Multimeter techniques for measuring DC Amps, Shunt vs. Feedback Ammeters

[nightfire]

Currently, I am in the process of building my own home lab- that will consist of a working table and some stuff in the corner of my living room and also double as general purpose workbench...
As I am a hobby photographer,  I also have to deal with lots of batteries and charging stuff, and want to do some measurements of batteries and later on also quality of LED Lights (PWM Signals and quality of colour index). In my 9-5 job I work as a system administrator and have a certification to work on standard electric equipment including doing safety checks, so I am expanding my knowledge to the world of electronics and the finer voltages...

Anyway, I am deciding what to put on my long wishlist and then building my small lab around that, so I want to have high-quality stuff, that also can be 2nd-hand.

As I want to check batteries whilst getting charged and discharged including logging, it is clear that basically i will need a multimeter that has a low burden voltage for measuring currents in the range of 100mA-2A.
(Other options would include a shunt and some amplifier like the eevblog µcurrent, that I also will likely explore, but want to focus on a good quality bench DMM first...)

Whilst searching, i stumbled upon the Fluke 8808A, that claimed as a "Feedback Ammeter" it has a drastically lower Burden Voltage than most other DMM in that price range.
So I did some homework and looked for some documentation regarding the differences.
So the technique behind this is clear, in a feedback Ammeter there is an opamp that does not need to have a big voltage drop over a shunt like in most other Multimeters, which is a good thing when the Voltage in that circuit is around 1.4 Volt to charge some NiMh battery...

Question here for the more experienced multimeter folks: Which possible disadvantages (other than price) would a feedback multimeter possibly have?
And, also important: Which Multimeters that are affordable for hobby use are build that way?

[TimFox]

Feedback ammeters are commonplace in "electrometer" applications for measuring very small currents.  One of my favorite units is an old Keithley "logarithmic picoammeter" which allows reading a wide range of tiny current on a single meter scale, suitable for measuring grid current in small-signal amplifier tubes.
On the other hand, if you want to measure 1 ampere with a feedback circuit, the output of the feedback amplifier must supply 1 A through the feedback resistor to balance the current being measured.
Another way to measure DC current without a high burden voltage is to use a flux gate magnetic device, such as in the -hp- 428 series of "clip-on milliammeters", with full-scale ranges from 1 mA to 10 A.  Here, the DC feedback current goes through many turns of wire on the magnetic core to balance the high current through the single turn of a wire through the center of the probe.

[Kleinstein]

The feedback current measurement mode is not very practical for higher currents. It has mainly advantages when the voltage at the resistor can be higher than the usual 100-200 mV. This is only practical for relatively low currents. With high currents self heating is often more limiting than measuring small voltages.  Especially if used for higher currents it would need extra power - so rare with battery powered instruments and often only for low currents, like < 10 mA.

With a current source that has a complex impedance there can be complications, possibly even oscillation if not carefully designed. Besides the normal drop there may be an offset voltage - this can be problem in cases with very low voltage in the circuit. With a superimposed AC part, especially higher frequency or fast transients there can be a significant drop due a dynamic impedance, that usually is inductive.

With higher currents a large part of the drop is due to the fuse, not necessary the shunt.

[MosherIV]

Welcome to the forum.

If you want to measure current without burden voltage, try looking at clamp ameters which can measure dc currents.

Many here (forum) have ut210e.

Disadvantage of these is that the clamp must go round wire with current in 1 direction, ie  cannot have send and return next to each other.

[nightfire]

Thanks so far for all the input- I also have found an abstract from Keithley to that topic I will work my way through...

As I additionally want to do data logging of those values, I was looking for some multimeter with the possibility to get records out via RS232 oder LAN/LXI- the UT210E looks good, price is ok, but I am afraid of the accuracy measuring DC current in the typical range of around 500 mA I have estimated.

Batteries I want to torture involve some NiMh types as Eneloop AA Cells oder some stuff later on with the popular (NP) F970/ F550 Clones for lighting and video monitor purposes- so the charge/discharge currents will be typically between 100mA and 1A.
I also thought of some way to measure via DC probe to a oscilloscope and do the logging there, but at some point all this gets very expensive 

[robert.rozee]

consider something like the ACS712 hall-effect sensor, available as a module here:
https://www.ebay.com/itm/191991920122

you generally supply these with 5v, and then with no current flowing the output sits at 2.5v. hook this up to an arduino nano (or similar) and write a small program that sits in a loop grabbing samples from the module, offsets and scales to give a float, and then outputs this value to the serial port every few seconds.


cheers,
rob   :-)

[Brumby]

You beat me to it.

Welcome to the forum.

If you want to measure current without burden voltage, try looking at clamp ameters which can measure dc currents.

Many here (forum) have ut210e.

Disadvantage of these is that the clamp must go round wire with current in 1 direction, ie  cannot have send and return next to each other.

A very useful critter.

The other day I wanted to do a quick check of my alternator charging on my car and fretted about shunts, connections and so on - until I remembered I had one of these.  Then it was so easy - which incuded finding the right wire in the loom.

Mine doesn't get used a lot - but when it does, it makes life SOOOOOOO much easier!

[bdunham7]

There are different methods, as previous posters have mentioned, but there is always a tradeoff between burden voltage and precision.  I don't know your budget, but for the range you mention (500mA or so) you should be able to just use a DMM that uses a 10mR (.01 ohms) shunt, which will give you 5mV of burden voltage (not including test leads, internal wiring and a fuse, which could double that quite easily), which is pretty much as good as it gets.  This is typically the 10 amp range, which means a good 4.5 digit DMM will give you a reasonably accurate reading with a resolution of 1mA.  I don't know what options you have if your budget is lower.

[HKJ]

Batteries I want to torture involve some NiMh types as Eneloop AA Cells oder some stuff later on with the popular (NP) F970/ F550 Clones for lighting and video monitor purposes- so the charge/discharge currents will be typically between 100mA and 1A.
I also thought of some way to measure via DC probe to a oscilloscope and do the logging there, but at some point all this gets very expensive 

For testing batteries get an electronic load and use TestController ( https://www.eevblog.com/forum/testgear/program-that-can-log-from-many-multimeters/ ) to log data.
For measuring current use an external resistor and a sensitive meter, then you can get well below 0.1V drop (0.01V is possible) over the resistor.

[shakalnokturn]

If it's only the batteries you're evaluating and not the charger's method you may want to consider some of the "hobby" charger's used for remote controlled cars, planes, drones... It may be good enough to get the job done and although not a high-end solution work out easier and cheaper than populating a full lab for the task.

Our fantastic Dane (HKJ) has also done a page in that subject that summarises things much better than I would:
https://lygte-info.dk/info/HobbyChargers%20UK.html

[Wytnucls]

'For measuring current use an external resistor and a sensitive meter, then you can get well below 0.1V drop (0.01V is possible) over the resistor.'

Isn't that the description of a multimeter on the amps range?

[HKJ]

'For measuring current use an external resistor and a sensitive meter, then you can get well below 0.1V drop (0.01V is possible) over the resistor.'

Isn't that the description of a multimeter on the amps range?

Mostly, but you can optimize the resistor and connections for lower voltage drop.
A multimeter will usual have a uA and a mA range with low voltage drop (0.02 to 0.06V) and a uA and mA range with high voltage drop (0.2V to 0.6V). If you meter has a good mV range you can basically stay with the low voltage drop for any current range and if you accept lower resolution you can get even lower voltage drop (That may also be true when using the multimeter current ranges).

[Wytnucls]

'As I want to check batteries whilst getting charged and discharged including logging, it is clear that basically i will need a multimeter that has a low burden voltage for measuring currents in the range of 100mA-2A'.

The Fluke 8808A uses a feedback resistor on low ranges only (2000uA and 200uA).

https://assets.tequipment.net/assets/1/26/Documents/AppNotes_DigitalMultimeters_Functions_BenchMultimeter_MeasuringNanoamperes.pdf

[Kleinstein]

Similar to the clamp on meters there are also DC current transformers. There are 2 models relatively cheap available in Germany
https://www.pollin.de/p/aktiver-stromsensor-vacuumschmelze-t60404-n4646-x66282-15-a-5-v-180070
One would still need a DMM with logging or interface function.

For measuring the battery currents I don't see a real need for super low burden. With a meter with a rather high resolution one could always use a higher range to reduce the burden. If one gets some 500 mV drop at the full range (including the fuse), one would have 50 mV at 1/10 the full scale range, effectively giving up one digit. For 1 A this may mean using an external shunt, originally meant for some 10 A.

Especially lower cost meters may not have a shunt for every range and may have low burden on some ranges and a high burden on others.

[nightfire]

First: Thanks to all for the input given!
To further dive down in the situation I want to dive in:
I want to both measure the behaviour of batteries whilst charging/discharging, and to control the behaviour of some budget chargers, and to be able to verify if some high-quality charger really reports to me the truth or what else it is doing...
As I also want to share those findings via a blog, I will need a practicable way of getting the measurements out of the test equipment. I already have an Agilent U1272A, that I got as a good deal on ebay, but still have to get the data cable u1173B- will also need that for firmware upgrade.
And as an exerienced computer guy, I know (and have kinda a hatred) against proprietary drivers I would need to get something going- which means that I would like to invest in some interfaces that still will be supported in the future and where a windows update will not break things- with my Agilent DMM, you have 2 versions of data cable: U1173A for Windows up to 7  and U1173B for Windows 10 and 8.1- Due to that considerations I would rather have my future equipment equipped with some classic RS232 and ASCII output or LAN with scpi/LXI. Or maybe Modbus, thats also some quite common and understood protocol...

For precision, yes- I have taken some time to look at shunt values in different multimeters and also looked at the higher ranges. I am yet in the process to define some requirements and test scenarios, so that the required ranges can be concluded upon.
Looking at standard NiMh cells with 1.2 V nominal voltage and the equipment they will power, the currents I will have to take into account probably will be between 100mA and 500mA. This can also be very well handled by a multimeter on its 3A range (as in my u1272), where the internal 0.01 Ohms shunt will be used then- with a display accuracy of +/- 1mA I am very satisfied, as the measuring accuracy in that range is about 0.3%- so it should give me in this regard a tolerance of overall around 1% of the measured value- thats something I can very well live with...

With Lithium batteries, as used in cameras oder portable lighting fixtures, like the Sony NP-F550 and clones, world is a bit different and possibly a bit easier, as the overall voltage is bigger. Here I also have to account for rapid power draws like that of a camera taking pictures and drawing for some Sec. some big power spike and then return to idle.
Fuji cameras are reported to draw 500 mA from their small internal battery in those scenarios.

Logging: Yes, ultimately I will want to log that data to a PC for further use. Due to some considerations (and to have an interesting journey, which also may be of interest for other folks) the OS on that box will be FreeBSD. Also I want to use mostly open source software in that regard.
(I am also pragmatic- my current box where I type this on, and do my photo editing consists of a Windows machine, and I use Photoshop and Lightroom and some other plugins...)
I have currently a deep look into sigrok, which looks promising, also will do some evaluation of lxi-tools in the next few weeks.

Budget: As this is currently a hobby, it may cost money... I think I am happy when I can put something around 1000-1500 €€€ over a time of maybe the next 4 months into equipment, also including some good quality tools and 4mm wires etc., to simply be able to build some adapters needed to accomplish the tasks I will want to do.
As a multimeter often is the centre of the bench, I would want to spend some reasonable money on that, simply because once you buy cheap, you buy twice. And I intend to use that stuff I will buy now over the next 10 years with good care- and when I determine to cease that hobby, I would want to get a decent price on selling them...
A look at equipment points in my opinion so Siglent- the SDM3045X would quite well fit into my picture, as well as the SPD1168X  power supply. As I am not quite sure where this journey goes, I am a bit reluctant to jump initially on the electronic load Siglent provides due to its price tag, but this certainly is on my wishlist. Also a scope is on my list, as I will want to determine via a photodiode to measure the short burning duration of an electronic flash and PWM cycles of LED lights, because this is somewhat critical in some applications.
These have the advantage of being new and hopefully not needing repairs for some time- whereas some DMM like an Agilent 34401A or so, which goes on ebay for about the same price range, probably will need some cleaning and capacitor replacement in the near future...

[nightfire]

Similar to the clamp on meters there are also DC current transformers. There are 2 models relatively cheap available in Germany
https://www.pollin.de/p/aktiver-stromsensor-vacuumschmelze-t60404-n4646-x66282-15-a-5-v-180070
One would still need a DMM with logging or interface function.

Looks nice, maybe I can get that into some adapter of some sort- basically will save me the hassle to get a µcurrent gold here in germany ;-)
Such devices would have been my Plan B, to measure via a shunt and then feed the voltage (drop) through an amplifier- at that price, it almost sounds to good to be true and I think I will order some of them simply to tinker around a bit- a logging multimeter is also on my TODO-list, that can read those values and report them...

[bdunham7]

The current in and out of batteries may be pulsed or otherwise vary at quite a high rate.  For charging, some chargers may have pulsed outputs but generally you can measure charging pretty well with a regular DMM that integrates over a fairly long period.  I'm not so sure that applies to the types of loads that you might see, such as what you mentioned when a camera takes a photo.  You may want to consider something like the Keithley DMM6500 that has a fast digitizing system and, IIRC, has seven different current shunts for low burden and accuracy.  It is more money, but take a look at what it does, including (nearly) simultaneous voltage/current measurements and the possibility of adding a scan card.

[nightfire]

Thanks again for all the input. I am currently evaluating lots of possibilities and basically the level with which those measurements are to be taken.
Also binge-read the pages of HKJ which gave me also some valuable input for consideration.

I now have to evaluate by whether starting low, using some brain cells and rather cheap multimeters with some adaptors, or go directly for the highend-equipment and the amounts of €€€ to spend...

And by looking at some chargers, it is clear that I need some sort of whether fast sampling bench DMM oder a real oscilloscope here. Well, I want a scope anyway, but I had hoped that i could stretch those investments over some time...
Another solution could consist of pressing an oscilloscope as a DAQ into service- combined with some µcurrent Adapters and some homemade wiring magic could a scope do lots of that stuff- question here would be the precision- I will spend the next days defining some numbers and checking them with real hardware and reality in general...