Monitor connections

[Rick C. Hodgin]

Is there a way to monitor signals in existing
wires? For example, with an oscilloscope and
probe I can watch voltage changes. Is there a
standard way to connect to an existing, working
device, and monitor and record its switching
over time? Such would seem to be desirable for
peeking at proprietary "wake up" chirps, and to
monitor device communications to establish its
protocol interface.

On old phone systems, a coil device could be
used to "copy" the phone signal without tapping
the voltage. I guess I'm looking for something
similar for general purpose wire use.

Best regards,
Rick C. Hodgin

 

[rickman]

On 12/12/2014 2:22 PM, Theo Markettos wrote:

Rick C. Hodgin <rick.c.hodgin@gmail.com> wrote:
Is there a way to monitor signals in existing
wires? For example, with an oscilloscope and
probe I can watch voltage changes. Is there a
standard way to connect to an existing, working
device, and monitor and record its switching
over time? Such would seem to be desirable for
peeking at proprietary "wake up" chirps, and to
monitor device communications to establish its
protocol interface.

Xilinx has ChipScope and Altera has SignalTap. These are logic analysers
that you can put inside the FPGA and attach to signals to monitor the state
of your design.

There are some caveats:
Typically changing the probe state requires a recompile of your design
(which can take hours)
The amount of state you can record is limited by the internal memory in your
device

but it's still far better than trying to route signals outside and using a
real logic analyser.

I've done the logic analyzer thing before. A board I have in production
has 9 signals it can drive on a 10 pin connector to see internal state.
To be most useful I used a mux to select which internal signals drive
these outputs. There is an undocumented control register in the FPGA to
control the mux. But even with this if I need a signal that I hadn't
planned for, I have to reprogram the FPGA.

To Rick C.'s question, I have never seen a probe intended to save you
the trouble of accessing one end of the wire. A needle can be used to
pierce the insulation if needed.

--

Rick

 

[Tom Gardner]

On 16/12/14 03:24, rickman wrote:

To Rick C.'s question, I have never seen a probe intended to save you the
trouble of accessing one end of the wire.

A current probe can do that, with significant limitations, of course!

 

[GaborSzakacs]

Tom Gardner wrote:

On 16/12/14 03:24, rickman wrote:

To Rick C.'s question, I have never seen a probe intended to save you the
trouble of accessing one end of the wire.

A current probe can do that, with significant limitations, of course!

or a needle stuck in the middle of the wire (assuming we're
talking discrete wires and not a trace inside an IC). There
are also insulation displacement quick-connects made for
tapping into automotive wiring. On the other hand I've heard
the OP has moved on, so we'll probably never really know what
the intent of the question really was.

 

[Arne Pagel]

Not sure what you are really looking for,
there are some methods around trying to decrypt the bitstream:

https://www.cosic.esat.kuleuven.be/publications/article-737.pdf

http://perso.uclouvain.be/fstandae/PUBLIS/12.pdf

 

[rickman]

On 12/16/2014 6:56 AM, Tom Gardner wrote:

On 16/12/14 03:24, rickman wrote:

To Rick C.'s question, I have never seen a probe intended to save you the
trouble of accessing one end of the wire.

A current probe can do that, with significant limitations, of course!

A current probe measures current, not voltage. It is a pretty poor way
to measure signals.

--

Rick

 

[Theo Markettos]

rickman <gnuarm@gmail.com> wrote:

A current probe measures current, not voltage. It is a pretty poor way
to measure signals.

You can also capacitively couple signals into a probe, but you only get the
higher frequencies due to the filtering effect. That might be OK for GHz
signals, but not for DC.

Theo