Bakon BK3200 soldering station with a claimed power rating of 120 W

I came across a rather interesting soldering station. Its claimed 120 W power caught my attention, as did mentions online that it was an induction station.
Let's see what it's all about and take it apart while we're at it.


I've already reviewed a soldering station from this manufacturer, but recently I found out they make several other models that are marketed as induction stations. At least, that's what I've repeatedly seen mentioned on forums. Interestingly, the store page doesn't say anything about this. Then again, I don't tend to rely on information from sellers, since they often list features that aren't there or leave out ones that are.

The claimed specifications are:
Model - BK3200
Power - 120 W
Power supply AC220V 50HZ
Temperature range 200-480°C
Typical grounding circuit resistance - no more than 2 Ω
Potential between the tip and ground No more than 2 mV
Temperature control accuracy ±2°C (at 25/45%)
Weight 1.5 kg (excluding the soldering iron)
Dimensions 155x130x90mm
Power cord length 1.2 m

It comes in a fairly large, sturdy box.


Most of the different station models are listed on the side:
BK2000
BK2000A
BK3000
BK3200
BK3300
BK3300L

The list is actually longer; for example, there are the BK2000+ and BK2000B. I don't really understand the differences between some of them. Maybe they have different enclosures but the same power, or maybe there's something else.

For example, here are four models with obvious differences:
1. BK2000A 90 W (some sources say the BK2000 is rated at 120 W)
2. BK3300A 150 W
3. BK3500 120 W with a fume extractor and solder feeder
4. BK3600 150 W with a solder feeder

The first model is almost identical to the one I'm reviewing.


This is the BK3200 model. As I mentioned above, it's designed to be a 120 W soldering station.
The operating voltage is listed on the other side of the box, and, surprise... it's 110 V. The store page said 220 V, but we're used to this by now. Let's figure it out.

Everything is packed really well. Inside the main box are smaller boxes containing all the included items.
1. Soldering station
2. Soldering iron
3. Soldering iron stand
4. Wire for connecting the ESD protection
5. Power cord
6. A pile of paperwork.


First, the paperwork. There's quite a lot of it: a manual, a warranty card, and a couple of things I can't identify.


Unfortunately, the manual is entirely in Chinese, but I found a PDF with an English version, which is much easier to understand.
The manual covers quite a lot: specifications, a temperature control graph, how to operate the station, and the tip options available for it.
I don't usually read manuals, especially for simple devices like soldering stations, but this one turned out to be much more complicated. I'll explain why in the section on operating the station.


There's not much to say about the stand or the grounding wire: the stand is completely standard, as is the wire, which has an alligator clip on one end and a plug for connecting to the station on the other.


The second box contained the soldering iron and power cord. Unfortunately, the power cord has a flat-pin plug, but the other end uses a connector familiar to many, so you can use a standard “computer” power cord. They cost next to nothing, and you may already have one at home.


The soldering iron looks different from the well-known “Bacon” 950, since it doesn't use the now-familiar T12 tips.
The cable is flexible and about one meter long. The distance from the tip to the start of the handle is about 65–70 mm. That's quite a bit compared with a T12, of course, but in actual use the difference isn't particularly noticeable.


There's no point in taking a closer look at the stand and the other accessories, but I will take a closer look at the soldering iron, especially since there's plenty to see.
1. The soldering iron is marked LF302. As far as I can tell, it's part of the same series as the LF201/LF202/LF305.
2. The connection is roughly similar to that on the other models, but it uses a much more convenient connector. I'll show it with its matching half a little later to make the idea clear.
3. The tip is protected from damage by a silicone tube.
4. And here’s the “needle” tip. It’s my least favorite, perhaps because of the kind of work I do, but more likely simply because I’m used to working with other tip shapes.


The tips are changed in much the same way as on stations that use 900-series tips and similar ones. Unscrew and remove the front section, then replace the tip itself.


Interestingly, as far as I can tell, the stations are actually tested before being sent to the store. The discoloration on the heating element is evidence of that.
As I mentioned, the tip is a “needle” type. I couldn’t identify its model, but its appearance and, to some extent, its markings match tips for the Quick203 / Quick203H induction stations.


The heating system is also different from the familiar 900-series design, where a ceramic heating element fits into a hole in the tip. Here, things work differently. The tip is heated from the outside (at least in theory), while a thermocouple is inserted inside. The assembly is spring-loaded, so the end of the temperature sensor presses against the inside of the tip, not far from its point.


The soldering iron is very easy to disassemble: unscrew the plastic front section, then slide the handle backward.
The cable is secured with a metal strip, which is attached to the front section by a piece of steel wire with hooks. It’s a simple but quite reliable design. Overall, the soldering iron’s construction is very similar to the Quick 203.


Interestingly, the cable consists of three ordinary wires and a coaxial cable, all inside a common sheath. The heating element is powered through the coaxial cable.


I measured 0.86 Ω for the heating circuit; the manual specifies about 1 Ω. This includes the resistance of the power wires. In this case, that suggests the station really is induction-based, since a conventional heating element with this resistance is unlikely—especially given that the resistance doesn’t change with temperature.


A thermocouple is used as the temperature sensor, as indicated by its low resistance. The manual says the resistance is less than 10 Ω, although talking about the resistance of a thermocouple isn’t really correct, since its job is to produce a voltage that depends on temperature.
During the test, though, I measured the resistance, because due to the measurement method it “drifts” in one direction or the other. When cold, the thermocouple circuit measures about 2.5 Ω; when hot, about 1 or 4 Ω, depending on the polarity of the meter probes. The thermocouple had already cooled down a little by the time I took the photo. I can, of course, measure the voltage produced by the thermocouple if needed.


I’ll definitely come back to the soldering iron later, but for now I’ll talk about the soldering station itself, which is essentially a power supply and a temperature controller.
The front panel is extremely minimalist: three control buttons, a three-digit display, a power switch, and a connector for the soldering iron. Though, to be fair, I often use a station with just two buttons, and it’s perfectly convenient. This one has three, but as it turns out, sometimes two are better than three :)


The case looks quite nice, although it could be a bit smaller. I have some reservations about the finish on the edges of the front panel—it was clearly done by hand—but overall it looks fine.


Above, I mentioned the unusual design of the soldering iron connector. I’ve come across connectors with threaded locking collars before. This one is much simpler and more convenient; let me explain.
To connect it, simply insert the plug into the socket, and the collar locks in place with the help of balls on the instrument side.
To disconnect it, just pull the collar on the plug, and it comes free. It’s a simple and genuinely convenient system.

You could point out that the station has a male connector while the cable has a female one, even though for safety it’s customary to do it the other way around. But there are two small mitigating factors.
1. The soldering iron’s supply voltage isn’t very high (specified as 36 V)
2. When the soldering iron is disconnected, the station displays an error message and cuts power to it.


On the back is the power cord inlet, combined with a fuse holder that also contains a spare fuse.
Above the connector is a socket for a grounding wire or an antistatic wrist strap. Keep in mind, though, that this will only work if the station itself is grounded.
For safety, the connector is wired through a 1 MΩ resistor.

The label also says the station is designed for a 110 V supply. Unfortunately, there’s no getting around a modification here. But the product page says 220.


There are several ventilation holes on the bottom, which are basically unnecessary since there’s hardly any heat to dissipate.
I have a minor gripe with the feet: they’re a little slippery, though it’s not particularly bothersome.


For obvious reasons, there’s no getting around taking it apart. To the manufacturer’s credit, though, they provide photos of the internals as well as the exterior.
The station isn’t particularly easy to disassemble: four screws at the back, four at the front, and then comes the trickiest part—sliding the chassis backward. The controller board itself gets in the way if you try to slide it forward. The wires to the connectors, buttons, and so on also get in the way, but fortunately all the connections are plug-in, so they’re not much of a problem.


I’ll start by looking at the power board, especially since that’s the one we’ll be modifying.
The board is quite large. It carries all the power components, as well as the power supply for the “brains” board.


I’m pretty sure the first thing you noticed was the pair of capacitors wrapped in heat-shrink tubing.
Not long ago, I posted an article about converting a cordless drill charger from 110 V to 220 V. That design used a voltage doubler with two capacitors. This time, things are a little more clever, and at the same time simpler.


The station was originally designed for 220–230 V, while operation from 110–115 V is a modification made by the manufacturer.
As you can see in the photo, the capacitors are soldered in place of the bridge rectifier. Our job will be to restore the original configuration, before the modification. Interestingly, I’ve come across lots of references to this modification online, to the point where it feels like the 220–230 V version doesn’t exist at all.


We’ll come back to the modification later; for now, let’s continue the inspection. The inverter uses IRFP460 transistors, and both even have the same date code. In my power supply reviews, I’ve pointed out transistors from different batches. Here, everything is as it should be.
The transistors are mounted on fairly large heatsinks, especially considering the output power is only 120 W. In my opinion, the inverter was designed for more than 120 W—perhaps about one and a half times as much. And since the soldering iron never draws full power continuously, there’s a huge margin. I really don’t understand why they used such a powerful inverter.


There’s also an auxiliary power supply on the board. It powers the control board, the PWM controller for the power inverter, the driver, and other circuits.
It uses a well-known VIPer-series PWM controller with an integrated high-voltage transistor. This is a fairly reliable circuit, powered from the same high-voltage rail as the power inverter.


Essentially, it’s just a regular power supply designed to provide two voltages.


There’s a connector nearby that carries power to the control board and a control signal from the board to the power inverter.
Interestingly, the control signal passes through an optocoupler (visible on the right). Apparently, the goal was either to match the signal levels or to improve reliability—or perhaps both.
The markings also show that the controller is powered by a dual supply of +5 V and -5 V. The negative voltage is needed to power the op-amp that amplifies the thermocouple signal.


Here’s the inverter’s output stage. There’s no bridge rectifier or filter capacitors, unlike what we saw in the “Bacon” 950 series. In this case, the soldering iron—or, more precisely, the heating element—is powered by AC.
A large inductor and a capacitor are installed at the transformer output. The inductor is in series, and the capacitor is in parallel with the heating element.


To examine the power inverter circuitry, I’ll have to remove the extra capacitors. I had to cut the heat-shrink tubing, then peel it off the glue holding it to the board. If I didn’t need to modify the unit, I’d say that although the manufacturer made a modification, they did a pretty solid job of it.
There are two 120 µF, 450 V capacitors installed, supposedly made by Nichicon. But I’m about 99% sure they’re counterfeit: despite the stated capacitance of 120 µF, their actual capacitance is around 100 µF. In this case, though, it’s absolutely not critical, since there’s also a main filter capacitor. What is notable is that the manufacturer used 450 V capacitors where 200–250 V would have been enough. Apparently, they used whatever was closest.

By the way, similar capacitors are used in power supplies for modern monitors because of their small diameter, so I decided to desolder them and keep them as spares.


Beneath the capacitors were two transformers: a current transformer for protecting the inverter against overload, and a driver transformer for controlling the inverter transistors.
Even lower down, under a layer of glue, I found an unused connector marked AC OUT. Why it says "OUT" remains a mystery to me, but I strongly suspect this board is also used in a version with solder feed, with the feeder connected to this connector.


The power inverter is built around the rather interesting UCC3895 PWM controller. Its full name is BiCMOS Advanced Phase-Shift PWM Controller. For more information, here's a link to the datasheet. This is the first time I've come across this controller in a finished device.

Next to it is a TC4424, a driver for controlling MOSFETs. Here's a link to the datasheet. The driver isn't connected directly to the PWM controller; an HEF40818 is used in between. At first I thought it was simply acting as a buffer or level shifter, but it turned out to be a little different. The PWM controller is designed to control a bridge, so it has four outputs, while our inverter is a half-bridge with just two transistors. Apparently, the manufacturer had a particular reason for choosing this PWM controller, and to use it in a half-bridge inverter, they combined pairs of outputs using AND gates, which is what the HEF40818 contains.


A photo from another angle, which may be useful.


There are no components on the underside. The bridge rectifier you see in the photo isn't installed quite as intended.
The soldering is good, but I do have a few comments; I'll show them in a little more detail later.


I didn't redraw the schematic—it would have taken far too long and probably wouldn't have been very useful. But I did put together a rough block diagram; I think it gives you a general idea of what goes where.


But let's get back to adapting it for our mains voltage. Though really, this is more about restoring the original power supply circuit than modifying it.
The bridge rectifier was moved to the underside of the board, and it's used not as a bridge but simply as two diodes. This is needed for the voltage doubler circuit.

By the way, while I was poking around, I noticed the power connector. As you can see, it originally had three wires, but the middle one was cut off. The middle contact went to the junction of the EMI suppression capacitors. These are the capacitors that make device enclosures give you a tingle if there's no ground connection. I didn't restore this; I left it as is. But don't worry: the soldering iron tip and the rear grounding terminal are connected to the ground contact on the station's power connector.


If you desolder the extra capacitors, you'll see the thick leads of the bridge rectifier sticking up (I wasn't the one who scorched the capacitor with a soldering iron).
I didn't desolder the bridge rectifier; I simply clipped its leads. I desoldered the leads sticking out of the board, then cleaned out the holes.


After that, I installed the bridge rectifier back in its original position, following the markings on the board. That's essentially the whole modification.


While I was at it, I also replaced the filter capacitor. The original was an unknown brand with an actual capacitance of 84 µF; I replaced it with a name-brand one measuring 100 µF. I did this for two reasons:
1. The name-brand capacitor should be more reliable.
2. Since I removed the extra capacitors, the filter's total capacitance went down. It had been 84+(100/2)=134 µF. Now it's 100 µF, which is actually more than enough.


Speaking of soldering quality: the marks near the capacitor leads are mine, but the ones near the current transformer were already there.
There are also a few minor soldering flaws around the larger components, though otherwise I have no complaints.


My biggest concern was the soldering on the inverter's output inductor. It's a heavy component (though it's secured with glue), and the solder joints aren't very good. To be safe, I touched them up and washed the board afterward.


After all the work, the board looked like this.
I glued the new filter capacitor in place with silicone. Personally, I think the manufacturer could easily have installed it vertically; there's plenty of room.


The control board is probably the shortest part of this review.
This board is mounted parallel to the front panel. It carries the control buttons, display, and microcontroller.


Everything is controlled by an Atmega8A microcontroller, one of the most popular microcontrollers.
The thermocouple signal is amplified by an OP07C op-amp made by Texas Instruments.


The station can display error messages. There are only two, though: a thermocouple error and a heater error.


I first powered it up without the soldering iron connected, so, as expected, I got an error message. It didn't look like the one in the manual, though. I figured out the problem a little later: I had received a station with a faulty display. The middle segment of the third digit doesn't light up. Worse still, it turned out to draw more current from the processor output than the other segments. I tried to burn it out with a higher current, but only made things worse.
The display works in a way that uses three resistors, one for each digit, rather than eight resistors for each segment. When the middle segment is on, the brightness of the entire third digit drops. Anyway, I might replace the display, although it doesn't get in the way of using the station. Still, it's a defect.

Yes, powering it up without the soldering iron connected brings up a thermocouple error. If you connect the iron and restart the station, it starts up set to 350 degrees (the default value).


Then things got a lot more interesting. No matter what I tried, I couldn't adjust the temperature. I tried holding the keys in different combinations, but nothing happened. It felt as if the keypad was completely disconnected. I even started checking the buttons with a multimeter.
But it turned out to be much simpler and more obvious: people usually read the manual last, when they really ought to read it first.

I started reading the manual and came across a section about entering a password and changing it to the default password. Honestly, I got stuck at that point. A password? On a soldering station?


But never mind the password—how are you supposed to change the temperature? So I kept reading. I read it, thought I'd completely forgotten English, then read it again. No, I think I still remember it.
I have no idea what the developers were smoking, but I suspect they weren't just smoking it—they were washing it down with mushrooms.

Try to wrap your head around how to change the station's temperature:
1. Turn off the power.
2. Hold down the two outer buttons and turn the power on.
3. When the letter C appears on the display, press the middle button.
4. Enter the password! (three zeros by default)
5. When oI appears, use the up/down buttons to change the temperature.
6. Set the temperature, then turn off the power.
7. Turn the power back on; the station will start at the set temperature.

Maybe I just didn't figure it out, but I still couldn't work out how to change the temperature during normal operation. There are no presets at all, and the keypad doesn't work while the station is running. At all.



Now the same thing, but with an example.
1. Hold down the outer buttons and turn the station on.
2. Enter the password. The left/right buttons change the value; the middle button confirms it.
3. Enter four zeros. By the way, to reset the password, turn the station on while holding down all three buttons.
4. You get to the temperature setting menu.


1. Set the temperature.
2, 3. Press the middle button again to enter the temperature correction menu, where you can set a positive or negative offset. The photo shows -1 or +7. My unit was set to -3 by default.
4. Turn off the power, then turn it back on to get the set temperature.

I've used various soldering stations, some inexpensive, some basic, but I've never come across such a ridiculous control scheme.
There's a 0/1 setting. I thought it disabled the password, but no—as far as I can tell, it enables or disables automatic standby mode (I could be wrong).


All right, let's leave the clever controls to the developers and move on to the tests.
The specs claim a maximum power of 120 W. I tried turning the station on with the soldering iron cold, and peak power consumption was 109 W. Once it warmed up, consumption dropped to 2.2 W. So the actual power is nowhere near 120 W—it's no more than 105 W.


I was curious about what was happening at the station's inverter output, so I decided to measure the voltage and frequency.
The voltage varies from 30 to 55 V, then drops to zero once heating is complete.
The frequency also varies, reaching a maximum of about 160 kHz. According to the PWM controller's datasheet, it can't exceed 250 kHz (500/2), so those figures seem quite plausible. In general, I always thought induction stations operated at much higher frequencies—at least five times higher.


Out of curiosity, I also measured the heater's inductance with and without the tip. It was 9.6 and 9.0 µH, respectively.
Now I'm not even sure whether to call this an induction station, or whether it simply uses a heater and a switching power supply with AC output. I think the latter is more likely.


Let's discuss in the comments whether this is an induction station or not. For now, I'll move on to the actual tests.
First, temperature.
During testing, I found that the station holds its temperature well and really does reach the set 350 degrees. The reading in the photo is a little low because the thermocouple wasn't pressed firmly enough; with a bit more pressure, the tester showed 248–250.


I also plotted the warm-up and temperature regulation over time. The graph covers the first 2 minutes after startup: the temperature initially rises to 370 degrees, then drops quickly and starts holding steady. Here, too, the thermocouple was pressed down as in the previous photo, so the readings are low. I tried to press it down to get a more accurate graph, but even when I held it in place for a while, it first showed 350 and then dropped to 330–340. It's really awkward to press a round object against another round object :(


But tests are one thing; real-world use is much more interesting. I actually got to use the station for a real job: replacing capacitors in a server power supply with a double-sided board and large copper planes.

For this review, I took a couple of boards and decided to experiment on them. First I desoldered the capacitors, then got carried away and desoldered the transformer as well.


It went much the same way with the motherboard. I didn't run into any particular difficulties desoldering the capacitors, and I also desoldered an inductor, which is much harder to heat up. I can say for sure that it's noticeably easier to work with than even a modified Bakon 950. Desoldering is simple: add a little regular solder to dilute the lead-free solder, then the part comes out fairly easily. If the leads are close together, heat both at once and it's even easier to desolder.
The job was made much harder by the extremely awkward needle-style tip, which is poorly suited to this kind of work.


To demonstrate, I recorded a short video. First I desolder capacitors from a monitor board, then from a motherboard. Just so you know, the camera tripod made things a little awkward here. I wanted to get the board in the frame, so I had to solder at an angle :)


So, what can I say in conclusion? I won't list the pros and cons point by point this time; I'll just describe my impressions.
First of all, the design and overall build quality. I'd give it a four out of five, maybe a four-plus, since there are a few minor issues. For example, the output inductor is poorly soldered, some joints look messy, and there are traces of flux.
But the overall build quality is quite good, as you can see in the photos. For a Chinese device, I'd even say it's pretty decent.
The one thing that put a bit of a damper on the overall impression was having to convert the station from 110 V to 220 V, though it wasn't difficult.

Ease of use. The station solders quite well, and you can really feel that it has power to spare. But the included tip is very awkward for working with components that have a high thermal mass; they shouldn't have included it. The cable to the iron is flexible, and the iron itself is lightweight. Even the fact that the tip sticks out much farther than on T12 models—and especially compared with my Aoyue 2738, whose tip has very little overhang—doesn't make it uncomfortable to use. The station also does a good job of maintaining the set temperature.

As for the controls, I'd give them a two out of five. Honestly, I'd give them a one, but you can operate the station, even if the process is a complete pain. Why on earth would you make temperature adjustment require restarting the station and entering a password?

Of everything on the list, my biggest complaint is the controls. Converting it from 110 V to 220 V is no problem for a radio amateur, and the poor soldering on the power board is easy to fix too. But maybe I was just "lucky," as I was with the defective display.
Speaking of the display, I forgot to mention it earlier. On the plus side, temperature correction was enabled at the factory. But how did the inspector miss the display defect?

As I understand it, Bakon set out to make an equivalent of the Quick 203 station, and almost succeeded—if not for the price.


That's all for now. As usual, I look forward to your questions and comments. I hope you found the review useful.