Today’s post is mostly in black and white as we’ll take a look at the circuit of the 858D hot air rework station. Almost every part of the circuit is on the PCB, except for the mains switch, fuse and transformer. The mains voltage is switched by a single pole, single throw (spst) switch in the hot wire. It goes through a fuse and is connected to the PCB. On the PCB the mains voltage is connected to the transformer primary wires. This transformer has two secondary windings: a 10VAC for the low voltage supply and a 25VAC for the fan driver.
The wires connecting to the hand piece are connected directly to the circuit on the PCB through several connectors.
There are five parts to the circuit on the PCB of the 858D. These are:
- The mains powered heater driver circuit
- The low voltage power supply
- The temperature sensor amplifier circuit
- The fan driver
- The microcontroller with user interface
Let’s take a closer look at these circuits now.
- All images are clickable for a larger view.
- The complete schematic is attached as a PDF below this post.
Mains powered heater driver circuit
The heater inside the hand piece is mains powered. The triac Q1 is used as a switch to enable current to flow through the heater for a certain period. Q1 is driven by U2, a triac driver optocoupler. When the small triac inside the optocoupler is switched on, current flows though R23 and in/out the gate of Q1. R24 is used to drain the internal capacitances of Q1 when the triac is in the off-state. This prevents currents flowing inside the triac and possibly switching it on. R25 and C15 form a snubber circuit.
Inside the optocoupler is a zero-cross detection circuit. So it will only start conducting at the next zero-crossing of the mains AC voltage, after the low-voltage input is switched on. The low-voltage side is connected through R8 to an I/O pin of the microcontroller.
C20 is used as filter capacitor across the heater. R28 is the bleeder resistor to discharge C20 when the mains is switched off.
Low voltage power supply
This circuit is a straightforward dual rail linear power supply. It is powered from the 10VAC secondary of the transformer. The positive rail gets half-wave rectified through diode D5. The linear regulator U6 turns it into a stable 5V rail. D7 provides protection on the regulator in case the output voltage rises above the input voltage by accident or fault.
The negative rail is also half-wave rectified, using D6. There is no regulation, only a bulk/smoothing capacitor C11. An additional low pass filter is created with R22 and C6 (C6 is drawn near the op-amp in the temperature sensor circuit). The negative voltage rail is only used for this op-amp.
Temperature sensor amplifier
Inside the heater is a K-type thermocouple. The thermocouple is soldered to normal copper wires inside the hand piece, but there is no cold joint temperature measurement and/or compensation. The voltage created by the thermocouple is in the order of 10’s of µV to 100’s of mV. This needs to be amplified before the ADC can get a decent resolution. This is where the op-amp circuit comes in. The thermocouple signal is put through low-pass filter R9/C1 before entering the op-amp
The 6.8MΩ resistor R1 is a bit strange. The value is really large for one. Secondly a thermocouple doesn’t need a current drive, so it isn’t needed. Still it does create a voltage divider together with R9 and the thermocouple’s internal resistance of a few Ohms. This puts the op-amp’s non-inverting input at about 35µV + the thermocouple voltage. I have no idea why the designer’s chose to do it this way.
Note (December 16, 2016)
As par bill156+’s comment below it looks like the reason for R1 is fault protection. In the absence of the thermocouple it pulls the input to the positive voltage rail. The microcontroller should detect this and shut down the heater.
The circuit around op-amp U1 is a standard non-inverting amplifier. Trimmer pot P2 and resistor R18 form the feedback resistance. R11 the resistance to ground. Diodes D8 and D9 protect the op-amp’s inputs: they can never be more than a diode drop apart.
The op-amp itself is a Texas Instruments OP-07 Low Offset, Low Drift Operational Amplifier. What surprises me is that they use such a part, but don’t provide a stable regulated negative voltage rail. They also don’t use its capability to null the input offset voltage. The PCB layout doesn’t give any reason to suspect the designers had a clue how to use such a part. I wouldn’t be surprised if you could replace it with a plain TL061 without any significant change in the circuit’s behavior.
R19, C9 and form a low-pass T-filter. This is the anti-aliasing filter in front of the MCU’s ADC. R21 is used to limit the current flowing into the microcontroller pin in case of a circuit fault.
Fan driver and control circuit
The fan is powered from the 25V secondary of the transformer. Diodes D1 to D4 form a full-wave bridge rectifier. C3 is the bulk / smoothing cap. The DC voltage will be a bit below 36V.
The fan driver itself is just transistor Q2. Its base is connected to the positive rail through R5. Without the rest of the circuit this would put Q2 right into saturation. The fan is then driven with at full current and voltage.
The microcontroller can switch the fan completely off using Q7. When Q7 is on, it diverts the base current of Q2 to ground. Q2 is switched off and so is the fan.
The speed control is a simple, but clever analogue circuit around Q8. The idea is that Q8 can divert part of the base current for Q2 away to ground. The more Q8 conducts, the less Q2 conducts and the slower the fan speed.
The emitter of Q8 is connected to zener diode Z1. This zener created a steady voltage reference with its current limited by resistor R4. The base of Q8 now always sits one diode drop above this reference voltage. At least, that is the idea. Resistors R20, R6 and the fan speed dial potmeter P1 are a voltage divider. The voltage at the taper of the potmeter depend both on its setting and the actual voltage on the fan (emitter of Q2). If the fan voltage rises, and its speed increases, the voltage on the potmeter taper increases. The taper is connected to the base of Q8 and so the base voltage rises. This drives a current into the base of Q8 and is starts to conduct. This lowers the base current of Q2, which conducts less. In turn the fan voltage decreases again. The resistive voltage divider and Q8 form a feedback loop that control the conduction of Q2, and so the fan speed.
If you increase the potmeter’s resistance between taper and ground, a larger voltage occurs at the base of Q2 for any given fan voltage. So in turn the fan voltage and speed must be lower to keep Q2’s base one diode drop above the zener reference voltage. The opposite is also true of course: lowering the potmeters resistance between taper and ground increases the fan speed.
Between the driver Q2 and the fan the voltage is stabilized by capacitors C4 and C12. The microcontroller senses the fan voltage through resistors R29 and R30 followed by filter C24 and R3.
When you look at the fan’s specifications you’ll see that it actually a 24VDC fan. Yet the rectified 25VAC from the transformer gets to about 35V. How does the fan survive this? Well, this all comes down to the zener voltage and the resistors in the voltage divider of the speed control. I haven’t been able to determine the zener voltage yet, but I’m sure that even at maximum speed, the fan voltage will not exceed 24V.
This also means that at about 11V is dropped across the driver transistor at full speed and full fan current. The sticker on the fan in my unit says a maximum current of ‘0.35A’. If the fan does indeed draw this full current at full voltage, the driver transistor must dissipate 3.85W. That’s a 2nd heater circuit right there!
Microcontroller, switches and LED display
This circuit is based on the Youyue 858D which features an Atmel ATmega8 microcontroller. It has a separate analogue voltage reference input for its ADC. The reference voltage is created by U4, which is a T431 voltage reference. The voltage rail is filtered through R13, R14, C13 and C22 to create a stable reference for the ADC.
The ‘up’ and ‘down’ switches are directly connected to the microcontroller’s I/O pins. The same applies to the 7-segment LED display.
There is a RC filter shown, going nowhere. On the Youyue’s PCB this sits next to an unpopulated connector. It’s drawn in for completeness.