EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: ez24 on August 22, 2015, 07:23:58 pm
-
I came across this graph where current is on the Y axis
(from this link: http://sound.westhost.com/articles/external-psu.htm#kil (http://sound.westhost.com/articles/external-psu.htm#kil) furnished by Chris C on another post.)
When it comes to measuring current I am gun shy. I will not even use my $50 meter, I only use Harbor Fright $5 meters.
I guess that a low ohm resistor is in series with the load and the scope probes are put across the resistor and the output of the scope was to a CSV file and the author put in conversion equation for the Y scale to convert to amps and plotted the amps?
Am I right? Or did he put the scope probes in series with the circuit? I have a Rigol DZ1054 so if the author hooked up his scope in series - how would I hook up the Rigol like this.
-
No. you cannot measure current simply using an oscilloscope and the normal (voltage) probes. That would be like trying to measure current with a volt-meter.
You must measure the voltage across a series resistor ("shunt"), or, using something like a current transducer (a torrodial coil or a Hall-effect device).
-
Hooking the probes up is series is a sure fire way to blow up your scope.
The author most likely used a resistor is series and connected the probes in parallel across the resistor. Be careful if you also want to measure voltage in the same circuit with another channel. The best practice is to place the resistor on the low side of the circuit then hook up the ground of the probe to the ground point of the circuit. This way if you also want to measure a voltage as well the scope is already reference to the ground of the circuit.
Dave also did a few videos on the subject both directly and indirectly. The power line attack on the safe video shows how he measured current with a resistor. There is also how not to blow up your scope video.
-
Daves uCurrent is essentially a current to voltage converter. You can probe the voltage side with a usual scope probe. Be sure of the max current allowed by the uCurrent.
-
Hi,
I also asked this question recently, but with the point of view that I wanted to measure the lag and lead between voltage and current.
How can I do that given that dave's uCurrent is a shunt and introduces a resistive load. Or a torroid would introduce an inductive load. Either of which would inherently effect the lag/lead?
My scope is a Rigol 1054, perhaps it has some functionality which can automatically ofset the display to compensate?
Thanks,
James
-
Just use a 0.1 ohm shunt resistor, and measure voltage drop across it, do some ohm's law calculation to I = E/R
-
The shunt resistor is usually so small, that it doesn't have an impact in most applications. Your toroid's resistance is probably larger than the uCurrent's 10 mOhm (that's 0.01 Ohm) shunt. Also for lag and lead measurement, a pure resistive load, will not change the phase angle.
-
Accurate current measurements on a DSO are normally taken with a Current probe of which there are 2 main types: Current transformer (CT) and Hall Effect.
A combination of both allows both AC and DC measurements to be taken.
This is easily accomplished with a Current probe and DSO by changing the Ch input units from V to A(I) and then selecting the correct attenuation to match your current probe. eg. 10 mA/mV
Most scope owners are aware Current Probes are expensive items, often costing many times the value of entry level DSO's.
With complex Power analysis capabilities of some DSO's there are other measurement factors to consider: timing Skew between Voltage and Current probes.
This error is usually only nS or pS and does not affect basic measurement to any significant degree.
This arises from the propagation delays between the 2 probes as a result of differing cable lengths and risetime delays in Current probes. For accurate comparative measurements probes must be "De-Skewed" using the DSO processing to negate any time difference between probes.
Current measurements can also be taken using a shunt method and calculations should be made to determine the voltage amplitude as to how it converts to A/div.
Like any measurement techniques, all current measurements will affect the DUT and the resultant waveforms to some degree.
-
For this kind of job I'm using LEM transducers connected to scope ( this is poor man scope current probe with bandwith about 100kHz :). Based on your application select proper sensitivity ( 6 to 50 A)
http://se.farnell.com/lem/cksr-6-np/current-transducer-6a-5v/dp/2146828 (http://se.farnell.com/lem/cksr-6-np/current-transducer-6a-5v/dp/2146828) price is about 23 USD.
From transducer you can remove the current sensing wires and wind your own wire if you are not going to make PCB. This transducers can be powered from USB from scope.
On Agilent you can switch from voltage reading and adjust sensitivity and also power can be calculated and displayed, but I'm not familiar with Rigol.
-
the 1rst question : how much is the input resistance of my scope !!!!
Rigol DZ1054 1Mohm? ok you can use a shunt to obtain an image of the current (in volt)
on the old scope, sometimes you had low input resistance
the 2nd question is how i'll connect my probe
you risk to make a short-circuit : look at dave's video
-
Accurate current measurements on a DSO are normally taken with a Current probe of which there are 2 main types: Current transformer (CT) and Hall Effect.
A combination of both allows both AC and DC measurements to be taken.
This is easily accomplished with a Current probe and DSO by changing the Ch input units from V to A(I) and then selecting the correct attenuation to match your current probe. eg. 10 mA/mV
Most scope owners are aware Current Probes are expensive items, often costing many times the value of entry level DSO's.
With complex Power analysis capabilities of some DSO's there are other measurement factors to consider: timing Skew between Voltage and Current probes.
This error is usually only nS or pS and does not affect basic measurement to any significant degree.
This arises from the propagation delays between the 2 probes as a result of differing cable lengths and risetime delays in Current probes. For accurate comparative measurements probes must be "De-Skewed" using the DSO processing to negate any time difference between probes.
Current measurements can also be taken using a shunt method and calculations should be made to determine the voltage amplitude as to how it converts to A/div.
Like any measurement techniques, all current measurements will affect the DUT and the resultant waveforms to some degree.
Why are these current probes so expensive? What do they contain? Maybe their components aren't so expensive?
Do the oscilloscope need to support I too? Do oscilloscopes from companies such as Rigol, Siglent and Hantek support it too?
When I started to learn electronics, it always puzzled me that oscilloscopes measures voltage but not current. Why not? From my POV, measuring current in an oscilloscope form could help in many other ways too.
-
Current measurements can be made with ANY oscilloscope.
For current measurements the Ch signal input IS in volts/div but for accurate OSD it is converted to amps/div and at the correct attenuation factor by the DSO.
The cost of current probes is related to BW and accuracy and although a basic AC current probe is a passive probe it is of much more complex design than a passive voltage probe.
-
Accurate current measurements on a DSO are normally taken with a Current probe of which there are 2 main types: Current transformer (CT) and Hall Effect.
A combination of both allows both AC and DC measurements to be taken.
This is easily accomplished with a Current probe and DSO by changing the Ch input units from V to A(I) and then selecting the correct attenuation to match your current probe. eg. 10 mA/mV
Most scope owners are aware Current Probes are expensive items, often costing many times the value of entry level DSO's.
With complex Power analysis capabilities of some DSO's there are other measurement factors to consider: timing Skew between Voltage and Current probes.
This error is usually only nS or pS and does not affect basic measurement to any significant degree.
This arises from the propagation delays between the 2 probes as a result of differing cable lengths and risetime delays in Current probes. For accurate comparative measurements probes must be "De-Skewed" using the DSO processing to negate any time difference between probes.
Current measurements can also be taken using a shunt method and calculations should be made to determine the voltage amplitude as to how it converts to A/div.
Like any measurement techniques, all current measurements will affect the DUT and the resultant waveforms to some degree.
This ^.
The issues of propagation and phase delay are important if you are concerned with accurate power calculations. Look for "De-Skew" functionality in the scope's settings, or as Rigol calls it on the 1054z, "Delay-Cal" in the second page of the Channel settings menu. This is especially important if you are using some kind of inductive (coil) pickup for current sensing. Do you know the phase delay introduced by the coil? Can you synch it properly to your voltage probe, to give accurate power computations? Try doing a simultaneous measurement of current, by looking at voltage drop across a current-viewing resistor (little or no phase delay), and the coil pickup probe indication (unknown phase and propagation delay). Are they exactly synchronous, or is there some "skew" in the coil pickup that needs to be corrected to match the time constant of a voltage probe, as it must be for accurate instantaneous power computation?
Why are these current probes so expensive? What do they contain? Maybe their components aren't so expensive?
They contain _engineering_. The components themselves may be relatively cheap, but getting them to work together properly and accurately isn't so cheap.
Do the oscilloscope need to support I too? Do oscilloscopes from companies such as Rigol, Siglent and Hantek support it too?
All an oscilloscope is, is a time-domain voltmeter. But many DSOs can _display_ a voltage reading, such as the voltage drop across a current-viewing resistor, in units of Current (amps, mA, etc). Even the Rigol 1054Z can do this much... although the "amps" label doesn't carry through to the "measurements" displays, which can be a little confusing. The Rigol also has a ridiculously large number of "probe attenuation" values available. Using these values with the appropriate CVR one can essentially do the Ohm's Law calculation in the background so that the scope trace "current" values are accurate.
When I started to learn electronics, it always puzzled me that oscilloscopes measures voltage but not current. Why not? From my POV, measuring current in an oscilloscope form could help in many other ways too.
Since we know and understand Ohm's Law, we can easily convert a voltage drop through a known resistance, into a current value. With a multichannel scope one can use two channels and the math features to perform a differential voltage measurement of voltage drop across a current-viewing resistive element just about anywhere in a circuit, giving the current through that element. So the scope can measure current, with the correct use of current probes or current-viewing resistors.
-
Why are these current probes so expensive? What do they contain? Maybe their components aren't so expensive?
All probes are expensive considering the raw materials. They are expensive because they need to be reasonablly calibrated and stable, and because they are made/sold in low volume which makes them more expensive to manufacture. Current probes are much more complex to make than simple voltage probes. It doesn't seem at all unusual
Do the oscilloscope need to support I too? Do oscilloscopes from companies such as Rigol, Siglent and Hantek support it too?
I don't know of ANY scopes that measure current directly.
When I started to learn electronics, it always puzzled me that oscilloscopes measures voltage but not current. Why not? From my POV, measuring current in an oscilloscope form could help in many other ways too.
There are many different kinds of current that need to be measured. It is much more difficult than measuring voltage. There are many different ways of measuring current depending on the exact conditions of each situation. Dunno why this seems so mysterious.
-
Why are these current probes so expensive? What do they contain? Maybe their components aren't so expensive?
Do the oscilloscope need to support I too? Do oscilloscopes from companies such as Rigol, Siglent and Hantek support it too?
When I started to learn electronics, it always puzzled me that oscilloscopes measures voltage but not current. Why not? From my POV, measuring current in an oscilloscope form could help in many other ways too.
Really depends what you need to do. Like others, I use the Pearson transformers if I don't need DC. Some of these will work into the several MHz. I have also used the LEM LA-55 and an old Micro Switch. If you work from a common ground, the shunt/amp may work. I have several antique ones of various sizes. Normally, I only use these with a meter.
Also I have a 1960s Tektronix probe that I got for free (non working). These are not bad probes if you could find one and the sensor appears to be rated for much higher bandwidth. I was able to modify mine to work well above 80MHz.
https://www.youtube.com/watch?v=6f8zoyBxizs (https://www.youtube.com/watch?v=6f8zoyBxizs)
https://www.youtube.com/watch?v=RVY9q5JhHHU (https://www.youtube.com/watch?v=RVY9q5JhHHU)
-
I've got one of these -> http://www.dx.com/p/hantek-cc-65-ac-dc-current-clamp-meter-multimeter-with-bnc-connector-blue-black-197938?utm_source=GoogleShoppingAU&utm_medium=CPC&utm_content=197938&utm_campaign=428&tc=AUD&gclid=CNnM3Zj23rkCFUpZpQodBgcAYg#.VdwRkH3oSf4 (http://www.dx.com/p/hantek-cc-65-ac-dc-current-clamp-meter-multimeter-with-bnc-connector-blue-black-197938?utm_source=GoogleShoppingAU&utm_medium=CPC&utm_content=197938&utm_campaign=428&tc=AUD&gclid=CNnM3Zj23rkCFUpZpQodBgcAYg#.VdwRkH3oSf4)
Didn't break the bank, and good enough for most measurement.
-
I've been trying for some time (in a couple of other forums) to get someone to do a simultaneous measurement with one of those Hantek CC65 current clamps, and an ordinary current-viewing resistor, to see how accurate the clamp is (magnitude of current) and what is its inherent delay (skew, phase delay).
Are you able to perform such a comparison? It would be very interesting to see the results of such testing.
Obviously, for accurate power math where one is multiplying current x voltage to produce an instantaneous power curve, one needs to have the current probe synchronized with the voltage probe. Using the voltage drop across a low-value, noninductive CVR for current, and monitoring voltage with the same type of passive probe, you get fairly accurate synchronization.
Of course, properly using a CVR for current has its own set of pitfalls...
-
How can the Hantek CC65 measure DC?
-
As mentioned above, any oscilloscope can measure the voltage across a resistor and therefore can be used to measure current with a shunt.
In the original post, the measurement was done on a piece of mains powered equipment so a current sensing transformer or Hall effect sensor would be best, as they offer galvanic isolation from the mains. If these are unavailable, then you'd need to use differential probes with a suitable CAT rating or an isolating transformer to be safe.
-
How can the Hantek CC65 measure DC?
It uses a Hall effect sensor.
-
I've been trying for some time (in a couple of other forums) to get someone to do a simultaneous measurement with one of those Hantek CC65 current clamps, and an ordinary current-viewing resistor, to see how accurate the clamp is (magnitude of current) and what is its inherent delay (skew, phase delay).
Are you able to perform such a comparison? It would be very interesting to see the results of such testing.
Yeah, that's an interesting idea. I'll put a sinewave into an audio power amplifier and then into a 10 ohm resistor and look at it as you suggest.
-
As alsetalokin4017 says,you can use two channels of a 'scope to measure the voltage with respect to common on both sides of the current sensing resistor,then use the mathematical functions to give the voltage waveform across the resistor.
From this,we can derive the current waveform.
This us how I did it with an analog Oscilloscope on the few occasions I needed to.
I usually wasn't that worried about instantaneous values---------more the overall trend of the thing.
With a DSO you should be able to do considerably better.
-
Okay. Here is the current through a 10 ohm resistor BLUE (Hantek CC65 current clamp) and the voltage across it YELLOW. Resistor measured 9.7 ohms +resistance of connecting leads + amplifier output impedance. Mostly the voltage is about 15 volts peak but I forgot to bump up the amplifier output as the voltage fell away slightly at high frequencies. There is about 5uS phase shift.
Edit -> used an ordinary 10W wire wound resistor so maybe it is a little bit inductive. Will try it again tomorrow (10pm here) with 10 x parallel 100R 2W carbon, and also keep the test voltage constant as the frequency goes up.
-
Edit -> used an ordinary 10W wire wound resistor so maybe it is a little bit inductive. Will try it again tomorrow (10pm here) with 10 x parallel 100R 2W carbon, and also keep the test voltage constant as the frequency goes up.
I agree that's probably the case. A thin film resistor would probably be best at high frequencies.
-
Okay. Here is the current through a 10 ohm resistor BLUE (Hantek CC65 current clamp) and the voltage across it YELLOW. Resistor measured 9.7 ohms +resistance of connecting leads + amplifier output impedance. Mostly the voltage is about 15 volts peak but I forgot to bump up the amplifier output as the voltage fell away slightly at high frequencies. There is about 5uS phase shift.
Edit -> used an ordinary 10W wire wound resistor so maybe it is a little bit inductive. Will try it again tomorrow (10pm here) with 10 x parallel 100R 2W carbon, and also keep the test voltage constant as the frequency goes up.
Thanks for performing that test. 5 microseconds is a rather huge delay, as those things go. This is generally beyond the ability of scope "de-skewing" settings to compensate. For instance my Rigol 1054z only allows +/- 100 ns "Delay-Cal" adjustment, so it would be impossible to synchronize this current clamp to a plain passive voltage probe, even by setting one channel to +100 ns and the other to -100 ns "Delay". So, one would not be able to use this current probe in conjunction with a passive voltage probe for generating an accurate instantaneous power curve at anything but the very lowest frequencies.
You might be surprised to see just how much magnitude error is introduced at higher frequencies by using a wirewound resistor for the current-sense resistor. I recommend using a real, noninductive resistor such as the Ayrton-Perry-wound Ohmite WNE1R0FE or similar for the CSR, and properly connecting the sense leads to the resistor "Kelvin probe" style. Even the inductance of a few inches of plain wire (like a passive probe's Ground reference lead) is enough to seriously screw up measurements.
Thanks again for performing the test! You might also compare the responses using square wave vs. sine wave....
Here's a little video I did some time ago to prove a point to a "free energy" claimant who based her claims on faulty measurements. Please ignore the tone of the video, it is pitched to the "eighth grade" level of the claimant who really was out of her depth.
https://www.youtube.com/watch?v=-a1plHZwmWg (https://www.youtube.com/watch?v=-a1plHZwmWg)
-
Tried it with ten parallel 2W 100R and at 25kHz the lag is 4.5uS instead of 5uS with the wire wound resistor.
-
Tried it with ten parallel 2W 100R and at 25kHz the lag is 4.5uS instead of 5uS with the wire wound resistor.
Not really a significant difference at such low frequency. I just had a look at the specifications for the CC65 and its frequency range is stated to be DC to 20 kHz.
So I'd challenge the statement that it is "good enough for most measurement". It would be nearly useless as an AC current sensor for my own work. But a DC current clamp does come in handy sometimes.
Here are some current probes with reasonable bandwidths:
http://teledynelecroy.com/probes/probemodel.aspx?modelid=1130&categoryid=3&mseries=426 (http://teledynelecroy.com/probes/probemodel.aspx?modelid=1130&categoryid=3&mseries=426)
These probes are way out of my price range though. But I have borrowed one to perform this demonstration, again pitched to the lower-division student:
https://www.youtube.com/watch?v=KWDfrzBIxoQ (https://www.youtube.com/watch?v=KWDfrzBIxoQ)
-
So I'd challenge the statement that it is "good enough for most measurement". It would be nearly useless as an AC current sensor for my own work.
+1
Yes I looked at those Hantek CC65's a while back and discounted them as useless for most work.
But of course it all depends on just what you need, for examination of most switchers you'd want at least 100K BW, preferably 1M or more, then try as you might there's very little quality under US$1K.
As I own a Tek P6021, I keep using it as a standard to compare all others against; 120 Hz-60 MHz BW, 1mA min and 15A peak . It's hard to beat and been around for years.
Tek's selection and list pricing:
http://www.tek.com/current-probe (http://www.tek.com/current-probe)
-
Horses for courses. My main interest is automotive ignition systems and for this that clamp is just fine. I used to work in a SMPS design lab for ten years and used way fancier current clamps. For what I do nowadays though the CC65 is good enough at least for what I do.
-
Im not sure if this will help in any way
http://hackaday.com/?s=AD8428 (http://hackaday.com/?s=AD8428)
But the 3.5mhz bandwith of the opamp maybe adaptable to your application?
-
http://www.dorkbotpdx.org/blog/paul/measuring_microamps_milliamps_at_3_mhz_bandwidth (http://www.dorkbotpdx.org/blog/paul/measuring_microamps_milliamps_at_3_mhz_bandwidth) This is the acuall full instructions with scope shots, I dont know how much current your measuring but maybe ditch the shunt for a clamp or hall effect device, this should help speed up the AC skew, maybe not I dont know....
-
Just picked up a Hantek CC-65 current probe. I'm working on some motor sensing circuitry and have a precise onboard current sensing circuit. I clamped the CC-65 around one of the motor leads and connected it to my scope's CH4. The onboard circuit is connected to my scope's CH1. I've attached a screenshot of the two traces, offset a bit for comparison. (CH1's cutoff on the first peak is due to saturation - the motor was drawing more current than the circuit could report. It was within the CC-65's range, though, so its output isn't truncated.)
I'd say, for the money, the CC-65 does a pretty good job of tracking what the onboard circuit reports. Would I like to have a nice Tek current probe with its standalone amplifier? Sure! But for many applications I think the CC-65 is a heck of a deal.
-
Just picked up a Hantek CC-65 current probe.
I'd say, for the money, the CC-65 does a pretty good job of tracking what the onboard circuit reports.
Thanks for the info. Here is an Amazon link to it
http://www.amazon.com/Hantek%C2%AE-Current-Clamp-kHz-20mA/dp/B00BLD6FB8 (http://www.amazon.com/Hantek%C2%AE-Current-Clamp-kHz-20mA/dp/B00BLD6FB8)
so I added it to my wish list