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

Exact match. Not showing close matches.
PICList Thread
'[PIC] debug technique'
2008\11\10@134541 by Dr Skip

picon face

Seeing that some of the pc based logic analyzers are slow or small (like the
parallel port-with-only-a-buffer homebrew types), If one had a pic design that
was synchronous with the external hardware (same clock), or asynchronous
circuitry that didn't depend on timing at all from the pic, what would be the
cons in slowing the pic clock down to whatever freq would make tracing easy?
Would dropping from 10MHz to 10kHz be too much?

It obviously wouldn't show race conditions, capacitive loading effects, etc,
but even then, if the problems went away at slow speed and not at high speed it
would be telling. Does anyone do this? Any problems? If so, is it useful for
comm too, only scaled slower, out of the pic (as in SPI and such)?

-Skip

2008\11\10@180258 by Jinx

face picon face
> Would dropping from 10MHz to 10kHz be too much?

Skip, the PIC will run down as slow as you want, all the way down to
DC and single-stepping . I looked into this with regard to reducing power
consumption (below a certain frequency there's no further drop)

http://home.clear.net.nz/pages/joecolquitt/0f84rc.html

If you want really slow you'd have to use an external clock, not RC

> Does anyone do this? Any problems? If so, is it useful for comm too,
> only scaled slower, out of the pic (as in SPI and such)?

I tried it once and it did help sort out a problem. The PIC was running with
external RC, R being a pot. What also works is s/w switching in a different
R (eg with a 4066). This may have an application where the PIC is normally
running slowly to save power, but needs to speed up for comms. Not as
useful now as it once was, many current PICs having s/w selectable speeds


2008\11\10@233745 by Dr Skip

picon face
Thanks.

Jinx wrote:
>
> Skip, the PIC will run down as slow as you want, all the way down to
> DC and single-stepping . I looked into this with regard to reducing power
> consumption (below a certain frequency there's no further drop)
>
> home.clear.net.nz/pages/joecolquitt/0f84rc.html
>
> If you want really slow you'd have to use an external clock, not RC
>
>
> I tried it once and it did help sort out a problem. The PIC was running with
> external RC, R being a pot. What also works is s/w switching in a different
> R (eg with a 4066). This may have an application where the PIC is normally
> running slowly to save power, but needs to speed up for comms. Not as
> useful now as it once was, many current PICs having s/w selectable speeds
>
>

2008\11\11@074100 by Alan B. Pearce

face picon face
>Seeing that some of the pc based logic analyzers are slow ...

You think they are slow? Try using the current crop of Tektronix and Agilent
'scopes and logic analyzers!!! They seem to be based on windows XP, and take
an age to boot up. Then on the Tek TDS7000 unit we have, one has to have
administrator privileges for the software that handles the scope hardware to
run. The Agilent logic analyzer we have seems to use Excel as the interface
to the hardware, and to display the results.

take a measurement, let the thing trigger, then go make a cuppa tea while it
writes the display ...

I am most unimpressed with both instruments ...

2008\11\11@081208 by Xiaofan Chen

face picon face
On Tue, Nov 11, 2008 at 8:40 PM, Alan B. Pearce
<spam_OUTAlan.B.PearceTakeThisOuTspamstfc.ac.uk> wrote:
>>Seeing that some of the pc based logic analyzers are slow ...
>
> You think they are slow? Try using the current crop of Tektronix and Agilent
> 'scopes and logic analyzers!!! They seem to be based on windows XP, and take
> an age to boot up. Then on the Tek TDS7000 unit we have, one has to have
> administrator privileges for the software that handles the scope hardware to
> run. The Agilent logic analyzer we have seems to use Excel as the interface
> to the hardware, and to display the results.
>
> take a measurement, let the thing trigger, then go make a cuppa tea while it
> writes the display ...
>
> I am most unimpressed with both instruments ...

Is it really that bad? I've seen the old Tek running Windows 2000
and it was not too bad. We just bought a Agilent but I have not used
it yet.

The Agilent and Lecroy scopes running Windows XP are both
very good.

Xiaofan

2008\11\11@093316 by Alan B. Pearce

face picon face
>Is it really that bad? I've seen the old Tek running Windows 2000
>and it was not too bad. We just bought a Agilent but I have not
>used it yet.

I found it chronic. Maybe I am too used to 'scopes from the analogue days,
but these digital ones doo seem to take an interminable time to display the
result after a trigger. I used to think my DPO3000 series scope was slow,
and I was prepared to forgive that as having a slow processor to allow it to
operate on battery, but the DPO7000 just seems to take relatively ages. For
the price of it they could have included some decent CPU power.

2008\11\11@103225 by Joe Bento

face
flavicon
face
Alan B. Pearce wrote:
>> Is it really that bad? I've seen the old Tek running Windows 2000
>> and it was not too bad. We just bought a Agilent but I have not
>> used it yet.
>>    
>
> I found it chronic. Maybe I am too used to 'scopes from the analogue days,
> but these digital ones doo seem to take an interminable time to display the
> result after a trigger. I used to think my DPO3000 series scope was slow,
> and I was prepared to forgive that as having a slow processor to allow it to
> operate on battery, but the DPO7000 just seems to take relatively ages. For
> the price of it they could have included some decent CPU power.
>
>  
I use a Tektronix TDS-7104 scope at work, and really dislike it.  It's
Windows 2000 based, and tied to the company network.  To the network, it
is a computer.  The IT department does not readily grant administrative
privileges and cannot understand the difference between a piece of test
equipment and office computer.  So, to save or print waveforms is
problematic. That is an IT issue, though.

As Alan said, I do not like the perceptible delay it takes to trigger.  
I do not like the digitized waveform, as it does not have the sharpness
of an analog scope.  I intensely dislike the need to have to look at the
display essentially straight on, as there is virtually no side angle
vision.  The division grid is also all but invisible unless you look at
the display straight on.

Though only 100MHz bandwidth, give me my old reliable 465b.  Lovely scope!

Joe

2008\11\11@224022 by Xiaofan Chen

face picon face
On Tue, Nov 11, 2008 at 11:31 PM, Joe Bento <.....josephKILLspamspam@spam@kirtland.com> wrote:
> I use a Tektronix TDS-7104 scope at work, and really dislike it.  It's
> Windows 2000 based, and tied to the company network.  To the network, it
> is a computer.  The IT department does not readily grant administrative
> privileges and cannot understand the difference between a piece of test
> equipment and office computer.  So, to save or print waveforms is
> problematic. That is an IT issue, though.

That is an IT issue really. We have Windows XP based scope and
we have the admin right. We have the desktop and notebooks
and we are the admins. We do have separated lab network
and office network. The lab network is using static address.
The office network is using DHCP. The scopes are currently
not on the network though since it is quite convenient to use
flash drive.

> As Alan said, I do not like the perceptible delay it takes to trigger.
> I do not like the digitized waveform, as it does not have the sharpness
> of an analog scope.  I intensely dislike the need to have to look at the
> display essentially straight on, as there is virtually no side angle
> vision.  The division grid is also all but invisible unless you look at
> the display straight on.

I used a Tek running Windows 2k back in 2003 and I thought it
was fine (mainly doing DC/DC converter research then). Now
I have access to two Windows XP based Agilent 8000
series scopes and they are fine. The other lab has a
Lecroy which is even better. I am doing mostly analog
related work (DC/DC, ADC, DAC stuff) and a bit of debugging
for the MCUs (ARM7), so maybe they are not really high
speed compared to what you guys are doing.

Xiaofan

More... (looser matching)
- Last day of these posts
- In 2008 , 2009 only
- Today
- New search...