@114936 by olin_piclist

face picon face
Bob Axtell wrote:
> Gee, Peter there are a LOT of low-cost solutions for 5V to 15V
> up-switchers. Should be less than $3 for
> 50ma at 15V, with everything.

$3 is way too expensive for that.  Only 50mA and 15V output from 5V input is
about as easy as it gets.  You can use a jellybean NPN transistor as the
swithing element and still have decent efficiency.  A rough parts list:

 PIC 10F206 switching controller              $.50
 2N4401 switching element                      .05
 inductor                                      .50
 low ESR input and output caps                 .75
 a few jellybean resistors, etc                .20
                                          --------
 TOTAL                                        2.00

And I'm being rather pessimistic with most of the prices.  This was just off
the top of my head.  In reality, I think I could get that down to between
$1.00 and $1.50 in reasonable volumes with a couple hours total design time.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com
____________________________________________

2004\10\30@141942 by Matt Pobursky

flavicon
face
Bob,

The schematic is a cut 'n paste section of a client schematic so I
really couldn't post the whole thing. You should see on the left, R52,
R53, Q9 and Q10 as the input to the switcher circuit. TP6 is the output
(in this case ~16V for the values shown). The output of the circuit
goes to another section of the board that contains an electronic piezo
buzzer volume control.
This circuit is one part of an MSP430 based portable medical monitoring
device that runs off a 3 cell NiMH battery pack. It's efficient, small,
relatively low cost and generates very low electrical noise (both EMI
and noise on the VBAT supply) -- which is why we settled on this
circuit for our design. The LT3461 also contains the onboard power
switch and schottky diode, which eliminates two of the more costly
components in the circuit (and extran PC board space).
I should have probably given a better explanation of the circuit in my
original posting.

I hope that clears things up!

Matt

On Fri, 29 Oct 2004 22:42:03 -0700, Bob Barr wrote:
{Quote hidden}

> _____________________________________________

2004\10\30@144753 by Ken Pergola

flavicon
face

A few thoughts:
---------------

With regard to programming OTP PICmicros, you'll see that a lot of them call
out a maximum Vpp supply current specification of 50 mA (milliamps),
however, some OTP PICmicros (like the PIC16C5XX, for example) call out a
maximum Vpp supply current specification of 100 mA. So if you want to
support these OTP PICmicros, it looks like you need your Vpp generator to
source up to 100 mA at the Vpp voltage if you want to comply with the
programming specification.

Also pay very careful attention to the Vpp supply voltage maximum and
minimum specification windows in the programming specifications. Sometimes
you can get lucky and choose a fixed Vpp that covers everything, but the
moment programming specifications change (and they do change) with regard to
the Vpp max/min window, it can cause you a lot of grief if you have a fixed
Vpp device programmer. A variable Vpp solution would be to post-regulate the
15 volts and provide a variable Vpp generator under digital control with say
a digital potentiometer. Or you could use a boost switcher to start with
that has digital control of the voltage output -- the Texas Instruments
TPS61045 looks like it may be a nice part to investigate -- you might be
cutting close in meeting the 100 mA requirement -- I have not looked in
detail.

I'm not sure how sensitive the PICmicros are to noisy Vpp voltages from a
switcher, but using a linear LDO post regulator can help greatly in this
regard. But then you are back to a digital potentiometer to make a variable
Vpp generator from the linear post regulator. This all adds complexity and
cost, but having a variable Vpp generator is really nice if the programmer
needs to program Flash parts as well (event though it appears this is not
your goal with this programmer).

It seems that the tricky part of a USB port-powered programmer is that,
depending on how you approach things, you can eat a lot of the available USB
current (as in mA) that the USB port provides -- leaving you less current
that your target ICSP hardware can have access to.

What about the notebook/laptop people? Can their native USB ports source 500
mA without an external powered hub. I don't own a laptop/notebook so I can't
check for myself. For my PC I use an Adaptec DuoConnect and those USB ports
can source 500 mA, but my native motherboard USB can only source 100 mA I
believe (without an external, powered hub).

Best regards,

Ken Pergola


____________________________________________

2004\10\30@155419 by Matt Pobursky

flavicon
face
Good post Ken. As always, carefully setting your requirements based on
end unit requirements is necessary to avoid going off half-cocked on
any design...
Personally, if I were doing a USB PIC programmer, I'd definitely make a post regulated variable VPP generator capable of at least 100 mA. It gives you the option to cover all the bases and the increased cost is pretty minimal. I'd also make a variable VDD generator so you can do programming and verification over the entire range and conform to
Microchip's production programmer specs. There are already a zillion "cheap" PIC programmer designs out there, I'd want to have a unit that distinguished itself from the rest by being higher quality and only slightly more expensive.

I've done several commercial USB connected device designs and my informal research showed that *most* modern motherboards don't control USB +5V at all, they tend to just run the on-board 5V supply to the USB
+5V pins through a PTC thermistor for overcurrent protection. All the USB channels are usually fed from the same source, so glitching at
plug-in can be problematic for other USB channels on some motherboards (hence the USB maximum +5V bus capacitance specification). Most motherbaords I examined used 500mA to 1A PTC thermistors, although the USB specification requires you to draw only up to 100mA from the host until you negotiate for more power.
Most laptops I looked at limit USB +5V current to 100mA, period. My
Dell laptop is very unhappy if you try to draw more than that... :-(

Bottom line -- follow the USB specifications to the letter and you
won't have any problems. If you "stretch the specs" all sorts of wacky
and hard to diagnose problems can happen.

Matt Pobursky
Maximum Performance Systems

On Sat, 30 Oct 2004 14:47:43 -0400, Ken Pergola wrote:
{Quote hidden}

___________________________________________

2004\10\30@163655 by William Chops Westfield

face picon face
On Oct 30, 2004, at 12:54 PM, Matt Pobursky wrote:

> if I were doing a USB PIC programmer, I'd definitely make a
> post regulated variable VPP generator capable of at least 100 mA.

Disturbingly close to the 2.5W per-port power limit for bus powered
USB devices :-(  (but USB disks tend to ignore that limit all the time,
and frequently get away with it...)

BillW

____________________________________________

2004\10\30@165900 by Ken Pergola

flavicon
face

Matt Pobursky wrote:

> Personally, if I were doing a USB PIC programmer, I'd definitely make a
> post regulated variable VPP generator capable of at least 100 mA. It
> gives you the option to cover all the bases and the increased cost is
> pretty minimal. I'd also make a variable VDD generator so you can do
> programming and verification over the entire range and conform to
> Microchip's production programmer specs.

Yes, I totally agree with you on both points. When I developed the
Micro-bRISC Device Programmer back in 1994 there were only a handful of PIC
chips out there and things were pretty simple. Then, over the years, the
PICmicro lineage became like the grocery store cereal aisle -- many, many
choices. And with all those new parts came many changes to the programming
specifications with regard to the Vpp min/max voltage window and Vpp
programming currents. To be ahead of the game, you seriously had to consider
a variable Vpp generator -- and one that could source 100 mA to comply with
the programming specifications.

> Most laptops I looked at limit USB +5V current to 100mA, period. My
> Dell laptop is very unhappy if you try to draw more than that... :-(

That is what I was afraid of. I absolutely love the idea of a pure, USB
port-powered programmer (no wall warts!), but the fact that some native USB
ports will not be able to shell out 500 mA puts a real damper on things. I
definitely want a device programmer with some "muscle" that can power my
target hardware. It seems in cases where native USB ports can only supply
100 mA or current, it's probably best to design a USB programmer that uses a
wall-wart to supply the necessary current for the programmer and for the
target circuitry. It most likely would be cheaper for the end-user --
otherwise the end-user would have to shell out money for an external powered
hub (to get 500 mA) -- something I would imagine would cost more than a wall
wart.

Additionally, you really need to design for the minimum voltage that the USB
port can supply as outlined in the specification -- if I remember correctly
it can be as low as ~ 4.1 V (with respect to ground) or something like that.
I wish it was more of a tightly regulated specification like 5 V +/- 5% --
that would be nice. All in all it seems like it's just easier and cheaper to
use an external wall wart for power to a USB programmer. Having a choice of
port-powered or wall wart powered is even better yet. :)

Best regards,

Ken Pergola































____________________________________________

2004\10\30@180542 by steve

flavicon
face
> Most laptops I looked at limit USB +5V current to 100mA, period. My
> Dell laptop is very unhappy if you try to draw more than that... :-(

And some desktops have no active current limit on the USB supply
other than track melting and escaping smoke.
Yes, I learnt the hard way with a DLP module.

Steve.


____________________________________________

2004\10\30@180547 by steve

flavicon
face
On 30 Oct 2004 at 10:29, Peter Crowcroft wrote:

> There I found a small potted DC to DC
> converter, 22mm x 10mm x 12mm, 5V in, 15V out, 66mA or 132 mA versions
> available. No low ESR ecap needed. Nice price of $5.20 in bulk.

Is that an isolated module ?
Does it need to be ?

Steve.


____________________________________________

2004\10\30@182821 by Spehro Pefhany

picon face
At 10:29 AM 10/30/2004 +0800, you wrote:
>Tony Nixon has been designing a PIC programmer to do 'C' DIP PICs from the
>USB port. Getting the MAX761 chips, low ESR ecaps and choke for the
>stepping up of the 5V to 15V has been a bit of a hassle. But all this
>seems to have been solved during my visit to the Shanghai Electronics Show
>two weeks back. There I found a small potted DC to DC converter, 22mm x
>10mm x 12mm, 5V in, 15V out, 66mA or 132 mA versions available. No low ESR
>ecap needed. Nice price of $5.20 in bulk.
>
>http://www.mornsun-power.com

Is that $5.20 Hong Kong dollars or US dollars?

I come up with about 91 US cents and 9 parts (not counting the output
electrolytic cap)
US prices, off the shelf, multiple sources for everything, in 1K. Cheaper
in Asia,
perhaps half or less in 5K+.

In small quantities the DC-DC converters are a PITA. There are lots of
pinouts and they don't like making small lots. In my experience anyhow.
There are a few Taiwan-headquartered companies that are pretty competitive,
in huge volume anyway.

>But at this stage can anyone see why this DC to DC converter might not work?

Something to do with meeting the USB current draw spec exactly might be
a gotcha.

Best regards,

Spehro Pefhany --"it's the network..."            "The Journey is the reward"
RemoveMEspeffTakeThisOuTspaminterlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com




____________________________________________

2004\10\30@192644 by Jinx

face picon face

> Disturbingly close to the 2.5W per-port power limit for bus
> powered USB devices :-(  (but USB disks tend to ignore that
> limit all the time, and frequently get away with it...)

Silicon Chip have a USB "booster" in this month's issue

http://www.siliconchip.com.au/cms/A_102685/article.html

____________________________________________

2004\10\30@194554 by Matt Pobursky

flavicon
face
On Sun, 31 Oct 2004 11:05:32 +1300, spamBeGonestevespamBeGonespamtla.co.nz wrote:
> > Most laptops I looked at limit USB +5V current to 100mA, period. My
> > Dell laptop is very unhappy if you try to draw more than that... :-(

> And some desktops have no active current limit on the USB supply
> other than track melting and escaping smoke.
> Yes, I learnt the hard way with a DLP module.

Hehehe... yes, I believe it. I looked at as many motherboards as I
could get my hands in with my first USB design but didn't run across
any without some kind of overcurrent protection -- but I'm sure they do
exist.
I can almost hear the exchange in an engineering review...
Engineer 1: "OK, we have 2 USB host ports. Each port must supply 500 mA
output current and should probably have some sort of overcurrent
protection. The easiest is PTC thermistors -- they will protect us
against short circuits and cost only $.10 each"
Project Leader/Marketing Guy: "Hmmmm. The USB specification says the
user is not to draw more than 500 mA under any circumstances. Leave
them off - they are unnecessary and save us $.20!!!"
Engineer 1: "But.. but.. but..."

Matt Pobursky
Maximum Performance Systems

___________________________________________

2004\10\30@195251 by olin_piclist

face picon face
Ken Pergola wrote:
> Additionally, you really need to design for the minimum voltage that
> the USB port can supply as outlined in the specification -- if I
> remember correctly it can be as low as ~ 4.1 V (with respect to
> ground) or something like that. I wish it was more of a tightly
> regulated specification like 5 V +/- 5% -- that would be nice.

The specification is reasonably tight at the host USB port.  The loose
device specification is due to the voltage drop accross the maximum allowed
cable resistance at the maximum allowed current.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com
____________________________________________

2004\10\30@204012 by Bob Barr

flavicon
face
On Sat, 30 Oct 2004 13:19:40 -0500, Matt Pobursky wrote:

>Bob,
>
>The schematic is a cut 'n paste section of a client schematic so I
>really couldn't post the whole thing. You should see on the left, R52,
>R53, Q9 and Q10 as the input to the switcher circuit. TP6 is the output
>(in this case ~16V for the values shown). The output of the circuit
>goes to another section of the board that contains an electronic piezo
>buzzer volume control.
>

Thanks for the clarification. I hadn't realized that it had been cut
from a larger schematic.

Regards, Bob


____________________________________________

2004\10\30@214103 by Peter Crowcroft

flavicon
face
>From: "Ken Pergola" <TakeThisOuTno_spamEraseMEspamspam_OUTlocalnet.com>
>
>Hi Peter,
>
>Could you provide a DC-DC converter part number?


We are testing Mornsun

WRB0515S-1W

It certainly seems to power Kit 150 OK from the USB port.

Input 4.5V - 9V. Output 15V 66mA.

There is something to be said for the convenience of having a nice module
like this over a surface mount IC/ inductor/low ESR ecap.


regards,
                DIY Electronics (HK) Ltd
      PO Box 88458, Sham Shui Po, Hong Kong
     M/F, 97 Fuk Wa Street, Sham Shui Po, HK
Factory: voice 852-2304 2250    Fax: 852-2729 1400
Home: voice 852-2720 0255,      Mobile: 852-6273 2049
Web:  http://www.kitsrus.com    Email: RemoveMEpeterhkspamTakeThisOuTkitsrus.com

   Chinese/Thai language emails to  peter5998EraseMEspam.....netvigator.com
---------------------------------------------------------------

____________________________________________

2004\10\30@233953 by Andrew Warren

flavicon
face
Ken Pergola <EraseMEpiclistspammit.edu> wrote:

> you really need to design for the minimum voltage that the USB port
> can supply as outlined in the specification -- if I remember correctly
> it can be as low as ~ 4.1 V (with respect to ground) or something like
> that.

   4.35V, with transients as low as 4.07V.

   -Andy

=== Andrew Warren - RemoveMEaiwEraseMEspamEraseMEcypress.com
===
=== Principal Design Engineer
=== Cypress Semiconductor Corporation


____________________________________________

2004\10\31@005516 by Ken Pergola

flavicon
face

Olin Lathrop wrote:

> The specification is reasonably tight at the host USB port.  The loose
> device specification is due to the voltage drop across the
> maximum allowed
> cable resistance at the maximum allowed current.


Hi Olin,

Yes of course -- thank you for correcting me. Yeah, I incorrectly said "at
the port" but I was really thinking "after the cable" when I wrote it
because that is the voltage that the USB device programmer designer is
ultimately up against and has to deal with. That's why I'm trying to find a
nice heavy gauge USB cable and use the shortest cable that's feasible -- I
have not looked too hard at this point.

If, on the other hand, end-users were guaranteed +5 V +/- 5% *after* the
cable under all worst-case conditions, that would be great!

Veering off-topic to flash PICmicro programming...with the requirement that
many PIC18 Flash PICmicros need at least 4.5 V with respect to Vss for a
bulk erase, it seems that boosting the USB supply voltage should be done by
the device programmer to ensure that no angry customers come knocking...or
calling...or e-mailing. :)

So between boosting the incoming voltage and providing Vpp at 100 mA, it
really _____ on my Christmas pie! :(
It eats into the total available USB power budget and provides less for the
target board. So, as I said before, I think the best solution would be to
have a dual option for the user to cover both bases. If the user's target
does not draw that much -- USB port powered is really nice. But if the user
needs the programmer to provide some extra, he just connects a wall wart and
everyone is happy.

Best regards,

Ken Pergola




____________________________________________

2004\10\31@015117 by William Chops Westfield

face picon face
On Oct 30, 2004, at 9:55 PM, Ken Pergola wrote:

> If, on the other hand, end-users were guaranteed +5 V +/- 5% *after*
> the
> cable under all worst-case conditions, that would be great!
>
I would have been great, if cable manufacturers weren't so quick to
sell all sorts of USB cables that were never supposed to exist. :-(
You WERE supposed to be able to expect certain behavior of USB cables,
but I doubt whether you can anymore :-( :-(

BillW

____________________________________________

2004\10\31@071023 by olin_piclist

face picon face
Ken Pergola wrote:
> That's why
> I'm trying to find a nice heavy gauge USB cable and use the shortest
> cable that's feasible -- I have not looked too hard at this point.

Your device would not be USB complient if it required such a cable.  One of
the important features of the USB spec is that if you can plug something
together, it's OK to do so.  You can't expect customers to use just your
special USB cable, and you wouldn't be allowed to call it a USB device
anyway.

> If, on the other hand, end-users were guaranteed +5 V +/- 5% *after*
> the cable under all worst-case conditions, that would be great!

My point was that this would be impossible to guarantee except with short or
expensive cables or at very low maximum power levels.

Generally I think the USB spec was intelligently designed.  I was however
dissappointed by the way power to a device was specified.  It seems like
this was the exception where the spec was more for the convenience of the
host than the device.  It's going to be common to want to power 5V
circuitry, so why not specify a port output voltage such that the minimum at
the device is 5.5V?  That would allow for simple local regulation (a good
idea anyway) without wasting much power.  The other approach could have been
specify the maximum voltage just under 20V so that low cost parts and
topologies could be used for a buck regulator.  Ethernet does this, except
uses about 48V, although that is designed for significantly longer cable
runs.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com
____________________________________________

2004\10\31@112747 by Ken Pergola

flavicon
face

Olin Lathrop wrote:

Ken Pergola wrote:

> That's why I'm trying to find a nice heavy gauge USB cable and use the
shortest
> cable that's feasible -- I have not looked too hard at this point.


Olin Lathrop wrote:

> Your device would not be USB compliant if it required such a
> cable.  One of
> the important features of the USB spec is that if you can plug something
> together, it's OK to do so.  You can't expect customers to use just your
> special USB cable, and you wouldn't be allowed to call it a USB device
> anyway.


Hi Olin,

Oh No -- you misunderstood -- I'm not making a USB device or selling one --
this is just what I want for myself. I'm saying that I, as an end user,
would like to find the lowest AWG gauge cable that I can find -- I don't
mind spending extra money getting it. Like I said, I have not searched too
hard at this point.

>From what I can gather from the USB specification, as far as the power wires
are concerned, the *minimum* gauge is 28 AWG. I don't *think* there is a
problem with a cable manufacturer making a USB cable with 24, 22, or 20 AWG
wires instead of the minimum 28 AWG.


>From the USB specification:
---------------------------

Non-Twisted Power Pair:

Wire Gauge: Minimum 28 AWG or as specified by the user contingent upon the
specified cable length.



Best regards,

Ken Pergola


____________________________________________

2004\10\31@114800 by Ken Pergola

flavicon
face

After some 'googling' is looks like it is very easy to get a USB cable with
20 AWG for the power pair -- that's all I'm after. At least I can have a
choice and have control in the matter by avoiding excessive voltage drop
with the combination of 20 AWG for the power pair and a short cable length.

Best regards,

Ken Pergola



____________________________________________

2004\10\31@163012 by Peter L. Peres

picon face

USB is a cash cow. It is a set of protocols and hardware specifications
designed to sell chips and products made by a group of manufacturers with
a specific set of manufacturing and engineering capabilities. That's how
one makes money in a saturated market apparently.

The common points seem to be: most silicon that will run off usb will be
low power cmos at 3V or lower. It will be intelligent (no $2 homemade
interfaces are possible). Any need to extend the network beyond basic
capabilities will require the end user to purchase more equipment (hubs),
that will sell more chips.

For all this complexity the users will receive the benefit of automatic
device identification and driver loading, and a little more speed than
serial (but less than firewire, which was already available when usb was
invented).

This is just my opinion, feel free to comment.

Peter
____________________________________________

2004\10\31@170637 by Spehro Pefhany

picon face
At 05:21 PM 10/31/2004 -0500, you wrote:

>USB is a cash cow. It is a set of protocols and hardware specifications
>designed to sell chips and products made by a group of manufacturers with
>a specific set of manufacturing and engineering capabilities. That's how
>one makes money in a saturated market apparently.
>
>The common points seem to be: most silicon that will run off usb will be
>low power cmos at 3V or lower. It will be intelligent (no $2 homemade
>interfaces are possible). Any need to extend the network beyond basic
>capabilities will require the end user to purchase more equipment (hubs),
>that will sell more chips.
>
>For all this complexity the users will receive the benefit of automatic
>device identification and driver loading, and a little more speed than
>serial (but less than firewire, which was already available when usb was
>invented).
>
>This is just my opinion, feel free to comment.
>
>Peter

The major benefit of USB over serial for low-speed devices is that it
supplies power- reducing the cost of peripherals. It also functions
well (2.0+) for high-speed devices such as scanners and digital cameras
that would be useless with a serial interface. The cables and plugs are
potentially extremely cheap- only a few wires.

I don't think the hubs are any big deal, the companies that make both
the hubs and (I think) the chips for hubs are not huge companies. I think
they just wanted to remove all the legacy barnacles of PS2 and serial
ports and streamline future computers- while making peripherals cheaper
and lighter. Win-win.

Best regards,

Spehro Pefhany --"it's the network..."            "The Journey is the reward"
RemoveMEspeffspam_OUTspamKILLspaminterlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com




____________________________________________

2004\10\31@182322 by William Chops Westfield

face picon face

On Oct 31, 2004, at 2:21 PM, Peter L. Peres wrote:

> USB is a cash cow. It is a set of protocols and hardware specifications
> designed to sell chips and products made by a group of manufacturers
> with a specific set of manufacturing and engineering capabilities.
> That's how one makes money in a saturated market apparently.
>

Cynic!  I always saw USB (1.1) as a well-aimed and reasonably well
executed attack on the customer-support issues of typical low to
moderate speed peripheral devices.  Printers, modems, scanners,
cameras, dongles, keyboards, mice; all the crap that people were
connecting serial and parallel ports with arbitrary and frequently
incorrect cables, and never getting right.  Newer peripherals were
adding their own near-proprietary and/or mysterious interface cards
(bus mice, scsi scanners, HPIB), better matched to the tasks, but with
their own cabling and support issues.  Things were a mess.

It's too bad that the designers failed to predict some of the
applications,
or the rate at which the amount of data going into or out of a personal
computer would increase.  11mbps probably seemed perfectly reasonable in
the days of 1G disk drives and before the digital camera revolution, but
is pretty inadequate when a common application is to move 1G of pictures
from your camera to your 250G disk :=(

USB2 is considerably less noble.  Firewire was already well deployed
by that time, and overlapped a lot of the new applications that USB2
hoped to pick up.  I can't see a lot of reason for USB2 other than
typical NIH syndrome.  (USB OTG looks like a useful extension, but
that's not the part that's getting deployed...)

It doesn't look to me like USB peripherals or host cards are expensive
enough to attribute all of USB's goals to greed (compared to, say,
HPIB.)
One would have to be very paranoid to think that the elimination of
amateur (and/or "custom low volume") electronic devices was a goal of
USB, even though it seems to be having that result.  Some of that is
laziness on OUR part; it really shouldn't be that much more difficult
to develop a peripheral using a $5 16C745 now than it was to make one
using a $4 16C57 in "the old days."

BillW

____________________________________________

2004\10\31@195655 by Peter Crowcroft

flavicon
face
>Date: Sat, 30 Oct 2004 18:42:05 -0400
>From: Spehro Pefhany <RemoveMEspeffTakeThisOuTspamspaminterlog.com>
>
>At 10:29 AM 10/30/2004 +0800, you wrote:
> >Tony Nixon has been designing a PIC programmer to do 'C' DIP PICs from
> the USB port. Getting >the MAX761 chips, low ESR ecaps and choke for the
> stepping up of the 5V to 15V has been a bit >of a hassle. But all this
> seems to have been solved during my visit to the Shanghai >Electronics
> Show two weeks back. There I found a small potted DC to DC converter,
> 22mm x 10mm >x 12mm, 5V in, 15V out, 66mA or 132 mA versions available.
> No low ESR ecap needed. Nice price >of $5.20 in bulk.
> >
> >http://www.mornsun-power.com

>Is that $5.20 Hong Kong dollars or US dollars?

USD


>I come up with about 91 US cents and 9 parts (not counting the output
electrolytic cap)
>US prices, off the shelf, multiple sources for everything, in 1K. Cheaper
in Asia,
>perhaps half or less in 5K+.

There is alot to be said for the pure convenience of 1 component which i
can buy in lots of 100 as opposed to a reel of SM IC's, MOQ 1K low ESR
ecaps, 1K chokes etc.


>In small quantities the DC-DC converters are a PITA. There are lots of
>pinouts and they don't like making small lots.

For me in HK MOQ's are alot less than they are for overseas buyers. I pay
cash in Shenzhen and they deliver COD to Hongkong.


>In my experience anyhow.
>There are a few Taiwan-headquartered companies that are pretty competitive,
>in huge volume anyway.

Most Taiwan based companies are now operating out of China - even Sales
offices. The shift has been rapid.

>But at this stage can anyone see why this DC to DC converter might not work?

>Something to do with meeting the USB current draw spec exactly might be
>a gotcha.

Yes Tony is looking at this.

Thanks,

peter



regards,
                DIY Electronics (HK) Ltd
      PO Box 88458, Sham Shui Po, Hong Kong
     M/F, 97 Fuk Wa Street, Sham Shui Po, HK
Factory: voice 852-2304 2250    Fax: 852-2729 1400
Home: voice 852-2720 0255,      Mobile: 852-6273 2049
Web:  http://www.kitsrus.com    Email: EraseMEpeterhkspamspamspamBeGonekitsrus.com

   Chinese/Thai language emails to  RemoveMEpeter5998KILLspamspamnetvigator.com
---------------------------------------------------------------

____________________________________________

2004\10\31@230730 by Matt Pobursky

flavicon
face
On Sun, 31 Oct 2004 15:23:17 -0800, William "Chops" Westfield wrote:
> Cynic!  I always saw USB (1.1) as a well-aimed and reasonably well
..snip...
> One would have to be very paranoid to think that the elimination of
> amateur (and/or "custom low volume") electronic devices was a goal of
> USB, even though it seems to be having that result.  Some of that is
> laziness on OUR part; it really shouldn't be that much more difficult
> to develop a peripheral using a $5 16C745 now than it was to make one
> using a $4 16C57 in "the old days."

Consider that the amateur or custom low volume guy has to cough up an additional $1500 (minimum) for a vendor ID and I wouldn't really call that cynical! I have never understood how an "open standard" needs to charge for things like vendor ID's or MAC addresses for that matter. I
wouldn't call that an "open standard", I'd call it a "paid association"
only open to paying members. A look at the consortium member companies
on the USB specifications can't help but make you a little suspicious
and cynical. To them $1500 (or even $15,000) additional cost for a
given product is a fly spec on their budget -- to the small guy it's a
deal breaker.
Until FTDI came along, all the "big guys" were no help at all with
Vendor ID's (and without one you can't legally call yourself USB and use
USB markings).

I won't even get into the technical aspect of taking the venerable serial port which could be both a peer-peer interface or host-client
interface (your choice) and more or less forcing the new standard to be
PC-centric host-client only. Just look at how difficult it is to make
two PICs talk to each other over USB, i.e. no PC involved... Yes, USB
OnTheGo is making that somewhat possible, but very limited in
functionality.

Matt Pobursky
Maximum Performance Systems

___________________________________________


'[EE] 'C' PIC programmer from USB Port power'
2004\11\01@084414 by alan smith
picon face
Also....

Consider the case of using a DC/DC module that costs
maybe a few cents more than discrete but when it fails
you don't have to replace 6 components, just one.
Depends on the product your building tho.


               
__________________________________
Do you Yahoo!?
Yahoo! Mail Address AutoComplete - You start. We finish.
http://promotions.yahoo.com/new_mail
____________________________________________

2004\11\01@085223 by alan smith

picon face
Doesnt the USB spec say that the outputs have to be
protected??

--- Matt Pobursky <piclistSTOPspamspamspam_OUTmps-design.com> wrote:

{Quote hidden}

> ______________________________________________

2004\11\01@093219 by Spehro Pefhany

picon face
At 05:52 AM 11/1/2004 -0800, you wrote:
>Doesnt the USB spec say that the outputs have to be
>protected??

Yes, of course it does. Maximum 5A, must *not* require
mechanical user interaction to reset and the condition must be
reported to the USB software, Polymeric PTC or solid-state
switches are acceptable. Circuit breakers or fuses are not.

RTFM: 7.2.1.2.1 Over-current Protection


Best regards,

Spehro Pefhany --"it's the network..."            "The Journey is the reward"
KILLspamspeffspamBeGonespaminterlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com




____________________________________________

2004\11\01@100951 by Bob Barr

flavicon
face
On Sun, 31 Oct 2004 15:23:17 -0800, William "Chops" Westfield wrote:

>One would have to be very paranoid to think that the elimination of
>amateur (and/or "custom low volume") electronic devices was a goal of
>USB, even though it seems to be having that result.  Some of that is
>laziness on OUR part; it really shouldn't be that much more difficult
>to develop a peripheral using a $5 16C745 now than it was to make one
>using a $4 16C57 in "the old days."

That is all well and good as long as you're making personal devices
and have no intentions of commercializing them. If you ever do, the
USB logo becomes the sticking point.

The USBIF membership fee is on the order of US$2500 per year. Of
course, you can save money by buying the license to use the logo for a
mere US$1500 per product. And that doesn't count the cost of any
compliance testing done through an independent test lab.


Regards, Bob

____________________________________________

2004\11\01@132806 by Peter L. Peres

picon face

On Sun, 31 Oct 2004, Spehro Pefhany wrote:

{Quote hidden}

Yes, the manufacturers who make the devices in sufficient numbers (>10,000
pcs probably) win, and everyone else, 99% of the small shops and low
units/year developers, are lying low. 'small' is relative.

Peter
____________________________________________

2004\11\02@044507 by No Religion

flavicon
face
At 17.20 2004.10.31 -0500, you wrote:
{Quote hidden}

I believe that was the "intention" of the purchasers.. but those
"engineers" (if we dare to call them with this word) that designed USB
must have had other intentions, or must have been truly incapable.

Have you ever looked in depth at how the USB interface and protocol work
inside? I've never seen anything more "complex with no reason" in my life.
There's no technical justifications to all that complexity. Only market,
it seems..



>Best regards,
>
>Spehro Pefhany --"it's the network..."            "The Journey is the reward"
>EraseMEspeffspamEraseMEinterlog.com             Info for manufacturers: http://www.trexon.com
>Embedded software/hardware/analog  Info for designers:  http://www.speff.com
>
>
>
>
>_____________________________________________

2004\11\02@075833 by olin_piclist

face picon face
No Religion wrote:
> Have you ever looked in depth at how the USB interface and protocol work
> inside?

Yes, I've read the spec.

> I've never seen anything more "complex with no reason" in my
> life.
> There's no technical justifications to all that complexity. Only market,
> it seems..

I don't see much technical justification for your statement.  Only a
deliberate attitude problem it seems..

The main intention behind USB was making it idiot proof for the customer.
This comes at a price of some complexity, particularly in the protocol.  I
think most of it was well designed given that goal.  Like I said before, I
think they dropped the ball on supplying power, but that's not the part your
complaining about anyway.  Aside from your content free rant above, what
exactly do you think is needlessly complex?  It seems to me that every piece
has a purpose.

My biggest gripe is that the USB organization doesn't seem to care about the
little guys.  $1,500 is nothing for a large company, but despite what Andrew
Warren said (who works for a large company that produces USB chips), $1,500
can be a real issue for a little company.  The vendor IDs should be much
cheaper, since these obviously don't cost much more than a little
bookeeping.  If they want to provide plugfests, make those pay their own way
by charging what it takes to produce them.  Or, at least have a sliding
scale for vendor IDs based on company size or number of units shipped.  If I
remember right, some other standards (ethernet?) base the fee in part on
company size.  Or why not have the USB organization sell individual device
IDs within a general purpose vendor ID for $50 or something, just enough to
cover the registration costs.  Little guys could be accomodated if there was
only the will to do so, but this is not in the interest of the big guys that
are the influential "members" of the USB organization.  Besides, they all
got there with the existing rules, so why would they want to let others in
more cheaply?

The other dissappointment has been the lack of a really accessable USB
microcontroller for quick one-off or low volume projects.  I looked at the
Cypress line a year or two ago, and there wasn't the combination of what I
really wanted, in addition to the whole company presenting a strong attitude
against small volume designs.  In one case the chip looked reasonable, but
it was OTP only with no flash or UV erasable parts.  The only way to do
software development was to buy a case of parts and use a new one each
retry.  Duh!

Microchip's offering is on an outdated PIC and can only do low speed.
Pretty useless.  Fortunately this should all be fixed with the introduction
of the full speed USB 18F parts.  That took way too long.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com
____________________________________________

2004\11\02@095343 by Dave Tweed

face
flavicon
face
olin_piclist@embedinc.com (Olin Lathrop) wrote:

> The other dissappointment has been the lack of a really accessable USB
> microcontroller for quick one-off or low volume projects.

For low-volume/one-off projects, it's hard to beat marrying an FT245BM to
a generic microcontroller (or DSP or FPGA or ...).

It has large internal FIFOs and a simple 8-bit parallel bus that can
transfer about 1.5 MB/sec. to/from those FIFOs, making most small message
transfers essentially instantaneous for all practical purposes.

I've got a project right now that uses an FT245BM in combination with a
Xilinx FPGA and a Silicon Labs C8051F121 (high-performance 8051 clone
microcontroller), and I'm planning to get rid of the microcontroller and
do everything with dedicated logic in the FPGA because the interface is
so simple.

All parameters relating to the configuration of the USB connection (VID,
PID, power requirements, etc.) are stored in a small external serial EEPROM
(optional, if you can live with all of the defaults), and FTDI provides a
free utility for editing the content of the EEPROM from the host. You can
use either the "VCP" virtual comm port driver or the direct "D2XX" DLL
driver for the host-side application code.

-- Dave Tweed
____________________________________________

2004\11\02@103522 by No Religion

flavicon
face
At 07.58 2004.11.02 -0500, you wrote:
>No Religion wrote:
>> Have you ever looked in depth at how the USB interface and protocol work
>> inside?
>
>Yes, I've read the spec.

So you made your homework. Good.


>> I've never seen anything more "complex with no reason" in my life.
>> There's no technical justifications to all that complexity. Only market,
>> it seems..
>
>I don't see much technical justification for your statement.  Only a
>deliberate attitude problem it seems..

I expressed a personal opinion on the USB protocol being more complex than
it could have been (let away that to implement USB in a FPGA is a true waste
of logic, oppositely to implementing some other kind of (better) interface),
and you reply with a personal insult.. the ole good Olin with serious behaviour
problems is back! And you talk about others' attitude!! Lord, save us!! :DDD


>The main intention behind USB was making it idiot proof for the customer.
>This comes at a price of some complexity, particularly in the protocol.  I
>think most of it was well designed given that goal.  Like I said before, I
>think they dropped the ball on supplying power, but that's not the part your
>complaining about anyway.  Aside from your content free rant above, what

I'll make you a favour and not answer.. so I don't help you outletting YOUR
famous attitude for trolling, and you don't get kicked for the tenth time!

I do this because it seems to me you're doing your best to enter the world
records book in the "most kicked troll" page, and I ain't gonna help you on
your foul attempts.

I'm back on topic.. either if you follow or not (as most likely will be).

N.R.

____________________________________________

2004\11\02@111922 by Peter Moreton

flavicon
face
Oh jeez, here we go again.....

'No Religion', FWIW, I happen to agree with Olin - you really do have a
*major* attitude problem.



 

{Original Message removed}

2004\11\02@121647 by olin_piclist

face picon face
No Religion wrote:
> I expressed a personal opinion on the USB protocol being more complex
> than
> it could have been (let away that to implement USB in a FPGA is a true
> waste of logic, oppositely to implementing some other kind of (better)
> interface), and you reply with a personal insult..

Sorry, this was not meant as an insult.  Your statement really did sound
more like a religious argument than a technical one.  It would be
intereseting to hear what you consider "complex with not reason".  Yes, the
protocol is rather complex.  It was designed assuming dedicated silicon at
both ends.  I didn't design the protocol, but while reading the spec I
thought I could see the reason for most of it.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com
____________________________________________

2004\11\02@122546 by Alan B. Pearce

face picon face
>Oh jeez, here we go again.....
>
>'No Religion', FWIW, I happen to agree with Olin -
>you really do have a *major* attitude problem.

I tend to agree. I felt Olin's points were pretty right on the mark.

{Quote hidden}

Yes you expressed a personal opinion, but to say there is no technical
justification belies the fact that "it works, out of the box" as Olin
explained.

It seems to me that you do not have an appreciation of what is involved with
having an operating system connect and disconnect devices at any time,
without major screwups of what goes on inside. Remember this is not just a
disk file system (like floppy systems have been doing for ages) but ANY
device. Such internal complexity in the operating system requires some
complexity in the attachment protocol.

____________________________________________

2004\11\02@123217 by Omer YALHI

flavicon
face
Why can't we use the PICList as a means of exchanging ideas, asking
questions etc. rather than fight all the time (not all the time obviously,
but hey).  "No Religion" is obviously trying to take advantage of the fact
that Olin has been kicked out couple of times and delibaretly trying to test
his temper, and for what?  I, for one, get exteremely usefull information
from the list all the time, although my electrical eng. skills/knowledge
prevents me from replying most of the time.  

This list is great and let's keep it that way, it is entirely up to us.

Regards, Omer

____________________________________________

2004\11\02@124628 by Ken Pergola

flavicon
face

Dave Tweed wrote:

> For low-volume/one-off projects, it's hard to beat marrying an FT245BM to
> a generic microcontroller (or DSP or FPGA or ...).

Hiya Dave,

Amen! Amen!

Yeah, when I programmed a PIC via a FT245BM, it knocked my socks off! The
speed is incredible! I love that little bugger of a chip. I'm really looking
forward to working with the newest member of FTDI's family soon -- the
FT2232C. It's been out for a while, but I have not played with it yet.

If you are not aware of it, you might be really interested in the FT2232C as
well -- especially for your FPGA part of the project. Check out the
Multi-Protocol Synchronous Serial Engine interface part of the data sheet if
you have not.

It's a little bit more expensive than the FT245BM, but it sure packs a lot
of punch! Lots of options.

Best regards,

Ken Pergola


____________________________________________

2004\11\02@155243 by steve

flavicon
face
On 2 Nov 2004 at 7:58, Olin Lathrop wrote:

> The main intention behind USB was making it idiot proof for the
> customer. This comes at a price of some complexity, particularly in
> the protocol.  I think most of it was well designed given that goal.

I agree. I haven't looked too deeply at 2.0, but if you compare USB to
say,  Bluetooth or Devicenet, it's pretty well defined.

> My biggest gripe is that the USB organization doesn't seem to care
> about the little guys.

This seems to be the viewpoint of many posters in this thread, based on
the premise that USB is a replacement for the serial port. It's not the
USB.Org who have dropped the serial port. It's hardly reasonable to
blame them for the decision that PC makers have made.

> $1,500 is nothing for a large company, but
> despite what Andrew Warren said (who works for a large company that
> produces USB chips), $1,500 can be a real issue for a little company.

As a little company, I agree that $1500 is an issue. I can think of much
better things to do with that sum. At just over $1k for MPLAB C30 and
ICD2, you could say the same of Microchip in regards to getting started
with dsPIC. In the context of NRE's, a USB VID isn't that bad.

> why not have the USB organization sell individual device IDs within a
> general purpose vendor ID for $50 or something, just enough to cover
> the registration costs.  Little guys could be accomodated if there was
> only the will to do so

Like Andrew, I looked at doing something like this and selling product
IDs under a common Vendor ID. The killer catch is with drivers and
support in general. There's no direct correlation, but if someone doesn't
have the means to find $1500 for a VID, chances are they won't have
the means to create a device driver. That means a disproportionate
amount of support for a low, one-off, fee.

If the device is a straight HID, then it's pretty straight forward with
support already in the major OS's. Otherwise, the every device is unique
and needs to communicate via the OS.

If you are selling silicon, then the equation is a bit more balanced.
Cypress have their ram based micros and a common driver interface.
FTDI have PID blocks available on request and provide a matching
DLL. To the user, it looks like a custom device.

If you need USB 2.0, then the same economics come into play. The
only people who can make it work are silicon vendors, as they will get
ongoing returns. They are facing the same decision as we are - is it
worth pursuing at that cost ?

Steve.

____________________________________________

2004\11\02@164943 by Wouter van Ooijen

face picon face
> As a little company, I agree that $1500 is an issue. I can
> think of much
> better things to do with that sum. At just over $1k for MPLAB C30 and
> ICD2, you could say the same of Microchip in regards to
> getting started
> with dsPIC.

But an ICD2 is ~ $150, an the C30 full-functional demo is free!

Wouter van Ooijen

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


____________________________________________

2004\11\02@173931 by olin_piclist

face picon face
Wouter van Ooijen wrote:
>> As a little company, I agree that $1500 is an issue. I can
>> think of much
>> better things to do with that sum. At just over $1k for MPLAB C30 and
>> ICD2, you could say the same of Microchip in regards to
>> getting started
>> with dsPIC.
>
> But an ICD2 is ~ $150, an the C30 full-functional demo is free!

In addition an ICD2 or other PIC development equipment is used immediately
on billable hours, whereas buying a USB vendor ID is $1500 right when you're
trying to get a production run going and before you have sales.  Those not
in the position will probably never understand, but $1500 definitely keeps
little guys from just trying something.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com
____________________________________________

2004\11\02@182952 by Marcel Duchamp

picon face
At 02:39 PM 11/2/04, you wrote:

>In addition an ICD2 or other PIC development equipment is used immediately
>on billable hours, whereas buying a USB vendor ID is $1500 right when you're
>trying to get a production run going and before you have sales.  Those not
>in the position will probably never understand, but $1500 definitely keeps
>little guys from just trying something.

Well said! Those USB suits are a bunch of profiteers!

And it doesn't stop with USB either!

Like, I'm totally trying to put together a happening gamebox and by the
time it's all loaded up, the tag is like $3000! And that doesn't even
include the costs I'll be looking at to mod the bitch! It isn't
fair!  Like, what's that all about? I'm all "hey I wanna fast box" and
they're all "it'll be 3 grand boy"!  That definitely keeps little guys from
just trying something.

I once had a great idea for a world beater product but it was gonna need an
oscilloscope for development.  Would you believe the price of a top of the
line scope these days is around $5000?  It definitely keeps little guys
from just trying something.

Plus, don't get me started on the price of cars, man.  They're *really*
priced to keep little guys from just trying something!

There ought to be a law!
MD

____________________________________________

2004\11\02@185248 by Andrew Warren

flavicon
face
Olin Lathrop <@spam@piclist@spam@spamspam_OUTmit.edu> wrote:

> My biggest gripe is that the USB organization doesn't seem to care
> about the little guys.

   Yes, that's probably true, although of course I don't believe
   that they're deliberately trying to keep the "little guys" out,
   as a few people have suggested.

> $1,500 is nothing for a large company, but despite what Andrew
> Warren said (who works for a large company that produces USB
> chips), $1,500 can be a real issue for a little company.

   Of course, I haven't ALWAYS worked for Cypress... Before I became
   a sellout and started working for The Man, I owned a small
   company much like yours, so I think I'm pretty familiar with the
   financial constraints under which you work.

   When I say that $1500 should be affordable if someone's serious
   enough to be successful making USB devices, I'm not saying "Let
   them eat cake"; I'm helping that person decide whether what he's
   doing is a hobby or not.

> The vendor IDs should be much cheaper, since these obviously don't
> cost much more than a little bookeeping.  If they want to provide
> plugfests, make those pay their own way by charging what it takes
> to produce them.

   As I said in a previous message, I believe that the ID-only fee,
   which used to be $200, was raised to $1500 so as to narrow the
   gap between it and the ID-with-testing fee of $2500.

   Narrowing the gap encourages people to pay the $2500 and actually
   test their devices before unleashing them on the world. Testing
   is a good thing.

   Charging less for Vendor IDs and making people "pay their own
   way" by charging more for testing would DISCOURAGE people from
   testing their devices.  That would be a bad thing.

> Or, at least have a sliding scale for vendor IDs based on company
> size or number of units shipped.

   Device programmers are expensive.  Is yours priced on a sliding
   scale?

   Ok, maybe that's an unfair analogy.  How about this instead:
   Every once in a while, someone on the piclist asks about
   receiving a sliding-scale royalty for his firmware, based on the
   number of units his customer ships.  The response is almost
   always "Don't do it; it's too easy for your customer to cheat
   you."

   If the USB-IF had to verify company size and sales volume, I
   wouldn't be surprised if the extra overhead raised even the
   minimum license fee above $1500.  The USB-IF is a small
   organization; they don't already have a bunch of accountants
   sitting around looking for something to do.

> why not have the USB organization sell individual device IDs within
> a general purpose vendor ID for $50 or something, just enough to
> cover the registration costs.

   Because then the devices built by those people wouldn't get
   tested and certified, and the introduction of those untested
   devices into the USB "ecosystem" would probably disrupt what is
   currently a very pleasant user experience.  

> Little guys could be accomodated if there was only the will to do
> so, but this is not in the interest of the big guys that are the
> influential "members" of the USB organization. Besides, they all
> got there with the existing rules, so why would they want to let
> others in more cheaply?

   Think.  If we really wanted to keep you out, wouldn't we have
   conspired with the USB-IF to charge a lot more than $1500?

> The other dissappointment has been the lack of a really accessable
> USB microcontroller for quick one-off or low volume projects.  I
> looked at the Cypress line a year or two ago, and there wasn't the
> combination of what I really wanted, in addition to the whole
> company presenting a strong attitude against small volume designs.

   Well, you can't really blame us... If you're planning to sell so
   few USB devices that you won't be able to afford the $1500 VID
   fee, how much money can we expect to make from you on sales of
   our microcontrollers?

   Much as I'd like for everyone in the world to be using our
   parts, it just isn't cost-effective to support thousands of 10-
   to 100-unit/year USB customers when the same revenue can be made
   from one medium-sized customer.

> In one case the chip looked reasonable, but it was OTP only with
> no flash or UV erasable parts.  The only way to do software
> development was to buy a case of parts and use a new one each
> retry.

   Yeah, I've always thought it would be nice to have reprogrammable
   low-speed micros in addition to our reprogrammable full- and
   high-speed parts.

   Personally, though, I think your programming skills are good
   enough that it wouldn't take you more than a few dozen attempts
   before you got your code working... But if you think it would
   take more than 300 tries, it would be cheaper to buy our $500
   emulator.

   -Andy

=== Andrew Warren -- spamBeGoneaiwspamKILLspamcypress.com
=== Principal Design Engineer
=== Cypress Semiconductor Corporation
===
=== Opinions expressed above do not
=== necessarily represent those of
=== Cypress Semiconductor Corporation

____________________________________________

2004\11\02@193403 by Rob Young

picon face
> I once had a great idea for a world beater product but it was gonna need
> an oscilloscope for development.  Would you believe the price of a top of
> the line scope these days is around $5000?  It definitely keeps little
> guys from just trying something.

Try greater than  $100,000 for the latest Agilent multi-GHz real time
sampling scopes...

On the other hand, a good tool is worth what you pay for it.

I bought an Agilent 54622D (definately not a top of the line scope or logic
analyzer but still pretty good).  I waited until somebody sent me a 0% Visa
card application.  Bought the scope on that card and paid it off in 6-months
before interest charges applied.

I use the dickens out of that scope, definately paid for itself.

Rob Young
____________________________________________

2004\11\02@194833 by steve

flavicon
face
> > But an ICD2 is ~ $150, an the C30 full-functional demo is free!

The demo is only free for a limited time. After that, you buy it.

> In addition an ICD2 or other PIC development equipment is used
> immediately on billable hours, whereas buying a USB vendor ID is $1500
> right when you're trying to get a production run going and before you
> have sales.

Isn't that the same thing ?
You buy some tools and start selling your time.
You buy a vendor ID and start selling product.

One is a consultant and the other a manufacturer, but the bottom line is
the same. You need to spend money in order to make money. If the
amount isn't justified, the risk is too high or the amount is simply too
much, is a decision each person/company makes for themselves.

> Those not in the position will probably never understand,
> but $1500 definitely keeps little guys from just trying something.

That may be the case for some on this list, but I doubt it applies to you
or me. We are in a position to decide that we would rather spend that on
a vacation, the garden or other development tools. We can hardly
blame the usb org for that.

Steve.


____________________________________________

2004\11\03@010259 by William Chops Westfield

face picon face
>
>> My biggest gripe is that the USB organization doesn't seem to care
>> about the little guys.
>
>
This is an exactly accurate statement.  The USB organization DOESN'T
care about the little guys.  It's probably not even part of their
charter to care about the little guys.  They serve the PC and PC
peripheral markets, and the user and support base of consumer personal
computers.  It's somewhat unfortunate that a side-effect is the loss
of legacy ports (especially serial) from PCs, but those were only
present "automatically" for a relatively brief time period.  You can
still add a PCI-based serial card, or a USB->serial converter to your
PC for no more than it cost to do so before the serial port became a
standard part of the PC chipsets.  Likewise parallel.
It's somewhat MORE unfortunate that embedded hardware suppliers are
finding it harder to lunch off the leavings of the PC peripheral market;
serial modems, or printers that can talk ascii, or barcode readers that
talk to PS2 or serial ports are increasingly hard to find.

BillW

____________________________________________

2004\11\03@023040 by Wouter van Ooijen

face picon face
>  When I say that $1500 should be affordable if someone's serious
>  enough to be successful making USB devices,

Nonsense. I sell a programmer, serial interface. I'd like to make an USB
version, but I can't rely on it to be a big hit. Development costs are
small. Now that $1500 is a real show stopper.

Wouter van Ooijen

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


____________________________________________

2004\11\03@044440 by Alan B. Pearce

face picon face
>> > But an ICD2 is ~ $150, an the C30 full-functional demo is free!
>
>The demo is only free for a limited time. After that, you buy it.

Has anyone tried to see if it can be reloaded like the C18 compiler?

____________________________________________

2004\11\03@051106 by steve

flavicon
face
On 3 Nov 2004 at 8:30, Wouter van Ooijen wrote:

> >  When I say that $1500 should be affordable if someone's serious
> >  enough to be successful making USB devices,
>
> Nonsense. I sell a programmer, serial interface. I'd like to make an
> USB version, but I can't rely on it to be a big hit. Development costs
> are small. Now that $1500 is a real show stopper.

Which is why FTDI have created the ideal product and support structure
to suit your needs. You have no need for vendor ID.

Steve.
 


____________________________________________

2004\11\03@075028 by olin_piclist

face picon face
Andrew Warren wrote:
>     As I said in a previous message, I believe that the ID-only fee,
>     which used to be $200, was raised to $1500 so as to narrow the
>     gap between it and the ID-with-testing fee of $2500.
>
>     Narrowing the gap encourages people to pay the $2500 and actually
>     test their devices before unleashing them on the world. Testing
>     is a good thing.
>
>     Charging less for Vendor IDs and making people "pay their own
>     way" by charging more for testing would DISCOURAGE people from
>     testing their devices.  That would be a bad thing.

Perhaps, but I strongly resent this holier than thou attitude and the fact
that the decision is not up to me.  I'm not advocating bad products, but how
much testing I chose to do is between me and my customers.

>     Think.  If we really wanted to keep you out, wouldn't we have
>     conspired with the USB-IF to charge a lot more than $1500?

I'm not saying its deliberate, only that they don't care and have no reason
to care from their point of view.

>     Well, you can't really blame us... If you're planning to sell so
>     few USB devices that you won't be able to afford the $1500 VID
>     fee, how much money can we expect to make from you on sales of
>     our microcontrollers?
>
>     Much as I'd like for everyone in the world to be using our
>     parts, it just isn't cost-effective to support thousands of 10-
>     to 100-unit/year USB customers when the same revenue can be made
>     from one medium-sized customer.

The opposite of this thinking got Microchip to where they are.  Applications
will crawl out of the woodwork when you reduce the cost to entry.  Most of
them will stay low volume, but a few will break thru to higher volumes.  It
also creates a general groundswell with more people versed in your products,
so that they feel comfortable with them when they do have a high volume
design.

Face it, most microcontrollers are rougly similar.  To get market share, you
need a lot of engineers already comfortable with your products when they
start on that high volume design.

For example, if I were starting a mid to high volume USB design today, I'd
probably use the new 18F parts and not waste a lot of time looking around,
assuming they can do the job.  You probably have parts that could do the job
too, but I'm already familiar with PICs, have my tool chain all set up, and
have a great comfort level with them.  On the other hand, my limited
experience with Cypress has been somewhat negative.

For a very high volume design you spend time looking around, you pick
whatever is $.01 cheaper than the next solution, and tool costs and comfort
don't matter.  However, those kind of designs are rare indeed.  Of all the
many dozens of microcontroller projects I've done, only was designed for
that kind of volume up front.  By the way, that one ended up using a PIC
10F204.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com
____________________________________________

2004\11\03@114003 by Joe McCauley

picon face
Olin,

You come across to me as the kind of guy who would aim for 110%+ certainty
that your product would work as advertised. Sadly many others are not so
careful. In this situation bad products are inevitable and relativly few of
these would 'taint' the standard. I'm always amazed at the way USB stuff
works out of the box. It would be a pity if a flood of bad peripherals left
people leery of the standard.

Incidently, does anyone know of a reliable USB to 16 bit ISA slot adapter? I
have several old ISA PC acquisition cards I'd like to use in a new system.

Joe



{Original Message removed}

2004\11\03@120051 by No Religion

flavicon
face
At 16.39 2004.11.03 +0000, you wrote:
>Olin,
>
>You come across to me as the kind of guy who would aim for 110%+ certainty
>that your product would work as advertised. Sadly many others are not so
>careful. In this situation bad products are inevitable and relativly few of
>these would 'taint' the standard. I'm always amazed at the way USB stuff
>works out of the box. It would be a pity if a flood of bad peripherals left
>people leery of the standard.
>
>Incidently, does anyone know of a reliable USB to 16 bit ISA slot adapter? I
>have several old ISA PC acquisition cards I'd like to use in a new system.

Here it is:
http://www.arstech.com/usbisa.htm

Greets,
N.R.


>
>Joe

____________________________________________

2004\11\03@122341 by Bob Axtell

face picon face
I think Olin is right on the mark. The problem in the past is that there
were no REAL standards. $1500 is a lot of money to
small players, but $1500 to maintain a high standard seems reasonable to
me; SONY Mem stick specs originally needed
a $10K USD investment in 1998.

Look at the mess that serial and parallel ports became. But yes, the
protocol IS needlessly complex. I
don't have a binder large enough to hold my USB2.0 reference copy.

I think we can keep making RS232 tools that work well, then buy $19
adaptors to convert them to USB.
Thats what I do and it works like a charm (It also allows me to use DOS
machines, which are popular as
production test equipment across the border and all have 2 serial ports).

--Bob

Joe McCauley wrote:

{Quote hidden}

>{Original Message removed}

2004\11\03@144625 by Andrew Warren

flavicon
face
Wouter van Ooijen <.....piclistspam_OUTspammit.edu> wrote:

> >  When I say that $1500 should be affordable if someone's serious
> >  enough to be successful making USB devices,
>
> Nonsense. I sell a programmer, serial interface. I'd like to make
> an USB version, but I can't rely on it to be a big hit.

   Ah, sorry... "Serious enough to be successful" was unclear
   shorthand for "serious enough, with a product that'll be popular
   enough, to be successful".

   I was simply saying that it's no different from any other
   business decision:  If you don't think you'll make more than
   $1500 selling USB devices, then don't sell them.

   -Andy

=== Andrew Warren -- TakeThisOuTaiw.....spamTakeThisOuTcypress.com
=== Principal Design Engineer
=== Cypress Semiconductor Corporation
===
=== Opinions expressed above do not
=== necessarily represent those of
=== Cypress Semiconductor Corporation

____________________________________________

2004\11\03@153848 by Wouter van Ooijen

face picon face
>     I was simply saying that it's no different from any other
>     business decision:  If you don't think you'll make more than
>     $1500 selling USB devices, then don't sell them.

Right. So you agree that this policy has effectively killed small-scale
USB development.

Wouter van Ooijen

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


____________________________________________

2004\11\03@181204 by p.cousens

flavicon

Who is approving noreligion's posts?

____________________________________________

2004\11\04@130033 by olin_piclist

face picon face
Joe McCauley wrote:
> You come across to me as the kind of guy who would aim for 110%+
> certainty that your product would work as advertised.

I didn't say I wouldn't test my USB product, I certainly would and quite
carefully at that.  My objection is that the USB organization makes me pay
for some of their testing their way to get a vendor ID.  I think I could
test a USB product effectively without attending one of their plugfests, but
in any case that should be for me to decide.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com
____________________________________________

2004\11\04@133824 by Dave VanHorn

flavicon
face
At 01:00 PM 11/4/2004, Olin Lathrop wrote:

>Joe McCauley wrote:
> > You come across to me as the kind of guy who would aim for 110%+
> > certainty that your product would work as advertised.
>
>I didn't say I wouldn't test my USB product, I certainly would and quite
>carefully at that.  My objection is that the USB organization makes me pay
>for some of their testing their way to get a vendor ID.  I think I could
>test a USB product effectively without attending one of their plugfests, but
>in any case that should be for me to decide.

You certainly can. You just can't use the USB logo.

____________________________________________

2004\11\04@164528 by Peter L. Peres

picon face

On Wed, 3 Nov 2004, Andrew Warren wrote:

> Wouter van Ooijen <TakeThisOuTpiclistKILLspamspamspammit.edu> wrote:
>
>>>  When I say that $1500 should be affordable if someone's serious
>>>  enough to be successful making USB devices,
>>
>> Nonsense. I sell a programmer, serial interface. I'd like to make
>> an USB version, but I can't rely on it to be a big hit.
>
>    Ah, sorry... "Serious enough to be successful" was unclear
>    shorthand for "serious enough, with a product that'll be popular
>    enough, to be successful".
>
>    I was simply saying that it's no different from any other
>    business decision:  If you don't think you'll make more than
>    $1500 selling USB devices, then don't sell them.

You probably mean 'make more than $6000 at least', per year, likely in the
first year or two. You do not expect them to sell stuff just to break even
?

Peter
____________________________________________

More... (looser matching)
- Last day of these posts
- In 2004 , 2005 only
- Today
- New search...