Can my $15 DIY AC/DC Current Clamp keep up with a commercial one? || DIY or Buy
Vložit
- čas přidán 10. 07. 2024
- Visit Brilliant.org and get a 20% discount: Brilliant.org/GreatScott/
Previous video: • Producing lots of Manm...
Facebook: / greatscottlab
Twitter: / greatscottlab
Support me for more videos: www.patreon.com/GreatScott?ty=h
Altium Designer: altium.com/yt/greatscott!
You can find more information about my DIY solution here: / 52795400
The website shown with the nice 3D printed current clamp: matthewtran.dev/2019/08/diy-d...
You can get the commercial AC/DC current clamp here (affiliate links):
s.click.aliexpress.com/e/_AVZjBj
DIY solution parts list (affiliate links):
LM7805: s.click.aliexpress.com/e/_dXS...
Resistors: s.click.aliexpress.com/e/_dTP...
Capacitors: s.click.aliexpress.com/e/_d7d... s.click.aliexpress.com/e/_dU4...
S49E Hall Effect Sensor: s.click.aliexpress.com/e/_98ld1L
200k Trimmers: s.click.aliexpress.com/e/_AVTh3R
TL082 OpAmp: s.click.aliexpress.com/e/_9iaLp7
BNC connector: s.click.aliexpress.com/e/_AYumOt
PCB Terminal: s.click.aliexpress.com/e/_dYb...
Perfboard: s.click.aliexpress.com/e/_dXx...
In this episode of DIY or Buy we will be having a closer look at my broken AC/DC current clamp which can be also called a current probe. The job of such a current clamp is simple, visualizing the current waveform on an oscilloscope. So in this video I will explain how such a current clamp functions and how we can make a DIY alternative in order to find out whether it makes sense to DIY such a tool or whether we should stick to the commercial solution instead. Let's get started!
Thanks to Brilliant.org for sponsoring this video.
Music:
2011 Lookalike by Bartlebeats
0:00 Why current clamps are awesome
2:04 Intro
3:00 Why not use a current shunt?
3:55 How does an AC current clamp work?
5:17 Measuring a DC current with a current clamp
6:41 Building a DIY AC/DC current clamp
7:26 Building a circuit for the current clamp
9:07 Testing my DIY solution
10:30 Verdict - Věda a technologie
I was just trying to understand how can I make a current clamp. And there my German friend just dropped a video! 🤩
Hope you enjoyed it!
7 likes
Me to! I recently bought an oscilloscope and realized accessories are crazy expansive. DIY is basically the only way
Same
@@greatscottlab Can we use a commercial AC DC clamp meter as Current probe ? Any Hack
You could add a 100nF from pin 3 of the opamp to ground to decouple the offset adjustment voltage, this may take out some of the noise. Also the 1Meg on the output of the opamp will make a low-pass filter together with the capacitance of the coax cable and the oscilloscope input, which could be limiting the bandwidth.
Agreed
I would add a buffer at the output before the coax and in my opinion that would solve the problem with bandwidth.
You need to be mindful of your amplifier's bandwidth. Depending on your opamp IC, there is a limit on the maximum frequency it can operate at. And also, for such small internal voltages, some EMI shielding and signal traces isolation might be desirable. Furthermore, as @MikrySoft has already said, there is another current clamp design, which uses an active secondary coil. The purpose of this coil is to induce an oposing flux such that the total magnetic flux being measured by the Hall-effect sensor is 0.
I agree with closed loop current clamp design, it can provide range beyond hall sensor can handle (since it measure 0 magnetic field). for the frequency respond I think he use cheap Chinese opamp clone which can't handle that amount of frequency. I think improving the design by using class A amp instead since it need precision and not efficiency.
The "buy" clamp had what seemed to be two hall effect sensors as well, does that help with accuracy?
Also, the maximum frequency of an op-amp is often stated at unity gain, or at gain of 1, so the thing you should look out for is the maximum gain-bandwidth product.
edit: using a 2 stage opamp amplifier isn't a bad idea. The max frequency should increase by about 60-70%
This is exactly right. A gain of 200 with a square-ish wave frequency of ~5kHz is pretty rough for a cheap-o op-amp
There is a second way such current clamp can work, utilizing both the Hall effect sensor and a secondary coil. Circuit drives the secondary coil to cancel out the magnetic flux in the core (as measured by the Hall effect sensors). That way you can get bigger dynamic range, but you need a more sensitive sensor that's linear around zero.
there is even a third way, only using coils: you use one coil as the compensation coil and another one as the probing coil, measuring its inductance and thus the magnetization of the core. But these are not really suitable for AC measurements, unless you make a switch for two different modes of operation.
There is the reason why the comercial current clamp uses two hall sensors.
@@VladoT
You, Sir, are a very good teacher. Your diagrams are terrific, and with your dialogue, you easily get your lesson across. As a retired electronics tech, I find you are doing a great service. Thanks. I have subscribed.
I have to say; I always learn a lot from the videos on your channel, but the most impressive thing is this dedication to making so many drawings and explanations with the pen. Congratulations and thanks for sharing so much information and knowledge. I'm a fan. 🇧🇷
Thanks Scott! Now I understand a bit more about how my Hantek CC-65 works. Your videos always help clear things up, it's a nice service that you do by sharing this information ... Thank You!
I never knew the symbol for magnetic flux was a TIE Fighter. :)
The more you know ;-)
More like PHI Fighter ._.
More like the Green Lantern Symbol lol
|-0-|
I always wondered why my opamp circuits seemed like garbage. Never made the realization that my breadboard was the cause. Thanks for calling that out!
It's a great video. I'm starting to make my own measurement tools as the college lab is inaccessible due to covid. Thank you greatscott
Glad it was helpful!
This is great. I want to measure the current of 4 or 6 circuits in my RV, and this looks much easier than redoing the wiring to add Inline current sensors.
Great video!
I created, back in day's, AC current clamp by myself and didn't purchase factory one.. I just take old ferrite core and wind up my own windings and connect across secondary output, one resistor.. Worked as a charm, only problem was, that I didn't cut in half ferrite ring and connection was problematic because that.. 😂 But that DIY was quick and didn't cost anything and was made just for one job..
6:43 Best part! 😂
Awesome video, this topic I did not expect, but it is still surprising how cheap and effective the diy clamp is.
Hi Mr.Scott, I am a subcriber of your channel from India. Your CZcams videos are very inspiring and knowledge enhancing. Keeping rocking bro!!!
Thanks, will do!
The best as always, keep the good work @Great Scott
Man, I swear you have access to my Google search history. I've been looking for this exact build on and off for a week.
You are really great Scott! I was also thinking about this yesterday and you just made it, awesome!
Glad you liked it!
The AC/DC current clamp you made Rocks
XD
Great Scott, that's a decent circuit. To make it work on higher frequences, solder couple wires more to the spare opamp you have already there! Use it as a buffer, because in current setup you might have create some filter between virtual ground and coax. In my opinion, if its working great at low frequencies, this would be the reason.
BTW I have looked for a while for nice inexpensive current clamp and for sure going to build that.
Another option for even cheaper way of measuring the current would be to use small value resistor (0,1 ohm for example) and amplify the output with opamps to suitable level. Good for electronics and more accurate. Regards.
Honestly. Both you and Electroboom are the reason why I got passing grades in my electronics major. Thank you so much!
P.S. I'm also german and it's weird talking english to a fellow German guy 😂
Underrated comment!
Happy to help!
@GreatScott! I saw from the datasheet of the opamp you used in the DIY version that it has an f unity of 4 MHz . An op amp with a bigger funity (ad8001 perhaps) would make the diy clamp able to function at higher frequencies ?
The frequency is also limited by max frequency of hall sensor
I just recently made a signal booster coil for an automotive DC clamp meter. The problem was trying to chase down a parasitic draw. The inexpensive DC clamp meter just didn't have the necessary granularity for mA currents in the 1's an 10's range.
For a quickie solution, I created a 10 loop (10x) coil of wire which was placed in-line with the circuit to be measured. The amp clamp was then applied to the 10x coil. The hack allowed me to locate the majority of the parasitic draw. In the back of my mind, though, I was thinking a 100 loop (100x) coil would be nice.
Just this week I ordered a spool of 20 gauge magnet wire and wound the 100x version. Works like a charm!
Granted, if one has access to electronic bench style tools and equipment, such an effort is a waste of time. But, if one is just limited to a generic amp clamp and multimeter, well, the ability to extend those tools has a pay-off.
The music of ac/dc plays faintly in the distance...
Sound of the drums
Beating in my heart
The thunder of guns
Tore me apart
You've been
Thunderstruck! 🎶 🎵
finaly. i was wating alot of time for that video
There you go :-)
Great content as always
Glad you enjoyed it
Pretty impressive work, dude! Fantastic! 😃
Stay safe and creative there! 🖖😊
Thanks, you too!
Cool little project. Thanks for sharing with us.
Thanks for watching!
Let's get started can never get old
I also tried the DIY version a few months ago. My problem was already visible with pure DC. At some point the core became magnetized so that no sensible measurement was possible any more. Either demagnetize or wait a day. I tested various core materials and hall sensors. With some combination the effect was earlier, with others later, but always the same problem with the residual magnetism. I then bought this cheap current clamp. It works perfectly, although it wasn't that cheap anymore ;-)
I always wondered how they worked, thanks
Finally you uploaded
I have to write 30 pages of electronics assignment till tomorrow
I am running low on time
But I got time to watch it
This 😅
Hope you give me your blessing to do the assignment in little time
Hahaha, just like me... "Ah, i'll start this the night before".
Good luck to you :-)
Aha I am done thanks :)
Yours will allways be the best!!!
Thanks mate :-)
I think the reason that the "DIY" didn't perform well on higher frequencies is due to eddy current losses, that's why the "buy" one (as seen in the video) is using laminated thin plates to minimize the effect of Eddy current losses. And the core you were using on the DIY was single piece core.
Thanks for video it was great.
@10:07 he used the core from the buy one and still had performance issues. Its more than likely an issue with the hall effect sensor itself not being designed for such applications since it was the cheapest linear hall effect sensor you can buy (I literally bought the exact some ones like a week ago).
@@bestcreations4703 I think he use the hall effect sensor only for DC current not for AC.
Great video, sir. Thank you.
I just bought this thing yesterday!
quite nice and useful. Thank you.
Hey buddy I just enjoy watching ur videos as i love electronics
6:09 you soldered an SMD capacitor to a THT component in air. My mind is blown 🤯!! Also, at 8:20, all you had to do to improve the rail splitter was to buffer the 10k-10k voltage divider (without the electrolytic caps) with an op amp in a unity gain configuration.
But I still think the dual voltage rail was not the culprit. So adding a voltage follower would not have improved much. But of course feel free to try it out.
Thank you for listening!
I have learned a lot from this video
Hmm, okay, 5kHz is no way good enough for investigating switch mode power supply operation. The other problem with current sense resistors is that they have inductance. A more accurate equation for voltage across a sense resistor is V = iR + Ldi/dt, and if the current is changing fast, the voltage signal is useless for observing the current waveform.
The hardest part of measuring current accurately is accounting for the change in core characteristics with a gapped core. Rather than break the flux path, it is better to break the current path. Less convenient to be sure, but far more accurate and cheaper than a high quality high bandwidth current probe.
The solution I use is an LEM current transducer, with a split supply from a dc-dc converter, all powered from a 12V plug top adapter. I mounted this all in a box with a 2.1mm dc power connector and a BNC for the signal output. Yes I have to break the current path, pass the wire through the sensor and reconnect it, but I get a superb signal with a bandwidth of up to 200kHz with the LA 55-P transducer, with an output resistance calibrated for 10mV per amp. It works REALLY well!
Thanks you for the information
I always like your videos, you are great
Got the same Hantek clamp....It is really really good for its price! Bought mine for 35$
An interesting article. Thanks.
Helpful information
Very good... Thank you guys🙏🙏🙏
Great project!
Great Job! Scott
Thanks!
Wideband AC/DC current probes are hard to do. I bought the ICP5150 from AliExpress and am quite pleased. Not cheap but better than the thousands of $ wanted by Tek, HP, etc.
Looking at a datasheet of Honeywell SS49E, there's a graph suggesting that this kind of Hall sensor is suitable up to 30 - 40 kHz 😉
the bottle neck definitely on the op amp, it just a slow op amp.
Loved the "desoldering" job... ;D
Great!! Great Scott 😉
Consider buying other solder tips if it takes you one hour for such a small circuit. A small conical tip has a low mass and their for heat capacity and a bad geometry for transfer of heat to the board. I use almost only wedge shaped ones. To reduce thermal stress, I use bigger ones. The higher mass means the solder don't get solid as easy, and I don't have to stay on the point as long or even increase the temperature.
The current clamp is a nice idea. I rarely use one at home so I am used to borrow the one from work when I need it.
The king is back
Oh stop it😊
@@greatscottlab You deserve it
Thanks for sharing! Literally just commented regarding a few days back on Kerry Wong's latest video where I'll be needing for the hybrid CVT MG's design inspection and performance characterization. Now am even more-so planning to make and appreciate the insight. Wondering if those plates are Metglas, Permalloy, MuMetal or what? Was thinking about just using a cheapo current clamp meter hacked out and I think you went even more cost effectively. Really awesome to watch as always!
I love AC/DC
I already bought one and payed through my teeth for it, but for my application (mostly audio) I would definitively have used your DIY version.
I'm pretty sure your issue is caused by four factors. 1 You are using a dual op-amp. The TL082 ideally needs unused inputs tied to ground via a 10k resistor. 2 the GBW for a TL082 is 4MHz which sounds like it should be enough but really is not for anything that has a square wave (a square wave can be thought of as a collection of sine waves all constructively and destructively interfering and some are at radically high frequencies). 3 it's a FET input op-amp. Bipolar would be a better idea due to FET's having a gate charge (I thought you might need reminding of this) which really does make a difference in situations like this. 4 It's a TL082. They are an old design and renowned for distortion in audio circuits. In fact in audio circuits you can be better off using something like a 741 in these cases. Ad that to everything else and you did well to get it up to 5KHz. An OPA228 should boost performance significantly.
Great video man.
Glad you enjoyed it
@@greatscottlab Really, loved your content man. It really inspires young youth like me.
Watching you gives me a courage to join this field. That's why right now I am pursuing b.tech in electronics and communication engineering.
Great thanks again.👍
You are a hope and inspiration to a lot of people.
I want to make myself something great, so that I will be also able to have this glorious mind and instruments like yours.
Till now, I am just able to buy mastech mas830l multimeter, that costs around 7us bucks or 500rupees in India and trying very hard to anyhow buy a fluke one with all features.
One day I will be able to do that and your videos will keep inspiring us and maybe at that time, we both will become great youtubers.
Great video, maybe the material of core contributed to limit the bandwidth for high frequency, many iron core only works in specific ranges and attenuate in others. Thanks for share
I have some much fun with this video
Do frequency response of DIY clamp. From that you can exactly find What's the problem with higher frequencies. It might be wrong chosen IC for Opamp or any other. Replace it & do frequency response again.....Your Keysight oscilloscope must have this feature.
Don't forget me when oscilloscope giveaway starts because I also need it😌😁
Apart from choosing a more suitable opamp, the circuit's performance could also easily be increased by using an active railsplitter with an opamp. While the circuit still is very simple and requires minimal components, the output voltage will become more stable. Also adding small decoupling capacitance to the opamp and properly terminating the unused opamp will probably help performance.
Anyways interesting topic. It blows my mind how expensive current clamps are while the principle isn't really that complicated... :-)
Thank you.
Very cool video
I'd posit two areas for investigation
1. the voltage rail splitting. As you mentioned, resistor-capacitor network is ugly. I don't know how well it copes with higher frequency demands. I've seen rail splitting ICs for DIY audio devices, which have a couple of op-amps packaged in them.
2. The OpAmp's frequency response curves. The output curves really look like they're being filtered.
2.1 oooooh, the perfboards' traces leading to stray capacitance and other unwanted features.
It's highly likely that a designed PCB could bring this excellent proof of concept up to being very functional.
Nice work!
Thanks for the feedback :-)
Great channel!!
Thank you!
interesting. Thanks for sharing.
Thanks for watching!
Awesome I will try to make one with arduino. ❤️
Great Video thanks!
One comment to the bad AC performance. If I saw it correctly you almost had a Gain of 200 if the 1A current ceated as shown a voltage of 500uV this would decrease your Opamp Bandwith from 4Mhz to 20kHz. (GBP 4MHz Typical) which would decrease the Amplitude of the Harmoics incredible creating potential the bad shapes. Possible solution at least to try it different Opamp with higher bandwidth. May take a look at the one used in the commercial clamp. What do you think?
The SS49E has a frequency response that drops off sharply at 40khz.
It probably also uses internal chopper stabilization.
I think it's the hall sensor that is giving you trouble. The one you picked is not recommended for high frequency applications if i am not mistaken.
Amazing video anyways! Tho i suspect that, if have money for an oscilloscope, buying a current clamp is not that expansive XD.
The datasheet says 3us respond time which means above 100kHz should be possible. But that was not the case. Not sure if the sensor was the problem.
@@greatscottlab 3 microseconds gives us ~ 333KHz. For something like a square wave signal to be reproduced with a reasonable approximation, we would need at least five harmonics IMO, so this device could perform at best to 60KHz. With PWM circuits, the waveform will not be 50/50, and that implies an even lower maximum frequency of operation.
Just my $0.02 worth.
I love this so much. Do you think you could do a video using parts from an apc ups to make a sine wave inverter?
Nice project - but you should have considered using a dual opamp where you feed your voltage-divider into one section which is configured as a voltage follower. Then you can use the output of this opamp as your virtual gnd. Regarding your bandwidth problem: what is the amplification you used for your inverting amplifier section, and have you considered the opamp's gain-bandwidth product? A 1 MHz opamp in a 100x connection can only deliver 10 kHz of bandwidth.
The gain bandwidth product of the op amp is 2~4 MHz. This means that you have a 3dB band worst case 20 kHz. With capacitive cables between sensor and opamp input you even have a lower bandwidth due to the integrating effect. Better check your circuit again. Also, why did you not use your opamp #2 in the IC to perform the virtual ground with it?
Current clamps are nice
Can it be used as clump meter adapter for DC current? Can it be connected to a multimeter and to "measure" amperes in the area of volts?
Electronics engineering at its best!
Superb video . Can u make a video upon good books related to electronics
Perhaps next time it's a good idea to make "fork" style DC current clamp ! there're some DC current meters exist came with a "fork" instead of a "clamp" that need to circle around the cable, I'm really curious about how they works! (ex. KEW MATE 2012RA multimeter from Japan company Kyoritsu )
You may want to look into another JLCPCB sponsorship, as perfboard + high frequencies is not a good combination. A good designed board for your electronics could make the difference ;)
You should do a DIY or Buy of amateur radio! (with the proper disclaimers of course haha)
Maybe you should use compensated probes to do some measurements directly at the sensor (or amplifier) output. Also the sensor could be fake. According to the datasheet, frequency roll off shoud start above 30 kHz, so there ís something to gain. Just my 2 cents! Great project!
Most current probe's I'm familiar with use a 50 ohm output impedance and you have to set the scope to 50 ohms input impedance. You've set yours up for 1 MOHM output impedance. I'm not sure that's the best choice here.
I think we have to use a special ferrite core for that
Because at higher frequency some cores doesn't demagnetize properly
I think that the culprit is the lower frequency response or high recovery time of amplifier ic as well as hall effect sensor...but for this price its acceptable..👍👍👍👍👍
If I just want to measure dc current it is possible just making a tip with a 49E sensor and placing the sensor face in parallel with the cable and touching it this assures the magnetic flux pass orthogonally to the sensor having a good current reading. (I think this will work just when you have isolated cables from others).
I could avoid the ferrite core and just use a pen with the sensor in the tip.
I've got one of these, which work well - it's just the calibration button is a bit like voodoo - sometimes it works and sometimes not. You really have to be aware each time that your "zero" may not be zero.
I think the frequency problem might have to do with the op amp's frequency limit (bandwidth?) and maybe parasitic inductances and capacitances. It may also have to do with the voltage divider, noise, and other stuff. Just some random educated guesses.
Few things to check: some of analog sensors have chopper stabilization and thus narrow bandwidth. So make sure that your sensor is not of this type. The SS49E has response time of ~3us, so it should go to 100kHz easily. Your opamp has 4MHz GBW product, so at gain of 200 your corner frequency will be only 20kHz, thus not so good picture. Not sure what was the intent of R4C5? Input voltage divider? I'd rather use 50ohm in series to avoid oscillations to drive a cable, but you will guarantee the frequency linearity. Finally, you have second opamp - why not use it for active virtual ground? The best would drop the gain to ~10 (400kHz BW) and if needed, run wire through the core multiple times to increase conversion gain.
The response time of your hall effect sensor is the problem it is probably in the micro seconds thus it is too slow for higher frequencies
3us according to the datasheet. Should be able to perform than what I measured.
I know such device but now I know working
Nice 👍
I would love to see repair added to the list of diy or buy as in your case where you had a commercial unit, is it worth repairing the unit even if you have to replace most of the circuit.
Hi Scott sir this is awesome,you should make version 2 where measure value shown on Arduino with LCD display 😌😌
So a DIY oscilloscope? ;-)
@@greatscottlab yaa😍😍
Amazing
Thanks
since there's a solenoid built in, why not try to design the closed loop current clamp model?
TL08x opamp family is quite old and might be the cause for your problems, since its offset voltage and noise are quite high. Since the amplitude of voltages you are trying to measure is very low, you might want to use a more recent precision amps to achieve a better result, like Texas Instruments OPAxxx families (like OPA210, OPA228, ...)