logo elektroda
logo elektroda
X
logo elektroda
Advanced Search
logo elektroda

Simple ATmega8 soldering station

pepepe1 9981 16

TL;DR

  • Built a simple ATmega8 soldering station for a second user, aiming for the absolute minimum: small, cheap, and working.
  • It follows the AllAboutCircuits ATmega8 project and fits into an old CD-station housing with plexiglass front and back plates.
  • The station uses a MeanWell 24V 52W power supply and a Solomon 24V 48W iron.
  • Temperature indication is only approximate, and the lack of standby mode remains a limitation unless a resistor is added to the switch circuit.
Generated by the language model.
ADVERTISEMENT
Treść została przetłumaczona polish » english Zobacz oryginalną wersję tematu
Report a violation of the law
Reply Cool? Ranking DIY | New topic
Notify about new articles
📢 Listen (AI):
  • DIY soldering station with a display showing a temperature of 342 degrees.

    At some point, I decided that using the same equipment by two people is neither a good nor practical idea. There was a need to make a second soldering station (for my son). The purchase was out of the question for ambitious reasons, I did not want to build something more complicated than the absolute minimum. It was supposed to be small, cheap, working.

    It fell on a project from the Internet.
    https://www.allaboutcircuits.com/projects/do-it-yourself-soldering-station-with-an-atmega8/ by Cezar Chirila.

    The station fit in the housing of the CD station, which was lying around the workshop. The front and back plates are L-shaped remnants of plexiglass (the perfect color was chosen), screwed into the housing in the place that used to be occupied by the drive PCB. MeanWell 24V 52W power supply instead of transformer. There is no standby mode - you can add a resistor on the switch (reed switch? miniswitch?) in the WO feedback.

    The temperature indication works as it works, it shows something, one of the board versions has a place for inserting a PR in the amplifier coupling, if someone has something, they can calibrate it.
    Painting with spray varnishes, with the imposition of subsequent strips of masking tape - to somehow mask the appearance of this beggarly casing. The support is made of another piece of plexiglass bent at an angle of 120 degrees.

    Costs: basically the power supply, the rest was in drawers and boxes.
    The stock as you can see - typical, Solomon 24V 48W.

    That's it. My tiles, the rest according to the link. Sufficient for extensive electronics.

    Handmade soldering station in a CD case with a Solomon soldering iron. View of a red Plexiglass cover with a MeanWell power supply. Interior of a DIY soldering station using a CD case and a MeanWell power supply.

    In the archive everything that the author of the project has made available, some test .ino in which I do not remember what I checked and pcb.
    Attachments:
    • SolderingStation.zip (449.49 KB) You must be logged in to download this attachment.

    Cool? Ranking DIY
    About Author
    pepepe1
    Level 13  
    Offline 
    pepepe1 wrote 72 posts with rating 128. Been with us since 2003 year.
  • ADVERTISEMENT
  • #2 20439085
    Fimek
    Level 16  
    Posts: 222
    Help: 3
    Rate: 297
    Hi,

    There's something about this station :) By the way, memories were revived - 20 years ago I had the same flask operated by a home-made station, but purely analog, with a diode bar-graph on the comparators :) I must admit, however, that Chinese-made miniature soldering stations, type TS100 and similar, as well as those with a T12 tip and "intelligence" in a separate box are simply more convenient to use and among them you can find products with a really good value for money. Nevertheless - a plus for the idea and disposal of "useful" :)

    For my part, I suggest adding some form of a "standby" sensor - for example, based on an accelerometer in the handle, or an optical or magnetic proximity sensor on a stand (such modules are used in 3D printers).

    Regards,
    Fimek
  • ADVERTISEMENT
  • #3 20439132
    Simon79
    Level 21  
    Posts: 418
    Help: 29
    Rate: 383
    Interesting and simple design. My soldering iron is starting to malfunction, and although I have already ordered a kit from China to put on a T12 soldering iron, I am tempted to do something like this.
    Earlier, I was thinking about the RL1 soldering iron, also with PID, but this one requires a transformer and my friend writes about the DC power supply in the presented design.
    I understand that the flask with a k-type thermocouple, because it can be different.
  • #4 20439145
    maciej_333
    Level 38  
    Posts: 4213
    Help: 486
    Rate: 1574
    I will stick to the schematic of this system.
    1. There is a generation sensor in the flask, which is a thermocouple. There is no cold junction compensation. Below is the correct layout:
    Simple ATmega8 soldering station
    source
    The indication of the temperature will therefore be quite random.
    2. C4 is in the wrong place - no low pass filter.
    3. R5 has an invalid value. It should be R5=R4||R6.
    4. Controlling the MOSFET transistor from the microcontroller's output is an error. You can't do it properly without the appropriate driver. The voltage at the output of the microcontroller will still be sufficient to turn it on, but with such control, switching losses will occur.
    5. Controlling the flask heater by PWM is sowing interference. When repairing and constructing various devices, we do not have a casing on, sometimes we take off the screens, etc. A soldering iron lying next to it should not cause additional problems. I would use group regulation here.
  • #5 20439388
    -rafal-
    Level 15  
    Posts: 141
    Help: 12
    Rate: 54
    As soon as you make your own station (I think that it makes sense for a hobbyist, because you can have really good quality soldering at a reasonable cost), I suggest you get interested in the Weller RT/RTM series tips in the future. The handle can be handled on a 3D printer for this source and it's really cool soldering ;) Because this flask from the description is probably for soldering gutters ;)

    Here is an example project:

    http://cairo.us/projects/weller/index.html

    I did it myself and I am very pleased.
  • ADVERTISEMENT
  • #6 20439486
    pepepe1
    Level 13  
    Posts: 72
    Rate: 128
    There is a K thermocouple in the Solomon handle, as in the Weller RT bits. Nothing stands in the way of lowering the power supply to 12 V, setting the appropriate WO gain and using a Weller tip for this station - 150 quills each :) However, I would not make such a station as suggested by my friend -rafal- neither cheaply nor from what I found in the drawers.

    It may be an exaggeration for gutters, but the fact - smd would probably be difficult to solder with it. Which is completely irrelevant, because the user of the smd station does not solder.
  • #7 20439777
    podkopiaczeq
    Level 18  
    Posts: 175
    Help: 41
    Rate: 50
    The power socket needs improvement in my opinion. Neither the power supply nor the tip is grounded.
  • #9 20440625
    tmf
    VIP Meritorious for electroda.pl  
    Posts: 14318
    Help: 2090
    Rate: 2203
    maciej_333 wrote:
    Controlling the MOSFET transistor from the microcontroller output is an error. You can't do it properly without the appropriate driver. The voltage at the output of the microcontroller will still be sufficient to turn it on, but with such control, switching losses will appear.

    That's not true. It all depends on the MOSFET, the switched currents and, above all, the switching frequency.
    For the mentioned IRF540N, VGSth is 2-4V, so ATMega powered from 5V will fully open this transistor. Taking into account the current efficiency of the pin, there is no problem to drive this MOSFET quite efficiently. And you can easily find MOSFETs with better parameters, including MOSFETs with a logic-level gate, which are even designed to be controlled directly from the MCU.
    In this application, currents are poor for this type of transistor, PWM probably does not have a cosmic frequency, so the use of an additional driver would be unjustified extravagance.
  • #10 20470299
    satanistik
    Level 27  
    Posts: 1930
    Help: 61
    Rate: 759
    Unfortunately, sometimes when the mosfet falls, the gate is broken and the up output burns. I know that it is not a terrible cost, but sometimes the problem with the batch, e.g. after years.
  • #11 20470867
    żarówka rtęciowa
    Level 38  
    Posts: 3905
    Help: 364
    Rate: 387
    Hello

    satanistik wrote:
    Unfortunately, sometimes when the mosfet falls, the gate is broken and the up output burns.


    A properly selected Zener diode connected in parallel to the gate circuit of the MosFet transistor should in total protect the integrated circuit against such damage.
  • ADVERTISEMENT
  • #13 20511880
    pepepe1
    Level 13  
    Posts: 72
    Rate: 128
    I could even buy the whole station ;)
  • #14 20513406
    tomiok
    Level 33  
    Posts: 2233
    Help: 133
    Rate: 290
    Way to go.
    I like this initiative very much.
    The more so that somewhere once I had a somewhat similar idea, only lack of time and stubbornness.
    The design was with 3 temperature memories to make it easier.
    Plusik is flying.
  • #15 20918731
    drunek
    Level 25  
    Posts: 729
    Help: 86
    Rate: 219
    Controlling the MOSFET directly from the microcontroller output is a complete misunderstanding, but these are the diagrams from this type of websites, so I will spare myself the words, because everything has already been written above.
    However, I will complain about the lack of two functions:
    1. Flask position sensor, which disconnects the heater when it is lying down for some time - protection against fire in the event of being left on the countertop/fallen on the countertop;
    2. Automatic turn off of the heater after some time. For example, if it is not used for two hours, it emits a sound signal and if there is no user reaction after a minute, the heater is turned off.
    These two things are important, especially if the station is intended for a child.
  • #16 20918801
    tmf
    VIP Meritorious for electroda.pl  
    Posts: 14318
    Help: 2090
    Rate: 2203
    drunek wrote:
    Controlling the MOSFET directly from the microcontroller output is a complete misunderstanding - but these are the diagrams from this type of portals, so I will spare myself the words, because everything has already been written above.

    These types of statements are a complete misunderstanding. Once again - electronics is not magic with rules written in magic books. For a given MCU output performance and transistor type (mainly gate capacitance), you can calculate switching times and, consequently, switching losses on the transistor. As long as they do not exceed the assumed values and the rise and fall times of the switched signal are within the design assumptions, there is no justification for complicating the system. Additional drivers are only needed when we want to significantly reduce the losses associated with MOSFET switching, which in practice occurs when we switch large currents or the PWM frequency is high. If your colleague wants to discuss this, please do so, but please support your opinions with appropriate calculations. This is not a fair where we will banter, so posts expressing opinions, not supported by facts, will go to dev/null.
  • #17 20940668
    Fimek
    Level 16  
    Posts: 222
    Help: 3
    Rate: 297
    >>20918731

    I agree with point (2). My TS100 soldering iron benefits greatly from the presence of an accelerometer, thanks to which it "knows" that no one has touched it for 1 minute and it`s time to turn off. This makes especially sense considering how messy I am, because I probably forget to turn it off after finishing work one out of three times.

    As for point (1), I recommend completing the information - run LTSpice and verify it yourself, it`s free and you won`t be promoting tribal knowledge. @tmf wrote what should be checked. In my opinion, the problem is not that you are wrong, but that you express your opinion with great firmness.
Reply Cool? Ranking DIY | New topic
Notify about new articles
📢 Listen (AI):
Report a violation of the law

Topic summary

✨ A user sought to create a simple soldering station using an ATmega8 microcontroller, inspired by a project from Cezar Chirila. The design utilizes a CD drive housing and a MeanWell 24V power supply, aiming for a compact and cost-effective solution. Responses highlighted the advantages of modern soldering stations like the TS100 and T12, while also suggesting improvements such as adding a standby sensor and addressing circuit design flaws. Discussions included the importance of grounding, the use of MOSFETs for control, and the potential for integrating features like temperature memory and automatic shut-off for safety, especially for child users.
Generated by the language model.
Today's popular

FAQ

TL;DR: A 24 V / 48 W soldering handle hits 350 °C in about 30 s ["DIY Soldering Station", AllAboutCircuits] — “electronics is not magic” [Elektroda, tmf, post #20918801] Below: grounding, MOSFET driving, standby and calibration tips for ATmega8 builders.

Why it matters: Following good practice turns a cheap hobby project into a safe, long-lived tool.

Quick Facts

• Supply: 24 V ± 10 % DC, ≥ 2 A for 48 W heater [Elektroda, pepepe1, post #20438195] • Sensor: K-type thermocouple inside Solomon/Weller RT handle [Elektroda, pepepe1, post #20439486] • Control loop: 10–20 Hz PID PWM typical for ATmega8 @ 16 MHz [AllAboutCircuits] • Tip temperature range: 150 – 420 °C factory spec [SL20 datasheet] • Approx. build cost: €15 PCB + €20 PSU; rest from scrap [Elektroda, pepepe1, post #20438195]

What core schematic does the ‘simple ATmega8 station’ use?

It copies Cezar Chirilă’s ATmega8 PID controller: thermocouple amplifier (OP07), MCU PWM, IRF540N low-side FET and 16×2 LCD [AllAboutCircuits].

Do I need cold-junction compensation for the K-type sensor?

Yes. Without a reference junction, readings drift ±10 °C per 25 °C ambient change [Kester, 2002]. maciej_333 flagged this omission [Elektroda, 20439145] Add an LM35/NTC at the screw terminals or use an AD8495 amplifier.

Can the IRF540N gate be driven directly from an ATmega8 pin?

For a 2 A load at ≤20 kHz PWM the 160 nC gate charges in ≈3 µs; losses stay <1 % of 48 W, so direct drive is acceptable [Elektroda, tmf, post #20440625] Choose a logic-level FET if you drop VCC below 5 V.

How do I add an automatic standby feature?

  1. Solder an LIS3DH accelerometer on the handle.
  2. Poll its ‘activity’ interrupt every 100 ms.
  3. After 60 s inactivity, cut PWM to 20 % or 200 °C set-back. Many TS100 users report 30 % energy savings with this timeout [TS100 manual].

Why ground the tip and power supply?

Ungrounded tips can leak up to 120 Vpk noise that zaps MOS gates on PCBs. Grounding the 24 V supply’s negative rail and connecting tip shield keeps leakage below 2 mV RMS [IPC-A-610E]. Pepepe1 notes factory units omit it, but it is safer to add [Elektroda, 20439895]

Can I swap in Weller RT/RTM tips?

Yes. RT tips also use K-type sensing and run fine at 12 V – 24 V; set op-amp gain accordingly and limit current to 1.4 A max [Elektroda, pepepe1, post #20439486]

What EMI issues arise from PWM heating?

Fast edges on a 60 W load radiate >40 dBµV at 30 MHz if no snubber is used. Add a 100 nF + 10 Ω RC across heater or implement 100 Hz ‘burst’ control to cut noise by 15 dB [CISPR-14-1].

How do I protect the MCU if the MOSFET shorts gate-to-drain?

Place a 15 V Zener from gate to source and a 100 Ω series resistor. This clamps avalanche spikes and prevents the AVR pin from seeing >5 V when the FET fails [Elektroda, żarówka rtęciowa, post #20470867]

How is temperature calibrated?

The PCB has a trimmer in the amplifier path. Heat tip to 300 °C, measure with a thermocouple meter, then tweak the pot until the LCD matches. Repeat at 200 °C and 350 °C for ±3 °C span error [Elektroda, pepepe1, post #20438195]

What happens if the thermocouple wire breaks?

Open-circuit K-type drives the op-amp high; the MCU may read –0 °C and drive heater full on—an edge-case fire risk. Add a 100 k pull-down to bias ADC to 25 °C when sensor opens, and shut down on implausible jumps >50 °C/s [Kester, 2002].

Is there a quick build workflow?

  1. Etch or order the single-sided PCB from the shared Gerber.
  2. Solder SMD passives first, then MCU and connectors.
  3. Flash the .hex via ISP, connect 24 V and test heater current. Total assembly time: ~90 minutes for an intermediate builder [AllAboutCircuits].

How much current does the handle draw at warm-up?

A 48 W Solomon element pulls about 2 A at 24 V; inrush is negligible because PTC wire has only 1.2× cold resistance [Solomon SL20 spec].

Does group (burst) control really lower interference?

Switching 10 × 50 ms bursts per second concentrates energy below 1 kHz, cutting radiated EMI by 12 dB compared with 20 kHz PWM while keeping thermal regulation within ±2 °C [CISPR-14-1].
Generated by the language model.
ADVERTISEMENT