Dave, I just noticed 400K+ subs, so wanted to say congrats to you mate. You work really hard for your videos and your subs, and in an increasingly broken RUclips full of click-bait thumbnails & titles and mindless 5 minute videos with 5 adverts in them, it gives me faith to see hard work paying off.
The only complaint I have about most of your videos that you go a little fast sometimes, but the beauty of video is that I can re-watch it. thanks for all the great work!
It's long, rambling, detailed and goes into lots of detail and at least mentions points that one should know about even if you don't go into it so it dosen't become a 4 hour video titled 'how to measure way too much about a DC-DC converter' Good video!
I have nothing but love and total respect for your devotion that you put into your videos!... I'm willing to bet you bailed a lot of young players out that where having trouble with there given semester based projects!.
Great video, I only wish you'd made it sooner! Last year about this time, I built a 12v power station/obnoxiously bright LED lantern for camping & stuff like that. (I haven't found a single retail electric lantern that's worth a damn, so my goal was to make an LED lantern with a nice, warm white color and a light output comparable to a propane lantern.) I used a buck converter with a constant output current & variable voltage to control the brightness of the LEDs. In order to get the best compromise between efficiency & light output, I set the current to a value I thought wouldn't be too hard on the LEDs, then ran it until the batteries dies. I repeated this process until I got everything dialed in. It took several days.
5:35 Dave, you're talking about the voltage drop on the wires from the supply to the converter, but what about the drop on the multimeter wires and the wires hanging down of the yellow bread board, which are longer?
Idjles Erle the multimedia has an impedance in the range of mega ohms. so very, very little current wil go through these test leads of the multimeter. So there will be also ben very little power loss or voltage drop in the test leads of the dmm.
Another plus for using Libreoffice as well, it pretty much does the same thing as Excel, just is not linked to a subscription model for pay as you use calculation.
It is not. Excel/Office can still be purchased like a one-time payment, but also available in monthly subscriptions (but is not limited to x amount of calculations..)
Nicely done! I have been automating my switchmode converter tests using an SMU and a DMM... but now using two Keithley SMUs and LUA scripting. I was wondering about how you did your graph with the X axis log, and then you answered it, good on ya!
I am a young player and I am confused by one of the traps that you mentioned. Can you explain why the voltage drop from the wires coming from the power supply matter? Why is the voltage drop from the wires connecting to the multimeter not considered? Is it because the multimeter has a really big impedance that we can treat it as an open circuit?
flowakfg yes, exactly. voltage drop is proportional to the current flowing through the wire. high impedance means almost no current flows through the test leads so there is effectively no voltage drop for these measurements.
If you measure the voltage at the power supply terminals and not the voltage at the pins on the regulator then power lost in those wires would be not accounted for, you wouldn't be purely measuring the loss caused by the regulator which is what you are interested in when the regulator will be used in different applications where the loss caused by outside factors will be different to your testing setup.
flowakfg its the basics of a resistor. every wire is a resistor, and when current flows, a voltage drop occurs. The amount of current is what determines the voltage drop. A voltmeter draws close to no current and thus the drop is so low that it can be ignored (less than a millivolt).
You're spot on. Multimeter input impedance will be in mega ohms, if not giga ohms, depending on it's design. It's small enough that the current flow is negligible. That said, with a high enough resolution meter, you could theoretically measure a voltage drop in that scenario. Don't know what you could do about it though. Six wire measurement? :P
Dave, could you do a short video introducing the SMU 2400 capabilities as a trade off with a multimeter and power supply. Looks so similar features that a more details presentation will help for sure to understand the differences.
Having done the single multimeter / oscilloscope DC converter measurement circus with a junk bin supply and hackjob load there is something else to remember: be careful about what your circuit ground actually is (understand your power supply) and don't get sloppy or you'll connect a fat switching node directly to ground and mosfets will explode in your face
I have to say I am shocked and amazed at the 0.3+ V drop on the output side, at only 80 mA. That's like 4 Ω in the connections, which can't possibly be the cables. Are breadboard contacts that bad?
since your wire resistances will most probably stay constant, you could just measure their influence once (at a high-enough current) and then correct for the resistive loss from the measured currents. But .34 V drop for just 250 mA of current seems high for your wires - do these contain some copper at all? I am asking, because I recently got crocodile cables from China which are magnetic - not the crocodile clamps, the cables!
So you're saying let the voltage drop stay and just compensate for it in calculations? but then you have a varying voltage at the DC/DC converter input pin itself which will cause it to change its switching characteristics to maintain the output and might change the efficiency vs a constant input voltage. Thus your graph would still be different to another graph recorded from a test with different input wires.
Solderless breadboards don't make the best of connections, especially when they are well used. Most of the voltage drop is probably due to the contact resistance of the breadboard. The wires should only drop around 10mV if they are pure copper.
Exactly, this is complete bullshit. Nice demonstration, but utter bullshit. That 10 cm of wire should never, ever have four ohms of resistance, as for other connection issues well that's on the tester - and I absolutely will laugh out of existence any "tester" that somehow manages to introduce four ohm's worth of resistance in his test cabling, no matter where he chooses to measure the voltage. Ridiculous!
Attila Asztalos for one its an educational video, if he used thicker cables he wouldn't be able to demonstrate so clearly why you measure the voltage right at the pins of the regulator, secondly since he's measuring the voltage at the pin of the regulator making sure the voltage is constant it doesn't matter what the resistance of the cabling is anyway?
Dave, try using QTI plot (google for unofficial builds), way better for plotting (you can zoom into graphs 8-)... Next step (if you desire) is Paraview.
As you knew you wanted a logarithmic graph, you could have measured data in values of current equally spaced in the logarithmic domain. For example, with 8 values per decade, the base numbers would be [1-1.334-1.778-2.371-3.162-4.217-5.623-7.499] (multiply by 10^(1/8) at each step).
Can you make a video of how to use 4 wire measurement method to measure very very small resistance with 4 multimeters? I was trying to measure the resistance of a current sensing resistor, which normally has a resistance between 2mOhm to 5mOhm....
The easiest way to do that is with a constant-current supply & a single multimeter. The procedure is to put a known current through your resistor, then probe the resistor terminals with your multimeter. If you don't have a supply with a constant-current mode, it's a bit more complicated, & requires 2 meters - one for current, & another for voltage across the resistor terminals.
I think the reason you cannot use line charts with horizontal log scales is because excel doesn't treat the horizontal axis as numerical but as categorical.
Hi Dave, Great video! It did, however, get me thinking. If both your input wires and output wires were of the same type and same length, would you still need to perform the 4-wire measurement to compute the efficiency? Specifically, my thinking was that if efficiency is computed as the ratio of the output vs. the input and if your wire lengths were the same for the input and output, the voltage drops should be proportional and therefore the computed efficiency should be similar to what you computed using the 4-wire measurement. Am I incorrect in my thinking?
We can tell Dave's quite animated when he speaks even though we can't see him because of the way the metal "flashes" as Dave moves around, blocking and unblocking the light in the first minute and 40 seconds or so of the video.
Tommy Oz Gnumeric users with xichart or such will probably save time to have the thumb free. That said, how'd he get the dvm to talk to the digital load?
You don't need to account for the length of the voltmeter leads; for voltage measurement there is negligible current going through the voltmeter leads due to Megaohm+ resistance of the voltmeter being several magnitudes larger than the source/load resistances - just be sure to measure voltage on input after the ammeter in your circuit to account for potential voltage drop through the ammeter on the input (and likewise before ammeter on the output).
He does go on too long at times, so I have to skip parts, but I also have to skip Dave's DSO, and most other teardowns of newish equipment, because you don't really learn much. ElectroBOOM forever.
I love your videos, I was just wondering since you were saying that the voltage was being dropped across the wire, so you were measuring right at the input, wouldn't there also be a drop across the wires going to the meter?
[ remembers his City & Guilds 2391: Test and Inspection ] It is just to have a 'known resistance' of the input cabling to isolate any 'current effects' to the circuit under test. [ language may be out-of-date/English ]
Hello Dave, Excellent this video, could you share the Excel Worksheets that I use in your Presentation? Thank you for sharing your knowledge with the community.- Greetings from Argentina.- Antonio.-
Thanks a lot for such a great informative video. can you tell me, how to measure efficiency in DCM mode? I should consider RMS value of current or not? Because in DCM, as current reaches to zero RMS value becoming very low and thus input power is also very low.
Instead of using 4 meters, you could just use the sense connections that should be on any decent psu and load, certainly are on mine and they're aren't top brand name stuff.
Dummy question: If I power a device with a 12 volt car battery, let's say a car stereo, but the cables are extremely thin and voltage drops significantly, does that mean the battery will drain faster?
It slightly depends. To the battery the load is lighter, but if the voltage is dropping to the load (car radio) the audio might be quieter (assuming it still worked at lower power) and you might crank the radio up. If you put the radio with good thick wires to a battery, set the volume, then swap out the thick wires with thin you might get a longer battery life. The best way to think of that radio scenario is you replaced the radio with a fixed load (resistor), then added another series resistor (thin wires) and reduced the current drain. A car radio might be a bad example for this because a lot of new electronics use switch mode for many things (including the output amplifier). Switch mode power supplies (and class "D" amplifiers) actually start drawing more current as the input voltage decreases, so adding a series resistance (thin wires) might actually drain the battery faster with increased current as the radio tries to compensate for the voltage drop.
Would like to have seen a clearer explanation of 4-wire sensing, without assuming even the basics, since its for "young players". eg:- * All wires will have some resistance, and there will be a voltage drop when they carry significant current. * Therefore the voltmeter wires must not be shared with any part of high current path.
Dave, I really like your videos, but can you make shorter versions if possible? Like a 10-min maximum summary version and 30-40 minute detailed version for people interested in more in-depth look. I think you would get a lot more views like this, I myself, for instance, find myself skipping a lot every week since I have over 40 subscriptions.
I am a bit confused how lousy efficiency doesn't equal heat. I guess it is because the output has less amps so the relative heat is higher, but smaller.
Well, you *might* get away with two multimeters, assuming that input voltage and output voltage is sufficiently stabilized. Otherwise, this might be the situation where even you can't get me away from cheap multimeters, Dave!
couldn't you use one meter? just measure the voltage in, voltage out, then the current in and the current out in series instead of all at once. besides you will get inaccuracies if you measure both at the same time as well. you will get a voltage drop over the few ohms of sense wire for current measurement, or you'd get a few mA leak across the voltage measurement resistor, depending on which way your meters are put..
"I'm sorry Mr Edison, your DC electrickery is very nice but our giant resistor rooms are glowing and we have used all the cooling fluid for beer. Thankfully Mr Tesla has got something he calls "AC/DC" which means we can transform from high Voltairs to low Voltairs." "And stay away from the Zoo. Electrocuting elephants has just made you the enemy of history."
I have never used a line chart. I have always considered it bad practice to draw a line between points. It is far better to apply a line of best fit and describe the equation compared to your hypothesis. Perhaps it is just because I publish scientific papers.
what i'd like to see would be a how-to on making a dc-dc/switching power supply as "quiet" as a linear one :/ I hate the little bastards because apart from conversion efficiency they have no other benefit vs a linear one
George Tsiros they can be completely silent (inaudible to humans), but that would drastically lower their efficiency. Still more efficient than linear ones tho.
In my 5.8Ghz video Transmission systems, clean power was mandaory. My solution was a buck converter set to 4.1V which feeds a linear regulator set to 3 3V. Perfectly clean output and good efficiency.
You never fucking ever put a spline trough a real word data points unless you know you have the resolution to back it up.All so, if you do have the resolution, then there is no need for a spline fit.
Crashy McSplodey Oh, just link the Maxim paper on it. They make a point of running 4 days of tests to note every inflection in operation, to find that point of quiescence near 0V and 13% improvement at 5.2v in, etc.
Yeah, right. I will now say "beacuse I don't have youtube channel and I am not into making videos" and you will say "THATS IT. SHUT UP". But this will not make this particular video any more interesting. I mean, like measuring voltage and current with 4 multimeters... How can you babble for half an hour about just that?!
Never Mind its perfect, i enjoy to hear every details how and why he choose a particular setting or option!! You become only a expert of because of the experience you have learned on the field over the years and if your work is your passion, it's great to talk about it!! I enjoy the details. Then again I am not no any way of an expert in electronics, is good to see what is behind of everything to make a good measurement.
That's my thing. And if you cared to actually listen there were lots of tips in there explained in detail. If you want 5 minute tutorials you subscribed to the wrong channel.
Come on, Dave. I love you and your channel. And I bloody wathc EVERYTHING from here like a maniac. Bit this one is just too long. It looks like you're getting out of topics(which is not true, right?)
I don't want to sound offensive, or like an asshole; But could you please start running your mic through an EQ, and lowering the "tone"? Watching these videos hurts my ears, especially if I was watching another video before these and it switches to one of yours while I am doing something, or if I am wearing headphones (which I have to promptly remove and switch to external speakers for your vids). Your voice tone is so high, and you speak so loudly, that it just hurts (literally), even when turning the volume down. I love your vids Dave, and as I stated, I am not trying to be offensive, do please do not take it that way.
EEVblog No No, Im not saying "modify" your voice at all, but you can use an EQ to cut the frequency down on the high end, and kind of "normalize" it in the videos, without changing your voice (I know, technically it DOES change it, but it isn't enough that you wouldn't still sound the same to the human ear). I don't think the problem is your "natural voice", as much as the way the mic pics up the higher frequencies with the mic being so close (lav mic), and the way they are reproduced through "tweeters", resulting in the sound being amplified like someone is screaming in your ear, it is actually "un-natural". If I were sitting in a room with you, this problem wouldn't exist, unless you were talking to me at this loudness, with my ear exactly where your mic is. As I said, I didn't mean it in a disrespectful way.
Again, no, not going to happen. I'm not going to piss around with post-edit audio just because someone has a problem with it on their audio setup. I can fiddle audio until the cows come home and people would *still* bitch about my voice. It's called a High Rising Terminal BTW.
EEVblog Its the placement of your microphone, not my audio setup; And it is my ears, they are rather sensitive to certain frequencies, which is why I cannot use television speakers or laptop speakers.
InXLsisDeo I'm sorry your audio equipment is shit then. As for EQ, as I stated, and you obviously didn't read, the biggest problem is when RUclips is playing whatever the fuck it plays, and I am working or something like that, then I have to switch EQ just for Daves vids. I also said I wasn't trying to be an ass, but you ignored that part as well and apparently thought you would do that for me. I had a legitimate suggestion, due to his specific circumstances with his voice, his Mic placement, etc., and stated WHY I had this suggestion. Not once was I rude, as asshole, or disrespectful, so there is no need for it from small people like yourself either.
Dave, I just noticed 400K+ subs, so wanted to say congrats to you mate. You work really hard for your videos and your subs, and in an increasingly broken RUclips full of click-bait thumbnails & titles and mindless 5 minute videos with 5 adverts in them, it gives me faith to see hard work paying off.
Thanks.
ForViewingOnly
The only complaint I have about most of your videos that you go a little fast sometimes, but the beauty of video is that I can re-watch it. thanks for all the great work!
It's long, rambling, detailed and goes into lots of detail and at least mentions points that one should know about even if you don't go into it so it dosen't become a 4 hour video titled 'how to measure way too much about a DC-DC converter'
Good video!
Dave, it's people like you that makes the world a better place to live! Thank you! Genies
I have nothing but love and total respect for your devotion that you put into your videos!... I'm willing to bet you bailed a lot of young players out that where having trouble with there given semester based projects!.
Dave - Great detail and perspective. Your videos invariably help me understand better or remind me of things I have forgotten or take for granted.
I appreciate the mini LibreOffice charting tutorial towards the end!
Hi Dave, Merry Christmas to you and your family, thanks all the videos you gave us to the whole year😊
Small error at 4:05 (and elsewhere) - Efficiency curve starts at 100uA, not 10uA
Great video Dave! Thanks for watching out for us young players!
Hi Dave, how is the Pi cluster project and Nixe Tube display going? I love seeing how other peoples minds work with these sort of projects.
Dave Jones - Australia's revenge for letting AvE onto the internet.
The test results are with 12v output, but the charts labelled 9v?
Oh, oops. I have a 9V one as well.
I'm confused too. dyslexia?
Great video, I only wish you'd made it sooner! Last year about this time, I built a 12v power station/obnoxiously bright LED lantern for camping & stuff like that. (I haven't found a single retail electric lantern that's worth a damn, so my goal was to make an LED lantern with a nice, warm white color and a light output comparable to a propane lantern.) I used a buck converter with a constant output current & variable voltage to control the brightness of the LEDs. In order to get the best compromise between efficiency & light output, I set the current to a value I thought wouldn't be too hard on the LEDs, then ran it until the batteries dies. I repeated this process until I got everything dialed in. It took several days.
nice video to help people test their Batteroo/Batteriser ;)
That's the idea!
Those haven't been ripped from existence yet?
I wanted to write almost the same :D Are they even a real thing? I hope they never release the Batteriser. Sounds like cancer in my ears...
Interesting video Dave, I learned a lot about accurate measurements! (also a great way to show off all the gear you have, haha)
5:35 Dave, you're talking about the voltage drop on the wires from the supply to the converter, but what about the drop on the multimeter wires and the wires hanging down of the yellow bread board, which are longer?
Idjles Erle the multimedia has an impedance in the range of mega ohms. so very, very little current wil go through these test leads of the multimeter. So there will be also ben very little power loss or voltage drop in the test leads of the dmm.
I'd be interested in a "traps for young players" video. That detail about the voltage drop in the power lead was really good.
For point six six ....... SIX
5:50 If you use a thicker wire does not help to decrease the voltage drop?
Yes, and it's possible to get away without compensating because the losses are so low.
EEVblog Thanks, i love your channel, it's very interesting.
Another plus for using Libreoffice as well, it pretty much does the same thing as Excel, just is not linked to a subscription model for pay as you use calculation.
Is what Excel is these day? I haven't used Excel/office for a long time now.
It is not. Excel/Office can still be purchased like a one-time payment, but also available in monthly subscriptions (but is not limited to x amount of calculations..)
Excel looks much nicer, and is actually more functional. But Libreoffice is great for being totally free.
Why not use Google drive? Just as good as libre office and nothing to install :).
Excel is a very powerful tool, but for most things Google Spreadsheets is more than enough.
Great demonstration on how to do a complete characterization.
Thank you very much!
What a good idea to use open office calc to do that Graph, didn't think of that use but hey, it's very usefully. Thanks Dave.
Your channel is awesome!
Greetings from Austria!
Great video. A lot of very good info!
Nicely done! I have been automating my switchmode converter tests using an SMU and a DMM... but now using two Keithley SMUs and LUA scripting.
I was wondering about how you did your graph with the X axis log, and then you answered it, good on ya!
I am a young player and I am confused by one of the traps that you mentioned. Can you explain why the voltage drop from the wires coming from the power supply matter? Why is the voltage drop from the wires connecting to the multimeter not considered? Is it because the multimeter has a really big impedance that we can treat it as an open circuit?
flowakfg yes, exactly. voltage drop is proportional to the current flowing through the wire. high impedance means almost no current flows through the test leads so there is effectively no voltage drop for these measurements.
If you measure the voltage at the power supply terminals and not the voltage at the pins on the regulator then power lost in those wires would be not accounted for, you wouldn't be purely measuring the loss caused by the regulator which is what you are interested in when the regulator will be used in different applications where the loss caused by outside factors will be different to your testing setup.
flowakfg its the basics of a resistor. every wire is a resistor, and when current flows, a voltage drop occurs. The amount of current is what determines the voltage drop. A voltmeter draws close to no current and thus the drop is so low that it can be ignored (less than a millivolt).
You're spot on. Multimeter input impedance will be in mega ohms, if not giga ohms, depending on it's design. It's small enough that the current flow is negligible. That said, with a high enough resolution meter, you could theoretically measure a voltage drop in that scenario. Don't know what you could do about it though. Six wire measurement? :P
Dave, could you do a short video introducing the SMU 2400 capabilities as a trade off with a multimeter and power supply. Looks so similar features that a more details presentation will help for sure to understand the differences.
Having done the single multimeter / oscilloscope DC converter measurement circus with a junk bin supply and hackjob load there is something else to remember: be careful about what your circuit ground actually is (understand your power supply) and don't get sloppy or you'll connect a fat switching node directly to ground and mosfets will explode in your face
Awesome tutorial. Make some more tutorials like this. Like measurement related
Dave should do the Rigol DP832 remote sense wire mod !
Great video Dave, any chance you could look at DC to mains inverters in the new year, thanks in advance. - Martin.
Dave, thanks for this great and educative video. This ones are the best!
I have to say I am shocked and amazed at the 0.3+ V drop on the output side, at only 80 mA. That's like 4 Ω in the connections, which can't possibly be the cables. Are breadboard contacts that bad?
Very good informative video. Very well done. I really enjoy these and find them valuable.
since your wire resistances will most probably stay constant, you could just measure their influence once (at a high-enough current) and then correct for the resistive loss from the measured currents.
But .34 V drop for just 250 mA of current seems high for your wires - do these contain some copper at all? I am asking, because I recently got crocodile cables from China which are magnetic - not the crocodile clamps, the cables!
So you're saying let the voltage drop stay and just compensate for it in calculations? but then you have a varying voltage at the DC/DC converter input pin itself which will cause it to change its switching characteristics to maintain the output and might change the efficiency vs a constant input voltage. Thus your graph would still be different to another graph recorded from a test with different input wires.
Solderless breadboards don't make the best of connections, especially when they are well used. Most of the voltage drop is probably due to the contact resistance of the breadboard.
The wires should only drop around 10mV if they are pure copper.
rocketman221projects And that's why you measure the voltage using separate sense terminals.
Exactly, this is complete bullshit. Nice demonstration, but utter bullshit. That 10 cm of wire should never, ever have four ohms of resistance, as for other connection issues well that's on the tester - and I absolutely will laugh out of existence any "tester" that somehow manages to introduce four ohm's worth of resistance in his test cabling, no matter where he chooses to measure the voltage. Ridiculous!
Attila Asztalos for one its an educational video, if he used thicker cables he wouldn't be able to demonstrate so clearly why you measure the voltage right at the pins of the regulator, secondly since he's measuring the voltage at the pin of the regulator making sure the voltage is constant it doesn't matter what the resistance of the cabling is anyway?
Dave, try using QTI plot (google for unofficial builds), way better for plotting (you can zoom into graphs 8-)... Next step (if you desire) is Paraview.
Another excellent video bro. Love this topic.
As you knew you wanted a logarithmic graph, you could have measured data in values of current equally spaced in the logarithmic domain. For example, with 8 values per decade, the base numbers would be [1-1.334-1.778-2.371-3.162-4.217-5.623-7.499] (multiply by 10^(1/8) at each step).
Nice job! Would be nice to see the automated version of this :) Like an arduino based logger and auto Vin adjustments.
No need for arduino, he has proper lab equipment with programming interface. Just connect them to PC, write software and that's it.
Thank you so much! I'm designing a buck converter for a class and this really helped!
Can you make a video of how to use 4 wire measurement method to measure very very small resistance with 4 multimeters? I was trying to measure the resistance of a current sensing resistor, which normally has a resistance between 2mOhm to 5mOhm....
The easiest way to do that is with a constant-current supply & a single multimeter. The procedure is to put a known current through your resistor, then probe the resistor terminals with your multimeter. If you don't have a supply with a constant-current mode, it's a bit more complicated, & requires 2 meters - one for current, & another for voltage across the resistor terminals.
I think the reason you cannot use line charts with horizontal log scales is because excel doesn't treat the horizontal axis as numerical but as categorical.
Hi Dave,
Great video! It did, however, get me thinking. If both your input wires and output wires were of the same type and same length, would you still need to perform the 4-wire measurement to compute the efficiency?
Specifically, my thinking was that if efficiency is computed as the ratio of the output vs. the input and if your wire lengths were the same for the input and output, the voltage drops should be proportional and therefore the computed efficiency should be similar to what you computed using the 4-wire measurement. Am I incorrect in my thinking?
So the shadetree seller specs are a 5-9V converter, 300mA rated,
"They're not gonna bloody read it" lol, made my day
You were getting 12v out not 9, did you have 2 converters or something?
Pilot G2 right? Nice choice of pen.
Yep.
Pilot Metropolitan is my choice.
We can tell Dave's quite animated when he speaks even though we can't see him because of the way the metal "flashes" as Dave moves around, blocking and unblocking the light in the first minute and 40 seconds or so of the video.
LOL nicely spotted. Yes, I flap my arms around behind the camera.
Can to characterise a PSU like this with a load that is a high power resistor instead of a CC load from an electronic load?
Thumbs up to all Libre Office users! :)
Tommy Oz Gnumeric users with xichart or such will probably save time to have the thumb free. That said, how'd he get the dvm to talk to the digital load?
Does that voltage drop also include the length of the multimeter leads
You don't need to account for the length of the voltmeter leads; for voltage measurement there is negligible current going through the voltmeter leads due to Megaohm+ resistance of the voltmeter being several magnitudes larger than the source/load resistances - just be sure to measure voltage on input after the ammeter in your circuit to account for potential voltage drop through the ammeter on the input (and likewise before ammeter on the output).
Very interesting Dave, Thanks.
Don't tell Julian Ilett, he literally has 1000's of converters. Like Dave is nutso for DSOs, Julian is nutso for converters. lol
He has recently tried to try live streaming, imagine if he did all the calculations for all his converters live!
Yea, it would be at least twice as boring as his recorded stuff.
He does go on too long at times, so I have to skip parts, but I also have to skip Dave's DSO, and most other teardowns of newish equipment, because you don't really learn much. ElectroBOOM forever.
SidneyCritic ComedyHound todays item is another dc to dc coverter
In that situation proper lab equipment and automated test software would do the trick.
I love your videos, I was just wondering since you were saying that the voltage was being dropped across the wire, so you were measuring right at the input, wouldn't there also be a drop across the wires going to the meter?
Builder6275 no because the meter has a high input impedance (10 MOhm) and hence there is only a tiny current. But Dave explained it in the video...
There is a very small (basically insignicant) voltage drop across the probe leads because the current flowing into the multimeter is very small
No, I explained that. The meter is 10M input, so no current flow = no drop
Ohh, Ok I understand, Thanks!
can I use an electronic load like that bk precision one as a PSU regulator? Say I have some beefy transformer + bridge rectifier + smoothing caps?
Looks like an interesting project for two INA219 boards and an Arduino. Stay tuned..
I'm guessing that the 'sense' lines are used to measure the resistance of the input lines and factor this in. [ good old Ohms law ]
Philip Rowney
more accurately, the sense lines are attached "after" the wire resistance so that it can be ignored.
[ remembers his City & Guilds 2391: Test and Inspection ]
It is just to have a 'known resistance' of the input cabling to isolate any 'current effects' to the circuit under test. [ language may be out-of-date/English ]
Dave, are you sure you have enough multimeters?
Hello Dave, Excellent this video, could you share the Excel Worksheets that I use in your Presentation?
Thank you for sharing your knowledge with the community.-
Greetings from Argentina.-
Antonio.-
Thanks a lot for such a great informative video. can you tell me, how to measure efficiency in DCM mode? I should consider RMS value of current or not? Because in DCM, as current reaches to zero RMS value becoming very low and thus input power is also very low.
wohoo...under thousand views...
The inductor is the first thing to observe when it comes to efficiency....
For low output voltages a proper Schottky diode can help as well.
Active rectification!
Instead of using 4 meters, you could just use the sense connections that should be on any decent psu and load, certainly are on mine and they're aren't top brand name stuff.
Pieh0 he does and explains it around 12:00 in.
Most people will not have a PSU with sense terminals, or maybe even an electronic load with that. I mentioned that.
Why not just use bigger guage wires?
agilent, digilent, siglent. what's next, soylent?
ms3bani only if it's green
So what we doin for eevblog #1000?
How is that interesting? The real question is what's going to happen for #1024.
The list on 31:47 is not in decades ;)
9 ... 2 are missing :)
100,
90, 80, 70, 60, 50, 40, 30, 10,
9, 8, 7, 6, 5, 4, 3, 2, 1,
.9, .8 ...
Dummy question: If I power a device with a 12 volt car battery, let's say a car stereo, but the cables are extremely thin and voltage drops significantly, does that mean the battery will drain faster?
It slightly depends. To the battery the load is lighter, but if the voltage is dropping to the load (car radio) the audio might be quieter (assuming it still worked at lower power) and you might crank the radio up. If you put the radio with good thick wires to a battery, set the volume, then swap out the thick wires with thin you might get a longer battery life. The best way to think of that radio scenario is you replaced the radio with a fixed load (resistor), then added another series resistor (thin wires) and reduced the current drain.
A car radio might be a bad example for this because a lot of new electronics use switch mode for many things (including the output amplifier). Switch mode power supplies (and class "D" amplifiers) actually start drawing more current as the input voltage decreases, so adding a series resistance (thin wires) might actually drain the battery faster with increased current as the radio tries to compensate for the voltage drop.
Would like to have seen a clearer explanation of 4-wire sensing, without assuming even the basics, since its for "young players". eg:-
* All wires will have some resistance, and there will be a voltage drop when they carry significant current.
* Therefore the voltmeter wires must not be shared with any part of high current path.
24:25 +9 V -9V ?? :)have a nice day:)
Good job!
Add four (4) EEVBlog™© brand multimeters, stirring gently!
Dave, I really like your videos, but can you make shorter versions if possible? Like a 10-min maximum summary version and 30-40 minute detailed version for people interested in more in-depth look. I think you would get a lot more views like this, I myself, for instance, find myself skipping a lot every week since I have over 40 subscriptions.
I am a bit confused how lousy efficiency doesn't equal heat. I guess it is because the output has less amps so the relative heat is higher, but smaller.
At the low current end, correct.
use nickle wire?
Well, you *might* get away with two multimeters, assuming that input voltage and output voltage is sufficiently stabilized.
Otherwise, this might be the situation where even you can't get me away from cheap multimeters, Dave!
that's why a decent power supply have the sense terminals...
Excellent.
Remember kids, always tweak your knob just right!
And you gotta have your tongue at the right angle...
couldn't you use one meter? just measure the voltage in, voltage out, then the current in and the current out in series instead of all at once. besides you will get inaccuracies if you measure both at the same time as well. you will get a voltage drop over the few ohms of sense wire for current measurement, or you'd get a few mA leak across the voltage measurement resistor, depending on which way your meters are put..
"I'm sorry Mr Edison, your DC electrickery is very nice but our giant resistor rooms are glowing and we have used all the cooling fluid for beer.
Thankfully Mr Tesla has got something he calls "AC/DC" which means we can transform from high Voltairs to low Voltairs."
"And stay away from the Zoo. Electrocuting elephants has just made you the enemy of history."
thank you
2:57 calculator-later hehe
4:59 666 Illuminati confirmed!
SATAAAAN
I have never used a line chart. I have always considered it bad practice to draw a line between points. It is far better to apply a line of best fit and describe the equation compared to your hypothesis. Perhaps it is just because I publish scientific papers.
what i'd like to see would be a how-to on making a dc-dc/switching power supply as "quiet" as a linear one :/ I hate the little bastards because apart from conversion efficiency they have no other benefit vs a linear one
George Tsiros they can be completely silent (inaudible to humans), but that would drastically lower their efficiency. Still more efficient than linear ones tho.
drkastenbrot talking about EMI here, not sound.
In my 5.8Ghz video Transmission systems, clean power was mandaory. My solution was a buck converter set to 4.1V which feeds a linear regulator set to 3 3V. Perfectly clean output and good efficiency.
9V Brick? it output 12V...
multimeter fetish??!! i need some of those....
you can buy them ;)
I don't think that a converter is that efficient if it's running at 100 °C.
You never fucking ever put a spline trough a real word data points unless you know you have the resolution to back it up.All so, if you do have the resolution, then there is no need for a spline fit.
Get over it, I didn't do it deliberately.
Crashy McSplodey Oh, just link the Maxim paper on it. They make a point of running 4 days of tests to note every inflection in operation, to find that point of quiescence near 0V and 13% improvement at 5.2v in, etc.
Such a long video about like couple measurements. Come on, make a long video about something more complex/interesting. This one should be BLAB
Why don't you do a 5 minute of this video then?
Yeah, right. I will now say "beacuse I don't have youtube channel and I am not into making videos" and you will say "THATS IT. SHUT UP". But this will not make this particular video any more interesting. I mean, like measuring voltage and current with 4 multimeters... How can you babble for half an hour about just that?!
Never Mind its perfect, i enjoy to hear every details how and why he choose a particular setting or option!! You become only a expert of because of the experience you have learned on the field over the years and if your work is your passion, it's great to talk about it!! I enjoy the details. Then again I am not no any way of an expert in electronics, is good to see what is behind of everything to make a good measurement.
That's my thing. And if you cared to actually listen there were lots of tips in there explained in detail.
If you want 5 minute tutorials you subscribed to the wrong channel.
Come on, Dave. I love you and your channel. And I bloody wathc EVERYTHING from here like a maniac. Bit this one is just too long. It looks like you're getting out of topics(which is not true, right?)
Four point six six ... ... ... ... volts.
I don't want to sound offensive, or like an asshole; But could you please start running your mic through an EQ, and lowering the "tone"? Watching these videos hurts my ears, especially if I was watching another video before these and it switches to one of yours while I am doing something, or if I am wearing headphones (which I have to promptly remove and switch to external speakers for your vids). Your voice tone is so high, and you speak so loudly, that it just hurts (literally), even when turning the volume down. I love your vids Dave, and as I stated, I am not trying to be offensive, do please do not take it that way.
Sorry but no, I'm not going to modify my natural voice. I've done a thousand videos like this. My voice is my voice, take it or leave it.
EEVblog
No No, Im not saying "modify" your voice at all, but you can use an EQ to cut the frequency down on the high end, and kind of "normalize" it in the videos, without changing your voice (I know, technically it DOES change it, but it isn't enough that you wouldn't still sound the same to the human ear). I don't think the problem is your "natural voice", as much as the way the mic pics up the higher frequencies with the mic being so close (lav mic), and the way they are reproduced through "tweeters", resulting in the sound being amplified like someone is screaming in your ear, it is actually "un-natural". If I were sitting in a room with you, this problem wouldn't exist, unless you were talking to me at this loudness, with my ear exactly where your mic is. As I said, I didn't mean it in a disrespectful way.
Again, no, not going to happen. I'm not going to piss around with post-edit audio just because someone has a problem with it on their audio setup. I can fiddle audio until the cows come home and people would *still* bitch about my voice. It's called a High Rising Terminal BTW.
EEVblog
Its the placement of your microphone, not my audio setup; And it is my ears, they are rather sensitive to certain frequencies, which is why I cannot use television speakers or laptop speakers.
InXLsisDeo
I'm sorry your audio equipment is shit then. As for EQ, as I stated, and you obviously didn't read, the biggest problem is when RUclips is playing whatever the fuck it plays, and I am working or something like that, then I have to switch EQ just for Daves vids. I also said I wasn't trying to be an ass, but you ignored that part as well and apparently thought you would do that for me. I had a legitimate suggestion, due to his specific circumstances with his voice, his Mic placement, etc., and stated WHY I had this suggestion. Not once was I rude, as asshole, or disrespectful, so there is no need for it from small people like yourself either.
dave this was too long winded. 15-150mA. fin.
Aww Dave, splines are a big no-no in any scientific graph ;)
suh dude
Just goes to show how many of the breadboards and the corresponding jumper wires available today are just pure garbage.