Micsig Current Probe CP2100A/B Tests and Comparing

[Algasman]

No complaints so far ...
Looks like I got an A-Version with almost B-Specs. I’d say my unit has a bandwidth > 2MHz.
Maybe the binning Elasia suggested in an earlier post, isn’t so unlikely.
Would be nice to see a similar measurement with a real B Probe or other A versions.
Suggestions for a better test setup are also welcome.
Here are my measurements and an updated graph:

[bicycleguy]

I have a B model to compare.  Did you re-adjust the SDG output at each level.  I noticed the voltage across the resistor drops a little with frequency.

[Elasia]

I'd still put my money on that they would bin them

Solid design.. some flux in the components so just mass produce in one step with common components for cheap then tune after -> then a grading bin and done

[EEVblog]

After following the discussions and trying to decide if I should go with the A or B version, I decided for the economic more bang for the buck A-Version.
Last week I got it delivered so I wanted to check the performance myself. The results are almost too good to be true, so I would like to know if you agree with my test setup.
(Attachment Link)
SDG1032X as source, feeding a BNC-connector through 5 loops on the current probe clamp into 3x150 Ohms. SDS1204X-E Channel 1 Voltage of 50 Ohm load resistor, 10x Probe, Channel 4 CP2100A 2x (because of the 5 turn coil)
(Attachment Link)
During measurement, I maintained 2,50V RMS over the load resistor, and measured the current as RMS (nominal 50mA)
(Attachment Link)
This picture shows the voltage and current @800kHz, the nominal bandwidth of the CP2100A.
(Attachment Link)
To me this looks much better than the specs, is my setup suitable for bandwidth measurement?

The bandwidth of the A model is 2.3MHz? 

[Algasman]

I have a B model to compare.  Did you re-adjust the SDG output at each level.  I noticed the voltage across the resistor drops a little with frequency.

Yes
During measurement, I maintained 2,50V RMS over the load resistor, and measured the current as RMS (nominal 50mA)
I measured the voltage at the resistor with channel 1 of the scope and adjusted the output voltage of the signal generator accordingly. Constant voltage on the load resistor leads to a constant current through the loop (“RMS constant” since we're using AC)
The signal generator does not see a pure resistive load, there is also the inductance of the five turn coil. The dent around 4MHz could be some resonance effect.

[bicycleguy]

Here's a B
It doesn't look a whole lot better.

Code: [Select]

HZ,current(RMS)ma
          1 turn,    5 turns
50000,50,50
100000,48.6,48.7
200000,47.6
300000,48.1
400000,49.1
500000,49.6
600000,49.8
700000,48.8
800000,48.7,51.5
900000,48
1000000,47,50.2
1300000,44,52.2
2000000,37.7,42.6
2300000,35.6,40.9
2500000,34.6,40
3000000,33,38.2
3500000,31.9,37.0
4000000,33,38.4
4500000,31.2,38.4
5000000,23.5,31.7
6000000,10.7,13.1

Perhaps you can plot it with yours to compare.  I only used a single pass through probe.  Have redone some with 5 turns and it looks better.  Don't understand why it would matter as I rebased 50ma @ 50kHz,

[Martin72]

I´ve used the bodeplot function of my scope:

https://www.eevblog.com/forum/testgear/micsig-current-probe-cp2100b-tests-and-comparing/msg3072460/#msg3072460

(B-Model)

[Algasman]

Here's a B
It doesn't look a whole lot better.
# removed data #
Perhaps you can plot it with yours to compare.  I only used a single pass through probe.

Here’s the tough comparison:

The same shape and even the dent at around 4 MHz, only slightly better in the 100 to 600kHz range.

[Martin72]

And now, what are the conclusions ?

Please show your test setup, especially which resistor you´ve used for.

[bicycleguy]

Redid some data with 5 turns. See edit above.  Maybe should toss the 1 turn for consistency.  Edit maybe shouldn't compare spec to 5 tun.

[Hydron]

When I've done this kind of test I've used a 50R terminator on the scope input as both a well-matched load and the I-to-V conversion for the reference channel - basically have the co-ax from the sig-gen split into 2 wires at the half-way point (BNC to banana adapters are probably good enough here for <3MHz, I use a shielded box for higher frequencies), clamp onto the non-ground wire, then terminate into 50R on the scope. This gives 1V/A or 0.1V/A on the probe and 50V/A as the reference.

[Algasman]

After following the discussions and trying to decide if I should go with the A or B version, I decided for the economic more bang for the buck A-Version.
Last week I got it delivered so I wanted to check the performance myself. The results are almost too good to be true, so I would like to know if you agree with my test setup.
(Attachment Link)
SDG1032X as source, feeding a BNC-connector through 5 loops on the current probe clamp into 3x150 Ohms. SDS1204X-E Channel 1 Voltage of 50 Ohm load resistor, 10x Probe, Channel 4 CP2100A 2x (because of the 5 turn coil)
(Attachment Link)
During measurement, I maintained 2,50V RMS over the load resistor, and measured the current as RMS (nominal 50mA)
(Attachment Link)
This picture shows the voltage and current @800kHz, the nominal bandwidth of the CP2100A.
(Attachment Link)
To me this looks much better than the specs, is my setup suitable for bandwidth measurement?

The bandwidth of the A model is 2.3MHz? 

Yes, I was skeptical too, that’s why I put it up to discussion on your excellent forum!
The similar data from bicycleguys B-type probe pretty much confirms my claim and also Elasias binning theory.
Would be nice to see other A-type measurements (and also B ones)

[bicycleguy]

Algasman
I'm not sure that multiple turns thru the clamp are within the spec.  Sure it makes it work better as I show here:

Redid some data with 5 turns. See edit above.  Maybe should toss the 1 turn for consistency.

But how many turns? Where would you stop.  I now think the 5 turn should be dropped and if you re-did with just 1 time thru the clamp your results would match the spec.

[Martin72]

Would be nice to see other A-type measurements (and also B ones)

I´v turned a wire 10 times through the coil, for load I´ve used the integrated 50ohm input and then start bode plot.
Result, see link above.

[Noy]

Currently i have a CC-65 (working ok but its more a toy, but it has 10mA resolution...)
Normaly Im measuring in the below 5/3 A range.. So the micsig with 50mA resolution is a bit less attractiv but 800kHz (maybe more ;-) ) would be nice for switching regulator stuff these days (CP2100B would be nice but to expensive for the "sometimes" use case... ) and i can use multiple turns to get a higher resolution...

I found CP2100A for 140€ inkl. Shipping in Ali Express to Germany. Good offer or are these things fake / doesnt meet binning specs?
Should i order better from Batronix?

[Martin72]

Hi,

Please give a link to this offer(I´ve searching on Ali for it, best price was 233€..).
I guess, taxes will be added to the price so I would buy it from Batronix - No stress in case of warranty or money back guarantee...

working ok but its more a toy, but it has 10mA resolution...

Could you please show a pic with this resolution (10mA/Div.) ?

[Algasman]

Algasman
I'm not sure that multiple turns thru the clamp are within the spec.  Sure it makes it work better as I show here:

Redid some data with 5 turns. See edit above.  Maybe should toss the 1 turn for consistency.

But how many turns? Where would you stop.  I now think the 5 turn should be dropped and if you re-did with just 1 time thru the clamp your results would match the spec.

Using multiple turns is a common way to increase the probes sensitivity.
In my example the five turns increase the sensitivity by a factor of five so the signal to noise ratio is five times better than with just one turn.

But as usual, there is no free lunch.

Whereas the sensitivity increases linear with every turn, the insertion impedance increases by the square!
So five turns give you 25 times the impedance, 10 turns 100 times the insertion impedance.
It is like a transformer.

[Noy]

€ 131,71  18%OFF | BNC Port Strom Sonde Breite Verwenden Clamp 10A 100A Oszilloskop Zubehör USB Powered DC Testing Universal Mini Elektrische Meter
https://a.aliexpress.com/_mryzsHn


CC-65 Manual:
http://www.hantek.com/m/en/productdetail_77.html

[Martin72]

First thought, much too cheap to be true..

[Noy]

So asked the Ali sellers (found 2 with below 140€) some questions and it looks like "better don't touch!"
1. Thing the devices are without official case.
2. I asked them if these parts are genuine and no production "garbage".  If they met the official specification...

Both seller replied in a "talk around the corner" style :-D Don't know the exact english synonym for this speaking style..

The first replied nothing uptil now, only that it is without the case.
The second one replied "I can confirm the item will be the same as you can see on ads".

So both didn't say anything about genuine or specification... Maybe they didn't know... Or they knew and will not reply in an useable way (because they knew... ;-))

But without case / maybe without calibration sheet the devices are likely "production garbage / fake ones"...

I will order from Batronix i think. 5% off is possible if you ask for it....

[2N3055]

So asked the Ali sellers (found 2 with below 140€) some questions and it looks like "better don't touch!"
1. Thing the devices are without official case.
2. I asked them if these parts are genuine and no production "garbage".  If they met the official specification...

Both seller replied in a "talk around the corner" style :-D Don't know the exact english synonym for this speaking style..

The first replied nothing uptil now, only that it is without the case.
The second one replied "I can confirm the item will be the same as you can see on ads".

So both didn't say anything about genuine or specification... Maybe they didn't know... Or they knew and will not reply in an useable way (because they knew... ;-))

But without case / maybe without calibration sheet the devices are likely "production garbage / fake ones"...

I will order from Batronix i think. 5% off is possible if you ask for it....

Good decision...

[Algasman]

So asked the Ali sellers (found 2 with below 140€) some questions and it looks like "better don't touch!"
1. Thing the devices are without official case.
2. I asked them if these parts are genuine and no production "garbage".  If they met the official specification...

Both seller replied in a "talk around the corner" style :-D Don't know the exact english synonym for this speaking style..

The first replied nothing uptil now, only that it is without the case.
The second one replied "I can confirm the item will be the same as you can see on ads".

So both didn't say anything about genuine or specification... Maybe they didn't know... Or they knew and will not reply in an useable way (because they knew... ;-))

But without case / maybe without calibration sheet the devices are likely "production garbage / fake ones"...

I will order from Batronix i think. 5% off is possible if you ask for it....

I got mine from Batronix but without the 5% you get for some other equipment. But they ship very fast.

[Algasman]

As for my load, I did the following measurements using a LCR Meter @ 100kHz.

As a Resistor 50.12 Ohms, Theta 1.9 degrees

The series inductance 2,58uH with the clamp attached (5 turn coil with core)

and the series inductance 0.99uH without the clamp (5 turn air coil)
Regarding bicyleguys question

But how many turns? Where would you stop.

It depends on the device under test, putting the clamp around a conductor influences the DUT more or less. It’s the transformer rules you have to take into account. One can do as many turns needed as to get a good S/N ratio as long as the influence to the DUT is acceptable.

Pretty much as if a shunt resistor is inserted to use the voltage drop to measure the current. This burden voltage also influences the DUT. The reason why Dave designed the uCurrent Gold to keep this influence as small as possible.

[bicycleguy]

...
Regarding bicyleguys question

But how many turns? Where would you stop.

It depends on the device under test, putting the clamp around a conductor influences the DUT more or less. It’s the transformer rules you have to take into account. One can do as many turns needed as to get a good S/N ratio as long as the influence to the DUT is acceptable.

Pretty much as if a shunt resistor is inserted to use the voltage drop to measure the current. This burden voltage also influences the DUT. The reason why Dave designed the uCurrent Gold to keep this influence as small as possible.

What I was getting at is this thread is about comparing the performance of various probes against each other, including A vs  B Micsigs.  It doesn't make any sense to compare A with 5 turns vs B with 0.  Seems that the best is to compare no turns on every probe.

[Weston]

The official manual for the CP2100 does not give any noise values. Has anyone measured values for the peak to peak and RMS noise values for the probe?