Anybody knows about electronic loads?

[e9999]

been thinking about getting one for battery testing. I'm considering going 400W in case I want to test biggish solar panels at some point, but those are pricy. So, 200W it may be. (Although it appears that some may be upscaled in power with just some software instructions.) I'd want external voltage sensing, and preferably computer control.
Do you have or work with one? Happy with it? Issues?
TIA

 

[ntsqd]

If I understand the question correctly then I use an Autometer BVA-300 (I assume it's 300A?). Been quite happy with it as a battery tester. Have not used any of it's other functions.

 

[e9999]

Thanks but that Autometer looks like something different and specific to 12V automotive systems. An Electronic Load is a device that allows you to generate an arbitrary power load in various modes like CC, CV, CW, or CR over wide voltage and current ranges to test electronic devices like power supplies, batteries etc. Sort of like a super duper programmable adjustable power resistor of sorts with data recording. More an electronics bench tool than a portable instrument from what I know. Like the Siglent SDL1000 series or the Rigol equivalent, for example. Very versatile. I suppose one could use them to test a 12V car battery or some vehicle electronics, and I may well do that, but I don't think that is the primary target. I'm most interested for now in testing smaller batteries for capacity loss, some DIY circuits, and possibly solar panels (if the latter even makes sense, not sure yet).

 

[ntsqd]

I wasn't sure that I did understand and it's clear that I didn't.

 

[e9999]

TBH, I didn't know much at all about these until I started to research them a few days ago. And -sadly- realized I could not be without one now, since I've been doing half-@ssed capacity measurements for years in various DIY modes (like with light bulbs etc) that really were just that... I could measure discharge energy for those conditions, of course, but could never actually control the process, so the measurements were always questionable. With an electronic load, you can just set a specific discharge current, for example, and it will be just that for the whole process. With a light bulb, the current will vary with time, unfortunately. And when it comes to most batteries, the discharge capacity is a function of discharge current, of course.

 

[1911]

paging @Engineer8000 ...

 

[Engineer8000]

sorry, never had a need for one of these

 

[george_tlc]

I have and use a Manyuo electronic load. It's well proven and there's a few similar units out there. Fairly old school in terms on 'UI', but it works just fine. I've used for a bunch of years. Mostly as an electronic load for LED driver testing and also for smaller battery runtime tests. They have a PC program for logging and controlling the unit.

https://www.dc-electronic-load.com/product/m98120-150v-0-30a-300w-m98-series-programmable-led-dc-electronic-load/?portfolioCats=10

This class of test equipment has its uses, though overkill for 'something that may come in handy'....

Electronic loads aren't perfect in some use cases since they are not going to behave as arbitrary weird loads (active loads with fast transient load changes). The reason is there is a control loop that has to change the load demands to maintain, for example, constant current, constant resistance etc. Depending on your power source and its control loop that is maintaining its regulation (e.g. a LED driver that has a fast response to load changes) the two control loops can do 'battle' and not achieve steady state. Sometimes adding a bunch of capacitance across the electronic load's inputs can modify the control loop interactions.

So, it's one of those products that you need to understand why you need it, what you plan to use it to do and how well it will 'behave' with your power source.

cheers,
george.

 

[e9999]

thanks George, will look at that brand. Not too worried about fast transients or LEDs, though. Do you know if some of those electronic loads can do semi-continuous scanning over ranges of current or apparent resistances well enough? I'm thinking about ways to generate IV curves for solar panels.

 

[george_tlc]

You can program steps in the various modes. Below shows Constant Resistance and changing the resistance values every second. It will log current, voltage, power etc so you can get curves. You can save the data and then pull it into excel for instance and do more with it.

I'm sure most electronic loads have similar features and new models like have fancier and more modern interfaces etc.

The Manyuo unit is surprisingly accurate, I see excellent correlation with my fluke 87III for voltage and current. The load also has sense connections allowing 4 wire operation - two that carry the load current and two that are the sense lines.

cheers,
george.

 

[e9999]

Every second is pretty good, that's nice (well assuming everything stabilizes within a second, that is). Good to know. Thanks. With solar panels, I'd indeed want to get the scan done fast enough that insolation changes with time are not much of a factor.
Yes, from what I've read, many of these electronic loads are very accurate with both current and voltage, the former being less obvious.

 

[george_tlc]

1 second was just an example. The fastest is 0.1 sec per change

cheers,
george.

 

[e9999]

excellent, I will check that with whatever model I end up interested in. And if there is a limit to the number of steps one can take.

 

[george_tlc]

I'm sure there's a limit, don't know offhand what it is. Something to verify in the product's documention/spec. Fancier units that gave scpi may have pretty fancy capabilities...

cheers,
george.

 

[LandLocked93]

Late to the party, but...you're looking for a variable resistor?

 

[trainbufftony]

I use a dummy load for testing audio amplifiers.
Personally I'd forego the ui, software and boards and use a simple fluke meters to monitor voltage/current across your load then use ohms law to calculate anything you need. Use multiple resistors with tap off points if you need to vary the load.
You don't need second by second monitoring. Anything that discharges in a matter if seconds isn't useful anyway.
My dummy load consists of several high wattage resistors mounted to an aluminum heat sink using heat sink compound. It gets hot enough to burn you. The resistors should be available through mouser, heat sinks are expensive so salvage one. This way you'll actually learn what you're doing instead of learning to trust someone else's gadget ...

A fluke i410 current clamp or equivalent should give you all of the data that you need, in conjunction with a Voltage measurement of the pack under load.
Have fun. You'll be stepping out of the typical electronic gizmo world and into the world of making real current do real work.
The amp clamp will be very useful.

 

[ntsqd]

My work uses stuff like this and some custom written for D/A&A. We use air cooled Ni-Chro wire and ribbon for the resistive loads. I think that largest that we've tested, so far, is about 5 MW. Those ribbons and wire occupy a couple small rooms. A couple of the small systems do use a Fluke clamp-on current sensor. I've occasionally helped with some of the related tooling, but am not usually involved with the testing itself and don't know much about it.

 

[e9999]

I do of course have a big dummy load as well, a rack of light bulbs in my case, which works great when needed. And meters out the wazoo. The problem with that load, though, as with the ones just above, is that there is no control over the discharge rate so you can't really do good quantitative tests or do scanning of DUTs. So it may be good enough for quick aging tests and maybe some simple relative comparisons, but nothing more controlled. For example, you can't do a fixed current discharge, like a C/20 etc, so can't do very meaningful discharge curves. That's why I will be getting an electronic load before too long, hopefully.

 

[Tesota]

Most of the time I don't need all the fancy features of an electronic load, this was my quick and dirty solution since I only need to test a single voltage. The mosfets are controlled via an external uC and each of those resistors will draw about 20A @ 12V. Two of these will easily draw 200A and cost substantially less to build than an equivalent electronic load.

 

[e9999]

Nice! Doesn't that backing plate get awfully hot at full load?