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:

[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]

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 ;-)

[Bob Eager]

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

[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 -

makes several types, again, not for disk drives.

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

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

Cheers

[Bob Eager]

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

[tabbypurr]

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!

[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.

[tabbypurr]

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]

8<

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.

[tabbypurr]

A current transformer

NT

[dennis]

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

[The Natural Philosopher]

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

That means hall effect usually.

[whisky-dave]

[dennis]

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.