Reply by Mark Borgerson●January 25, 20092009-01-25
In article <6u3ut3Fdhh53U1@mid.individual.net>, Faweglassenlk@falk-
willberg.de says...
> Mark Borgerson schrieb:
> ...
> > I've also done a bit-banged uart for debug output on an early
> > small-package 8051-compatible processor. It didn't need
> > a timer or interrupts. Getting input is many times more difficult
> > and is much easier with timers and interrupts.
>
> Input can be quite simple, if you make the uC under Test a SPI-master.
> Needs little code only and can handle Rx and Tx reliably.
>
> As "industry standard PCs" are becoming expensive arcade games with lots
> of Giga-something, that can run crashbangflashIII with 500fps, but do
> not have any useful interfaces any longer, any uC with SPI and UART plus
> a RS232/USB converter can do this tough SPI-job for you ;-)
>
> > I don't know of any bit-banged USB output implementations. If
> > there is one, I suspect it takes many KBytes of code and a
> > very fast processor.
>
> 12MHz and 1400 Bytes: http://www.obdev.at/products/avrusb/index-de.html
>
Very interesting. Although it's only a low-speed link, it could be
useful as a debugging interface. Still, I'm glad my 8KB, 8MHz
MSP430 system has a UART!
Mark Borgerson
Reply by Falk Willberg●January 25, 20092009-01-25
Mark Borgerson schrieb:
...
> I've also done a bit-banged uart for debug output on an early
> small-package 8051-compatible processor. It didn't need
> a timer or interrupts. Getting input is many times more difficult
> and is much easier with timers and interrupts.
Input can be quite simple, if you make the uC under Test a SPI-master.
Needs little code only and can handle Rx and Tx reliably.
As "industry standard PCs" are becoming expensive arcade games with lots
of Giga-something, that can run crashbangflashIII with 500fps, but do
not have any useful interfaces any longer, any uC with SPI and UART plus
a RS232/USB converter can do this tough SPI-job for you ;-)
> I don't know of any bit-banged USB output implementations. If
> there is one, I suspect it takes many KBytes of code and a
> very fast processor.
Reply by Mark Borgerson●January 25, 20092009-01-25
In article <_b6dnS8mN5fKA-HUnZ2dnUVZ_sDinZ2d@posted.usinternet>,
grante@visi.com says...
> On 2009-01-25, Gene S Berkowitz <first.last@verizon.net> wrote:
>
> > The real point is, virtually any micro with a timer and an I/O port can
> > at least bit-bang a PC-acceptable serial port. That isn't true of USB.
>
> You don't even need a timer if you're willing to put up with a
> few restrictions. On some very low-spec uControllers, I've
> done a "UART" output for debugging/logging purposes that relies
> on nothing but repeatable instruction timings and takes no more
> than a couple-dozen lines of code. (Receiving is a bit more
> work, but still trivial compared to USB.)
>
>
I've also done a bit-banged uart for debug output on an early
small-package 8051-compatible processor. It didn't need
a timer or interrupts. Getting input is many times more difficult
and is much easier with timers and interrupts.
I don't know of any bit-banged USB output implementations. If
there is one, I suspect it takes many KBytes of code and a
very fast processor.
Mark Borgerson
Reply by Mark Borgerson●January 25, 20092009-01-25
In article <MPG.23e65d6c9f0bdb06989949@news.verizon.net>,
first.last@verizon.net says...
> In article <MPG.23e244f8c37ac8929896ac@news.motzarella.org>,
> mborgerson@comcast.net says...
> > In article <sg4NenAMkCeJFA4h@phaedsys.demon.co.uk>, chris@phaedsys.org
> > says...
>
> > > Many (most?) mcu , and certainly the smaller ones don't have USB on
> > > board. But the all have a serial port (or two)
> >
> > I hate to burst your bubble---there are whole series of MCUs with no
> > hardware serial interfaces. Take a look at the Microchip PIC10 and
> > PIC12 series, for example. I'm sure other processor vendors have
> > low-end chips with no serial interfaces also.
>
> The real point is, virtually any micro with a timer and an I/O port can
> at least bit-bang a PC-acceptable serial port. That isn't true of USB.
>
LOL! So many of these threads would be a lot shorter if people
would just state their real point properly in the first place!
Mark Borgerson
Reply by Grant Edwards●January 25, 20092009-01-25
On 2009-01-25, Gene S Berkowitz <first.last@verizon.net> wrote:
> The real point is, virtually any micro with a timer and an I/O port can
> at least bit-bang a PC-acceptable serial port. That isn't true of USB.
You don't even need a timer if you're willing to put up with a
few restrictions. On some very low-spec uControllers, I've
done a "UART" output for debugging/logging purposes that relies
on nothing but repeatable instruction timings and takes no more
than a couple-dozen lines of code. (Receiving is a bit more
work, but still trivial compared to USB.)
--
Grant
Reply by Gene S. Berkowitz●January 25, 20092009-01-25
In article <MPG.23e244f8c37ac8929896ac@news.motzarella.org>,
mborgerson@comcast.net says...
> In article <sg4NenAMkCeJFA4h@phaedsys.demon.co.uk>, chris@phaedsys.org
> says...
> > Many (most?) mcu , and certainly the smaller ones don't have USB on
> > board. But the all have a serial port (or two)
>
> I hate to burst your bubble---there are whole series of MCUs with no
> hardware serial interfaces. Take a look at the Microchip PIC10 and
> PIC12 series, for example. I'm sure other processor vendors have
> low-end chips with no serial interfaces also.
The real point is, virtually any micro with a timer and an I/O port can
at least bit-bang a PC-acceptable serial port. That isn't true of USB.
--Gene
Reply by Falk Willberg●January 25, 20092009-01-25
bitbarf@gmail.com schrieb:
> On Jan 21, 10:34 pm, Paul Keinanen <keina...@sci.fi> wrote:
>> There are expensive special USB/RS485 converters that generate the
>> data direction control internally, but this is not of much use when
>> controlling a radio transceiver.
>>
> I found a USB/RS485 converter for $49.95 at
> http://www.usconverters.com/index.php?main_page=product_info&products_id=197
After looking around, I made my own USB/RS485 converter. FTDI's FT232
has a TXDEN-output. This is designed for RS485 Rx/Tx switching. Works
perfectly.
> It has automatic send control, which probably means the first start
> bit that comes along triggers a one-shot that turns on the driver.
>
> There is a Win32 call that controls RTS etc. but I don't have the name
> of it handy. Have you measured the delay between issuing the call and
> seeing the change in the RTS level?
I tried the RTS-line as well. The delay was changing, so I found it
unusable, as the device, my application was talking to, responded within
<100ms which leaded to collisions.
HTH,
Falk
Reply by ●January 24, 20092009-01-24
On Jan 21, 10:34=A0pm, Paul Keinanen <keina...@sci.fi> wrote:
>
> There are expensive special USB/RS485 converters that generate the
> data direction control internally, but this is not of much use when
> controlling a radio transceiver.
>
I found a USB/RS485 converter for $49.95 at
http://www.usconverters.com/index.php?main_page=3Dproduct_info&products_id=
=3D197
It has automatic send control, which probably means the first start
bit that comes along triggers a one-shot that turns on the driver.
There is a Win32 call that controls RTS etc. but I don't have the name
of it handy. Have you measured the delay between issuing the call and
seeing the change in the RTS level?
-Brad
Reply by Don Seglio●January 23, 20092009-01-23
On 1/23/2009 3:47 PM, Albert van der Horst wrote:
> In article<4978D795.130F9E3@yahoo.com>,
> CBFalconer<cbfalconer@maineline.net> wrote:
>
>> Chris H wrote:
>>
>>> Paul Carpenter<paul@pcserviceselectronics.co.uk> writes
>>>
>>>> chris@phaedsys.org says...
>>>>
>>>>
>> ... snip ...
>>
>>>>> The problem is that many modern laptops don't have a serial
>>>>> port. Most desktop PC's still have one AFAIK but lots of USB
>>>>> so many consumer items have USB.
>>>>>
>>>> About 6 months ago (maybe 9) one of my custoemrs deliberately
>>>> bought a HP laptop because it had a Serial port and XP.
>>>>
>>> I did exactly the same because I needed a serial port.
>>>
>> And I bought a used IBM Thinkpad T30 for the same reason.
>>
>
> And it was one of the must-have's when I finally choose a
> motherboard as a successor to the 90 Mhz Plato board.
>
>
>> --
>> [mail]: Chuck F (cbfalconer at maineline dot net)
>> [page]:<http://cbfalconer.home.att.net>
>>
Not a laptop but I just put together a PC for use in a Linux programming
project using the Intel Atom Dual core 330 mother board, it has USB,
1GHz Ethernet, parallel, and a serial port, the whole PC, motherboard,
small case, 2GB RAM, 500GB HD, and DVD burner cost me $175. The
processor is X64 compatible and has Virtual Technology support. I have
Ubuntu X64 version installed and will be using IForth for the project.
It's a nice little quiet machine on the cheap with old port support,
sometimes that comes in handy.
--
Cecil
K5NWA
www.k5nwa.com www.qrpradio.com
< http://parts.softrockradio.org/>
"Blessed are the cracked, for they shall let in the light."
Reply by Albert van der Horst●January 23, 20092009-01-23
In article <4978D795.130F9E3@yahoo.com>,
CBFalconer <cbfalconer@maineline.net> wrote:
>Chris H wrote:
>> Paul Carpenter <paul@pcserviceselectronics.co.uk> writes
>>> chris@phaedsys.org says...
>>>
>... snip ...
>>>
>>>> The problem is that many modern laptops don't have a serial
>>>> port. Most desktop PC's still have one AFAIK but lots of USB
>>>> so many consumer items have USB.
>>>
>>> About 6 months ago (maybe 9) one of my custoemrs deliberately
>>> bought a HP laptop because it had a Serial port and XP.
>>
>> I did exactly the same because I needed a serial port.
>
>And I bought a used IBM Thinkpad T30 for the same reason.
And it was one of the must-have's when I finally choose a
motherboard as a successor to the 90 Mhz Plato board.
>
>--
> [mail]: Chuck F (cbfalconer at maineline dot net)
> [page]: <http://cbfalconer.home.att.net>
> Try the download section.
--
--
Albert van der Horst, UTRECHT,THE NETHERLANDS
Economic growth -- like all pyramid schemes -- ultimately falters.
albert@spe&ar&c.xs4all.nl &=n http://home.hccnet.nl/a.w.m.van.der.horst