T12 soldering station

[RoGeorge]

Many cheap soldering iron handles and 900M soldering tips, a few T12 cartridges, plenty of parts from the junk box, and a nice 220/2x(20+4) Vac transformer.

Former tests (for 900M only):
https://rogeorge.wordpress.com/2016/09/23/zero-parts-thermostated-soldering-station/

The goal, this:



 
(but with T12 this time)

[RoGeorge]

The T12 heating element together with the wires from the handle has 8.69  at room temperature.

The measured DC voltage generated by the thermocouple sensor inside the T12 when the tip was heated by an external hot air gun:

Code: [Select]

0.360mV at  50*C
0.950mV at 100*C
1.453mV at 150*C
2.252mV at 200*C
2.960mV at 250*C
3.578mV at 300*C
4.058mV at 350*C
4.863mV at 400*C

For the highest temperatures, the cold junction side of the T12 thermoucouple was at about 40-50*C.

This is the transformer, and has enough power to drive two soldering irons at the same time, e.g. for soldering tweezers.  I have a second soldering handle for T12, but no soldering tweezers.



The red "knob" is in fact a cap from a plastic bottle, and was filled with melted plastic using a gluegun.  Seems very fragile, but it took the form of the potentiometer's axle.  Can be removed and put it back, like a normal knob. 

Not sure if it's a good idea to complicate things from the start.

Should I double the control circuits, add another connector, and made it to support two independent soldering irons from the start?

[RoGeorge]

The most challenging part will be to properly measure the thermocouple voltage of the T12 type soldering tips (like the ones for Hakko FX-951 solder stations).



The black handle in the middle is Bakon LF005 while the blue one is a Hakko FM-2028 clone.

T12 is a monoblock design, which means the heating element and the temperature sensor are coupled in series, and the different metals that are making the tip itself of a T12 are forming a thermocouple.



There is no separate connection wire to the thermocouple alone, so the only way to read it's voltage is to turn off the heating power, then to connect the heating wires to a very sensitive and very low thermal drift amplifier, in order to measure a few mV produced by the thermocouple.

For the T12 type, that will be to measure about 4mV DC, with \$\mu\$V accuracy, while on average the same wires are powered by 24V AC.   

Some solder stations are using 24V DC, and PWM control to regulate the temperature.  The plan for this solder station is to use 50 Hz AC at 24V RMS, and to turn off some of the semi-alternances of the 24V AC in order to control the temperature.



The advantage of using AC will be to have no PWM, so no spurious RF (also some sources recommend AC powering, and not DC, for a prolonged life of the heating element, not sure if this is true or not). 

For the series ensemble of AC powered heating element + thermoucouple, the only certain moment to measure the thermocouple voltage will be during the zero crossing of the AC, so once at 10 ms, at most.

However, during the zero crossings, the 24V AC that feeds the heating element will override the 4mV produced by the thermocouple in only a couple of microseconds.   

Looks like a difficult task to measure 4mV DC while 67882mV of peak to peak AC is powering the same circuit.  Some tricks will be required.

On top of that, the other side of the T12 soldering cartridge can heat up to about 50*C (120*F).  When compared with the room temperature (about 24*C/75*F) this heating difference is far from negligible.  Ideally, this temperature will need to be measured, too, and taken into account for properly cold junction compensation.  Not sure what tricks to use about this one either.

[RoGeorge]

I am completely bamboozled by these T12 tips.

First, the thermocouple voltage is very small, then, apparently there is no cold junction compensation, yet the cold junction in my tests went as high as 50*C (120*F), which, if left unmeasured will add a significant temperature error to the tip.

The only way this can work properly would be if inside the handle would be some sort of absolute temperature sensor, to measure the cold junction, like a thermistor or an RTD.  So I opened the handles, and all I could find was a capacitor, 100nF for the Hakko type, and 10nF for the Bakon type.







Maybe the capacitor is used as a temperature sensor.  The capacitance of the 100nF one (inside Hakko handle) indeed varies a lot with temperature, but the 10nF (Bakon handle) doesn't vary much.

Measuring the T12 cartridge alone is even weirder, the capacitance between the metal shield and the 2 inner pins (from the heater/thermocouple) varies widely with temperature, between a few hundreds pF up to hundreds of nF, while if I measure the resistance, it stays at about 10-12M .  Apparently it generates some voltage, too.  I need to experiment more with this, in better controlled conditions.

Another thing, by looking at the temperature reading in the Bakon soldering station schematic, it doesn't make much sense to me, and the OpAmp used is a LM358.  The offset alone, and the thermal drift of the LM358 are big enough to make LM358 unusable to amplify 3-5mV.
 

Would have been interesting to see the schematic of the Hakko FX-951, too, but I couldn't find it.

Does anybody have the schematic of Hakko FX-951, please, or some other Hakko model that uses T12 tips?

[Alti]

Maybe the capacitor is used as a temperature sensor.

I really doubt it. I do not think you need to know cold junction temperature to 5degK in such application. You can safely assume:
Thandle=Troom+k(Ttip - Troom)
that is: Thandle will be at same fraction in between Ttip and Troom.

However, during the zero crossings, the 24V AC that feeds the heating element will override the 4mV produced by the thermocouple in only a couple of microseconds.

If you drive it through triac then current flowing through triac after zero crossing is 0A unless you trigger it. Or close to 0A, maybe several nA.