Real world electronic load knowledge needed

[pdx_soft]

I am pretty new to electronics and am working a few of my projects that would really benefit from an electronic load tester. I've been googling and I am unable to find a definitive answer to this question. I am looking for a tester that will work in the under 10 ma range as well as the lowish amp range.  My voltage requirements would be easily satisfied by 150V (honestly 60V). I purchased a Rigol DL3021 (updated to latest firmware) and was disappointed when I hooked it to my power supply at 5 volts and set the load to 500 ohms and no current flowed. I had to lower the unit to 450 ohms to get a reading at all and to 400 ohms to get the current to be somewhat stable. After reading here found that this was normal behavior for the unit. I will be returning. Will the Maynuo M9711/M9811 allow me to do a test at lets say 3ma? How low will the current go in real terms (either based on CC or CR) and still work with some accuracy? One of my projects sends a 2-4 hertz square wave with about a 4 millisecond pulse width (again at about 3-8ma) that I would like to test against a variable load, will this unit work?  Is there another unit I should look at in the $500-$650 usd range? Thanks for your valued input (and a buddy, that doesn't exist in your price range is a valid answer)...

[HKJ]

My Itech load 120V/30A (That is about the same as the Maynuo) will work at around 3mA, but the tolerance is about +/-2 mA.
For low current I usually use a SMU, but it is rather expensive equipment.

[tautech]

Is there another unit I should look at in the $500-$650 usd range?

Yep.
https://siglentna.com/dc-electronic-load/sdl1000x/

[tv84]

Would an upgrade to the DL3021A model improve the minimum?

[Elasia]

I just tried using my Siglent SDL1k load units.. it gets a bit unstable sub 10ma load and its settling time would take to long i'd think

Best i could do is put it into constant power mode for 50mW and when supplied with 5V of clean power its still got about a 5-7% bounce to it of around 505-535ohms update rate of around 10Hz on the display


You would be better off using a SMU for the ultra low stuff

[Electro Fan]

Yesterday I tested an old portable phone battery.  The battery is spec'd for 3.6V, with a DMM the battery showed 4.0V, but when I tried to use the battery in the phone, no go. 

Out of curiosity I set a Maynuo 9712 to .00001 Amps (10 uAmps) and found that the battery would quickly drop from 4V to just under 2V but remain close to 2V.  When I set the Maynuo to .001 Amps (1 milliAmp) the battery sank from 4V to nearly 0V within a second or two.

I haven't used the Maynuo very often but whenever I do it seems to be accurate and precise.  I remember several years ago when I consulted one of the veteran forum members who was an early adopter of the Maynuo - he said he was very pleasantly surprised with the accuracy and precision of it's measurements.

YMMV but hope that provides some useful insight.

Edit:  I just tried again and paid a little more attention.  Today the battery is down to around 3.45V.  When I tested I set the current for .001A.  When I have the Maynuo powered on but the front panel operate button set to off and attach the test leads the Maynuo acts like a DMM reading voltage with no (or very little) load.  As soon as I turned on the Maynuo operate button the Maynuo shows the voltage sink to under 2V but then stabilize with a current draw of .00009A  In other words, it shows the .001A set point and the actual draw of .00009A.  It's a pretty neat piece of test equipment.

[pdx_soft]

@HKJ Never heard of an SMU. Very cool, you're right exactly what I am looking for if it were in my budget!

[pdx_soft]

@TV84 From what I've read the upgrade doesn't make any difference. I have a ticket open with Rigol, I will ask to make sure.

[pdx_soft]

@Electro Fan Yeah, that is very helpful information. Thanks for taking the time to test.

[MarkL]

I have an Array 3723A.  To really understand what's happening with any particular load you also need to look at the dynamic characteristics with your intended pulse.

I thought this was an interesting question since I didn't know how the Array 3723A behaves at this low level.  I captured a few measurements:

3723A_4mA_steady_state.png: Steady state @4mA.  The Array 3723A can be set at 100uA increments, but it doesn't really start drawing anything until 200-300uA.  The *average* value is consistent with the setting (minus a small offset), and it goes up incrementally with the setting.  But in terms of the noise at this low level it's pretty bad.  Maybe it doesn't have enough internal resolution and it's dithering around the result.  Dunno.

3723A_4mA_4ms_pulse.png: This is the load with a 4ms pulse applied at 4Hz.  The voltage level out the pulse generator was 9V.  You can see the current is way off.  This is likely due to some combination of the internal circuitry trying to suddenly adapt to the change in input, and also charging of the input capacitor on the 3723A.  I suspect you're going to find on a lot of programmable loads with input capacitors that will result in non-ideal behavior.

3723A_4mA_40ms_pulse.png: Increasing the pulse width allows the 3723A to settle to the correct average value, but of course the noise component is back.

BJT_4mA_4ms_pulse.png: For kicks, this is a BJT biased just enough to have a collector current of 4mA.  Depending on your requirements, perhaps this would be sufficient, or some other simple current sink circuit.

If you really need a full range programmable solution, I would agree with the others that you're probably looking at an SMU.

[DaJMasta]

@HKJ Never heard of an SMU. Very cool, you're right exactly what I am looking for if it were in my budget!

Harder to find, but battery simulators would probably cover your lower current/lower voltage requirements.  There are a fair number of loads out there, and as mentioned they are for all sorts of things - often high power and fast response speed are emphasized, but there's a ton of configuration and outside of SMUs, the very fine gradations of current and voltage are harder to get (since the dynamic range to then do tens of amps at a high voltage would be insane).

I don't really know what project you're describing, but are you looking for instantaneous power measurements and a variable load?  If you have a repetitive signal you don't really need to characterize every moment of it to get a good idea of power consumption, often an average or something will do (if you know your idle current and your repetition frequency and pulse width, you can work out the current draw for the pulse with just an accurate average).  Especially if you don't need the variable load (if a constant resistance could work, for example), you can just measure current and voltage with a fast enough rate to catch your signal of interest and do the math yourself.  You can also apply a known signal and look at the current and voltage measurements from it - more akin to an LCR meter or vector analysis - to characterize a load.... I'm just not sure what you're actually trying to measure and whether you actually need a load for it or whether some other setup would be a better choice.

[Electro Fan]

I ran some tests but didn't take notes on each photo so feel free to ask questions about what they are showing.  I'd be happy to re-run some if you want.  In general, I set the power supply for 5V and 3 milliAmp (except maybe one or two where I used 4 mA on the PS to see how it would change the Maynuo measurements).  When the PS output is on the PS current display is actual (vs. the set current when the PS output is off.)  I set the Maynuo for various current and resistance settings as you can see on the Maynuo display.

Edit:  when the Maynuo is "Off" it is mostly just acting like a DMM.  I had two sets of cables each about 3' connecting the Maynuo and PS, plus some jumpers and connectors I could probably shorten down or eliminate; was just going fast.

[Tjuurko]

In experiments with Maynuo MN-M9812 (CC mode), it was found that the current step is about 50 μa at a limit of 3 amperes and about 500 μa at a limit of 30 amperes (a total of about 60,000 steps per limit).

[Dubbie]

My electronic load (B&K Precision) is useless below about 20mA. It's noisy and inaccurate.
For loads below that, I have in the past built up a custom circuit which is magnitudes more precise.
I too would love an SMU, but they are veeeeery expensive!

[PedroDaGr8]

The older Keithley 236/237/238 source meters can occasionally be found for relatively cheap (~$600-700 if you watch eBay). They have a fully detailed schematic and are capable of sinking as low as 1.0000nA (100fA resolution) and as high as 100.00mA (or 1.0A for the 238).

[pdx_soft]

@DaJMasta I have built a pacemaker for a local university heart research department. It is a constant current pacer it would be great to be able to just punch in (or program) different resistances in and make sure the unit is properly adjusting the voltage to compensate. Right now I attach different resistors to the leads while connected to my scope. The battery tester is a great idea though, never thought of that either. Right now I am thinking about building my own "micro" load unit. Sounds like a fun new challenge.

[pdx_soft]

@MarkL Honestly your Array unit is heads and shoulders better then the DL3021 in performance. I would be so pleased if the Rigol had performed even close to that well! Thank you so much for trying out your unit. Real world test can be so different from what you get from a spec sheet.

[pdx_soft]

@Electro Fan Wow, thanks again for taking the time to do these tests. So interesting to see what these device will do in the low range!

[pdx_soft]

@Everyone! Thank you so much to everyone that posted to my question. I am overwhelmed by the support given to a newbie on eevblog. I haven't heard back from Rigol yet but from what I have heard here almost all of the other load units do a better job at lower currents. I am still leaning toward the Maynuo unit but I will research the other suggestions before pulling the trigger. Cheers and have a great weekend!

[pdx_soft]

Wanted to follow up. I did return the DL3021 and recieved a Maynuo M9811 DHL express from Hong Kong (Amazon).
The unit is really amazing. Works down to less than 1ma. Smaller, quiet with no surprises. Performs just as I thought the Rigol would have out of the box (well after I figured out it was set for 220, doh!) From the tests I've done seems super accurate and an all around great tool. Very pleased especial at the price point. Just to be fair the fit and finish is not as nice as the Rigol, the manual is pretty useless, the interface is most certainly not high tech but works well. Haven't tried the Software/TTL interface yet, building the isolated usb/ttl adapter for fun tomorrow. Anyway, thanks again for the help on this! Probably going to have to look into the Test Gear Annonymous thread soon...

[Mr. Scram]

Wanted to follow up. I did return the DL3021 and recieved a Maynuo M9811 DHL express from Hong Kong (Amazon).
The unit is really amazing. Works down to less than 1ma. Smaller, quiet with no surprises. Performs just as I thought the Rigol would have out of the box (well after I figured out it was set for 220, doh!) From the tests I've done seems super accurate and an all around great tool. Very pleased especial at the price point. Just to be fair the fit and finish is not as nice as the Rigol, the manual is pretty useless, the interface is most certainly not high tech but works well. Haven't tried the Software/TTL interface yet, building the isolated usb/ttl adapter for fun tomorrow. Anyway, thanks again for the help on this! Probably going to have to look into the Test Gear Annonymous thread soon...

What about the fit and finish makes you say that?

[pdx_soft]

Don't get me wrong, it's not bad. The plastic just seems like it's not quite as high quality as the Rigol unit. The rotary encoder seems a little flimsy but works great. Honestly nothing I care about. It just feels like the case/controls were designed 15 years ago.  It's performance has been flawless, no bugs in the interface that I have found. Calibration looks to be excellent. The unit is easy to use once you push through the manual. I am impressed with the unit and would recommend to anyone, especially at the price point (<$500usd express shipped from Hong Kong).

[eevcandies]

Since you are talking about very low power levels , just grab some opamps & mosfets & build your own.  The  big cost typically comes from big power, like a 300 Watt  or 800 Watt load.  Sounds like you need low currents & less than a watt loading.  Breadboard ready in a few hours.

[jemangedeslolos]

My electronic load (B&K Precision) is useless below about 20mA. It's noisy and inaccurate.
For loads below that, I have in the past built up a custom circuit which is magnitudes more precise.
I too would love an SMU, but they are veeeeery expensive!

I just tested with my BK Precision 8601 and my Rigol DP832 and it works like a charm down to 1mA.
I quickly checked with a 60000 point DMM.
Much better than I expected.

What model of BL Precision do you have ?

[TurboTom]

When working with "classic" DC loads in conjunction with micropower circuitry, one should always consider the capacitance at the load's terminals that's usually in the range of 1~10µF. Especially in the op's situation where he's applying square pulses of a few milliseconds duration, the capacitance can cause erroneous results.

At the mentioned low power levels, and if only a resistive load has to be "simulated" and no CC CV or CP operation is required , I'ld go for a power potentiometer / rheostat, and if it's got to be controlled by a computer, attach a cog belt drive and a stepper motor (maybe with an index switch to identify CW or CCW position) and once do a calibration of that setup. KISS concept 

Just my two (euro-) cents...

Cheers,
Thomas


Edit: P.S. Or use a bunch of fixed registors in 2's exponential steppings and and switch them into the circuit by a bunch of relays -- poor man's DAC.