DIY Current Probe

[Chris]

Current probes for oscilloscopes are nice to have, but very pricey. I
vaguely recall someone somewhere sometime saying that you can make a
pretty good one up from certain computer scrap parts (some component in
the disk drive, perhaps?) and there may also be a sensor in the power
supply section that deals with power factor correction, I gather. I need
something that can sense down to say 10mA or better that's (ideally) fine
enough to sample individual pcb traces a few mm apart.
Any ideas?

[Lee]

Problem I can see is that the hall effect IC, which I presume is the
device suggested for the "probe" is likely to be a switch rather than an
analogue sensor.

There are some cheap sensors about:
http://arduinosensors.com/index.php/interfacing-a-hall-effect-current-sensor-with-an-arduino/

[Bob Minchin]

The read/write head from a disc drive would be a good start to
experiment with an AC current probe.
DC will be harder and need a linear Hall sensor.

[Chris]

On Sun, 07 Feb 2016 22:14:39 +0000, Bob Minchin wrote:

> The read/write head from a disc drive would be a good start to
> experiment with an AC current probe.
> DC will be harder and need a linear Hall sensor.

Yeah, that's it: the r/w head. Does anyone know what these things
comprise of?

[The Natural Philosopher]

No 'of' with 'comprise...'

A consists of lots of B's.
A comprises lots of B's.

GrammarNazi ;-)



--
If I had all the money I've spent on drink...
..I'd spend it on drink.

Sir Henry (at Rawlinson's End)

[Bob Eager]

On Mon, 08 Feb 2016 08:06:24 +0000, The Natural Philosopher wrote:

> On 08/02/16 00:24, Chris wrote:
>> On Sun, 07 Feb 2016 22:14:39 +0000, Bob Minchin wrote:
>>
>>> The read/write head from a disc drive would be a good start to
>>> experiment with an AC current probe.
>>> DC will be harder and need a linear Hall sensor.
>>
>> Yeah, that's it: the r/w head. Does anyone know what these things
>> comprise of?
>>
> No 'of' with 'comprise...'
>
> A consists of lots of B's.
> A comprises lots of B's.
>
> GrammarNazi ;-)

I hate it when people get that one wrong. The university PR people had
this error on theor website the other day.

[The Natural Philosopher]

"Their".

:) :)


--
He who shits in the road, will meet flies on his return.

"Mr Natural"

[Syd Rumpo]

Any reasonably new drive - say 10GB up - will use a magnetoresistive
read head. It's resistance changes with magnetic field strength.

I've used several different magnetoresistive devices, but not from disk
drives. The newest devices can be very sensitive - www.nve.com makes
several types, again, not for disk drives.

They have an application note which may be of interest...

http://www.nve.com/Downloads/SB-SA-001_Current_Measurement_using_GMR_Sensors.pdf

"GMR Sensors" - Giant magnetoresistance - is a good search term.

Cheers
--
Syd

[Bob Eager]

That's not a grammar error - it's a typo. Pot, kettle, black! :)

[tabb...@gmail.com]

On Monday, 8 February 2016 08:06:27 UTC, The Natural Philosopher wrote:
> On 08/02/16 00:24, Chris wrote:

> > Yeah, that's it: the r/w head. Does anyone know what these things
> > comprise of?
> >
> No 'of' with 'comprise...'
>
> A consists of lots of B's.
> A comprises lots of B's.
>
> GrammarNazi ;-)

obviously it's Bs not B's.


NT

[The Natural Philosopher]

What is the theorem/law that says for every attempt to correct a
spelling/grammar mistake, two more will be made...:-)

I stand corrected of course!

>
> NT
>


--
Truth welcomes investigation because truth knows investigation will lead
to converts. It is deception that uses all the other techniques.

[The Natural Philosopher]

smiley darling. You missed my spelling/grammar errors too.

I was merely highlighting the issue that is that the more you try to get
stuff right the more errors creep in.

[tabb...@gmail.com]

On Monday, 8 February 2016 12:13:57 UTC, The Natural Philosopher wrote:

> On 08/02/16 11:27, tabbypurr wrote:
> > On Monday, 8 February 2016 08:06:27 UTC, The Natural Philosopher wrote:
> >> On 08/02/16 00:24, Chris wrote:
> >
> >>> Yeah, that's it: the r/w head. Does anyone know what these things
> >>> comprise of?
> >>>
> >> No 'of' with 'comprise...'
> >>
> >> A consists of lots of B's.
> >> A comprises lots of B's.
> >>
> >> GrammarNazi ;-)
> >
> > obviously it's Bs not B's.
> >
>
> What is the theorem/law that says for every attempt to correct a
> spelling/grammar mistake, two more will be made...:-)

it's always the way!


NT

[Bob Minchin]

It is possible that the electronics used to interface to the magneto
resistive sensor in the HDD might be usable but could well be too
optimised to the application and so not linear enough. The interface and
maybe the sensor itself might only have a narrow bandwidth linked to the
data rate off the disk which will be fixed.
If you roll your own interface, the key will be optimal matching to the
impedance of the sensor to get best signal to noise as well as some form
of equalisation. Something like a RIAA curve turntable pickup preamp
might be a good starting point.

After a while you might be excused for starting to think that what you
get in a scope current probe is actually quite good value!

[dennis@home]

On 08/02/2016 16:49, Bob Minchin wrote:
8<

> After a while you might be excused for starting to think that what you
> get in a scope current probe is actually quite good value!

I have a ~£25 DC clamp on multimeter, it might be easier to start with
one of those and dismantle the clamp to see what sensor + amp is in there.

[tabb...@gmail.com]

On Monday, 8 February 2016 19:26:46 UTC, dennis@home wrote:

> I have a ~£25 DC clamp on multimeter, it might be easier to start with
> one of those and dismantle the clamp to see what sensor + amp is in there.

A current transformer


NT

[dennis@home]

On DC?

[Syd Rumpo]

That'll probably be a Hall effect sensor, but you need the conductor to
go _through_ the clamp, which is difficult for a PCB trace.

As others have said or implied, there's a reason why you can't buy a
cheap Chinese device on ebay - it's a hard problem.

Cheers
--
Syd

[The Natural Philosopher]

Not if Dennis is not mistaken, and it is DC...

That means hall effect usually.

>
> NT
>


--
Canada is all right really, though not for the whole weekend.

"Saki"

[whisky-dave]

On Monday, 8 February 2016 23:20:44 UTC, Syd Rumpo wrote:
> On 08/02/2016 19:26, dennis@home wrote:
> > On 08/02/2016 16:49, Bob Minchin wrote:
> > 8<
> >
> >> After a while you might be excused for starting to think that what you
> >> get in a scope current probe is actually quite good value!
> >
> > I have a ~£25 DC clamp on multimeter, it might be easier to start with
> > one of those and dismantle the clamp to see what sensor + amp is in there.
>
> That'll probably be a Hall effect sensor, but you need the conductor to
> go _through_ the clamp, which is difficult for a PCB trace.

http://www.rapidonline.com/test-measurement/tti-thurlby-thandar-iprober-520-positional-current-field-probe-85-4563

[dennis@home]

On 09/02/2016 09:37, whisky-dave wrote:
> On Monday, 8 February 2016 23:20:44 UTC, Syd Rumpo wrote:
>> On 08/02/2016 19:26, dennis@home wrote:
>>> On 08/02/2016 16:49, Bob Minchin wrote:
>>> 8<
>>>
>>>> After a while you might be excused for starting to think that what you
>>>> get in a scope current probe is actually quite good value!
>>>
>>> I have a ~£25 DC clamp on multimeter, it might be easier to start with
>>> one of those and dismantle the clamp to see what sensor + amp is in there.
>>
>> That'll probably be a Hall effect sensor, but you need the conductor to
>> go _through_ the clamp, which is difficult for a PCB trace.
>
> http://www.rapidonline.com/test-measurement/tti-thurlby-thandar-iprober-520-positional-current-field-probe-85-4563
>

For a DIY approach I can't help thinking two pins spaced apart on the
track and measure the volt drop along the track.
Needs a very hi impedance meter though.

[Bob Minchin]

Why on earth does that method need a high impedance input????

[whisky-dave]

Getting a good connection is the first problem, then thre's the small currents involved and hence the small voltage drop, so you'll end up measuring microvolts and microamps.

[tabb...@gmail.com]

On Tuesday, 9 February 2016 13:14:55 UTC, Bob Minchin wrote:

> dennis@home wrote:

> > For a DIY approach I can't help thinking two pins spaced apart on the
> > track and measure the volt drop along the track.
> > Needs a very hi impedance meter though.
> >
> Why on earth does that method need a high impedance input????

due to an excess of drugs maybe. Track drop measurement can be out by an order of magnitude, it has its uses but is largely too inaccurate.


NT

[The Natural Philosopher]

Needs a very LOW impedance very low voltage meter, though


--
All political activity makes complete sense once the proposition that
all government is basically a self-legalising protection racket, is
fully understood.

[dennis@home]

On 09/02/2016 13:34, whisky-dave wrote:

>> For a DIY approach I can't help thinking two pins spaced apart on
>> the track and measure the volt drop along the track.
>
> Getting a good connection is the first problem,

That's a lot easier with a high impedance, even a few hundred ohms isn't
going to have much effect on a 10 Mohm input.
You could experiment with some conductive plastic rather than metal probes.

[dennis@home]

On 09/02/2016 13:55, The Natural Philosopher wrote:
> On 09/02/16 12:47, dennis@home wrote:
>> On 09/02/2016 09:37, whisky-dave wrote:
>>> On Monday, 8 February 2016 23:20:44 UTC, Syd Rumpo wrote:
>>>> On 08/02/2016 19:26, dennis@home wrote:
>>>>> On 08/02/2016 16:49, Bob Minchin wrote:
>>>>> 8<
>>>>>
>>>>>> After a while you might be excused for starting to think that what
>>>>>> you
>>>>>> get in a scope current probe is actually quite good value!
>>>>>
>>>>> I have a ~£25 DC clamp on multimeter, it might be easier to start with
>>>>> one of those and dismantle the clamp to see what sensor + amp is in
>>>>> there.
>>>>
>>>> That'll probably be a Hall effect sensor, but you need the conductor to
>>>> go _through_ the clamp, which is difficult for a PCB trace.
>>>
>>> http://www.rapidonline.com/test-measurement/tti-thurlby-thandar-iprober-520-positional-current-field-probe-85-4563
>>>
>>>
>>>
>>
>>
>> For a DIY approach I can't help thinking two pins spaced apart on the
>> track and measure the volt drop along the track.
>> Needs a very hi impedance meter though.
>>
> Needs a very LOW impedance very low voltage meter, though
>
>

Don't be silly he is measuring voltage not current.

[whisky-dave]

On Tuesday, 9 February 2016 15:01:39 UTC, dennis@home wrote:
> On 09/02/2016 13:34, whisky-dave wrote:
>
> >> For a DIY approach I can't help thinking two pins spaced apart on
> >> the track and measure the volt drop along the track.
> >
> > Getting a good connection is the first problem,
>
> That's a lot easier with a high impedance,

the impedence needs to be very low and good connection to the surface of the copper without damaging it. Sure stick a pin in it why not, or course usually the track is covered in an insulative materail to protect it from oxidisation.

>even a few hundred ohms isn't
> going to have much effect on a 10 Mohm input.

irrelivent to the probe as it has very low impedence.

> You could experiment with some conductive plastic rather than metal probes.

Well when I see this sort of device in lidle we'll both understand.
I've never seen a plastic probe for measuring low voltages/current I have a 20KV one for high voltages.

[whisky-dave]

from the manual.


Unlike a conventional current probe, the I-prober 520 has an extremely low
insertion impedance and negligible stray capacitance.
This enables current observations and measurements to be made without
disturbance to the circuit.

[dennis@home]

On 09/02/2016 16:04, whisky-dave wrote:
> On Tuesday, 9 February 2016 15:01:39 UTC, dennis@home wrote:
>> On 09/02/2016 13:34, whisky-dave wrote:
>>
>>>> For a DIY approach I can't help thinking two pins spaced apart
>>>> on the track and measure the volt drop along the track.
>>>
>>> Getting a good connection is the first problem,
>>
>> That's a lot easier with a high impedance,
>
> the impedence needs to be very low and good connection to the
> surface of the copper without damaging it. Sure stick a pin in it why
> not, or course usually the track is covered in an insulative materail
> to protect it from oxidisation.

Why does it need to be low impedance?
Te current flow into the meter needs to be as low as possible, its the
voltage you are measuring not the current.

>
>
>
>> even a few hundred ohms isn't going to have much effect on a 10
>> Mohm input.
>
> irrelivent to the probe as it has very low impedence.

See above.

>
>
>> You could experiment with some conductive plastic rather than metal
>> probes.
>
> Well when I see this sort of device in lidle we'll both understand.
> I've never seen a plastic probe for measuring low voltages/current I
> have a 20KV one for high voltages.
>
>

Well you won't see one for current.

Just think about it..

you want to measure the voltage so you want as little current flow in
your measuring device as possible ie. you want high impedance.

So if you start with a 10 Mohm probe and you add a few hundred ohms in
series using conductive plastic its going to make sod all difference.
However it will prevent damage.

If you still don't believe me then look at a DVM...
low impedance shunt and high impedance meter to measure current and high
impedance meter to measure voltage.

[whisky-dave]

On Tuesday, 9 February 2016 16:15:05 UTC, dennis@home wrote:
> On 09/02/2016 16:04, whisky-dave wrote:
> > On Tuesday, 9 February 2016 15:01:39 UTC, dennis@home wrote:
> >> On 09/02/2016 13:34, whisky-dave wrote:
> >>
> >>>> For a DIY approach I can't help thinking two pins spaced apart
> >>>> on the track and measure the volt drop along the track.
> >>>
> >>> Getting a good connection is the first problem,
> >>
> >> That's a lot easier with a high impedance,
> >
> > the impedence needs to be very low and good connection to the
> > surface of the copper without damaging it. Sure stick a pin in it why
> > not, or course usually the track is covered in an insulative materail
> > to protect it from oxidisation.
>
> Why does it need to be low impedance?

Not sure but small impedences will affect the current as we are NOT measuirng voltage or current in the normal sense.

We were going to buy one here so I read the manual first,
and decided it'd be too cpomplex for what we were going to use it for,
so decided it wasnt worth the teaching budget money.
But said it could be good for research so if they wanted to buy it on their research grant that would be OK, but I'm not going to push for us to get one on the teaching grant.


from the manual availble on-line at rapid E.

Unlike a conventional current probe, the I-prober 520 has an extremely low
insertion impedance and negligible stray capacitance.

The I-prober 520 in use
The magnitude of the signal from a positional current probe is critically related
to its position relative to the conductor. The size of the conductor (e.g. the
width of a PCB track) also has a significant effect.
This means that the sensitivity of the I-prober has to be adjusted to match
the track width when quantitative measurements are required. A calibrator
within the control box enables sensitivity adjustment in conjunction with a
calibration graph.
The measurement result will also include other field effects present at the tip
of the probe and not just that coming from the current through the conductor.
This may include DC effects from adjacent magnetised components and from
the earths magnetic field, plus AC effects from transformers and other field
radiating sources.
Current in adjacent tracks, or tracks on
the opposite side of the PCB will also
affect the measurement.
There are solutions to these potential
problems. The unwanted DC can
be nulled out by observing the
measurement without power to the
circuit, whilst AC interference can be
attenuated using bandwidth filters. The
I-prober control box includes a wide
range DC offset control and switchable
filters.
Nevertheless, the use of the I-prober
520 requires interpretation based upon
a proper understanding of circuits and
systems. It is a tool for the professional engineer.

> Te current flow into the meter needs to be as low as possible, its the
> voltage you are measuring not the current.

doesn't that depend on how it works.

> >> even a few hundred ohms isn't going to have much effect on a 10
> >> Mohm input.
> >
> > irrelivent to the probe as it has very low impedence.
>
> See above.
>
> >
> >
> >> You could experiment with some conductive plastic rather than metal
> >> probes.
> >
> > Well when I see this sort of device in lidle we'll both understand.
> > I've never seen a plastic probe for measuring low voltages/current I
> > have a 20KV one for high voltages.
> >
> >
>
>
> Well you won't see one for current.
>
> Just think about it..

I did and that's why we don;t have one.

>
> you want to measure the voltage so you want as little current flow in
> your measuring device as possible ie. you want high impedance.

Only when using old style meters. Thsi uses a difernt methode and that is one iof teh reasons is expensive and overly complex to use for us.

>
> So if you start with a 10 Mohm probe and you add a few hundred ohms in
> series using conductive plastic its going to make sod all difference.
> However it will prevent damage.
>
> If you still don't believe me then look at a DVM...

I have over 200 DVMs well DMMs really. have 1/2 dozen LCRs too.

[The Natural Philosopher]

Which is why you will always be a technician.


--
Ideas are more powerful than guns. We would not let our enemies have
guns, why should we let them have ideas?

Josef Stalin

[dennis@home]

Don't start being an idiot again if you know any good reason for us to
believe what you said then spit it out.

Otherwise I am correct.

[dennis@home]

On 09/02/2016 17:11, whisky-dave wrote:
> On Tuesday, 9 February 2016 16:15:05 UTC, dennis@home wrote:
>> On 09/02/2016 16:04, whisky-dave wrote:
>>> On Tuesday, 9 February 2016 15:01:39 UTC, dennis@home wrote:
>>>> On 09/02/2016 13:34, whisky-dave wrote:
>>>>
>>>>>> For a DIY approach I can't help thinking two pins spaced
>>>>>> apart on the track and measure the volt drop along the
>>>>>> track.
>>>>>
>>>>> Getting a good connection is the first problem,
>>>>
>>>> That's a lot easier with a high impedance,
>>>
>>> the impedence needs to be very low and good connection to the
>>> surface of the copper without damaging it. Sure stick a pin in it
>>> why not, or course usually the track is covered in an insulative
>>> materail to protect it from oxidisation.
>>
>> Why does it need to be low impedance?
>
> Not sure but small impedences will affect the current as we are NOT
> measuirng voltage or current in the normal sense.

That might be why I said use high impedance then.

We are talking about a DIY solution to measuring the current.
Ie. the track is a low impedance shunt across a high impedance volt
meter. Its not rocket science.


The only variable is the impedance of the track.
With the current probe there are dozens of variables that have to be
taken into account.
I would use a constant current source to inject a reference current down
the track to calibrate whichever method I was using if it had to be
accurate.

>>>> even a few hundred ohms isn't going to have much effect on a
>>>> 10 Mohm input.
>>>
>>> irrelivent to the probe as it has very low impedence.
>>
>> See above.
>>
>>>
>>>
>>>> You could experiment with some conductive plastic rather than
>>>> metal probes.
>>>
>>> Well when I see this sort of device in lidle we'll both
>>> understand. I've never seen a plastic probe for measuring low
>>> voltages/current I have a 20KV one for high voltages.
>>>
>>>
>>
>>
>> Well you won't see one for current.
>>
>> Just think about it..
>
> I did and that's why we don;t have one.
>
>
>>
>> you want to measure the voltage so you want as little current flow
>> in your measuring device as possible ie. you want high impedance.
>
> Only when using old style meters. Thsi uses a difernt methode and
> that is one iof teh reasons is expensive and overly complex to use
> for us.
>
>
>>
>> So if you start with a 10 Mohm probe and you add a few hundred ohms
>> in series using conductive plastic its going to make sod all
>> difference. However it will prevent damage.
>>
>> If you still don't believe me then look at a DVM...
>
> I have over 200 DVMs well DMMs really. have 1/2 dozen LCRs too.

Having them and understanding them is different AFAICS.

[tabb...@gmail.com]

On Tuesday, 9 February 2016 20:03:29 UTC, dennis@home wrote:

> Don't start being an idiot again if you know any good reason for us to
> believe what you said then spit it out.
>
> Otherwise I am correct.

have you ever been correct?

[dennis@home]

I have always been correct when TNP disagrees.

[The Natural Philosopher]

On 09/02/16 20:03, dennis@home wrote:
> Don't start being an idiot again if you know any good reason for us to
> believe what you said then spit it out.
>
> Otherwise I am correct.

Which is why you will never be a logician.


--
No Apple devices were knowingly used in the preparation of this post.

[The Natural Philosopher]

I think he's about random - one in around 20

[The Natural Philosopher]

And he will never be a philosopher either.

[dennis@home]

On 09/02/2016 21:38, The Natural Philosopher wrote:
> On 09/02/16 21:21, dennis@home wrote:
>> On 09/02/2016 20:15, tabb...@gmail.com wrote:
>>> On Tuesday, 9 February 2016 20:03:29 UTC, dennis@home wrote:
>>>
>>>> Don't start being an idiot again if you know any good reason for us to
>>>> believe what you said then spit it out.
>>>>
>>>> Otherwise I am correct.
>>>
>>> have you ever been correct?
>>>
>>
>> I have always been correct when TNP disagrees.
>
> And he will never be a philosopher either.
>
>
>

Up to your usual wriggles I see.
Say I am wrong and when I ask you to show where you can't and just spout
cr@p to hide your stupidity.
Others have spotted that you keep doing this so you can give up now.

[Cursitor Doom]

On Tue, 09 Feb 2016 12:47:52 +0000, dennis@home wrote:

> For a DIY approach I can't help thinking two pins spaced apart on the
> track and measure the volt drop along the track.
> Needs a very hi impedance meter though.

I like the idea (in principle) of measuring the volts drop to derive the
current. But the impedance of the meter is of secondary importance to the
meter's ability to *differentiate* between two very, very, very similar
voltages. So.... how about using a top-notch precision op-amp to vastly
multiply this voltage differential, then calibrate with a precision
current source? Might need a bit of tinkering but better for a DIY level
job than lashing out 500 quid or more (in some cases a *lot* more).

[tabb...@gmail.com]

On Tuesday, 9 February 2016 22:31:24 UTC, dennis@home wrote:
> On 09/02/2016 21:38, The Natural Philosopher wrote:
> > On 09/02/16 21:21, dennis@home wrote:

> >> On 09/02/2016 20:15, tabbypurr wrote:
> >>> On Tuesday, 9 February 2016 20:03:29 UTC, dennis@home wrote:

> >>>> Don't start being an idiot again if you know any good reason for us to
> >>>> believe what you said then spit it out.
> >>>>
> >>>> Otherwise I am correct.
> >>>
> >>> have you ever been correct?
> >>>
> >>
> >> I have always been correct when TNP disagrees.
> >
> > And he will never be a philosopher either.
> >
> Up to your usual wriggles I see.
> Say I am wrong and when I ask you to show where you can't and just spout
> cr@p to hide your stupidity.
> Others have spotted that you keep doing this so you can give up now.

It's nice he has a sense of humour, even if unintentional.


NT

[Cursitor Doom]

Much interesting info here that might be of use:

http://www.intersil.com/content/dam/Intersil/whitepapers/amplifiers/
current-sense-measurements.pdf

[dennis@home]

Well it depends on what meter/scope he has.
Some digital meters can work down to microvolts without help.

[Syd Rumpo]

On 07/02/2016 18:21, Chris wrote:
> Current probes for oscilloscopes are nice to have, but very pricey. I
> vaguely recall someone somewhere sometime saying that you can make a
> pretty good one up from certain computer scrap parts (some component in
> the disk drive, perhaps?) and there may also be a sensor in the power
> supply section that deals with power factor correction, I gather. I need
> something that can sense down to say 10mA or better that's (ideally) fine
> enough to sample individual pcb traces a few mm apart.
> Any ideas?

Cut the track with a scalpel, scrape the coating off, solder an sm
resistor across the break and measure the voltage drop. Obviously 1 ohm
will give 10mV at 10mA and 10mV may not matter, depending on your
circuit. When finished, take off the resistor and solder across the cut.

Messy, but cheap and potentially (!) accurate.

Cheers
--
Syd

[whisky-dave]

which is why it won't work.
if you have a DIY solution then post it lets see if your theory will actualy work.

> The only variable is the impedance of the track.

why is that the only variable , and how will you take accoiunt of that anyway.

> With the current probe there are dozens of variables that have to be
> taken into account.
> I would use a constant current source to inject a reference current down
> the track to calibrate whichever method I was using if it had to be
> accurate.

Depending on what and where you want to measure.
You have a PC computer board and you want to measuing the current
opassing through a particular track on teh PCB.
While the computer is on how are you going to achieve this ?

> >> If you still don't believe me then look at a DVM...
> >
> > I have over 200 DVMs well DMMs really. have 1/2 dozen LCRs too.
>
> Having them and understanding them is different AFAICS.

yes I do, enough to tell you're talking crap.

[whisky-dave]

On Wednesday, 10 February 2016 09:41:37 UTC, Syd Rumpo wrote:
> On 07/02/2016 18:21, Chris wrote:
> > Current probes for oscilloscopes are nice to have, but very pricey. I
> > vaguely recall someone somewhere sometime saying that you can make a
> > pretty good one up from certain computer scrap parts (some component in
> > the disk drive, perhaps?) and there may also be a sensor in the power
> > supply section that deals with power factor correction, I gather. I need
> > something that can sense down to say 10mA or better that's (ideally) fine
> > enough to sample individual pcb traces a few mm apart.
> > Any ideas?
>
> Cut the track with a scalpel, scrape the coating off, solder an sm
> resistor across the break and measure the voltage drop.

but that doesn't give you non-contact probing.

Make measurements via non-contact probing

> Obviously 1 ohm
> will give 10mV at 10mA and 10mV may not matter, depending on your
> circuit. When finished, take off the resistor and solder across the cut.
>
> Messy, but cheap and potentially (!) accurate.

and not very practical for those needing this type of instrument.


Of course if anyone has a better method then they can do it can't they.

Then there's those DIYer's that when they can't find the crewdriver a hammer does the same job.

[Syd Rumpo]

On 10/02/2016 10:34, whisky-dave wrote:
> On Wednesday, 10 February 2016 09:41:37 UTC, Syd Rumpo wrote:
>> On 07/02/2016 18:21, Chris wrote:
>>> Current probes for oscilloscopes are nice to have, but very pricey. I
>>> vaguely recall someone somewhere sometime saying that you can make a
>>> pretty good one up from certain computer scrap parts (some component in
>>> the disk drive, perhaps?) and there may also be a sensor in the power
>>> supply section that deals with power factor correction, I gather. I need
>>> something that can sense down to say 10mA or better that's (ideally) fine
>>> enough to sample individual pcb traces a few mm apart.
>>> Any ideas?
>>
>> Cut the track with a scalpel, scrape the coating off, solder an sm
>> resistor across the break and measure the voltage drop.
>
> but that doesn't give you non-contact probing.

Fuck me, you're right! I hadn't thought of that!

<twaddle snipped>

Cheers
--
Syd

[Cursitor Doom]

If you're going to break the track, why bother with soldering in a sense
resistor? Might as well just stick an ammeter straight across the bridge!

[Syd Rumpo]

Yes, you could, depending on the circuit. The resistor is maybe easier
- it doesn't matter if the probes drop off or are intermittent.

Cheers
--
Syd

[whisky-dave]

It does if you're measuring something time related and you've pretty much damaged the track too,and then there's shorting things out while trying to measure others. Of course the tracks are a few mm apart which is quite a large gap on PCBs, there's also a problemm of shorting out other devices.



>
> Cheers
> --
> Syd

[Syd Rumpo]

Christ you're dim.

Cheers
--
Syd

[whisky-dave]

But now where near as dim as you. How can you say

[Syd Rumpo]

On 10/02/2016 15:51, whisky-dave wrote:

<wittering snipped>

> But now where near as dim as you. How can you say
>
> "it doesn't matter if the probes drop off or are intermittent."

As compared to using a current meter across a track break, you
blithering fool. The circuit will continue to work. Bloody hell.

Cheers
--
Syd

[whisky-dave]

On Wednesday, 10 February 2016 15:55:22 UTC, Syd Rumpo wrote:
> On 10/02/2016 15:51, whisky-dave wrote:
>
> <wittering snipped>
>
> > But now where near as dim as you. How can you say
> >
> > "it doesn't matter if the probes drop off or are intermittent."
>
> As compared to using a current meter across a track break,

I wouldn't use a current meter across atrack.

> you
> blithering fool. The circuit will continue to work. Bloody hell.

not when you're shorting tracks out it won't, fucking idiot.

[dennis@home]

Go on then tell us the rest of the variables?

>
>
>> With the current probe there are dozens of variables that have to be
>> taken into account.
>> I would use a constant current source to inject a reference current down
>> the track to calibrate whichever method I was using if it had to be
>> accurate.
>
> Depending on what and where you want to measure.
> You have a PC computer board and you want to measuing the current
> opassing through a particular track on teh PCB.
> While the computer is on how are you going to achieve this ?

Why are you having trouble understanding any of this?
A track is a resistor, you measure the voltage across it just like your
digital multi-meters measure the voltage across a shunt resistor to
measure current. You sure as hell don't put it through a meter coil like
TNP thinks.
You can't measure the current with it off BTW so why ask?

>
>
>>>> If you still don't believe me then look at a DVM...
>>>
>>> I have over 200 DVMs well DMMs really. have 1/2 dozen LCRs too.
>>
>> Having them and understanding them is different AFAICS.
>
> yes I do, enough to tell you're talking crap.
>
>

You just don't understand it.

[whisky-dave]

Oi dip shit RTFM

The signal from a positional current probe is critically related to its position relative to the conductor.

The size of the conductor (e.g. the width of a PCB track) also has a significant effect.

The measurement result will also include other field effects present at the tip of the probe and not just that coming from the current through the conductor.

But don't worry I'm sure you'll get the result you want without needing to worry about itermittant contacts and shorting out between tracks.

>
> >
> >
> >> With the current probe there are dozens of variables that have to be
> >> taken into account.
> >> I would use a constant current source to inject a reference current down
> >> the track to calibrate whichever method I was using if it had to be
> >> accurate.
> >
> > Depending on what and where you want to measure.
> > You have a PC computer board and you want to measuing the current
> > opassing through a particular track on teh PCB.
> > While the computer is on how are you going to achieve this ?
>
> Why are you having trouble understanding any of this?

I'm not, you are.


> A track is a resistor, you measure the voltage across it.

So what is the restance of a track ? They aren't all the same you know.

[dennis@home]

You are either thick or deliberately trying to wind me up.

Why do you think I said you need a constant current source to calibrate it?
Just to ensure you understand..
you connect the constant current source to the track, you measure the
current using whatever method you choose like a current probe or the DIY
method.
You correct for the error it reads.
You can then turn the equipment on and do accurate measurements.

End!

[Cursitor Doom]

I think you're at cross purposes with everyone else here. We're not
talking about shunt measurements in this instance. The circuit will
continue to work because although we break the trace, the meter (ammeters
are very low resistance) will bridge it (series not shunt).

[whisky-dave]

Maybe it's you that;s the thick one.

>
> Why do you think I said you need a constant current source to calibrate it?

yuo don;t it's inbuit in that unit that's what the calibration is.

> Just to ensure you understand..
> you connect the constant current source to the track, you measure the
> current using whatever method you choose like a current probe or the DIY
> method.
> You correct for the error it reads.
> You can then turn the equipment on and do accurate measurements
>

> End!

you obviously don't get it so give it up, all the information and the use of that device is there for everyone to see, just read it.

[whisky-dave]

Are you saying it's what this meter does is break the tracks on the PCB think again.
This device is expensive partly because of it's ability to do in circuit testing i/e you don't have to destroy the tracks on your PCB to measure the current. it's NOT as simple as using an AVO or even a clamp meter.

[Chris]

On Thu, 11 Feb 2016 03:55:29 -0800, whisky-dave wrote:

> Are you saying it's what this meter does is break the tracks on the PCB
> think again.
> This device is expensive partly because of it's ability to do in circuit
> testing i/e you don't have to destroy the tracks on your PCB to measure
> the current. it's NOT as simple as using an AVO or even a clamp meter.

I still don't see how this answers the original question. This is a
proprietary current probe costing 500 odd quid. A DIY solution was
specifically requested.

[whisky-dave]

I do NOT have any idea how to make one of these devices up from scrap computer parts any more than I know how to make a DIY Mars or moon rocket.
Like others I could come up with crap and spout it or pretend a simple meter with a couple of probes can do it, but I know better.
If someone wants to help the OP then please post all the details rather than just say it doesn't matter if you have intermittent contacts or short the tracks out. If you've ever done PCB work you should realise that intermittent results due to poor contacts are NOT a good way to test anything.

As yet I've seen no solution to the OP problem which was to "sense the current runing through a PCB track", I'd suggest a mind meld with the PCB but unless the OP is vulcan that won't work either.
I DO NOT believe the OP wanted to cut up his PCBs or the tracks and put resistors in series with the tracks.

[dennis@home]

On 11/02/2016 11:51, whisky-dave wrote:
> On Wednesday, 10 February 2016 17:13:18 UTC, dennis@home wrote:

8<

>>
>> You are either thick or deliberately trying to wind me up.
>
> Maybe it's you that;s the thick one.
>
>>
>> Why do you think I said you need a constant current source to
>> calibrate it?
>
> yuo don;t it's inbuit in that unit that's what the calibration is.

So you actually agree that I was correct!
What are you arguing about?

>> Just to ensure you understand.. you connect the constant current
>> source to the track, you measure the current using whatever method
>> you choose like a current probe or the DIY method. You correct for
>> the error it reads. You can then turn the equipment on and do
>> accurate measurements
>>
>> End!
>
> you obviously don't get it so give it up, all the information and the
> use of that device is there for everyone to see, just read it.
>
>
>

We don't want to use that device, we can't afford £500 is the simple
answer you fail to understand.

[whisky-dave]

On Thursday, 11 February 2016 19:50:28 UTC, dennis@home wrote:
> On 11/02/2016 11:51, whisky-dave wrote:
> > On Wednesday, 10 February 2016 17:13:18 UTC, dennis@home wrote:
>
> 8<
>
> >>
> >> You are either thick or deliberately trying to wind me up.
> >
> > Maybe it's you that;s the thick one.
> >
> >>
> >> Why do you think I said you need a constant current source to
> >> calibrate it?
> >
> > yuo don;t it's inbuit in that unit that's what the calibration is.
>
> So you actually agree that I was correct!

yuo don;t need a constant current soucre ius part of the probe.

> What are you arguing about?

That's there's not a DIY solution using something from a disc drive you can cobbled together by a DIYer.

>
> We don't want to use that device, we can't afford £500 is the simple
> answer you fail to understand.

then what;s asked can't be done then can it, you'd be wasting uor time messing about taking disc drives apart in the hope of finding something you can use to teast the current flow in a PCB track, which is what the OP was hoping to do.

if he;'s prepared to break a track and use an ordinary DMM then that's fine but that set-up IS NOT the same as the £500 probe.

[dennis@home]

On 12/02/2016 11:35, whisky-dave wrote:
> On Thursday, 11 February 2016 19:50:28 UTC, dennis@home wrote:
>> On 11/02/2016 11:51, whisky-dave wrote:
>>> On Wednesday, 10 February 2016 17:13:18 UTC, dennis@home wrote:
>>
>> 8<
>>
>>>>
>>>> You are either thick or deliberately trying to wind me up.
>>>
>>> Maybe it's you that;s the thick one.
>>>
>>>>
>>>> Why do you think I said you need a constant current source to
>>>> calibrate it?
>>>
>>> yuo don;t it's inbuit in that unit that's what the calibration
>>> is.
>>
>> So you actually agree that I was correct!
>
> yuo don;t need a constant current soucre ius part of the probe.
>
>> What are you arguing about?
>
> That's there's not a DIY solution using something from a disc drive
> you can cobbled together by a DIYer.
>
>
>
>>
>> We don't want to use that device, we can't afford £500 is the
>> simple answer you fail to understand.
>
> then what;s asked can't be done then can it, you'd be wasting uor
> time messing about taking disc drives apart in the hope of finding
> something you can use to teast the current flow in a PCB track, which
> is what the OP was hoping to do.

Of course it can be done, just because you don't have the skills doesn't
mean someone else doesn't have the skills. The fact the current probe
exists is evidence that it can be done.

>
> if he;'s prepared to break a track and use an ordinary DMM then
> that's fine but that set-up IS NOT the same as the £500 probe.
>

Its also not what i suggested.

[whisky-dave]

It needs a lot of skills perhaps that's why the unit cost £500.
why havent; you proved it camn be doen by a DIYer ?

FFS come up with a ciruit and component list then.

[dennis@home]

Most people would know that a probe like that would have a limited
market and would cost a lot to buy even if it had £1 of components in it.

> why havent; you proved it camn be doen by a DIYer ?


>
> FFS come up with a ciruit and component list then.
>
>
>

Why should I, I haven't suggested using a disk drive and I don't have a
scrap one with magnetorestive heads.

[Johnny B Good]

On Tue, 09 Feb 2016 13:55:25 +0000, The Natural Philosopher wrote:

> On 09/02/16 12:47, dennis@home wrote:
>> On 09/02/2016 09:37, whisky-dave wrote:
>>> On Monday, 8 February 2016 23:20:44 UTC, Syd Rumpo wrote:
>>>> On 08/02/2016 19:26, dennis@home wrote:
>>>>> On 08/02/2016 16:49, Bob Minchin wrote:
>>>>> 8<
>>>>>
>>>>>> After a while you might be excused for starting to think that what
>>>>>> you get in a scope current probe is actually quite good value!
>>>>>
>>>>> I have a ~£25 DC clamp on multimeter, it might be easier to start
>>>>> with one of those and dismantle the clamp to see what sensor + amp
>>>>> is in there.
>>>>
>>>> That'll probably be a Hall effect sensor, but you need the conductor
>>>> to go _through_ the clamp, which is difficult for a PCB trace.
>>>
>>> http://www.rapidonline.com/test-measurement/tti-thurlby-thandar-
iprober-520-positional-current-field-probe-85-4563

>>>
>>>
>>>
>>
>> For a DIY approach I can't help thinking two pins spaced apart on the
>> track and measure the volt drop along the track.

>> Needs a very hi impedance meter though.
>>

> Needs a very LOW impedance very low voltage meter, though

That's counter to requirements, notably the LOW impedance. What you
*don't* want is a meter path impedance that becomes comparable to the
contact resistance where you're trying to pick a voltage reading from.

In this case, what you need is a meter impedance that's 3 or 4 orders of
magnitude greater than any contact resistance you might expect to see
(including the meter lead resistance). In this case, half an ohm contact
resistance plus another half ohm test lead resistance suggests that a
voltmeter impedance of 10,000 ohms should be sufficiently high enough to
make any such contact and test lead resistance insignificant factors in
measurement errors.

Compared to the 10 or 11 Meg Ohm DMM voltmeter impedance, this *is* a
very low impedance voltage measuring instrument indeed but that doesn't
make LOW impedance a desirable feature of such a voltmeter method of
using a short section of circuit trace as a current shunt where you need
to measure a volt drop virtually independent of the effects of contact
resistance and measuring instrument impedance.

Making a voltmeter sensitive enough to detect microvolts might require a
lower input impedance circuit than the standard 10/11 M Ohms 200mV FSD
DMM circuit but this is only an acceptable compromise when we're talking
about a 10 to 100 K ohm range of input impedance (about 2 or 3 orders of
magnitude smaller than the standard DMM voltmeter impedance) for
measuring microvolts in an extremely low impedance circuit.

--
Johnny B Good

[Johnny B Good]

If using contact spikes to pierce the solder resist mask is acceptable
(you can always apply a blob of varnish over the holes afterwards), then
a galvanic connection such as this represents the the cheapest and most
accurate alternative to a 500 quid current probe.

However, the effectiveness does depend on what frequencies may be
included in the DC component you're trying to measure and whether or not
the trace in question is also carrying high frequency voltage transients
which would require that the probe leads be effectively isolated so as
not to introduce additional loading.

The classic solution to such galvanic connected probing is to use
suitable stopper resistors at the probe tip to isolate the rest of the
probe circuit and its leads from the circuit trace under test. You might
need to use an op-amp in differential mode, one with a very high common
mode rejection ratio where you tie it to a reference ground rail of the
circuit under test, if necessary via an offset voltage to centre the
working voltage range of the circuit to the mid point voltage of the
differential amp's input voltage range.

You might even need to use a couple of DC to HF RF amps to create a
differential probe amp with very high CMRR, again tied to the ground
reference via a suitable offset voltage when you need to include DC
measurements with very high frequency transients as per the op-amp
example.

The complexity of your probing circuit rather depends on what sort of
signals you're expecting to measure. As always, you're using a short
section of circuit trace to act as an ammeter shunt where you're
measuring the volt drop between the contact points, requiring that the
voltage measuring circuit be relatively high impedance in order to
diminish errors due to contact resistance variations as I'm sure you're
only too aware. In this case, a measuring impedance of 10000 ohms or
higher should more than suffice (no need to go to the 10/11 M ohm
standard of a DMM).

If you really need a non-invasive means of probing, I doubt the idea of
repurposing a GMRR read/write head is a viable proposition since I'm
almost certain the heads include signal conditioning amplification that
has no interest whatsoever in handling DC signals. It's either a galvanic
probe or that 500 quid contactless probe with, afaics, nothing in between
(BICBW - you might be able to google up a cost effective contactless
probe solution if you search hard enough).

--
Johnny B Good