NY-D08, spot welder controller

I've already written a couple of posts about spot-welding controllers that use supercapacitors, but today we're looking at a more "classic" setup with a powerful 50 Hz transformer.
I'll show you the controller itself, as well as a small example of how it's used in a real spot welder.

As usual, a reminder: to keep up 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


I didn't order this controller for myself. I already have a welder using supercapacitors, and I don't work with batteries very often anyway. But a friend of mine uses one a lot. I'd previously helped him build a version with a controller from Yurok, and it works quite well.

But he needed another machine, so he bought a microwave oven transformer, an enclosure, some other bits and pieces, and the controller itself. He didn't really want to order a PCB, buy the components, and solder everything together. At the same time, he wanted a larger screen, so he went with a model from AliExpress.

There are quite a few different spot-welding controllers on AliExpress. Some are very basic, while others cost more. We wanted one that could generate two adjustable pulses and, as I mentioned above, had a large screen.

Among the better-known options, there are probably two models worth mentioning: this inexpensive one for around $15–16, and the one I'm reviewing here.


1. I also came across another one. It looks like a decent controller, with a transformer already installed and a triac mounted on a heatsink.
2. While searching, I found this controller, apparently for a supercapacitor spot welder, but with the option to reverse polarity. That should actually be useful. I once mentioned that I'd noticed a difference in how the electrodes performed.



The controller arrived relatively quickly, but the seller took their time shipping it... I was starting to think they wouldn't send it at all, and even had to extend the shipping deadline.
The packaging was basic: the boards were wrapped in bubble wrap, while the rest of the accessories were just in a zip-lock bag.


The package includes:
1. Controller board with display.
2. Power board.
3. Connecting cable with connectors
4. USB-UART converter
5. Temperature sensor
6. Button cap
7. Encoder knob.


As for the small accessories, the only really important one is the cable connecting the modules. Without it, the unit won't work.
They also included a USB-UART converter for some reason, and even went so far as to remove the marking from its chip.


The cable is supposedly for working with software, which I never managed to find. But the software only lets you set three parameters: input voltage, presumably output voltage, and transformer power. What's more, as I understand it, you can set all of these through the module's main controls anyway. But again, it's not clear how—and more importantly, why...


Power board.
It has galvanic isolation, with the "cold" and "hot" sides clearly separated. There's also a warning not to touch the hazardous section with your hands.


The board has a bunch of connectors:
1. Switch - connect a foot pedal or control button
2. AC12V - board power supply; use only a conventional "iron-core" transformer.
3. TEMP - temperature sensor connection
4. State - status; I haven't looked into it, and it's not described in the schematic, but I can see an optocoupler connected to it.
5. FAN - 12 V fan
6. VALVE - valve, or more precisely, the valve for the pneumatic electrode-clamping mechanism.

You can also see a couple of PWM controllers, a bridge rectifier, and transistors that switch the fan and valve.


1. The high-voltage section is much simpler: optoisolation, a powerful triac, a few other components, and holes for soldering the wires.
2. The terminal blocks are quite good and convenient to use.


Controller board.
On the front are an LCD, a button, and an encoder. Here I have to criticize the manufacturer. Not only did they use an encoder without a built-in button, they also made it so stiff that I thought it was about to start squeaking.
I don't really see the point of separating the encoder and button when push-button encoders cost next to nothing and would actually be more convenient to use.


All the other components are on the back, including the microcontroller.


Main components. The USB-UART converter pins are also visible on the right, but as I wrote, there’s no point in using it.


Board dimensions and spacing between the mounting holes.


The seller’s page didn’t include a wiring diagram, but it’s easy to find in similar listings from other sellers.


For a test run, I used only the power board, controller, and transformer; I didn’t even connect the temperature sensor. Still, everything worked just fine.


When power is applied, a white glow sweeps across the screen, followed by the word Loading. If your device happens to display information in Chinese, hold down the button before switching it on to disable language selection mode.


1. After startup, the screen shows the device model, settings, temperature, number of “welds,” output status, and even prompts.
2. There are six settings screens:
Memory slot number
First pulse duration
First pulse power
Pause between the first and second pulses
Second pulse duration
Second pulse power

Switch between settings screens and adjust values by turning the encoder; press the button to enable or disable adjustment mode. Keep turning the encoder to enter the settings menu, then press the button to select it.
3. The settings only let you adjust the valve on/off delay and the fan activation thresholds.
4. Valve settings: how long before current is applied the valve turns on, how long afterward it turns off, and how many on/off cycles to run in automatic mode.
5. Fan control: the turn-on temperature can be set from 1 to 99 degrees, with 3 degrees of hysteresis.
6. The fan can be set to off, on, or automatic, depending on the temperature.

There’s also an option to restore the factory settings, but I couldn’t figure out exactly how—either start up while holding the welding button, or hold down two buttons at once. The “weld” counter, however, can be reset by holding the button for about 3–4 seconds in the operating mode..



The description for another listing with the same controller included a photo of the settings menu with a third item, Information, but it wasn’t there on mine.


If the temperature sensor isn’t connected, the display shows 99 degrees. So if you select Auto in this state, the fan will turn on, as indicated by the LED lit next to the corresponding terminals.


Once the temperature sensor was connected, everything returned to normal. I didn’t check the measurement accuracy, but the temperature seems to match the actual value.


Rather than show a bunch of pictures, I’ll show the minimum and maximum values for all the settings. Since each screen can be configured completely independently, these are the overall minimum and maximum values.


Now let’s move on to the practical part. I can’t show the assembly process for the spot welder, and there isn’t much point anyway. In this case, my friend did practically all the work himself, assembling everything in the enclosure and even preparing mounting points for the controller.


The enclosure is from an old Soviet voltmeter, which served as a donor for the transformer.
Fuses were installed on both lines, and a fan was added just in case, although it’s not easy to get a transformer like this to overheat.
The front panel has:
1. Mains switch
2. Controller display, button, and encoder
3. USB socket for powering something at 5 V.
4. Switch for the power supply that powers the USB output
5. Backlight and backlight switch.
6. Socket for connecting the welding current trigger button.

The electrode mounting arrangement is also very well thought out—more precisely, the fixed electrode. But more on that later.


The layout inside is relatively spacious. The power transformer is visible at the top, with the auxiliary transformer for powering the controller to its right. A 5 V power supply is located under the power transformer.


The transformer donor was apparently a Moulinex microwave, and I can say for sure that the transformer turned out to be excellent: quiet and efficient.
The secondary winding is made of flexible wire with a cross-sectional area of 48 mm², wound in four turns.


Along with the spot welder itself, my friend also gave me various welding strips, about a meter of each size, backlight LEDs, a copper plate, and some washers.


There were six strips: three steel and three nickel. They also varied in both thickness and width.
Dead batteries came in handy too. They're great for experiments, since most of them are completely discharged as well.


Next came attempts at welding with different settings. As I later realized, though, I made a mistake that caused the welder to heat the weld area more than it actually welded it.

First, I tried steel nickel-plated strips, which are the easiest to weld. Two of the three settings worked reasonably well, but with the thickest strip, it wouldn't weld properly.

I also noticed a strange effect. It's not very clear, but you can still see it in the second photo on the middle battery. There are three welding attempts, so there should be six spots, but three are clearly visible while the other three are barely visible. All three faint spots are marks from the stationary electrode. I don't know why.



Next I tried nickel strips. The situation was basically the same: the thin and medium strips welded, but the thick one only pretended to. I tried several times and eventually just burned through the strip without managing to weld it properly.


I figured out the reason later: I was welding with the wrong settings.
For welding, you want a high current and the shortest possible welding time, but I was doing almost the opposite. In the end, I was just heating the strip until it glowed, without properly welding it.
From left to right, these are the different settings I tried with strips of different thicknesses.


To prove the point, when I tried welding two strips together, I got plenty of smoke but not much else :)
Actually, the strips did weld together in this case, but they got so hot that the piece of laminate I was working on nearly turned to charcoal.
Below are three stages. To the right of the strip is the mark it left on a scrap of laminate. You can see that shortening the pulse noticeably lowers the temperature at the weld point.


Then, just out of curiosity, I decided to weld a piece of nickel strip to a pair of pliers. Why did I do that... It took ages to pry the strip off afterward.
By the way, these welding settings are almost optimal: a very short first pulse and a longer second one. To improve them, I'd reduce the power of the first pulse.


This isn't welding, though—it's what happened when I accidentally switched it on without pressing the electrodes firmly against the workpiece. I was aiming to weld a couple of plates, but accidentally hit the switch and applied current before pressing them down.
I ended up burning right through the strip and damaging the electrodes. The second electrode was damaged quite badly.


I'm sure many people are wondering what the numbers shown in the settings mean. For example, what does the 5/60/5/5/60 setting mean?
It's very simple. There are six values in total, but the first is the memory slot number, so you can ignore it.
The rest are more interesting. Below are the waveforms for four settings:
5/60/5/5/60
20/60/5/5/60
5/60/5/30/60
5/60/20/20/60

1. First pulse duration
2. First pulse power
3. Pause between pulses
4. Second pulse duration
5. Second pulse power

The time is set in full cycles of the mains sine wave, so each cycle is 20 ms, i.e. 5 in the settings is 100 ms or 0.1 s.


Power is even simpler: essentially, it’s set as a percentage of the maximum. Below are the waveforms at the welding contacts with no load, at settings of 30, 50, 75, and 99.


And of course, I was curious about the current produced during welding. To measure it, I used a clamp meter with a peak measurement mode.
At a power setting of 60%, the clamp meter went into overload. The predicted maximum welding current is around 1000 A.



The stiff encoder had been annoying me from the start, and after I installed the board in the case and the knob started rubbing a little, it became a complete nightmare. In the end, I suggested replacing the original stiff encoder with the standard one I use, removing the button, and using the encoder’s own push button instead.
Done and done: I removed the original encoder and button and installed a standard one. It turned smoothly and was comfortable to press. I put everything back together and gave it a try—turns out I’d rushed things.
I took it apart again, remembered the Chinese guy who’d installed this encoder and all his relatives, because it turned out to be the reverse type and the rotation direction was wrong. I fixed it properly and put everything back together.

All in all, a standard push-button encoder is much better. I recommend making the change right away.


I also removed the original standoffs, which were about 8–10 mm long, and replaced them with the shortest plastic ones I could find, 5 mm long. This works better, and the screen sits almost flush with the case.


My friend printed a plug for the unused hole right away, but it’s still at his place.



Now let’s move on to the welding tests. It turned out there were two key cases:
1. 0.1 mm steel strip
2. 0.2 mm nickel strip.

In the first case, I needed to find out whether the welder would burn through the strip. I set the power close to minimum and started testing.


It turned out that this was actually a little too low, which means there’s room to go lower still. Here are the results for these settings:
1. 1/35/5/1/35
2. 1/35/5/1/45
3. 1/35/5/1/55
4. 1/35/5/1/65
5. 1/35/5/1/75
6. 1/35/5/1/90

The strip welds properly starting around the second or third setting.


I welded a second strip to the positive terminal and another to the negative terminal using the 1/35/5/1/90 setting—that is, the shortest possible first pulse and almost maximum power for the second pulse.


Then I took some thick nickel strip and started welding it with the last settings I’d used: 1/35/5/1/90.


1. 1/35/5/1/90
2. 1/35/5/2/90
3. 1/35/5/3/90
4–6. 1/35/5/3/90

Here I had to increase the second pulse duration from 1 to 3. The last setting worked great: the strip didn’t burn through, but it was welded so firmly that it simply tore when I tried to pull it off.


Results in macro photos


Now it’s time to show some additional features.
I immediately installed an LED in the designated spot using a specially printed holder. Unfortunately, you can’t see how brightly it shines in the photo, but I can say that it illuminates the welding area nicely. It has its own switch.


But the strip-welding option may be of greater interest. For this, there’s a copper plate that attaches to the stationary electrode.


The principle is very simple: if you need to weld two strips together, for example at an angle or in a cross, place the pair of strips on the plate, press them down with the second electrode, and apply the welding current.


For comparison, I removed the stationary electrode and did the same thing the usual way. The difference between the welds is immediately apparent.
In the first case, there are just neat spots on the strips, since the current passes straight through along the shortest path. The photo shows six single weld spots.
In the second case, the current follows a longer loop, so you have to increase the power. This causes more heat and can even overheat the strip. The photo shows two double spot welds.

The only thing I didn’t really like about the first mode is that the electrode sticks noticeably to the strip.


And here are all the test subjects, which I’ll now have to dispose of somehow.


I can already hear the question: what is the welding cable made of?
I usually use large-gauge wire, but it tends to be quite stiff. In this case, my friend decided to use special PSH wire, with a cross-section of 4 mm².
We used 12 lengths, each 206 cm long, for a total of 24.75 m. Four turns, calculated voltage of 3.83 V, and a cross-section of 12 × 4 = 48 mm². The wires were bundled together and insulated with soft heat-shrink tubing.

I’ll say this: the cable is really flexible, and its cross-section of almost 50 mm² made it easy to get a decent current through the load.


That’s about it. I’ll keep the conclusions brief: I liked the board. Sure, there are a few minor flaws, like the encoder, and it’s not clear why they included a USB-UART adapter, but otherwise it’s convenient, practical, and seems pretty reliable.

I think I’ve covered the welder itself too. I hope this was useful and helps you build your own convenient, reliable device :)