Would be nice if everyone used the exact same PSU

I sure hope the PCB guy won't be angry with me if I don't bother with proper measurements. Plot twist - I'm also the PCB guy

Exactly!

Sounds like a BGA PCB design challenge.

That's what I was also saying

Affecting

Cool story!

Almost sounds like my time working on AB 1771 IFE when I was servicing them.

If the current is different with every multimeter why bother? Just use the analog input it's connected to

@@bene5431 The measurement/calibration/verification procedures in place at that time required 4-20mA measurements. Once a specific meter and current range was defined the problem was addressed.

In case you didn't think of it, a similar idea applies to measuring voltage, although you are placing a resistor in parallel with the measurement instead of parallel. However, the impact is usually negligible in voltage measuring.

How much "higher" we're talking about? Anything over the maximum value of miliampere range?

@@offspringfan89 Yes, I typically use a clamp meter when I need to measure more current than the milliamp range can measure unless I need more accuracy or resolution than the clamp meter has. There are some clamp meters like the Uni-T UT210E that can measure milliamps, but they are very sensitive to magnetic fields when measuring DC.

I haven't used a multimeter to measure amperage in ages. I always use a clamp meter and a special clamp meter designed for 4-20mA signals.

@@offspringfan89 I only use a clamp meter for currents I expect to be > 100mA. Unless you have a clamp meter specifically designed for low current, most are designed for measuring high currents and tend to be really jumpy.

​@@rocketman221projectsI see, thanks.

It's common.

"short videos" are fine. Shorts? Yeah nah

beats a 1/2 hour multimeter infomercial

Your suggestion works when you are dealing with low resistance circuits, but goes wrong when making measurements on high resistance circuits because the current drawn by the voltmeter is then an appreciable part of the total current measured. Once the circuit impedance starts to be significant compared to the input impedance of the voltmeter used, then you are better off measuring the voltage across the ammeter and the circuit together since the burden voltage will be insignificant and the current will be accurate.

@@RexxSchneider but part of the roll of beening an engineer is ķnowing the affects of the test equipment on the circuit and maybe having to calculate the correct value. When I was training it was part of what we had to learn but that was pre digital age. Where ohms/V had a big affect on readings and you had to use the mirror on the meter.

@@TheEmbeddedHobbyist I doubt you were trained in the "pre-digital" age, since I bought my first logic ic around 1965. Nevertheless I'm well aware of the limitations of analogue meters. It doesn't change the basic principle that in low impedance circuits you can ignore the current drawn by the voltmeter and in high impedance circuits you can ignore the voltage dropped by the ammeter.

I meant the digital voltmeter, we used Avo 8's and worked on valves ;+) in day release to the local college from school. Well I've always had to calculate the affects of test equipment placement on the circuit, and taken this in to account when writing the test specifications. I need to be able to set the limits so any component or equipment being out of tolerance is detected in the test results

There are some interesting ways of dealing with noise. For an AC signal, a lock-in amplifier may be used to modulate the input signal and then filter out the resulting side bands generated. There’s a good video presentation on the subject by Zurich Instruments available on KZfaq titled, _Boost your Signal-to-Noise Ratio with Lock-in Detection_ . The system won’t allow me to paste the URL.

It looks like my previous reply was auto deleted. I am not an expert in the field of metrology. I wrote that lowering input resistance is a way of combatting thermal (Johnson) noise generated in the input network. For an ideal resistor, the noise power generated in the resistor is P=kTB, k=Boltzman’s constant, T=absolute temperature, B=bandwidth. Thus, the RMS noise voltage generated is v=sqrt(4kTBR). People who work in RF often use the approximation for the noise voltage added by a 50 Ohm input at room temperature as -174 dBm/Hz of input bandwidth.

Probably actually something like Nichrome.

Good luck with getting any sensible MilliAmp measurement at 3.3V with that setup.

@@johncoops6897 At 1.0V Dave demonstrates in the video getting a more accurate reading from the highest range (mid value 10.05mA, than the 9.793mA original one. Latter has extra digits but they are useless beyond the first unless you have time/motivation to do the correction calculation. Starting at the highest range also protects the meter (lower fuse!) from any unexpected high currents.

@kevinwalters5160 - that may, or may not, be the case in THIS video with THAT meter. However what you are not grasping is that a typical handheld multimeter on 10A range is using a 0.01 Ohm shunt. To measure 1mA that requires the meter to accurately measure 10 MicroVolts (uV) and it takes some serious ADC power to get a sensible and accurate measurement at those tiny voltages. The WHOLE idea of having ranges in a multimeter is so you can adjust your settings get your measurements into the sweet spot between accuracy, precision, resolution and burden voltage. Yes, it's always wise to _START_ at the highest Amp Range, but that's just the starting point.

@@johncoops6897 It's quite easy if you have a precise enough meter. The hard part is having a precise enough meter that's still handheld. But I suppose you *can* hook a power tool battery up to an inverter and duct tape it all to a 5.5-digit bench meter, and then hold the whole setup in your hand with a bit of exercise, so for certain silly versions of "handheld" it's possible.

and now it does? 0.3ma error?

Remember back in the day, multimeters were specified in Ohms per Volt? Now that even cheap meters have multi megohm inputs, you don’t see that quaint specification. My first Radio Shack meter was 50K Ohms/Volt.

No but spelling in the schematic did.

​@@wtmayhew- 50k/V = Luxury!! I am currently trying to find a cheap analogue meter for temporary use while I repair/modify my AVO Meter. The best that I can find at a reasonable price is only 20K per Volt. 😢

@@johncoops6897 Yes, the venerable Simpson 260 was rated 20K Ohm/volt. The Radio Shack meter I had as a teenager was the Radio Shack Micronta 22-240a Range Doubler. I just found one on eBay under $20. The seller says it looks nice, but that doesn’t mean it works. 🙂

When I teach Digital Circuits, I’m surprised how many students have escaped learning much about accurate measurement technique. My first lab has students measure and compare the switching current and maximum operating frequency characteristics of LS-TTL and CMOS devices. Students are frequently surprised that input capacitance of oscilloscopes is not negligible and needs to be compensated out by way of a capacitive and resistive voltage divider in a 10X probe to make acceptable high frequency small signal measurements. Several times I’ve had students comment that now they understand why their measurements were off from what they computed in the analog devices course which precedes my course. Apparently the students were just using coaxial cable with alligator clips to measure their filter and op-amp circuits.

@@wtmayhew there are some people which they fixing airplanes with ~5 $ "multimeter".

@@norbert.kiszka Some of them are at Boeing, maybe! At least if those workers took my course, they’d probably be a little careful with their technique.

​@@wtmayhew BTW. Airbus (at least A320 series) have small "issue" with "redundancy". Three computers and all of them can make bad calculations. But if nr. 2 (I dont remember exact name) will fail or lose power, then pilots cant control aircraft. There was one incident because of that. Anyway, today Boeing is worse than Airbus.

@@norbert.kiszka Thank you for the reply. That’s good question - how much redundancy is enough? It is not a good day when there aren’t enough matching computations for agreement. The scary thing is now control by wire is reaching consumer automobiles. Throttle by wire has been around for almost 20 years, but if the throttle goes crazy, the driver can still control the car. Now Tesla is doing completely electric steering on the Cybertruck. That seems not so well advised. I believe it is Brembo who have a fully electric brake system. I like the idea of having hydraulics in parallel with electric assist. I had a Nissan with electric brake boost of hydraulic brakes and it had problems with the backup ultra capacitor going bad - my car had issues. In later years Nissan when back to vacuum assist boost with an electric pump to produce vacuum.

On the first equation you subtract 100Ohms to make it match the mV/mA, example (1V/9.8mA) - 100R = 2R 2mV/mA = 2R But on the second equation you don’t subtract the 100R or it doesn’t make sense to me example, (1V/4.996mA) - 100R = 101R 0.2mV/uA = 200R Can you help?

@@shawnphillips8707 I think you have to remember that there are tolerances on the power supply voltage, the multimeter display and on the 100R resistor. Nevertheless, let's see what we get in the first case if we take the readings at face value: We have a displayed supply of 1.008V and a current of either 9.795mA (from the PSU display) or 9.793mA (from the multimeter display). So the total circuit resistance is somewhere between 1.008/0.009795 = 102.91R and 1.008/0.009793 = 102.93R. That means that the shunt resistance (which has to be the total resistance minus the load resistance) works out to be 102.92 - 100 = 2.92R, give or take 1 digit in the last place. The table at 4:10 indicates a shunt resistance of 1R but I'm afraid I can't tell you why the total circuit resistance appears to be close to 103R, not the 101R that would be expected from that table. Either there is a 2% cumulative error somewhere in the voltage, current or resistor measurements (or maybe even contact and lead resistance is significant), or the table at 4:10 is wrong. Sadly there was no explanation of where it came from. For the second case on the microamp range, we get 1.008V driving a current of either 4.969mA (PSU display) or 4.966mA (MM display), which corresponds to a total circuit resistance somewhere between 1.008/0.004969 = 202.85R and 1.008/0.004966 = 202.98R. The load resistance is still 100R, so the shunt resistance has to be 102.9R, give or take 1 digit in the last place. That at least, has a reasonable correspondence with the table shown at 4:10 although closer examination seems to indicate that it may be for a different multimeter with a full-scale reading of 5, 50, 500 rather than the BM786 which clearly has FS readings of 6, 60, 600. Checking back to the table shown at 2:52 it indicates that the BM786 has shunt resistances of 200R (0.2mV/μA on the μA ranges), 2R (2mV/mA on the mA ranges), and 0.03R (30mV/A on the ampere ranges). Why these don't correspond to the measurements in the video, I can't tell you. That's really a question only Dave can answer.

Writing this down, thanks.

@@offspringfan89 you're welcome 🤗 . I am editing in these hours the video where I show the different readings and the effect on the other voltages in the circuit. There are only two kinds of electronic hobbyists: those who have been bitten by the burden voltage, and those who do not know they have been bitten by the burden voltage. 😁

I took a machines course called Energy Conversion as an EE undergrad. Much stuff in the lab was Victorian Era equipment. There were open frame 10 HP motors the size of washing machines. The Weston current meter on our bench was in a wood box with an 1897 patent date nameplate. We spent a lot of time compensating for the fact the measuring tools we had were quite non-ideal. I suppose it was a good leaning experience to have to endure the antique equipment.

That's how a lot of amp clamps work.

Yes, common technique. But if your current changes dynamically you could be in trouble. If your PSU has a sense terminal then you can use that.

I thought the same thing. Add it back if you needed accuracy. Last week I was doing parasitic drain testing on a BMW, it was 12mA, but as long as it was less then 50mA its ok, so accuracy in that instance isn't necessary. I hope Dave responds to your question

They aren't very accurate for low currents. And can also be susceptible to interference. Burden voltage is not an issue as long as you know its there and can compensate for it.

...And Dave replies, "But, I AM a fun guy!"

The bartender says "We don't sell fungi here ...AND you sound like you are under-age" 😂

@@Markus0021 However, remember that when someone says "you must be fungi at parties" that's sarcasm. What they actually mean is that you're sporing.

Yes, come-a-guts-er. As in, to spill your guts, or in other words, have an accident.

I did not have major design input into the 786.

It probably depends on the particular flavor of electrical engineering. I took a US machines course called Energy Conversion in the 1970s and the term, “burden,” was used. As an aside, Fluke has on their web site an article titled _Can you live with the burden_ ? They mention the burden voltage of the 87V is 1.8 mV/mA on the 400 mA current range. The gist of the article is about the same as what Dave presented in the video.

I forgot to mention it can display the burden voltage as well.

I bought it in 2019 and still haven't unboxed it. I don't even know if it works

@@rasimbotHa, ha, ha.. Well I'll tell you one thing in this world I'm certain of, and it isn't the sun rising in the morning... It is, Dave's 121GW WILL work.

Yes it does. You can measure it if you have two multi-meters. It's very low current and the current is limited as to not damage whatever the unit under test is.

Less information due to the 1 minute time limit for videos there?

Are you talking about my Short , or the video I posted from a customer? With the customer video, he did the same thing using a 470ohm resistor.

@@EEVblogI'm talking about the customer video. By watching it in my phone I wasnt able to understand what was in the breadboard, to me i seemed that the meter was directly connected to the wires coming from the top of the screen. I completly missed that there was a resistor in there. Now it makes more sense!

You need to buy an EEVblog "Mouldy Meader" to accurately measure those pesky Burton voltages.

You right mate?

No actually, I hated the 1 minute limit. HAd to leave stuff out and speed it up 10%.

@@EEVblog Shame. Celebrate the accomplishment and move on! It is a very good video!

When you make datasheets for electrical components where you want to determine the current draw, you want it to be correct for example, and not artificially lowered because there was a sneaky resistance in series.

Most people don't have a high end 6.5 digit bench meter

My $22 clamp meter works great.

I think not. The 1v is across the combined resistance, 100 ohms + shunt resistance. Therefore 1v / 0.0098A = 102 ohms. 102 - 100 = 2 ohms. Someone correct me if I'm wrong.

@@rogerwilco5187 You might be right.. I haven't messed with ohm's law since 1999.. lol

@@calholli Actually I'm pretty much in the same boat, but my old brain was niggling at me that something wasn't quite right with 5k and a drop of only 0.2mA.

Interesting.