Yet another controller for JBC T210/T245

[shangaoren]

Is there a reason to power heater throught thermocouple ?

For the C245, no.
For C210, the internal wiring is a bit different, and the heater and the TC are in series without a tap in the middle, so there is no way to power the heater without passing current through the TC.

Note that this isn't a problem - the thermocouple is an integral part of the cartridge's construction, and not a separate, discrete part inside of it. It can handle the current just fine.

That explains why this is not a standard thermocouple

[SiliconWizard]

Is there a reason to power heater throught thermocouple ?

For the C245, no.
For C210, the internal wiring is a bit different, and the heater and the TC are in series without a tap in the middle, so there is no way to power the heater without passing current through the TC.

Note that this isn't a problem - the thermocouple is an integral part of the cartridge's construction, and not a separate, discrete part inside of it. It can handle the current just fine.

That explains why this is not a standard thermocouple

It's definitely not. And it also explains why the temp coefficients are different in C210 and C245 cartridges.

[shangaoren]

how do you proceed to measure characteristic of thermocouple?
 i used the reading of an another thermocouple to know the aproximate value of voltage vs temperature,
i was thinking about ploting voltage vs temperature to avoid some non linearity

[SiliconWizard]

Thermocouples usually give low voltages (so beware of any noise), but OTOH they are fairly linear. I've used only linear equations. That's adequate IMO for this kind of application.

Well, you'll need a reference temperature to do this of course.
One way to proceed is to use another soldering iron, and a separate thermometer (probably based on a thermocouple as well). Set the other iron to a few distinct temperatures over the range of interest. Measure the tip temperature with your thermometer (wet the tip with a blob of solder first to improve heat transfer). Then put the thermocouple you want to characterize (here the tip of the cartridge) against the tip of the other iron, both wetted by solder, let it stabilize, and measure the thermocouple voltage. Rinse and repeat.

From memory, the C210 has a coeff. of a bit under 10µV/°C. For the C245, I think it's ~40µV/°C, but I'd have to check that back.

[1sciguy]

Interesting..  I am wanting to do the complete opposite.  I have a JBC station, but I want to make a cable that goes to a cheap TS100 "handle" and utilize the cheap TS100 type tips.  Was posting to ask if anyone knew the connector type, but now I see you have it listed in this thread.  JBC tips are WAY too expensive, but I consider the controller to be the finest soldering station I have ever used.

[Ribster]

Interesting..  I am wanting to do the complete opposite.  I have a JBC station, but I want to make a cable that goes to a cheap TS100 "handle" and utilize the cheap TS100 type tips.  Was posting to ask if anyone knew the connector type, but now I see you have it listed in this thread.  JBC tips are WAY too expensive, but I consider the controller to be the finest soldering station I have ever used.

The controller is only as good so long it has a good working element.
Since the 2 handles have different V/C specifications, and the controller auto detects between the different handle types, it will be a challenge to get your calibrated temperature right

[Hydron]

For C210 the heater current seems to flow through the thermocouple and for C245 the heater can be powered separately.
The C210 schematic is from Xyphro's Nanosolder.
C245 pinout:

I've been fooling around with my JBC station (the cheapie BT-2BWA) looking at how I might make an add-on box to autoswitch between the T210 and T245 irons I have (you can actually buy one but I already feel bad enough about buying the station instead of DIYing it), and have made some notes on the way it works and how the tips are wired that may be of interest:

First of all I can confirm the quoted diagram is accurate - you can measure the thermocouple voltage between C1 and C3 (otherwise a dead short) and the heater between C2 and either of the other connections (with C3 showing the TC voltage in series also). This is despite some images of actual JBC tips being ground down suggesting otherwise (possibly due to them being older 2245 series tips, though I have one of those too and it also matches the diagram!).

The C210 tip on the other hand has the same connections as shown for C2 and C3, but C1 is a dead short with no TC voltage to C2. I assume this is for physical construction reasons - these things are tiny!

The station itself uses C2 as it's internal common ground connection (via a current shunt), and applies the AC heater voltage via phase angle control between this pin and C3. The full heater current flows via the thermocouple in both C210 and C245 cartridges! C1 is only used for sensing and there is no substantial current path possible (other than to mains earth, to which it is hard bonded in this station).

Heater current is sensed by a 0.05R shunt, rectified by an op-amp based ideal diode bridge and then amplified before going into an ADC. The station will reject any cartridge with more than ~10-20% lower resistance than expected (didn't test properly with higher).

TC voltage sense is done between C3 and C1 using a OP07 precision op-amp at a gain of -370 (appropriate for the 5V micro used - 3V3 based stations use more like x245 gain). There is clamping on the output (opamp is running on dual rails outside the microcontroller ones) but I can't see anything other than 4k7/2k7 resistors protecting the input from the heater voltage, which seems a bit rough on both the opamp and the 0805 resistors (assuming there is some clamping to supply rails going on inside the OP07).

There is also a 100x gain OP07 amplifier between C3 and ground (i.e. effectively C3), again going to an ADC input - this saturates during heating operation and I haven't worked out exactly what it is used for.

There is some other sensing stuff going on with C3 and C1 - C1 has 1M ohms to 9VAC, C1 has 100k to ground and between them there are some more high value resistors and a multi-transistor circuit I couldn't be bothered tracing heading off to a digital I/O pin.

Finally there is some zero crossing detection stuff and inputs for tip-change, sleep and T210 detection (all with 10k pullup to +5V and 10k in series with the GPIO input).

Hope these rambling notes make some sense and are of use confirming how JBC themselves do things (also see a much better effort, albeit with possibly incorrect cartridge diagram, here: https://www.eevblog.com/forum/testgear/jbc-soldering-station-cd-2bc-complete-schematic-analysis/ )

Edit: maybe the C245 cartridge can be wired in a couple of different ways - series and separate TC depending on where the heater wire terminates. I don't think it would make a big difference in operation and would explain the tip teardown I saw and some notes on the Unisolder project saying "series or separate TC".

I will also join those expressing interest in the OP's implementation of the ITBH algorithm - I need to build a Weller micro-tweezer controller and working out a control algorithm that deals well with the small-mass tips is one of the things I have to look at.

[dreamcat4]

I remember there was one question about driving the heaters with DC instead of AC, and how it could possibly damage the tips much faster. I haven't noticed that in practice, but I don't use it heavily either (just for prototyping, not for any kind of production work of course!), so this is not a definite answer on this point. I can just say that it works fine for my use.

Hello. I am just wondering today, that since you have an H bridge, if it could be useful for a different purpose: To be swapping around the polarity of the heater. Between the inner pin and the middle pin. So as to avoid electromigration.

But you are using the H bridge in a different way. Which is meant for other reasons and not this reason. I would normally suggest "well then we could just add another H bridge". However the part seems pretty expensive to source. Just for 1 of them. Let alone 2.

Nevertheless the general idea of reversing the polarity every so often... it makes some sense to me (for DC driving). I mean - if it is indeed actually an issue, for the electromigration. I really don't know.

[edited - added]

The effect of reversing polarity, is described in more detail on page 18 of this PDF (section 2.4.3), of chapter 2 pdf:

https://www.ifte.de/books/em/em_chap2.pdf

Entire book: https://www.ifte.de/books/em/

What is useful (from that page 18) is that they explain how "Tao et al. [TCH93]" were in fact using a rectangular square wave (not AC). And it yielded a 500x reduction over a DC waveform.

So indeed - it would appear true (my guess) that AC was not in fact required to reduce electromigration. You might get it even lower with AC. However instead of that, merely switching round the polarity of the pulsed DC is can also be of a significant benefit.

The reference they cited [TCH93] was from this other research paper:

[TCH93] J. Tao, N.W. Cheung, C. Hu, Metal electromigration damage healing under
bidirectional current stress. IEEE Electron Device Lett. 14(12), 554–556 (1993).
https://doi.org/10.1109/55.260787

Which seems to be behind either a free or paid paywall. However presumably in that source paper, it may go into further details, to explain which specific frequency range is best. From the states "10hz upwards".

 

So then my question is can the H-bridge (or something else) do the switching at a rate faster than 10Hz? Is that a capability which we can actually design into the circuit, in a way that is simpler than doing AC ?

Another constraint (for choosing DC instead of AC) is that because cannot use AC on a portable battery operated version. So maybe this method of reversing the DC can be used there too eh?

===

Perhaps another (completely different) reason why AC is preferred over DC is a lower noise. Which may be desirable in a lab environment. When soldering on live electronics.

[thm_w]

You don't solder live electronics, or if you are crazy and still feel the need to do that, the device must be isolated as the iron tip is grounded. Then AC or DC does not matter.
The tip will die from plating wearing through before electromigration is an issue, afaik: https://www.eevblog.com/forum/projects/diy-jbc-nano/msg1469445/#msg1469445

AC is simply used because its cheaper.

[SiliconWizard]

I remember there was one question about driving the heaters with DC instead of AC, and how it could possibly damage the tips much faster. I haven't noticed that in practice, but I don't use it heavily either (just for prototyping, not for any kind of production work of course!), so this is not a definite answer on this point. I can just say that it works fine for my use.

Hello. I am just wondering today, that since you have an H bridge, if it could be useful for a different purpose: To be swapping around the polarity of the heater. Between the inner pin and the middle pin. So as to avoid electromigration.

Not possible easily using DC IMO unless you go for an isolated design, because of the ground problem. The tip of the cartridge is grounded. I wanted to keep things simple.
Electromigration may be an issue, but IMHO, oxidation will be a dominant issue in the lifetime of a soldering tip rather than electromigration, especially when frequently using lead-free solder.

[SiliconWizard]

So then my question is can the H-bridge (or something else) do the switching at a rate faster than 10Hz?

Of course. I'm switching @100Hz. It could easily switch at a faster rate, but I don't see any benefit? And since I'm doing a TC measurement every cycle, there is a hard limit on the switching frequency anyway (properly measuring the TC requires some time to stabilize.)

[SiliconWizard]

I will also join those expressing interest in the OP's implementation of the ITBH algorithm

Alright. I'll probably post that soon.

[dreamcat4]

Not possible easily using DC IMO unless you go for an isolated design, because of the ground problem. The tip of the cartridge is grounded. I wanted to keep things simple.
Electromigration may be an issue, but IMHO, oxidation will be a dominant issue in the lifetime of a soldering tip rather than electromigration, especially when frequently using lead-free solder.

Yes thank you both for answering my questions about this matter. It has set my concerns to rest about electromigration and using DC being an acceptable solution. Actually I am very happy not to have to worry about those aspect anymore. And very grateful for that.

I also understand now, it's really up to the PSU whether or not it is going to be floating. So for it to be a nicely behaved piece of lab equipment. Then ideally it is in fact desirable for the PSU to be floating DC +24v. However with clearly marked external bar of bridging metal on the front that is a removable ground link. So for those fewer situations where working on something live is required (exceptions). Then that grounding bar can be removed just for that specific job only.

To be clear the metal housing of the main unit would always be grounded. However the PSU inside would be floating, as would the C245 cartridge. Unless that removable ground link was in place. Which it would normally be. I suppose this then also means that the PSU board must be properly isolated from the case. Which is an extra consideration not to use metal standoffs for it, whilst remaining mechanically strong etc.

In addition, (when in floating mode), isn't it nevertheless still remain desirable to maintain a 1w 1mohm resistor to GND. For the purpose of anti static grounding?

Or instead of being permanently wired that way... should I instead make that another optional external feature of my device? Because I do have here some 'proper' anti static grounding points. Which I already include a 1mohm resistor inside them. Well that is all the are really. Could have just made some myself had I known that.

Sorry because my questions perhaps not unique to dealing with these C245 cartridge. However mentally it will help me to move past those concerns. An to know 'how to wire up the op-amp', when it's thermocouple input is potentially either floating or grounded. As the signal is only a few micro volts. Unfortunately none of my own equipment is good enough to measure that source signal. I would have to rely on checking what is on the output of the op-amp.

[SiliconWizard]

So here is the code for the regulation algorithm. Cleaned-up (the original functions contain more code and TempControl_t structure contains more fields to handle temperature control in general, I left only what was necessary for the regulation algorithm here.)

Code: [Select]

typedef struct
{
float fSetTemperature; // Set Temperature (°C)
float fActualTemperature; // Actual (measured) Tip Temperature (°C)

// Improved Take-Back-Half algorithm.
float fGamma;
float fBeta1;
float fBeta2;
float fTimeStep;

float fTemperaturePrev;
float fTemperatureErrorPrevZC;
float fHeaterControl;
float fHeaterControlPrevZC;
} TempControl_t;

void TempControl_Init(TempControl_t *pTC)
{
pTC->fTemperaturePrev = pTC->fActualTemperature;
pTC->fTemperatureErrorPrevZC = pTC->fSetTemperature - pTC->fActualTemperature;
pTC->fHeaterControl = 0.0;
pTC->fHeaterControlPrevZC = 0.0;
}

// One step: Improved Take-Back-Half algorithm.
//
void TempControl_ImprovedTakeBackHalf(TempControl_t *pTC)
{
float fTemperatureError, fTemperatureDeriv;

fTemperatureError = pTC->fSetTemperature - pTC->fActualTemperature;
fTemperatureDeriv = (pTC->fActualTemperature - pTC->fTemperaturePrev) / pTC->fTimeStep;
pTC->fTemperaturePrev = pTC->fActualTemperature;

// Integral term.
pTC->fHeaterControl += (pTC->fGamma * fTemperatureError);

// Improvement: use derivative to minimize overshoot/undershoot.
if ((fTemperatureError > 0.0) && (fTemperatureDeriv < 0.0))
pTC->fHeaterControl -= pTC->fBeta1 * fTemperatureDeriv;
else if ((fTemperatureError < 0.0) && (fTemperatureDeriv > 0.0))
pTC->fHeaterControl -= pTC->fBeta2 * fTemperatureDeriv;

// Output variable clipping.
if (pTC->fHeaterControl < 0.0)
pTC->fHeaterControl = 0.0;
else if (pTC->fHeaterControl > 100.0)
pTC->fHeaterControl = 100.0;

// Take-Back Half.
if ((fTemperatureError * pTC->fTemperatureErrorPrevZC) < 0.0)
{
pTC->fHeaterControl = 0.5 * (pTC->fHeaterControl + pTC->fHeaterControlPrevZC);

pTC->fTemperatureErrorPrevZC = fTemperatureError;
pTC->fHeaterControlPrevZC = pTC->fHeaterControl;
}
}

Key fields:
* fSetTemperature must be set before starting the controller (target temperature in °C - you can use °F as well, the coefficients will probably have to be adjusted);
* fActualTemperature must be set at each iteration before calling TempControl_ImprovedTakeBackHalf();
* fGamma, fBeta1 and fBeta2 are the algorithm coefficients (must be set before starting the controller);
* fTimeStep is the time step in s (period between each iteration - 10 ms in my case, thus 10.e-3) (must be set before starting the controller).

Operation:
* After setting the proper fields, and making a first tip temperature measurement, and assigning it to the fActualTemperature field, call TempControl_Init() to initialize the controller;
* At each iteration (period must be fixed and correspond to the value set in the dTimeStep field): measure tip temperature, set fActualTemperature to it, and call TempControl_ImprovedTakeBackHalf(). After it returns, you can use the fHeaterControl field to control the heater PWM (fHeaterControl will be between 0.0 and 100.0, representing 0% to 100%);

The coefficients I used for C210 cartridges: fGamma = 1.0, fBeta1 = 0.2 and fBeta2 = 0.4. Keep in mind it's for temperatures in °C and a DC voltage (for the heater) of 12V.

[dreamcat4]

here is the code for the regulation algorithm.

We appreciate this! It's pretty helpful. Can this bit of code (and it's accompanying commentary / explanation) be considered as public domain? Because we have existing open source projects and hardware. Reading online it would seem that a public domain work should be compatible with all other licenses. The GPL ones in particular can be really annoying like that. It would be sad not using it because of a license incompatibility. Thank you for considering this matter.

To be clear: I am asking for the PT Dreamer firmware (open source). Because I already have 2 controller PCBs for flashing it. Of course they have their own specific issues, those hardware. And require modification. At least it's not doing everything from 0.

Elsewhere I also see some new open source projects, they are often struggling with the performance. So having your algorithm be freely usable for any of them. It would be a great help. I say that because some of those specific hardware designs are different from each other. And have different types of usage.

For example one is very small, and it's designed for portable use, and can run from USB PD. Like a competitor to the TS-80. While yet another one (also for T245)... is over engineered to the maximum. Making it a really powerful unit, for 2 irons  simultaneous at once. AKA "dual wielding".

Wheras mine are these pre-made 'STM32 T12 OLED' pcb. Which is a more mainstream solution. Somewhere in the middle. So I believe this is best use for my own time. To see how far I can improve them.

[SiliconWizard]

Well, feel free to evaluate this for your project and report back if you can. I'll be glad to see if this works fine for others. I was not completely decided about putting this in the public domain or open sourcing it - but I guess you can consider it is made so though by posting it here without any specific mention. The algorithm itself, apart from the derivative term (which helps quite a bit!), is public, and the implementation is nothing special.

I was happy with this control algorithm though, as it was definitely much easier to tune than a regular PID, while giving equivalent or better results.

So all in all, feel free to include this in your project, don't bother with the license, but don't forget to chime in if you do!

[DavidAlfa]

Hi, can you confirm the TC voltage?
40uV/°C for C245 and 10uV for T210?

[DEHiCKA]

looking at how I might make an add-on box to autoswitch between the T210 and T245 irons I have

Did you succeed? TEC relay module can only do it for T115 and T210. T210 and T245 switching does not work.

[Hydron]

Did you succeed? TEC relay module can only do it for T115 and T210. T210 and T245 switching does not work.

I have not finished it yet (have obtained a second holder, connectors, case and drew up a state machine plan, need to CAD up a PCB now), but I see no reason why it couldn't work if you simulated a complete iron disconnection when swapping between T210 and T245 - no different than unplugging one handle and plugging the other. There is an extra pin in the T210 connector jumpered to one of the others to signal a T210 iron - this would obviously need to be handled as well to make detection work.

[Jesus]

How does this iron work? https://github.com/Jan--Henrik/Otter-Iron-PRO

[dreamcat4]

check the issues tracker on github

[hatte]

Yes, it relies on pin 6 of the RPC1 connector. In T210 handles, pin 6 is connected to the tip, in T245 handles, it's left open.

Attached is the connector pin-out from my schematic.

I know I am reviving a very old thread, but I am working on designing a (low cost) alternative to your controller with handle auto-sense, which I expect will be running a modified version of https://github.com/deividAlfa/stm32_soldering_iron_controller.

Would you, by chance, recall the pinout of the JBC Microtweezer in the Hirose RPC1-12P-6x connector, so I can accommodate that option in addition to Hakko T12, JBC T210, JBC T245 and possibly even JBC NT115.

[SiliconWizard]

Yes, it relies on pin 6 of the RPC1 connector. In T210 handles, pin 6 is connected to the tip, in T245 handles, it's left open.

Attached is the connector pin-out from my schematic.

I know I am reviving a very old thread, but I am working on designing a (low cost) alternative to your controller with handle auto-sense, which I expect will be running a modified version of https://github.com/deividAlfa/stm32_soldering_iron_controller.

Would you, by chance, recall the pinout of the JBC Microtweezer in the Hirose RPC1-12P-6x connector, so I can accommodate that option in addition to Hakko T12, JBC T210, JBC T245 and possibly even JBC NT115.

I only ever supported the T210 and T245, so unfortunately I don't know much about the JBC Microtweezer. Do you mean the PA120 model?

[dtmf]

2 EURO JBC C245 controller

[boneDragon]

dtmf works well despite simplicity.

here is a mod for any T12 handle with tilt sensor (switches off after 2..4min).

for JBC C210 change R2=4.7..5.1 Ohm (or 10 || 10 Ohm), R10 to 100 Ohm, R17 to 2..4.7M, and power from 8..12V.