Hi Bob,
And we are all still suffering from the lack of real I/O from modern processors !!!!! So we use AVR, PIC, etc. to do the real work :).
Lots of fun :).
John
toggle quoted message
Show quoted text
On Sun, 24 Jun 2018, Bob Macklin wrote:
Many years ago I built a SPI programmer using a standard serial port. It
worked fine with real serial port. But when I tried using a USB serial port
it failed.
This programmer worked by BIT BANGING the RS-232 handshake lines. To make it
work with USB serial port I had to loop back the clock line with one of the
RS-232 input lines. I had to set the clock output to the port then wait for
the clock to come back in on the input bit.
Bob Macklin
K5MYJ
Seattle, Wa.
"Real Radios Glow In The Dark"
----- Original Message -----
From: <jmh6@...>
To: <[email protected]>
Sent: Sunday, June 24, 2018 5:27 AM
Subject: Re: [QRPLabs] avrdude firmware writing question
Hi,
The point I was trying to make is that the unit on the far end may have
time constraints that need to be honored. So a random pause especially if
long enough in what is being sent can lead to unreliable operation.
Probably enough said at this point....
I am impressed with Han's work in QCX. A REALLY NICE piece of code and
very carefully done :).
John
On Sat, 23 Jun 2018, Old Dog wrote:
It? not ?real time? if it is not hung on an hardware interrupt. Anything
else and
the OS decides how to schedule things.
On Sat, Jun 23, 2018 at 2:18 PM, <jmh6@...> wrote:
I am a low level designer - assembly, etc. so I have seen a lot
of pretty heavy stuff. I try to use some independent device to
measure what is really going on. CPU usage stuff is mostly
fiction. There is nothing like seeing a nice real-time data
stream interrupted for a second or two while the CPU is off
doing something else. And to make things much worse, all the CPU
produced numbers look like all is well :) :).
Lots of fun :).
John
On Sat, 23 Jun 2018, post.marcel@... wrote:
Hi John,
> You may be charting territory that even the chip
maker hasn't figured out
:).
Possibly..
Many months ago I noticed that there were some
issues with UART over USB
when Debian went from version 8 (Jessie) to 9
(Stretch), but back then I
believed I also had a programmer that wasn't happy.
I then went with the
Sparkfun one and things 'just worked'. Well.. up
until now. Now that I can
find out if the problem lies with the newer version
of avrdude, I may see if
I can get an older version and compile it from a tar
ball and see if that
makes a difference. Timing could still be the
elephant in the room, but at
least I can confirm there's nothing wrong with the
programmer, the chips and
the files Hans provided. Sigh of relief, really..
:-)
--
