Continuing the topic of high-power electronic loads

Not long ago, I showed a new version of the electronic load. In a sense, it’s a downgrade—a step backward rather than forward—but it’s not quite that simple, so I decided to write a short article about it.

As a reminder, to keep up with new posts on the blog, I recommend following me on Instagram. I’ll post notifications about new articles there and might also write about something interesting from time to time - link to my account

You can also subscribe to updates and new articles on the Telegram channel - https://t.me/KirichBlog


As I’ve already mentioned, today I’ll talk a little about the further development of the TEC-80k controller-based load series, which I’ve already written about.
It uses the same power modules as before, but a different board handles control and connection to a PC.



Externally, there are almost no major differences from the DL24MP and DL24EW series loads. Unfortunately, it falls noticeably short of them in terms of functionality, but in this case that’s justified, especially when dealing with high currents.


For production units, I build loads rated at 500–600 W and 1000–1200 W. There was also a 1500–1800 W version, and everything would have been fine, except that all loads based on Atorch boards have a software current limit of 40 A. So you can increase the power, but not the current, unfortunately. :(


But people often asked me about loads capable of handling higher currents. So far, there were two options: a “dumb” one based on simple analog control, albeit with a PC connection and undervoltage cutoff, or simply connecting two 40 A units in parallel.
The first option is perfectly decent in principle; I even managed to push it to 1300–1400 W, but its software is pretty poor. The second is inconvenient to control because, although the software is better, it can’t work with two devices at once.


Not long ago, someone who had first seen my ad on OLX called me. It said that a high-power model was in the works. The request was a typical one: to load 6S battery packs at currents of up to 100 A.
In response, I immediately suggested a pair of DL24MP loads as the classic solution. But the person was very interested in a single unit with higher power and current. In the end, I said I could build one, but added that I don’t like selling “rough” versions, since the ones I usually sell are essentially production units that I know inside out. That didn’t put the customer off. I also mentioned that this load would cost a little more than two 1200 W units because it would be more complicated to assemble. The customer gave me the go-ahead, and I got to work.



I used the same TEC-80K controller, paired with four 500–600 W power modules, and about a week later this setup was born.


The enclosures made assembly quite awkward. I plan to solve this problem in the future; I’ll show you one option at the end. Overall, though, it turned out reasonably neat, at least for a first unit—really, a prototype. I had to work out the control, power supply for the whole thing, proper connections, thermal protection, and so on. The customer also wanted XT60 and XT90 connectors on the front panel in addition to the “banana” sockets, and a USB-C port on the back for connecting to a PC.


But if you remember from the previous review, the TEC-80K load is limited to 1600 W and 80 A. So I contacted the developer, who made an additional firmware version for me with limits of 120 A and 2700 W. Now, when it starts up, it displays the model name TEC-80K 2700.
The software also had its current limit raised to 120 A. I was particularly concerned about that, since I wanted more than just local control.


Next came a brief test, calibration, and a check that everything worked. Unfortunately, the highest power I was able to test was 1400 W, but the load performed flawlessly.


Next, I tested current rather than power. For this, I used a battery made up of four 230 Ah cells, for which a discharge current of 100–120 A is no big deal. Of course, I wanted to test it at full power, but my 24 V battery is only 88 Ah at most, and I didn’t want to abuse it with currents like that.


I had no trouble getting the 100 A we’d agreed on, and easily pushed it up to 109 A as well. Beyond that, my clamp meter went into protection. I could have rummaged through the boxes and found my 660 A clamp meter, but I decided to leave those tests for next time. Besides, the load performed flawlessly at 109 A, and I don’t think anything would have changed at 120 A. What’s more, the 120 A load is easier on the modules than my small 40 A unit, since each module is only handling 120/4=30 A—and even easier than the two-module 100 A version, where each module handles 50 A. That is, Technically, I can easily make the same version for 160 A, and possibly even 200 A.


I monitored the temperature during the test, but since the load was running at half power, even the fans never reached full speed.
I was more interested in the temperature of the connectors and how the whole thing behaved. Current stability is reasonably good, but I plan to improve that too. As for the temperatures, they were actually better than I expected.
1. Overall view, test at 100 A
2. The banana plugs were practically cold. The cables were hotter where the lugs were crimped; they’re old cables I happened to find at home.
3. Here you can see that the upper-left battery busbar is hotter. It turned out to be slightly loose. I always test batteries for hot spots after assembly, because issues like this can crop up if the contact pressure is poor or some debris gets under the terminals. Even an ordinary drop of rosin can cause trouble.

I’ll answer in advance the question of why I don’t use a higher-quality thermal camera. It’s simple: I have to get it out, unpack it, and connect it to my phone, which is sometimes almost out of charge. The one I used to take these may not be as good, but it’s always at hand, and all I have to do is turn it on.




The finished unit ended up looking like this. It measures about 270×250×185 mm and weighs around 4–4.5 kg. It’s not as neat as the production models based on Atorch units, but I tried to make it look reasonably professional.


Right after testing, the load went off to its new owner. I only had time to take a few photos, including a group shot with the smaller 80 A, 1200 W version.


I was worried the screw heads on the top cover would make it look untidy, but it turned out they can actually be useful: if you put a second load on top, its feet land right on the screw heads, effectively securing it in place.


The person who received the load has already tried it out, though so far only at 40 A and 80 A, and recorded a short video report.
There’s no point showing 40 A, but I’ll show you 80 A. By the way, in the second part of the video you can hear the load’s fans ramping up to their operating speed. Using something like this at home wouldn’t be very comfortable, but at work it’s perfectly fine.

The video clearly shows the cell voltage sagging after the test starts, but it’s more important to see the voltage of each individual cell.
The video was shot simply as a report, so many thanks for that. I may ask for a video specifically showing how it works later.

Sorry, the video is short, so it uploaded as a Short. Link: https://youtube.com/shorts/8UxowXD0vn8?si=S1MJRKJPzbFZ7WQv

Here’s the second part of the video. I assume this is closer to the end of the test. Please don’t be too critical; the video is just meant to show how the battery and load behave.
Link: https://youtube.com/shorts/yK06pZ21Sj0?feature=share



But as you can imagine, I’m definitely not stopping here. I’m currently assembling a load made up of six modules, with current up to 120 A and power of around 3500–3600 W continuous and 4000 W for short periods. I’ve been working on it for a long time, but I’ve finally arrived at a workable solution.
In the end, I chose a metal enclosure 4U high, which fits all six modules. It’s not an ideal fit—a height another 5–10 mm greater would have been better—but this is what we have. And the price of a case like this leaves a lot to be desired, especially considering that I had to order laser-cut front and rear panels. I mean, you can't exactly cut something this beautiful with an angle grinder....

Someone is bound to ask: why not make loads that feed power back into the grid, or at least charge another battery?
In principle, that's possible, but it adds complexity to the circuitry and also makes the load more complicated to use.

Feeding power back into the grid is a problem: if the mains power goes out, the load shuts down because there's nowhere for the energy to go. This way, you can simply power it from a UPS or even a cordless drill battery (I've built versions like that).
Feeding power into another battery is a little simpler in some ways, but you always need to have a large-capacity discharged battery on hand, which you'll then have to discharge. Or you could end up with a stockpile of batteries at different charge levels. In short, it comes with its own problems.

That's why, for now, it's easier to just heat the air—especially with winter coming.




Earlier I mentioned how awkward it is to assemble a high-power load in a double-width case. Since I got my hands on a factory-made case, I checked that one too. It turns out there's plenty of room to fit four of my power modules lengthwise, and if you pack them more tightly, there's even room left for connectors.




That's the latest news from me. Unfortunately, I've been swamped and haven't written much, but I'll try to make up for it a little in the near future.

I'd also like to hear what you think about this. I'm really curious to hear your suggestions and, of course, your criticism.
One more thing: if anyone works with Arduino or STM (or something similar) and has a little free time to help me solve a small problem (extensive professional experience isn't essential), please get in touch, either by private message or email.