Scopes / Debugging tools
Alex
Going to be doing mostly pic , 8051 , avr
and cpld work to start with.
Some dsp , fpga and digital tv work eventually (mid to end next year).
Looking at dual channel dso up
to 200 MHz, preferably something that
can talk to a pc.
(buying second hand)
Something like
Tektronix 2430
or
HP545xx series
Any recommendations or advice appreciated
Also
whats tools and tricks do you use the most for debugging ?
oscilliscope + passive probes ?
Thank you
Alex
CBFalconer
>
> Finally looking at getting an oscilliscope.
>
> Going to be doing mostly pic , 8051 , avr and cpld work to start
> with. Some dsp , fpga and digital tv work eventually (mid to end
> next year).
>
> Looking at dual channel dso up to 200 MHz, preferably something
> that can talk to a pc.
>
> (buying second hand)
>
> Something like
> Tektronix 2430
> or
> HP545xx series
>
> Any recommendations or advice appreciated
The most important thing in a scope IMO is the triggering
facilities. Make sure you have a delayed sweep available, which
can either trigger after delay or run after delay. Make sure you
have an external trigger input available. It is handy to have an
internal way of selecting the trigger from an input channel or the
external input.
You should always have at least two, preferably four, input
channels. Make sure it is easy to switch from alternate sweeps to
chopped sweeps. Even at high rates you can get funny results from
synchronization to alternate sweep mechanisms.
For logic work you can get by with 1x probes, but 10x are
generally safer. You usually want the 1x for the external trigger
input.
You don't really need fancy stuff like on-screen sweep rates and
v/div markers.
Now you are ready to tackle almost anything :-)
--
Chuck F (cbfal...@yahoo.com) (cbfal...@worldnet.att.net)
Available for consulting/temporary embedded and systems.
<http://cbfalconer.home.att.net> USE worldnet address!
Jonathan Kirwan
I'm thinking 54645D HP/Agilent mega-zoom, given what you are fiddling with. It's a 2ch DSO, with a 16 digital input logic analyzer. It's been discontinued and replaced by a newer model (with a floppy, standard.) So it may be available in the used market, now.
I have one and like it a lot. I also have a Tek 4ch analog scope, as well, with a collection of probes (several of the standard 10x, 1x, HV probe, and lots of grabber clips.)
There are times for having each, so I'm not planning on getting rid of either any time soon.
Jon
Roger DiPaolo
You have already got some good advice on scopes, so I will try to help with
the debugging question.
I have found that good debugging tools can make or break a project,
depending on the complexity of the project.
First, consider your budget. I strongly suggest to not scrimp on debugging
tools, but often one doesn't have much say over what management has decided
to allow you to spend for project tools (fortunately, my employer is not in
that group).
If you can afford it, an emulator (or background debug device), logic
analyzer, and oscilloscope are the best combination. If you can acquire and
master these three tools for your project then no bug should be able to hide
from you for long.
The biggest debugging aids that I have found are the ability to single-step
through your code, set breakpoints, and examine variable/register/memory
contents. All of these can be accomplished by using an emulator, background
debugger (for those micros which don't have emulator support but do have a
dedicated debug port), or target resident monitor (which is the weakest of
the three, because it is generally very intrusive - but still good if it is
all you have).
A logic analyzer is best for execution tracing (although some emulators do
this as well), and accurately measuring timing relationships both between
instructions and between instructions and digital events occurring in your
system. I have very successfully debugged some very complex high-speed
systems using only a logic analyzer. In most cases I consider the logic
analyzer to be the "big gun" to be brought in when all else fails.
Of course, I have also been in many situations where I have only had a
serial port to write printf() statements to, and maybe a few LEDs hanging
off of an I/O port. Don't discount the power of these methods to be able to
help you solve most bugs - in many ways they are the not only the cheapest,
but by far the biggest "bang for the buck" you can get in debugging.
Regardless of how well I am funded for tools on a given project I always
insist on numerous LEDs and test points (on key signals) on my target
systems, and a serial port to dedicate to debug output is worth its weight
in gold if you have a spare one available.
Well, this is the VERY abridged version of "Embedded Debugging Tools and
Tricks", I could keep going all day - but unfortunately I need to get back
to work on my own project.
I hope this helps. Best of luck Alex!
Roger DiPaolo
Staff Software Engineer
Affymetrix
"Alex" <alxx.@..ihug..com..au> wrote in message
news:apagis$si6$1...@lust.ihug.co.nz...
al
<posting via google due to isps news server being
it useful crappy self and partially crashing, so can't post>
Thanks guys for the advice.
Greatly appreciated.
Currently a uni student(3rd year).
Certainly beats shift work and crawling around in aircraft.
Or being inside an engine at 4am in the middle of winter.
Going to be doing some contract work over the uni summer holidays.
Code monkey is probably the best way to describe it.
For one place they want their assembly converted to c .
And don't want to pay that much.
Only want me to use free compilers.
At least if I have a payment issuse I can get the industrial liason
department at uni to chase it up for me.
Count the work as industrial experiance.
Got myself a second hand 400MHz 4 channel analog scope
looking around for a digital scope.
For a digital scope the very highest frequency I can
see my self needing to look at in the next year or few is a 150 - 200 MHz
with the average being around 4 to 20MHz.
What bandwidth would I need to capture this properly?
Been over ten years since I have needed to use a dso.
rule of thumb is 5x the actual frequency ?
so for 200MHz need 1GHz
Logic analyser wise what would you look for ?
what brands ? features etc ?
Also how do pc based instruments compare to standalone.
The second hand equipment market is rather limited
here (Australia) compared to what you have available
in the US.
Thank you
Alex Gibson
PaulH
when you get bored you can always spray your colleagues
on the back of the neck.
"al" <al...@ihug.com.au> wrote in message
news:9c0bc09.02102...@posting.google.com...
HugoHH
or up. Is any body have experience with it. I think its WEB is
WWW.linkinstruments.com I bought a programmer from them 15 years ago,
they used have very good support
Sorry ans. so late,
HugoHH
got experience with it. I think its WEB is www.linkinstruments.com I
bought a programmer from them 15 years ago. They got very good support then.
Sorry ans. so late. I always look this group when i got time
></PRE></HTML>
Chris Hills
<roger_...@affymetrix.com> writes
>Hi Alex,
>
>You have already got some good advice on scopes, so I will try to help with
>the debugging question.
>
>I have found that good debugging tools can make or break a project,
>depending on the complexity of the project.
>
>First, consider your budget. I strongly suggest to not scrimp on debugging
>tools, but often one doesn't have much say over what management has decided
>to allow you to spend for project tools (fortunately, my employer is not in
>that group).
>
>If you can afford it, an emulator (or background debug device), logic
>analyzer, and oscilloscope are the best combination. If you can acquire and
>master these three tools for your project then no bug should be able to hide
>from you for long.
>Well, this is the VERY abridged version of "Embedded Debugging Tools and
>Tricks", I could keep going all day - but unfortunately I need to get back
>to work on my own project.
Hi
I have a paper on embedded debuggers. It covers monitors to ICE its at
http://Quest.phaedsys.org
You can't beat a good ice... but that said not all projects require one.
Regards
Chris
/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\
\/\/\/\/\ Chris Hills Staffs England /\/\/\/\/\
/\/\/ ch...@phaedsys.org www.phaedsys.org \/\/
\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/
Rich Webb
I use USB version of one of their LA4000-series analyzers on Win98.
They are a reasonable value. It's quite handy to be able to work on the
source, upload the code to the target, and then put it through its paces
all from one keyboard/display. Saves valuable benchtop space. It's also
handy to be able to save a reference captures to disk or printer without
needing to sneaker net a floppy or find a way to get HPIB onto a
notebook PC.
The interface software could use some work, although these are
annoyances not major problems:
I've never gotten hex-valued groups to print correctly; it's just
garbage characters.
The display of group hex values in the timing window is dependent only
on the zoom level, not on whether or not there's really room for the
value. That is, if I grossly oversample a signal in order to see more
detail at the leading edge, when zooming out to get more of the pulse
train on the screen the hex values aren't displayed ("not available in
this zoom") even though there may, in fact, be plenty of room for them.
The timing location table (where's A, B, trigger, etc.) really should be
right-justified. It would also be nice if the displayed base (micro,
milli, or whole seconds) could be adjusted.
The display fonts can't be changed.
The pop-up menu to insert a channel in the timing window does just that:
inserts. You can't "append" via that shortcut.
Trigger delays are based on the sample clock and available only up to
255 cycles. There are circumstances where it would be really handy to
have a much longer delay between levels.
Each trigger step is a rising or falling edge; you can't ask for "either
edge" at a trigger point. That is "trigger on the eighth data edge in
either direction after enable goes low" isn't possible.
All in all, an ok instrument for the price. I do wish he'd spend more
time on tweaking the software (over a year since the last significant
release) and less on full page adds in Circuit Cellar.
--
Rich Webb Norfolk, VA
al
"HugoHH" <hug...@aol.com> wrote in message
news:20021110092055...@mb-mf.aol.com...
Thanks.
Anyone used any of the acute logic analysers ?
acute
http://www.acute.com.tw/indexeng.htm
PKLA-1116 pocket model
http://www.acute.com.tw/english/spec-pkla1000.htm
LA-1032P
LA-1132P
LA-1164P
Thanks
Alex