EEVblog #296 - AIM-TTi I-Prober 520 Current Probe Review

[EEVblog]

Dave.

[nitro2k01]

And here's Mike's teardown for those who may want to watch that as well:

[EEVblog]

I'm trialing a new menu annotation system at the start of the video, let me know what you think.

Dave.

[Mercator]

I never thought about current paths before, although I find it interesting. What determines the width of such a current path?

Regarding the menu annotation system: I like the idea, but it looks a little messy. Also I don't think I will ever use it. Maybe someone else will. It would be a good system for combined teardown / review videos. So maybe just two or three buttons on the bottom for major sections.
You could place all the links to websites and other videos at the end of the video. Took me quite a while to find the link for Mike's teardown again.

[david77]

I found the trick Mike used with the LED and long exposure very neat. Really helps to visualise the way the current takes.

Regarding the menu I think the idea isn't bad as such, just the way you did it looked awful. Maybe put it on a splash screen (for example the EEVBlog logo) at the beginning of the video for a few seconds instead of plastering it all over your talking head?

[T4P]

Dave, at about 30mins mark i started hearing some 1KHz switching noise ... Where did it come from?

[kc0ngu]

Re patent for Flux Gate Magnetometer - Not sure if this is the one but look for GB 2461437, Paul Robertson is listed as Inventor, and he worked at Camb. Engineering on this device. I have attached the pdf of the above patent.
Steve

[nitro2k01]

I'm trialing a new menu annotation system at the start of the video, let me know what you think.

Dave.

Because of my OCD tendencies, I'm bugged by the fact that the boxes are not perfectly aligned. But it is what it is I guess. I'm also not sure I like the embossed "impact" style annotation.

As for the instrument, I wonder how long it will take before the coils get cheap enough to put in a second coil to compensate for Earth's magnetic field.

[free_electron]

I never thought about current paths before, although I find it interesting. What determines the width of such a current path?

frequency.

for Dc they go wherever they feel like. as frequency goes up they follow the path of the signal above.

[SeanB]

Wonder how they make the sensors? They must be using some ultra thin wire and old core memory ferrites.

[free_electron]

more likely to be a vacuum deposited multilayer inductor

[G7PSK]

I thought solid state so I goggled solid state magnetic flux gate and found this.

http://micromagnetics.com/products_mtj_f_s.html?gclid=CLv5kp6W4LACFQ8htAodXzDpwg

Also found this now.   ftp://ftp.boulder.nist.gov/pub/pappas/CNRS-APS%20Sensors/References/fluxgate_files/Microfluxgate.pdf

[EEVblog]

I never thought about current paths before, although I find it interesting. What determines the width of such a current path?

frequency.
for Dc they go wherever they feel like. as frequency goes up they follow the path of the signal above.

Yes, and no.
As I showed, DC (1KHz I used is close enough) does not just go "wherever they feel like" into the dead end voids etc.

Dave.

[EEVblog]

Wonder how they make the sensors? They must be using some ultra thin wire and old core memory ferrites.

Read the link I put on the video notes.

Dave.

[notsob]

Back in the early 80's maybe earlier HP had a current tracer for logic IC pins on PCBs - the HP 547A Current Tracer, it didn't hook up to a scope, or provide current readings,
it was just used to find out where the current sincs were.
reason I mentioned it it that it had a small tip that was placed onto the track in interest.

www.hparchive.com/Bench_Briefs/HP-Bench-Briefs-1977-09-10.pdf

[amspire]

Back in the early 80's maybe earlier HP had a current tracer for logic IC pins on PCBs - the HP 547A Current Tracer, it didn't hook up to a scope, or provide current readings,
it was just used to find out where the current sincs were.
reason I mentioned it it that it had a small tip that was placed onto the track in interest.

www.hparchive.com/Bench_Briefs/HP-Bench-Briefs-1977-09-10.pdf

Inspired by the HP Journal article on the current tracer, I actually made one for myself. Filed down a piece of ferrite to make the core.  If I were making it today, I would probably try a hard drive head as the sensor.

http://www.hpl.hp.com/hpjournal/pdfs/IssuePDFs/1976-12.pdf

The big problem with the HP-type sensor is that its sensitivity increases with frequency, so it was pretty useless at seeing an accurate current waveform. HP used the sensor to detect transients (ie switching waveform edges) so they were not concerned with the frequency response of the sensor. The beauty of the fluxgate sensor is it has a flat frequency response all the way to DC.

Richard.

[free_electron]

for Dc they go wherever they feel like. as frequency goes up they follow the path of the signal above.

Yes, and no.
As I showed, DC (1KHz I used is close enough) does not just go "wherever they feel like" into the dead end voids etc.

Dave.
[/quote]
let me rephrase that.
for low frequencies they will attempt to follow the most direct line between source and destination. they travel in a relatively wide band. of course they won't go in sticking-out-swaths of copper if there is no point in going there.

as frequencies increase they will tend to follow the path of the source signal.

Code: [Select]

DC
                        ________________________________
                       |                                |

( top layer)  in  ------                                |
                               --------- out
                                                        |______________________________|

(plane)       GND ------------------------------------------------------------------------------- GNd return

RF
                        ________________________________
                       |                                |

( top layer)  in  ------                                |
                               --------- out
                                                        |______________________________|
                        ________________________________
                       |                                |

( plane)      in  ------                                |
                               --------- out
                                                        |______________________________|
 

kind of hard to do in ascii ..

if you have a meander trace over a solid plane : for DC or low frequency the ground return current ( current coming back from the load , through the ground , to the source ) will flow in a straight line. ( top drawing )

for RF : the return current will follow the shape of the meander above the plane. it will not go for the beeline but run exactly in the same shape as the trace above, even if there is a solid plane  and no reason to do that.

[scottwolf369]

Dave, if you have a known current flow through the PCB trace (100 mA as measured by the Fluke 87), why not just calibrate the probe to the trace instead?  Adjust the PCB Sensitivity knob until the o-scope reads 100 mV. Then you can bump up the current flow through the trace to 200 mA or 500 mA and see if the probe is accurate at those levels.

[EEVblog]

Dave, if you have a known current flow through the PCB trace (100 mA as measured by the Fluke 87), why not just calibrate the probe to the trace instead?  Adjust the PCB Sensitivity knob until the o-scope reads 100 mV. Then you can bump up the current flow through the trace to 200 mA or 500 mA and see if the probe is accurate at those levels.

Good point. I should have done that.

Dave.

[EEVblog]

let me rephrase that.
for low frequencies they will attempt to follow the most direct line between source and destination. they travel in a relatively wide band. of course they won't go in sticking-out-swaths of copper if there is no point in going there.

as frequencies increase they will tend to follow the path of the source signal.

Code: [Select]

DC
                        ________________________________
                       |                                |

( top layer)  in  ------                                |
                               --------- out
                                                        |______________________________|

(plane)       GND ------------------------------------------------------------------------------- GNd return

RF
                        ________________________________
                       |                                |

( top layer)  in  ------                                |
                               --------- out
                                                        |______________________________|
                        ________________________________
                       |                                |

( plane)      in  ------                                |
                               --------- out
                                                        |______________________________|
 

kind of hard to do in ascii ..

if you have a meander trace over a solid plane : for DC or low frequency the ground return current ( current coming back from the load , through the ground , to the source ) will flow in a straight line. ( top drawing )

for RF : the return current will follow the shape of the meander above the plane. it will not go for the beeline but run exactly in the same shape as the trace above, even if there is a solid plane  and no reason to do that.

Yes, it does that because at RF the loop inductance in the ground plane just below the trace is the lowest, hence more current will flow there instead of wider out in the ground plane.
Just wanted to clarify that at DC, current just doesn't flow everywhere equally, as I showed in the video.

Dave.

[vlf3]

Interesting ! a search through the UK TTi web-site shows no I-prober 520 being available... looks like they consider UK purchase of this device, not worth marketing in the UK;  or is this a patent distribution, partnership-consortium issue ?... as Dave say's Go Figure. 

Okay I-prober own up... just found the product on the their UK site, via the site-map; would have thought it should be easier to find, without deep searching.

[G7PSK]

Is this the sensor they are using, it is manufactured or at least sold in the US. One could certainly make a similar instrument by using one.

http://micromagnetics.com/docs/SpinTJ_TMR_magnetic_sensors_brochure.pdf

[vlf3]

Here is an article on the Fluxgate development...

http://www.sensorsmag.com/sensors-mag/cambridge-university-researchers-microfabricate-fluxgate-mag-894

[qno]

I remember experimenting some years ago using a cassette player record-playback head as a current tracer.

[reliables]

Interesting ! a search through the UK TTi web-site shows no I-prober 520 being available... looks like they consider UK purchase of this device, not worth marketing in the UK;  or is this a patent distribution, partnership-consortium issue ?... as Dave say's Go Figure. 



______________________________________________________________________________________

Tablet PC China
Tablet Android

[mikeselectricstuff]

Interesting ! a search through the UK TTi web-site shows no I-prober 520 being available... looks like they consider UK purchase of this device, not worth marketing in the UK;  or is this a patent distribution, partnership-consortium issue ?... as Dave say's Go Figure. 

http://www.ttid.co.uk/go/iprober/index.htm

I bought mine direct from TTi UK

______________________________________________________________________________________

Tablet PC China
Tablet Android

[D3f1ant]

So few years on, do people who purchased these things still find them useful and reliable?

[MarkL]

I purchased one before Dave's review and still find it useful.

I've found it most useful to track down shorted components, and once to find an extremely elusive short on a PCB that was poorly manufactured.

The method I use is to inject 1kHz on the shorted signal (typically the power supply) using a small signal transformer, and then use the iProber to track where the 1kHz current is going.  The small tip of the iProber allows you to get down to the exact pin on a 0.1" DIP.  You can easily identify the ingress and egress of the current through the shorted component.

I've also used it to get some rough measurements on in-circuit linear regulator performance without cutting PCB traces.  That's nice.


What I don't like about it:

- Absolute accuracy is poor, even after the calibration procedure.

- It's very noisy on low level signals.  That's why they have a 2Hz BW setting.

- It has an attachment for measuring current in a wire, but I can get better accuracy, much less noise, and much better frequency response with a real AC/DC current probe.

- I really dislike the sensitivity to the earth's magnetic field which can cause large DC offsets when you change the orientation of the iProber even a little bit.  I've been playing with the idea of strapping a 3-axis magnetic sensor to it to generate an opposing offset on the output.

- It would have been nice to provide a USB power option instead of the hassle with an external wall wart.


I've had no reliability issues.

All in all, I still think it was a good purchase.  Just don't believe all the marketing hype that it will fix a rainy day.

[D3f1ant]

Thanks. I need a conventional current probe, and hoped this would fulfil that and be useful for debugging too. Looking at datasheet the absolute accuracy didnt look to fancy, so fear it probably better as a debugging tool than concise  measurements. I'll probably pick one up if it's on promo sometime, but looks like have to spend a few $k on a traditional probe in the meantime.

[splin]

I've found it most useful to track down shorted components, and once to find an extremely elusive short on a PCB that was poorly manufactured.

The method I use is to inject 1kHz on the shorted signal (typically the power supply) using a small signal transformer, and then use the iProber to track where the 1kHz current is going.  The small tip of the iProber allows you to get down to the exact pin on a 0.1" DIP.  You can easily identify the ingress and egress of the current through the shorted component.

If that's your main usage then surely all you need is a tiny coil, and perhaps an amplifier, for tracing an AC signal - for very little money? The question then is can you justify the rest of the cost for the less common use cases?

[D3f1ant]

Exactly. I feel the Iprober is slightly gimmicky. As a proper current probe the accuracy isn't there to justify the price, so I feel it's better to spend 2-3x the money on a proper probe. Now I can't decide if should buy a Hioki and psu or save $ and limit myself to my scope branded ones which are powered direct from the instrument. I have a general distaste for proprietary stuff...but current probes seem to hold value remarkably well.

[MarkL]

I've found it most useful to track down shorted components, and once to find an extremely elusive short on a PCB that was poorly manufactured.

The method I use is to inject 1kHz on the shorted signal (typically the power supply) using a small signal transformer, and then use the iProber to track where the 1kHz current is going.  The small tip of the iProber allows you to get down to the exact pin on a 0.1" DIP.  You can easily identify the ingress and egress of the current through the shorted component.

If that's your main usage then surely all you need is a tiny coil, and perhaps an amplifier, for tracing an AC signal - for very little money? The question then is can you justify the rest of the cost for the less common use cases?

Sure, you could build something that would do similarly.  You could also buy an old HP 547A current tracer probe for less than an iPober.

We had an opportunity to buy an iProber as part of a consulting contract to assist in tracking down failures in a third party, battery-powered energy metering device.  The meter was encountering a number of ESD induced failures.  The major failure mode was CMOS latch-up on various parts, so it wasn't as easy as finding a dead short.

The iProber made it very easy to identify which chips were involved in failures, and down to the pin.  It was a board of about 20 SOICs (0.05" spacing), and a microcontroller with even tighter spacing (I think it was 0.65mm).  That might be difficult to achieve with a home brew coil, especially on a tightly packed board.  And add to that repeatability for the relative value, since it's not an "all or nothing" measurement (there's other legitimate current flowing besides the fault current).

The time saved in not having to build something, and the ease of using the iProber with its long insulated tip was justification enough.  It got the job done.

In the case of the voltage regulator I mentioned, there was a problem with it going into dropout for 50us which was causing a board to reset.  At first it appeared to be the voltage regulator was not up to spec.  It took about a minute with the iProber to figure out it was 1 of about 15 chips that was putting a huge spike on the power rail (it was a bad driver chip).  And without cutting any traces.

The iProber saved me time here too, but of course there's plenty of other ways to figure out this one too.

The iProber is handy to have around, and I use it now that I have it.  Without a specific purpose, my justification would be as a convenience.   It's certainly not a necessity.

[SunnySpring]

I had just got hold of this i-prober. The calibration to PCB trace is too troublesome and requires to know the exact current through the PCB trace (unless this trace is not surrounded by other trace and ground plane).  and you need to adjust the PCB Trace sensitivity [which means you need to know the exact current through the pcb trace anyway.  Then you need to hold the probe vertically and maintain the tip to be perperdically with the trace.  ANd I dont have a different probe to do this PCB trace current measurement with 1 ohm resistor. And then the uncertainty associated with the hand holding the probe.  putting the probe nearby a SMPS inductor coil certainly pick up a lot of field although insignificant with the peak current profile that I am only interested in.

 I wound a 24gauge wire 5 x around the provided wire clamp accessory that is provided with the i-probe and measure current the traditional way with this probe  The trace is very clean and no noise (because I position the clamp away from the board, leaving my hand totally free) and I solved my problem in less than 5 minutes to confirm the intermittent peak current  issue. which exceeded the specs of the SMSP IC momentary.