Owon CP-05+ Current Probe (Former Pintek PA622 Thread)

[Martin72]

Today measuring a 300µs pulse, reference was a tektronix probe.

300mA:



2 A, tektronix :



Owon :



Noisy and overshoots….

[Martin72]

Further tests will come….at last.
Because:



Now I got two…..I´m joking, my clamp was damaged and I asked them for repair.
They won´t repair it, they also don´t want the defective clamp back, today I got a new one from them...
So now I can do a full "teardown", first I took the cover from the jaw away:



Ah, only hall-sensors, no compensation coil….
I will clean up the pcb from any parts so I could have a better look to the layout, for completing the schematic.

Next week I take some measures on a "real DUT" with the new one.

[Martin72]

A Picture of the hall sensor, taken with the andonstar microscope:





Any idea which could be the brand/manufactor ?
Search with "4pins hall sensor smd" wasn´t succesful.

[analogRF]

Martin,
did you increase the supplies to +/-5 to see what happens to measurements?
or any other tests?

[Hydron]

Any idea which could be the brand/manufactor ?
Search with "4pins hall sensor smd" wasn´t succesful.

It may be from Asahi Kasei, they make 4-pin hall elements for this sort of thing:
https://www.akm.com/akm/en/product/detail/0004/

[Martin72]

Hi,

did you increase the supplies to +/-5 to see what happens to measurements?
or any other tests?

Yes, and a little bit higher  - Result was the damaged clamp…
Still I want to make further tests on work, the clamp is also there but time....

[schmitt trigger]

Very interesting tear down, please continue.

[analogRF]

Hi,

did you increase the supplies to +/-5 to see what happens to measurements?
or any other tests?

Yes, and a little bit higher  - Result was the damaged clamp…
Still I want to make further tests on work, the clamp is also there but time....

do you mean the hall sensor died? because the opamps shouldnt die with +/-5V or even higher
but the current passing through the hall sensor must be limited, so maybe when you increased the supplies the current through the sensor exceeded the limit

was there any change in noise level of the clamp?

[Martin72]

Hi,

Yes, the hall-sensor died 



Increasing the voltage in a region, where every op-amp would function proper was the intention.
And the hope, that the noise level would decrease, but it didn´t.
They claimed a bandwith of appx 200khz, which wasn´t so bad at all, but I can´t believe it, although the clamp did measure up to 10khz proper, as I test it a few month ago.

please continue.

I´ll desolder all the parts from the pcb of the defective one, to have a look at all connections( for finishing the schematic).

[Spacemarine]

I´ll desolder all the parts from the pcb of the defective one, to have a look at all connections( for finishing the schematic).

Any progress on that?

[Martin72]

Hi,

Unfortunately not, loose the interest about - But now it´s actual again as I did some measurements yesterday with poor results.

Last year I did only sinewave measurements, yesterday squarewave with horrible results.

At a frequency of 100hz, everything seems normal:



At 2khz Overshoots appearing:



And finally at 22Khz it all sucks up:



So what´s going on there...
Therefore I had a look in my selfmade schematic and today I open up the "new" current probe.
And must admit, there are at least one value of a resistor going through the hall sensors different as drawn from the first probe schematic.
Hm-hm...

[CDaniel]

You should probe the signal path inside ... to identify the bad stage or hall sensors 

[Martin72]

I´m on the way, but do it tomorrow.

[jrf]

Martin,
Still can't remember where I have seen your waveforms...BUT
THe third harmonic of 22KHz is being amplified 4x more than the first & inverted!

Just did some reading on Bode plots, filters etc..
Filters:
First order: Signal: no peak. 20db/decade slope.
Second order: Signal:always a peak, height dependent on Q. Slope 40db/dec.

This one, ~12db/dec so useless as a filter, not designed!
SO not a low pass filter as such but some sort of cicuit limitation effect. ie limit of amps.

Also stability:
"When the magnitude crosses from >0dB to <0dB at a lower frequency than the Phase=180degrees then the system is stable."
This does not! Hence Unstable.

It makes me wonder if there is not a fault with your clamp.  CP-05 is meant to have signal 3db at ~200KHz. Yours is close to that but with a 6db signal increase at 65kHz what use is it! & the phase lag makes any comparison circuit voltage measurements useless past 25kHz. That is meant to be at 200kHz. (One wonders if all those tiny blank capacitors on the board are the correct values. )

Increasing the capacitance in the feedback of the final amp would reduce the Q I suspect & move it back to stable.
I would try to get another to replicate your tests with their meter.

Cheers,
John.

[Martin72]

Hi,

As I opened it up, there was not only difference between the resistor values near by the hall sensor connector, also I could see a difference on the feedback-resistors.
On the first probe, there was a "tower" of capacitors soldered onto these resistors, here there is only one cap soldered.
But first of all I have to fix the probe:
While I´ve measured with scope on the op-amp U6, the probe tip came inbetween two pins (+Ub and output) - Now there´s an dc-offset on the output and inputs from the op-amp, I think it´s defective.
So I order two new ones, but not mc33028, other one which could work with voltages down to +/- 2.5vdc .
Also ordered a smd-soldering tweezer, an assortment of smd capactitors and resistors.....
To be continued in the next week. 

[jrf]

Martin,
Looking at your drawn circuit you have 560pF capacitors in the feedback of U3a & b in parallel with 10kohm.
Correct me if I am wrong but when the freq is such that the impedance of the capacitor = 10k then the gain is 1/2.
This assumes there is no feed capacitance. ie none shown.
Hence at about 28kHz the gain is halved for both amplifiers.
I suspect this affects the 90' phase shift at 25KHz but there must be a change in input impedance as well as the circuit gain is not affected??
Reading about ultra sensitive hall chips as made by AKM, linked in post above, running at 1V, have a output RESISTANCE 250-400 ohm.
No capacitance mentioned. ie HW-108A

I would unplug the hall sensors & apply a signal to R20 to analyse the circuit response. With those circuit values the signal at 28KHz should be attenuated by 1/4. At 400A switch it should have ~unity gain.
I note the Chinese use MC33078 at 3V in other circuits I have noted. Some newer Chinese chips have different specs to the 'US' originals. Supply voltage is often 'enhanced' with possibly some loss in distortion noted.

I also note the range switched resistor vales do not make sense? ie R28 1K 4A - R27 100 Ohms 40A & 400A open.
I would expect 100 ohms=4A  1K=40A 10K=400A or such.

Good luck
John.

[Martin72]

Also ordered a smd-soldering tweezer

In comparison to my weller tool at work, it´s really a piece of shit - But it does it´s business and the op-amp is desoldered from the pcb.
"Tomorrow" the replacement should arrive.

[Martin72]

This thing is so crappy...

The trimpot VR3 to adjust the offset of the op07C is.....missing.
You have to trim it, there are inputs for it on the op-amp - My former cp05 got it, this one here not (same pcb)...
A repaired layer, different resistor in the hall-circuit, different capacitor in the output circuit, missing trimpot...
Hands away from this piece of shit !