last updated: 2024-03-22
In a video I saw, bitluni solder SMD components on a preheating station. I've tested an reflow oven (https://www.weigu.lu/other_projects/baking_smd/index.html), but for small pieces or to unsolder things I thought such a preheating station would be cool (or hot). When buying the sation I did not pay attention and bougt a 110 V station instead of a 220 V station. So I ordered a second station and will hack the first (or both) to get a reflow solder station :).
As Denis Bodor (Hackable Magazine) wrote, there is no better thing to do during a WE than hacking or reverse engeneering a device :).
So I have 2 UYUE 946-1010 preheating station. Both station are labeld with 400W. By opening the stations I found:
110 V Station:
3 heating elements (8 mmx60 mm) 110 V/100 W
R = 110 V²/100 W = 121 Ω, Rtotal = 40 Ω
Measured R = 40.3 Ω
Hotplate 10x10x 8.3mm³
220 V Station:
2 heating elements (8 mmx60 mm) 220 V/150 W
R = 220 V²/150 W = 323 Ω, Rtotal = 161 Ω
Measured R = 155 Ω
Hotplate 10x10x 6.1mm³
Both station have the same PCB! labeled YYUVDJS-220V V1.0.
I could use the 110 V station with 220 V and a max 25% PWM, but this is a little risky, so I ordered 3 heating elements 220 V/100 W
By looking at the components I found a BTA16-800B Triac with an EL3083 zero cross TRIAC driver to switch the heating elements. A chip labeled SZ2525C (7-Pin) is an AC/DC converter like an LNK304GN or similar. Here I couldn't find a datasheet, but the circuit shows that we produce a small DC voltage that is followed by an classical 78L05 to produce 5 V.
A 5 pin header (not soldered) is connected to the 20 pin microcontroller. The µC is not labeled but after a little research I think it is an STM8103F3. I tried to read the firmware (windows with STM8 programmer), but was not successful. I don't like these chips because we don't have all the tools under Linux, so I will replace it with another chip.
The temperature sensor is an K thermocouple type (specs of the station) and the small voltage is amplified with an LM321 opamp.
The 3 digit 7 segment display has his series resistors in the common anode circuit wich is not professional because we get different luminosity for different digits.
And here is the reverse engineered circuit :)
Then I measured the voltage on pin PA1 (5) by changing the temperature
Temperature [°C] | Voltage [mV] | Temperature [°C] | Voltage [mV] | Temperature [°C] | Voltage [mV] |
---|---|---|---|---|---|
110 | 663 | 210 | 990 | ||
20 | 378 | 120 | 698 | 220 | 1013 |
30 | 401 | 130 | 732 | 230 | 1045 |
40 | 431 | 140 | 763 | 240 | 1080 |
50 | 464 | 150 | 797 | 250 | 1111 |
60 | 498 | 160 | 831 | 260 | 1143 |
70 | 531 | 170 | 856 | 270 | 1176 |
80 | 565 | 180 | 889 | 280 | 1207 |
90 | 600 | 190 | 928 | 290 | 1243 |
100 | 630 | 200 | 958 | 300 | 1274 |
Ok we got a linearity with about 3.2 mV/°C. After this I measered the PWM on pin PA3 (10) without heating. The ambient temperature is about 20°C:
Target temp. [°C] | Duty cycle [%] | Target temp. [°C] | Duty cycle [%] | Target temp. [°C] | Duty cycle [%] |
---|---|---|---|---|---|
20 | 0 | 30 | 40 | 40 | 71 |
21 | 11 | ||||
22 | 11 | 32 | 44 | ||
23 | 17 | ||||
24 | 21 | 34 | 51 | ||
25 | 24 | 45 | 87 | ||
26 | 27 | 36 | 58 | ||
27 | 30 | ||||
28 | 33 | 38 | 64 | 48 | 98 |
29 | 37 | 49 | 100 |
This can give us a hint for the PID regulation. Duty cycle increases about 3%/°C.
Everything is on github: https://github.com/weigu1/lora2mqtt_ethernet_gw.