/       /       /    Protection and switching board for Atorch load testers
Support the project on Patreon


Protection and switching board for Atorch load testers

450 ₴
Go to store
This story began a long time ago—I think more than a year ago, though its roots actually go back to 2023. Today, I’ll show you what it all led me to.

To stay up to date with new posts on the blog, I recommend following me on Instagram, where I’ll post notifications about every new article and maybe write about something interesting too: link to my account

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


The story behind this project goes back to my DL24EW electronic load review. I found it quite interesting because it could do more than just act as an electronic load: it could also manage battery charging—or, more precisely, simply connect an external charger. Still, that was already pretty good.
Back then, I tested the included switching board and remember criticizing it heavily for its high voltage drop, which in turn prolonged the CV phase.

Some time later, I got my hands on a new model, first the CL24, for which I ordered two switching boards, one for each electronic load, and then the BW150, which can also manage charging, though I didn’t order a board for that one.



While preparing the CL24 review, I overlooked the switching board—and I shouldn’t have. It turned out that the old and new versions are radically different.


First, here’s the schematic for the original board included with the 24EW.
As I wrote in the review, the circuit is extremely basic: a MOSFET is driven through an isolated DC-DC converter and switches the charging circuit.
Reverse-polarity protection is as simple as it gets: a fuse and a diode. Minimalism in all its glory.

The main problem is that the MOSFET used has an on-resistance of about 75 mΩ, which means a voltage drop of 0.75 V at 10 A.
That voltage drop interferes with charging, especially for single cells. On top of that, at 10 A the MOSFET will dissipate about 7.5 W of heat, which is a lot. So this board is suitable for currents no higher than 5–7 A.


Not long ago, though, one of my readers contacted me for help figuring out some strange behavior from a new electronic load. The complaint was that with an input voltage of 4.2 V and a battery voltage of 3.8 V, the current dropped to about 0.35 A.
At first, I couldn’t even see what the problem was. But when I got hold of an identical board and took a closer look, I realized the previous board wasn’t quite as bad as I’d thought.


The thing is, in the second version of the board, the manufacturer put the reverse-polarity protection diode in series....
Yes, of course, when charging single cells, you get exactly the mess that reader contacted me about: you simply cannot charge a 4.2 V battery from a 4.2 V charger.


Anyway, that message was the last straw, and I decided to take matters into my own hands and solve the problem.
In fact, I’d partly solved it a year earlier, when I started building electronic loads based on the DL24EW board, only I used my own switching board there.


I designed two versions. The larger one handles up to 30–40 A and is used in full-size electronic loads based on the 24EW.


There’s also a smaller version with TO-220 transistors, rated for a maximum current of 20 A. It’s used in battery testers, like this one, or in an electronic load based on the CL24 controller. That’s because the maximum current there is 20 A.


I’d been thinking for a long time about making my own switching board for electronic loads in these series, but kept putting it off. Then a reader contacted me about the problem I mentioned above, and that was the last straw.
After all that, I sat down, designed a new circuit and a new PCB, and added it to my next JLCPCB order. A couple of weeks later, I had the finished boards in my hands.
I really liked the service, by the way. I recommend it. If you have any questions, I can share what I know.


First, I checked whether the board fit in the stock location. It fits perfectly, although there is one small caveat, which I'll get to later.


Before placing an order, I usually make a test sample first. I did the same here, but never got around to assembling it—the problem showed up much earlier.
The thing is, I initially designed the board for the 24EW load, and everything would have been perfect if not for one catch: it switches the power supply line, leaving ground connected.
The CL24 and BW150 do the opposite: 12 V is always present, and ground is switched. In my view, that's wrong, because good practice is to switch the power line and leave ground alone. The same goes for current-sense shunts, though the situation there is more complicated.

So I had to sit back down at the computer and make a few changes to the circuit. The result was a new, universal version that works with both types of control, without any jumpers or the like.

By the way, the new version has another important difference from the very first one I used. The original circuit wouldn't let you connect the power stage if anything was connected with the wrong polarity. The new circuit allows connection only when everything is connected with the correct polarity.
They look the same, but they actually work differently.
The old circuit worked on a deny-by-default principle: if something was wrong, it blocked operation; otherwise, it allowed it, even if there was no voltage at the input.
The new circuit allows operation only if both inputs have voltage with the correct polarity, and that voltage is above about 1.5–1.8 V. In other words, it won't connect the battery to the charger when there's no voltage.



Now let's see how it turned out. I grab the components and solder them onto the board.


It turned out reasonably neat, although I’ve already tweaked a few spots to make it more ergonomic to use.


I ordered a couple dozen boards, but assembled just two as a test. While routing the board, I also tried to make it suitable not only for typical Atorch loads, but for my own projects too—in an embedded setup.

The terminals are the same as on the Atorch, and there’s also a 5521 jack for connecting lower-power chargers.



There are clearly more components here than on the Atorch. I also added LEDs to indicate that the charger is connected correctly; they light up whether or not the board is connected to a load.
I also made provision for either an SMD fuse or a standard 25 A automotive fuse. The former should have lower voltage drop.

By the way, about the fuse. It’s not really needed anymore, and essentially protects against just one situation: both the charger and battery are connected, and the load sends a switching signal—in other words, the battery is charging, and you short the charger's terminals. The switching stage is on, so when the terminals are shorted, all the current flows through the transistors. The circuit will try to turn them off, but it may not react in time, so if the transistors fail, the fuse should blow. In every other situation, it just gets in the way because it has nonzero resistance.




Besides the terminals and other details, I tried to match the boards' style to the Atorch boards—even the solder mask is the same color :)


Unfortunately, I did mess up one spot, though nothing fatal: I swapped the S and D labels (good thing it was only the labels).



As I mentioned, I added a charger polarity indicator. It starts lighting at around 2–2.5 V, and is already quite bright at 3 V. The maximum voltage is 50 V.


There’s also a yellow LED on the board. It lights only when all of the following conditions are met:
1. The charger input has the correct polarity and a voltage above 1.5–1.8 V
2. The battery input has the correct polarity and a voltage above 1.5–1.8 V
3. A control signal is present.

So this is an indicator that the switching stage is operating, not just that a control signal is present.


But it’s worth checking—and, more specifically, comparing the original boards with mine. For this, I’m using an RD3060 power supply and a load based on the CL24, which, incidentally, contains a similar switching stage.


Let's test it. 12 V, which also powered the control circuit here. Current: 1, 5 and 10 A. The old version of the Atorch board.
Test setup on the left; voltage drop across the transistor path in the middle; voltage drop across the fuse path on the right.
The total drops were 0.04, 0.2 and 0.44 V, respectively.


The second version of the Atorch board, at the same currents: 1, 5 and 10 A, although it was already struggling at 10 A.
Total voltage drop: 0.36, 0.55 and 0.78 V. Yikes...

Here's what's curious: according to the datasheet, the transistor used has a resistance of 75 mΩ, but based on the measurements, it's closer to 35 mΩ. So what kind of transistors are these, anyway...?


Now for my board in its basic version. Same currents: 1, 5 and 10 A, plus 20 A. The order and measurement points are the same as before.
Total voltage drop: 0.015, 0.076, 0.16 and 0.346 V. I think you can see even without a microscope that the drop at 20 A is lower than on the first-version Atorch board at 10 A, let alone the second version.


The second board has two changes: the fuse has been replaced with an SMD one, a single 30 A fuse is fitted, although the board is laid out for two 15 A fuses—that would have been better. The trace between the power transistors has also been reinforced.

Same test currents: 1, 5, 10 and 20 A.
Total voltage drop: 0.013, 0.115, 0.14 and 0.312 V, slightly lower than on the first version, so I may well stick with this design.


The actual results are a little worse. Both copper and transistors have a positive temperature coefficient, so their resistance increases as they heat up. That means the voltage drop will be higher when hot. I didn't test all the boards, since the change should be roughly the same for all of them; I tested the latest one.
At 20 A, the drop across the switching path went from 0.232 to 0.285 V, and across the fuse path from 0.0804 to 0.0826 V. The total increased from 0.312 to 0.367 V, but it's still better than the first Atorch board at 10 A.


The transistors—at least the one on the right—reach about 110°C under these conditions, which is quite a lot. So 20 A is the maximum for extended use. The board near the fuse gets to around 85°C; I think the fuse itself was also around 100°C.


You can improve things a little by using larger heatsinks. I may give that a try, perhaps replacing only the one on the right, since the left heatsink is already quite close to the terminals and could get in the way.
Another option is to force the load's fan to run at low speed. That would help a lot, and would also cool the load's shunt. Unfortunately, this isn't suitable for loads based on the DL24EW.


As I already mentioned, the module is fully compatible with loads in the DL24EW, CL24 and BW150 series, as well as all newer models that support charge control. This applies to both electrical and mechanical compatibility.


Now for the price, and what makes my version of this module better than the original.
First, the differences:
1. Voltage drop. As I showed above, the original module—especially the second version—is completely unsuitable for charging low-voltage cells and batteries.
2. Reverse-polarity protection. My version provides full protection: the board simply won't connect the charger to the battery if anything is connected with the wrong polarity.
3. Reverse-current protection. With the Atorch version, power from the battery will always flow toward the charger through the MOSFET's body diode, and there's nothing you can do about it. My version completely breaks the circuit, which also protects your charger from damage, especially if it's based on a bench power supply or a DC-DC converter. It also makes the setup safer to use.
4. Higher current. The first Atorch version is rated for a maximum of 7–10 A, the second for 5 A. This one handles up to 20 A, or 25 A with forced cooling.
5. Indicator showing whether the charger is connected correctly.
6. Switching voltage. Atorch has the advantage here, since my board switches up to 50 V, while I think Atorch can switch up to 200 V. Actually, that's not quite right: it can switch 100 V, but if the voltage is above 50 V and you reverse the polarity, the transistors will fail because they're rated for 100 V, and with reverse polarity the charger and battery voltages add together.


As for the price, I'm planning to charge 450 UAH for the module plus a mounting hardware kit. If you already have an Atorch switching board, it'll be 400 UAH; in that case, the terminals and mounting hardware aren't included, since you already have them.
Yes, it's not exactly cheap, but I'm not Atorch—I don't have their production volumes or wholesale prices :(

I'll post the listing on OLX later. For now, you can email me.




That's all for today. I hope my little board will be useful to you in everyday use, and to me as a small source of income for future projects. A schematic and PCB layout for making it yourself may be added later for Patreon subscribers.
As usual, I’d be glad to hear your comments and suggestions.
450 ₴
Go to store
This page was found when searching for:
atorch bw600 3d model, самодельная плата защиты светильника от 389, 1, модуль зарядки atorch atorch cl24r, atorch xh-cdc-mk, kirich

Questions and answers

Why does the stock Atorch board charge a single cell poorly?

The second version has a reverse-polarity protection diode in series. With 3.8 V at the battery and 4.2 V from the charger, current falls to about 0.35 A; a 4.2 V source cannot charge a cell to 4.2 V.

What current can the author's Atorch switching board handle?

It is rated for up to 20 A, or up to 25 A with forced cooling. The first stock board is suitable for about 5–7 A, while the second is limited to 5 A.

How does the author's board differ from the stock Atorch board?

It fully breaks the circuit to prevent reverse current from reaching the charger, and connects the battery only with correct polarity and voltage above approximately 1.5–1.8 V. It also indicates correct charger connection, but its maximum switching voltage is lower, at 50 V.

The figures from this review are in the measurement database: Other devices

Calculate manuallyAll calculators

Discounted products


You may be interested in


Discounted products


Support the project

All the content on the site and the advice are free, but we’d be grateful for your support of the project and the channel!

Comments1

Comments are published in their original language.

  1. 31 March 2025 20:36#1
    Любопытная платка, видел что у производителя с ней явно беда, и здесь я вижу не только достойную замену, а и замену, которая при этом и выглядит "как родная", респект.

    +3

Discuss on the forumHave a question or some experience to share? Discuss the article with the author and readers: the forum is a convenient place to share schematics and photos.
Go to the forum

Only registered readers can comment. It takes a minute to sign up, and you can log in with the same account you use on the forum.

Log inRegister