Hantek CC-65 AC/DC Current Probe Teardown and Testing

[dcac]

I found the ut210e schematics and it does not seem to have any frequency compensation in the diff amp. From picture of the PCB C13 does not seem to be populated.

[dcac]

Red line = Spice simulation without any caps - BW is now about 115Khz in high gain mode. Question is how much more noise this would let through.

[dcac]

And for good measures - 10Khz square wave.

Red trace = without any caps in the 272 diff amp.




Remember this does not show the actual response with the HALL sensors - only what the amplifier look like in a spice simulation.

[dcac]

New Schematic with caps values added and some notes.

EDIT: Newer version of the Schematics here: https://www.eevblog.com/forum/testgear/hantek-cc-65-acdc-current-probe-teardown-and-testing/msg3545210/#msg3545210

 CC-65 CLAMP4.pdf

[Zhao]

The Hall sensor marked SE is likely to be HE12AF1D12U, a knockoff version of HW101A. "E" indicates that its sensitivity is in bin E (228-274mv @ 50mT).

Marking of HE12AF1D12U chip:

Taken from http://blog.sina.com.cn/s/blog_7d2f0f0a0100tv93.html

Datasheet of HW101A:
https://www.akm.com/content/dam/documents/products/magnetic-sensor/hall-element/in-sb-ultra-high-sensitivity/hw101a/hw101a-en-datasheet.pdf

[Noy]

So now we can change to another Sensor / better bin grade to improve the clamp?

[myf]

Hello,

I want to test the mains and laptop power supplies. I already have multimeter and uni-t210 ampmeter clamp, and don't have oscilloscope and clamp for DSO yet.
Do you think I shall be able to use this clamp for this ? I don't know what frequency I shall see on the DSO-display. (my future DSO should be rigol-1054 or siglent-1104).

Many thanks for your answers to this too elementary question !

F.

[Zhao]

Even if you replace them with HW101A bin G, which is the highest sensitivity part available, you only increase the sensitivity by 40%, which translates to a slightly lower noise floor as less amplification is needed. There are other ways to increase sensitivity without replacing any parts, such as reducing the air gap between the halves of the cores. Although this will also reduce max current level due to earlier core saturation.

It will be very interesting if we can improve its bandwidth to at least a few MHz. Although I afraid the stock core (which appears to be made of laminated permalloy) is not gonna have any permeability at such frequencies.

So now we can change to another Sensor / better bin grade to improve the clamp?

[Zhao]

BTW, I suspect the same sensor is used in the legendary UT210E clamp meter (see the photos in joeqsmith's post, quoted below). I would not recommend converting a UT210E to an oscilloscope current probe though, as it appears to use a pair of silicon steel core instead of permalloy core, according to the teardown from DiodeGoneWild (photo attached).

John and Jesse, this was my first attempt.  As I mentioned, the first problem is in the clamp.    Jaw was glued but I was able to cut it apart with an X-acto knife and not damage the plastic too bad.  There are two sensors, one at each end.  Core is a little strange.

[jrf]

myf,

1/ A current clamp is only required if you want to measure current while isolated from the voltage, do not want to influence the circuit resistance or the current is very high.
ie it's more accurate & sensitive to measure current using a series resistance.
At mains potential a 1 volt drop across a series resistance to measure current is negligible but your OSC needs to be isolated from this voltage. Historically I have used an isolation transformer, in my case a small old 240-110V step down transformer I had. GREAT CARE MUST be taken!
I can now use a tablet on a battery powered USB OSC.
Alternatively there are now available isolated current sensor IC's that eliminate these issues. ie ACS712, 714 or 758 series. Available for +/- DC current.

2/ The RMS current drawn by a laptop supply is ~0.2A. I suspect most newer supplies have yet to include active power factor correction as they are small. So The current waveform should look like a standard bridge rectified capacitor charge circuit. ie large spike of current towards the peak of the AC voltage. Its peak value will be >1amp, under full charge rate conditions, so this current clamp will work well with it.  I just use a short mains extension lead with the 3 wires loose.

Basically if you want to look accurately at currents less than 100-200mA this is not the way to go. There could be 10-20mA of noise present. This clamp is sensitive to magnetic fields so for sensitive measurements it needs to be away from sources of interference! ie the mains! unless you are only interested in higher frequencies & the mains can be filtered out.

Good luck
John.

[mickab]

Hello,

I'm currently searching a solution to increase the bandwidth of the clamp. The hall effect sensor seems to response up to 3MHz but the mod is design for +/- 1Mhz (Power supply design)
My first glance on this topic was very helpful.

I have simulated the signal conditioning stage of the hall-effect sensor on Spice with a rudimentary model. An OPA2350 is used for U3 (high speed, low noise, single supply...) and a OPA2192 (high C-Load capacity) for the output stage.
The simulation is based on the schematic "cc-65-CLAMP.pdf" wich seems right except the capacitor value. I correct some cap value with few measurements. The correction is into the .pdf.

Push to the limit we can obtain 2.3MHz with +/- 170° phase shift and -3dB (see picture)

I will make a comparison soon with a new clamp vs modified. 

[jrf]

Mickab,

Very impressive, especially if the osc pages shown where actual measurements of your modified clamp in operation.
Simulation certainly has come a long way!
I would not have expected such a good result given the iron laminations in the head.
Shows how similar circuits from other suppliers can achieve good results, with better spec & correct components.
I am sure none of the original designs where optimised with a simulator.

PS: The version of cc-65-CLAMP.pdf used is not the latest on the site. ie some corrections & updates & most capacitor values updated, but not surprisingly a little different to your measurements. I suspect errors, changes or quality will change these with batches.

Cheers,
John.

[Mechatrommer]

The hall effect sensor seems to response up to 3MHz but the mod is design for +/- 1Mhz (Power supply design)

yeah good job. the reason i didnt pursue modding my clamp is because i saw some datasheets about cheap grade hall sensor their BW rating always in KHz region, the opamps used in the clamp is in MHz BW region iirc. so i concluded the bottleneck is the hall sensor and any mod attempt on circuit or IC will be futile. good to hear that i'm wrong.

[mickab]

Hello,

See in the .PDF the full test comparison. 
 

[Mechatrommer]

Hello,
See in the .PDF the full test comparison. 

in your pdf stated to change U4 with OPA2195. i cant find what is OPA2195, no datasheet in internet... but in your earlier picture, suggesting to change U4 with OPA2192, so i guess OPA2195 is a typo?... i'm currently adding stuffs into lcsc cart, so i want to add this stock as well. i can find OPA2350 for U1,U3 but i cant find OPA2192 in DGK (vsop/pdso) package, only soic8. is there any specific reason why we cant use all U1,U3,U4 = OPA2350? why must U4 = OPA2192? kudos to your work and cheers.

edit: i just noticed your OPA2192 high C-load remark. so we cant use OPA2350? i'll check on this thanks.
edit2: checking schematics, there is no C-load directly subjected to U4, at least 1Kohm impedance in the middle.

observation:
OPA2350 is 7V max opamp (5.5V recommended operating voltage), battery in CC-65 is 9V so i think  need to buck down operating voltage to 5V, this probably will reduce maximum current spec from 6.5A to maybe 3A in lower range and 65A to 30A in upper range. for myself, i dont think thats a big problem, i never need to measure high current so far.

[mickab]

Indeed this is a typo error for the OPA195.

In first place U4 was an OPA2350 but the response wasn't great due to the poor capability of driving capacitive load.
C15 represent the C-Load and is needed for noise attenuation.

Good observations Mechatrommer the OPA2350 can't work at 9V but the regulator of the clamp is set at 7V.
I think we can replace all op amp with OPA2192. This will need some simulation and testing but it is possible.
Tell me if you're interested. 

[Mechatrommer]

In first place U4 was an OPA2350 but the response wasn't great due to the poor capability of driving capacitive load.
C15 represent the C-Load and is needed for noise attenuation.

but there is R18 1Kohm separating them. this is a bit surprise for me. maybe i will figure out how to reduce noise from that output later.

Good observations Mechatrommer the OPA2350 can't work at 9V but the regulator of the clamp is set at 7V.
I think we can replace all op amp with OPA2192. This will need some simulation and testing but it is possible.
Tell me if you're interested. 

yes interested. as i'm going to order some parts from lcsc, i will add few parts for this mod as well (save shipping cost). but i can only find OPA2350 and another opamp that can get supply from ±12V, but 15MHz GBW only, i wont name it as its limited supply so i dont want somebody to outrun me  (its 2-3X the OPA2350 price anyway) if really the bottleneck is in opamp BW, i'm suspecting U3 will need high BW since its in sort of high gain differential (instrumentation) mode from schematics you guys provided. i cant figure out what gain, but lets say gain 10, OPA2350 38MHz GBW still can give 3MHz+ BW, so i think its a wise choice. where the rest of opamps are mostly for Voffset correction i guess. if my eye and mind is correct U4-1 is only gain 4 differencing opamp, so 15MHz GBW opamp still can give 15/4 = 3MHz+ BW. unfortunately, i cant make order for OPA2192 since lcsc doesnt stock this opamp, i will try with whats available in lcsc first. ordering separately from digikey for this particular mod will result in shipping cost damage on my side. cheers.

[mawyatt]

Hello,

I'm currently searching a solution to increase the bandwidth of the clamp. The hall effect sensor seems to response up to 3MHz but the mod is design for +/- 1Mhz (Power supply design)
My first glance on this topic was very helpful.

I have simulated the signal conditioning stage of the hall-effect sensor on Spice with a rudimentary model. An OPA2350 is used for U3 (high speed, low noise, single supply...) and a OPA2192 (high C-Load capacity) for the output stage.
The simulation is based on the schematic "cc-65-CLAMP.pdf" wich seems right except the capacitor value. I correct some cap value with few measurements. The correction is into the .pdf.

Push to the limit we can obtain 2.3MHz with +/- 170° phase shift and -3dB (see picture)

I will make a comparison soon with a new clamp vs modified. 

Interesting modifications, thanks for posting. I have a couple questions:

1) Since R18 isolates the output capacitance due to the load, is the added 22pf is to help stabilize U4-1?

2) Shouldn't another 22pf like capacitor be added across R20 to balance the differential amplifier?

3) With R18 isolating the capacitive loading to U4-1, then just about any good op-amp can be used that has rail to rail I/O at +- 3V rails and sufficient BW. The 22pf added capacitance may need to be changed depending on the op-amp selected.

4) With the leakage in C4 and R269 bleed resistor, wonder if another type cap and larger resistor value could work? Or removing R269 and placing back to back low leakage diodes (CB junction of bipolar) might work better drift-wise? If the drift is dominated by the Hall Effect sensor then this wouldn't matter.

5) What did you find in relative BW before and after the mods, maybe you could post images showing such if it's not too much trouble?

Anyway, great work and many thanks for showing

Best,

[masterx81]

Bought one of this, as soon as it arrives, i'll do the mods. Thanks for sharing! Nice result!
I've readed on other threads that this clamp have a lot of noise, but i see only ~1mv of noise (around 10ma), seem quite good.

[maxspb69]

Today I made all the modifications exactly according to the mickab 's description. The result is in the pictures below.

  1 - original device,
2 - modified device.

Test conditions: 6Vp-p amp, 50Ω load, 3 turns of wire on  hantek cc-65.

The band has expanded to up to 2.4MHz (-3dB).
If only to remove the amplitude jumps in the 1.0-2.5MHz range, the frequency response would be generally very beautiful! But this jumps within 1db are quite acceptable, I think...

(I think I can save a lot on the Micsig CP2100B current probe purchase  ) 

Thank a lot for the detailed instructions!

[Noy]

Argh.. currently i will also do the modification, but still waiting for the opamp..
Unfortunally i purchased a Micsig last year November already.. 800k one..

[Bluegizmo83]

I bought a CC-650 with the same unstable supply.  I'm my case to solve the problem I put a 0.1uF cap in pins 1&2 of the pads for U5.

Hey I know you posted this several years ago, but can you recall what type and voltage the capacitor was that you used for this for the CC-650?

[Bluegizmo83]

I bought a CC-650 with the same unstable supply.  I'm my case to solve the problem I put a 0.1uF cap in pins 1&2 of the pads for U5.

Just a heads up... I just did this mod on my CC-650 and I don't see any difference at all. I placed a 0.1uF electrolytic capacitor on pins 1 and 2 of U5. No change at all.

[maxspb69]

0.1uF electrolytic???

[Bluegizmo83]

0.1uF electrolytic???

Yes... Why? That's what someone else said they used...