Weller Soldering Iron WES51

[bostonman]

I'm trying (and try in general) to avoid deviating from the thread regarding this solder station and now PIC issue (I really, really appreciate the help Ian.M has been providing because he's not only provided help to me, but gone out of his way to decipher the code; and provide edited code to burn new PICs for anyone in the entire world to download), but want to comment on your message to avoid anyone having a negative opinion about me. From our last public exchange regarding connectors, you've certainly made a point to place emphasis about mistakes on my part.

As with any public forum and social media site (which I'm not a part of), any comment can be taken out of context. Certainly I'm not at all taken offense, nor do I think you're angry towards my statements, but feel I should elaborate a bit because I don't want anyone thinking I think I'm above them.

I've never stated I'm a novice. I feel this forum was created with the intent of sharing knowledge between beginners and well advanced individuals without the need to categorize; and I feel you're categorizing me.

If not for this site, my last few in depth repairs probably would never have happened. I think it's absolutely great that someone in one part of the world repaired something and can answer questions to someone with the same issue in another part of the world. Also the fact that people take time out of their lives to help someone they've never met, will never meet, nor may not repay the favor.

I'm sure you drive a car. When someone makes an illegal move which results in cutting you off, do you roll down your window and ask if they are okay and swerved because they have a health issue, or if they cut you off for a reason? Most likely you yell "you're a horrible driver". Does that mean you're a "novice" driver? I'm sure you've made similar errors too, however, you feel the need to think you're the more dominate/advanced driver by judging their driving skills without asking why they made a dumb move. Similarly, you've placed me in the category of accusing me of thinking I'm a "novice" with PCB design.

My point is: you're absolutely correct. Maybe Weller had millions of extra 1206 components and wanted to use them, maybe it was cheaper for the pick and place machine, etc... I made a judgement call on what I feel is a poorly laid out PCB much like you've yelled at some idiot cutting you off. I feel the PCB is laid out by an amateur because I've made many (more) mistakes in the few PCBs I've laid out. The difference is: Weller is (probably) a multi-million dollar company making (what I feel) are amateur mistakes and I'm an average guy who sees these issues. Some technical gurus would group together and laugh, not try to shun them from the conversation.

In some cases the Weller PCB designer had plenty of PCB space to run etches differently but chose to run them under components. In other cases they didn't have enough room, but could have made the room by spacing out components. They chose a 33uF, 35V, capacitor and could have gone with a higher voltage but didn't have the space. Should they have placed the capacitor on the other site of the board, they could have gone larger.

For respect of keeping the thread on topic and those who reference this topic in the future to repair their Weller, I think we should keep personal opinions to ourselves, and this goes for me as well. Maybe I stepped out of line by offering my opinion about the PCB layout and component selection. For that, I'll admit, I was wrong to deviate the topic by providing an opinion.

[bostonman]

I soldered in the new PIC and unfortunately the LED still doesn’t blink.

The iron temp was set to just above minimum, it got hot enough to melt solder, I waited a while, and the LED remained green. I also increased the temp knob and then decreased it which should turn off the LED until it reaches the cooler set point, and still the same results.

[bostonman]

Ian.M,

I was curious whether you saw anything in the code that shows what it takes to make the LED to blink.

At this point, the iron is new which eliminates a thermocouple issue, and the PIC is new. Due to not knowing what the PIC is doing, the only last shot in the dark is wondering whether the 220pF on pin 5 of the PIC has become resistive causing low amplitude thus not clocking the PIC.

I'm uncertain if the PIC needs this to perform basic functions. If so, then obviously this isn't an issue because the PIC is somewhat working, but, if maybe it only needs the 60Hz to blink the LED (obviously dividing the frequency first), then maybe this is the culprit.

Other than that shot in the dark, I don't know what else could cause this issue.

[bostonman]

Just for reference, here are two examples of what I consider poorly run etches.

The two wires are the 24V AC input and the other picture shows an etch going underneath a fuse (the green part).

I know this is low frequency and doesn't matter if they are under components, but the designer had plenty of PCB space and my point was: it looked like the design was to consolidate space, however, numerous amounts of empty space exists; the need to run etches under components seems like someone just ran things haphazardly. The board also doesn't have any silscreen which was probably done to save money.

[tooki]

Just for reference, here are two examples of what I consider poorly run etches.

The two wires are the 24V AC input and the other picture shows an etch going underneath a fuse (the green part).

I know this is low frequency and doesn't matter if they are under components, but the designer had plenty of PCB space and my point was: it looked like the design was to consolidate space, however, numerous amounts of empty space exists; the need to run etches under components seems like someone just ran things haphazardly. The board also doesn't have any silscreen which was probably done to save money.

Thanks for providing those examples. I agree with you that those seem like very sloppy layout.

P.S. They’re called “traces”, not “etches”.

[bostonman]

P.S. They’re called “traces”, not “etches”.

Hmmm is this opinion? I've heard both over the years and you sparked my interest. I've also heard various names for raw PCBs and populated. The owner of my last job insisted an un-populated PCB is a 'printed wire board'.

Trace seems more like a small segment of an entire etch. A printed wire board seems like some 60s term (ironically the owner was mid 80s and probably using an outdated term).

For me, I didn't work in manufacturing, so my terminology is usually raw PCB for an un-populated board, PCB for populated, assembly for a unit comprised of PCBs, and etches for 'traces'.

Anyway, yesterday I took (more) measurements on the solder station and there is something I don't understand. With the new iron plugged in (and heating), pin 7 of the PIC seems to act weird. If I set pin 6 (temperature set poteniometer) to 1V DC, pin 7 has a 0.5v p-p 60Hz sine wave with a DC offset that climbs until it reached 1.5V DC and then it's just a DC voltage. As the iron cools (or so I believe), the DC voltage drops slightly and then the sine wave reappears until it climbs to 1.5V DC again.

Setting pin 6 to 2V DC, pin 7 acts the same way (60Hz sine wave 1.5v p-p) until it settles to 3V DC. Once the iron cools a bit, same thing, the sine wave reappears, DC offset climbs, and settles to a DC voltage at 3V.

Setting pin 6 to 4V DC, pin 7 acts the same way and settles to about 4V.

What I don't get is why a sine wave keeps appearing. I thought maybe it was the 100nF capacitor on pin 2 of the op-amp causing some time constant, but I've simulated it setting pin 3 of the op-amp to a slow rising 0 to 7.34mV which is approximately 180 degrees C (approximately 360 degrees F) and the output (as I expected) is just a DC ramp.

I'm wondering if something is going on at the op-amp end causing the PIC not to begin blinking the LED.

Edit: 7.34mV is approximately 180 degrees for a type k thermocouple

[bostonman]

So far I feel all the circuits have been tested and confirm to work with the exception of not figuring out why I'm getting a 60Hz sine wave out of the op-amp when the iron is heating. Not sure if this could be a reason the LED isn't blinking when the iron reaches temperature, but seems odd to see the 60Hz. It tells me I have a bad ground but it's not like this solder station has many grounds; just one ground and Earth ground which both seem to be connected (I measured from the iron to the ground prong on the AC plug and the circuit wouldn't work if I was missing common ground).

Yesterday I took some scope measurements on the op-amp along with voltage measurements. Both inputs seem to measure correctly when the iron is set to approximately 350 degrees F and the output measures correctly compared to a simulation circuit I created; except it has a sine wave when heating.

I took a scope measurement on both, the output of the transformer and the output of the op-amp (see attached scope picture). Whenever the 60Hz occurs on the output of the op-amp, the output of the transformer drops in amplitude a bit and has some funky little noise near the zero crossing.

I don't know if this is something that contributes to the LED not blinking, but figured I'd mention it because I'm out of ideas.

[floobydust]

I think you've had a hard time figuring out what is going on with this soldering station.  To me it keeps looking like some problem with TC(-) either open wire (fuse too) or shorted to the heater. I believe PE ground only connects to the shield, not to anything on the 24VAC secondary side. Unless there is a ground fault. If there is a short to PE GND, your scope readings would be wrong as well and small chance you damage the scope by passing large current through its ground. There should not be so much AC hum. Get out your ohmmeter.

Next what I would do is roll back and make a thermocouple simulator, just a voltage source like a 1.5V battery and a resistor voltage divider or Wheatstone bridge I use, to get 0-20mV inject that and see what the controller is doing based on dialed in temperature it sees.

[bostonman]

Next what I would do is roll back and make a thermocouple simulator, just a voltage source like a 1.5V battery and a resistor voltage divider or Wheatstone bridge I use, to get 0-20mV inject that and see what the controller is doing based on dialed in temperature it sees.

Ironically, this is exactly what I plan to do.

I believe PE ground only connects to the shield

What is PE?

At one point I tried shorting the fuse just in case it went resistive (if it's possible with these types - but figured what do I have to lose?).

I agree, I shouldn't have 60Hz, but the circuit seems simple enough that I feel I've covered everything. Your idea (and what I've wanted to do) seems like a good approach because I'm no longer battling a soldering iron heating and cooling constantly.

[tooki]

PE = “protective earth”, as in the true ground connected to the Earth for safety (as opposed to “ground” in the sense of the fundamentally arbitrary 0V reference of a circuit).

[floobydust]

[...] At one point I tried shorting the fuse just in case it went resistive (if it's possible with these types - but figured what do I have to lose?). [...]

The fuse protects the wiring and PCB traces from melting when a fault occurs (heater to thermocouple short). It might have already happened with your soldering station. You'd have to do continuity checks to make sure they are intact.

[bostonman]

Not sure if you overlooked a previous posting, or suggesting I re-measure (which I plan to do), but this is a new soldering iron. I do, however, like that my idea to use a battery to produce a small voltage in place of the iron was suggested by you; it provided some reassurance my idea wasn't crazy.

Also, this morning over breakfast your posting popped in my head and I said 'PE' is the soldering iron part number (Weller PES51 Solder Iron) and said this is the reference in the posting. Afterwards I read your updated post and discovered my breakfast wasn't strong enough to start my brain and that PE is Positive Earth.

[bostonman]

Next what I would do is roll back and make a thermocouple simulator, just a voltage source like a 1.5V battery and a resistor voltage divider or Wheatstone bridge I use, to get 0-20mV inject that and see what the controller is doing based on dialed in temperature it sees.

I have been flat out busy with some stuff, but this is my next step. I'm wondering if injecting a voltage from an external source may cause a ground loop issue. Should I connect the battery ground to the soldering iron ground?

My other thought was to tap into the 5v regulator on the iron, go through an op-amp, etc... and then everything will be referenced to the same ground.

BTW, just to update: I did ohm everything and can't seem to find any open ground connections that would explain why I see 60Hz on the output of the op-amp that is fed from the thermocouple. The PCB itself only has two leads from the transformer, and, if the return was open, the whole circuit wouldn't work. Earth ground from the (new) soldering iron to the plug also measure correctly.

[Andy Watson]

Out of pure curiosity, I've been having a nose around the code presented previously and found the following:

As Ian.M has noted previously, the EEPROM is only 16 bytes in size and is essentially "bolted on" via a general purpose I/O port. The take-away that I get from this is that the EEPROM cannot be pre-programmed with data. Therefore the "initialisation" of the EEPROM data must be part of the permanent code. I.e. once programmed, the PIC must know how to initialise its own data from a "virgin" state. So there should be no problem in installing a fresh chip programmed with the existing code.

On power-up, there are three steps to the initialisation:
1. Initialise the internal data (mostly set the RAM to zero) and define the I/O.
2. Determine the mains frequency using Timer 0. This happens every power-up cycle so the heating control constants can be baked-in to the code for eeither 50 or 60Hz operation.
3. Read the non-volatile parameters from the EEPROM.

What does the EEPROM store?
There are only five bytes of the data stored in the EEPROM; six if we count the magic number. The first operation of step 3 is to read the first EEPROM location. If this does not return the magic number of 0xF0, the EEPROM is (re-) initialised to default values.

The EEPROM data is read as:
Location 0 :  The magic number that indicates the EEPROM contains valid data.
Location 1 :  Tip offset temperature. One byte that saves the tip-offset temperature in 2Cs Fahrenheit.
Location 2 :  Non-volatile flags. Only two flags appear to be used. One to indicate temperature locked or unlocked. One to enable flashing of the LED.
Location 3 :  High byte of set point temperature when the station is locked.
Location 4 :  Low byte of set point lock temperature (in Fahrenheit.)
Location 5 :  Dial calibration offset.

The default "virgin" reset values are:
Tip-offset = 0. Flags=0, i.e. unlocked and LED flash enabled. Dial calibration = 0 offset. The set point operating temperature does not default since the station will be unlocked and the temperature will be determined by the dial.

The default "factory" reset values are:
Tip-offset = 0. Dial offset = 600 - Dial setting (see later). Flags are not reset and the temperature lock is not affected.

There are two modes of using the magic pencil (i.e. the magnetically triggered
reed-switch.)
During normal operation or as part of the power-up sequence.

During normal operation:
As per manual. If the pencil is used for 0 to 2.5 seconds, indicated by a red flashing LED, the temperature lock is toggled. If the pencil is used for more than 2.5 seconds, indicated by the LED turning green, the dial can be tweaked to offset the tip-temperature of the iron. When the pencil is removed the offset, if within range, is stored as the new tip-offset temperature. Note, this "calibration" takes into account the present tip-offset temperature and adjusts the new offset accordingly. There is no way to force the tip-offset temperature to zero except via the "factory reset".

During power-up:
AKA Factory Reset: As the manual suggests, this resets the "tip offset" temperature to zero, but that's not all.

What the manual does not say.
The manual requires that the temperature dial is set to 600F. If, in the factory reset phase, the dial reading is within 75 degrees of 600, this is taken as a calibration point of the dial. I.e. It assumes that the user has set the dial to 600 and whatever the difference is, is used as an offset to calibrate knob and potentiometer readings. This offset is stored in non-volatile memory.

If the reading does not fall within the 75 degree window the follow happens:
Less than 400 A non-volatile flag is cleared.
More than 800. The non-volatile flag is set.
As far as I can tell the only effect of setting this non-volatile flag is to prevent the LED from blinking during normal operation.

[bostonman]

Wow, that is certainly a deep explanation of the code.

Some of your explanation gets a bit confusing, probably because you have fresh knowledge since you've been studying it; and I'm not a code guy. Still, it was a good explanation and shed some light on things.

In any case, you stated the PIC detects line frequency. Maybe this explains the purpose of pin 5? I believe the PIC has an internal oscillator, so the blinking rate of the LED may be a divided frequency of the oscillator; or maybe it's divided from the 60Hz and pin 5 serves more than just initially detecting frequency.

Also, on page 3 of the manual, it states setting the dial to 600 degrees F for a factory reset. At least once with the old PIC I did a "factory reset", however, I didn't put the face plate on to see where the dial is, but know the middle point of the poteniometer is 600 degrees F.

In any case, if I'm understanding your breakdown of the code, does a possibility exist that I've held the "pencil" (i.e. a magnet I had on the side of my tool box) too long and caused the PIC to think some tip offset exists?

Although, now that I think of it, when I installed the new PIC, I just turned on the iron and waited without seeing the LED blink. From what I believe you've stated, a new PIC wouldn't have values from the original PIC, so the new PIC should have set flags and blinked the LED (unless I misinterpreted your explanation).

Location 1 :  Tip offset temperature. One byte that saves the tip-offset temperature in 2Cs Fahrenheit.

What is 2Cs?

[Andy Watson]

In any case, you stated the PIC detects line frequency. Maybe this explains the purpose of pin 5? I believe the PIC has an internal oscillator, so the blinking rate of the LED may be a divided frequency of the oscillator; or maybe it's divided from the 60Hz and pin 5 serves more than just initially detecting frequency.

Yes, no, the latter . The purpose of pin 5 is to detect the mains cycle - specifically when the mainis cycle crosses close to zero. This event is used in the initialisation to determine the mains frequency. The period of the mains cycle is measured against the PIC's initernal oscillator (whihc is assumed to be about 4 MHz.) The PIC will remain in this initialisation until it obtains a measurement of the mains frequency that lies in the range 48 to 54 Hz, or 54 to 62 Hz.

Now that the PIC knows the mains frequency to be either 50 or 60 Hz, it defines a tenth of second count as either 10 or 12 - this is the number of half-cycles of mains in a tenth of a second. 

From here on, the PIC sits in a loop that is synchronised (via pin 5)  to the main half-cycles. It must be! It needs to know when to fire the triac.

So we have a loop of code that runs every half-cycle of the mains. We can use the value(s) determined in the initialisation to increment a tenths of a second timer. And the tenth of second counter can be used to determine seconds, and the seconds to minutes. The reason for the tenths of second counter (rather than going directly to seconds), is that the heating of the iron occurs in tenth-second bursts. The length of each burst can be divided into either 10 or 12 units, which gives a semi-proportional heat control. In normal operation the LED flashes (if enabled) with the heating of the iron; in other situations the LED flashes at the rate set by the least significant bit of the tenths counter.

All the timing and synchronisation is inherently linked to the mains cycles via pn 5. The actual frequency of the PIC's oscillator is secondary (although it must be "close-enough" to get through the initialisation stage).

Conclusion: Pin 5 must "wobble" at the mains frequency or the PIC don't go!

Also, on page 3 of the manual, it states setting the dial to 600 degrees F for a factory reset. At least once with the old PIC I did a "factory reset", however, I didn't put the face plate on to see where the dial is, but know the middle point of the poteniometer is 600 degrees F.

Ok,  let me clarify what I meant. The manual is not wrong, but it does not tell lthe whole story. If you power-up with the magic pencil in place, one of four things happen:
1. Dial at minimum ( less than 400F), the LED flash enable flag is cleared (i.e. the LED will flash when the iron is heating).
2. Dial at maximum (greater than 800F), the LED flash enable flag is set - the LED does not flash during normal operation.
3. Dial (as read by the PIC) is between 525 and 675F. Set tip-offset temperature to zero. Assume the operator has set the knob to 600F and take difference as an offset to the dial reading, i.e. calibrate the dial.
4.If none of the above - do nothing.

Although, now that I think of it, when I installed the new PIC, I just turned on the iron and waited without seeing the LED blink. From what I believe you've stated, a new PIC wouldn't have values from the original PIC, so the new PIC should have set flags and blinked the LED (unless I misinterpreted your explanation).

Yes. I think you have the gist of what I was saying/thinking. All the "information" to create a fully working soldering station is baked-in to the firmware of the  PIC. 

Location 1 :  Tip offset temperature. One byte that saves the tip-offset temperature in 2Cs Fahrenheit.

What is 2Cs?

Two's compliment. I.e. the values in the one byte are assumed to range from -128 to 127.

[bostonman]

Initially you stated you had a "nose around" in this. You've done more than just (what I initially assumed) dabbled in the code, you've basically dug far deep.

I'm uncertain how much of this code is necessary for a soldering iron and seems it's much longer than needed.

One thing I dislike about pin 5: the amplitude going in is extremely high and (I believe) maintained to safe voltage levels via the input protection diodes.

I need to re-re-read your last posting to make sure I understand the steps, however, from my interpretation, it seems the LED isn't blinking for possibly two or three reasons:

1: If my mains frequency isn't 54Hz - 62Hz. I find this hard to believe, but will remeasure. Other than my house not being 60Hz (which I'm sure it is), then maybe the transformer or capacitor (220pF) is doing something funky to the frequency.

2: The new PIC got its program from the original PIC causing the firmware to lock out the blinking mode. But you've clearly stated the PIC initializes flags on its own and doesn't matter what it had for flags.

3: I set the temperature knob greater than 800 degrees F, held a magnet near the Reed switch, and powered the iron. This probably isn't the case since I installed a newly programmed PIC and turned on the iron without doing anything; but anything is possible. I can try setting the dial again to 600 (which would be 2.5V at pin 6, hold a magnet to the switch, and power the unit thus resetting the iron.

2 and 3 above seem like they can't the issue since I explained the reasons leaving just 1 above (bad transformer, mains frequency, or capacitor).

Have I missed anything in your explanation that conflicts with my three assumptions on what could be the issue?

[bostonman]

I took some scope measurements of pin 5 of the PIC to see how the 60Hz looks.

When I measured pin 5 alone, it looked like the phase kept jumping 180 degrees. I checked the triggering on the scope, and, no matter what I did, the phase kept jumping; hence why I named the files phase 1 and phase 2.

When I measured pin 5 and the secondary of the transformer together, it seemed both channels would flop back and forth in phase by 180 degrees, but, when I changed the trigger to pin 5, the phase seemed to stay the same.

I'm uncertain whether the phase hoping is causing the PIC not to see the correct line frequency, or if it's just poor triggering on the scope (it's a higher end digital 600MHz scope).

If I had to guess, I'd agree with someone who said a ground issue, but, I can't figure out where a bad ground would be. Earth ground from the prong all the way to the iron measures low resistance, and, all the grounds on the board measure low ohms back to the secondary of the transformer return.

The primary of the transformer measures 22 ohms and the secondary measures 71 ohms if this means anything.

[bostonman]

I'm digging up my old thread because I finally got some time to build a dummy thermocouple circuit which I can adjust.

Since I wasn't sure how to calculate the gain of the thermocouple op-amp circuit, I simulated it. My dummy circuit basically requires I remove the 1M resistor between the 5V and input of the + thermocouple, take that 5V, goes through four diodes (to drop the voltage), and two pots in series. Initially I wanted to use a resistor, but, since I wasn't sure how much the diodes would drop, I figured using two pots would make things easier.

Turns out, it seems I accidentally made it so one pot acts as sort of a course adjust and the other acts as a fine tune.

In any case, comparing the simulation circuit to my measurements, I set the virtual thermocouple to various voltages including 12.8mV (315 degrees C or 600 degrees F) and the output of the op-amp seems to match quite closely with the simulation values.

I also tried simulating a somewhat cold iron by turning down the voltage, recycling power, and slowly bringing up the voltage until it reached 600 degrees F (which is what I had the dial set to) and the green LED still didn't blink.

From what I can tell, the gate to the TRIAC controls. When I increase the thermocoupe voltage to within 600 degrees F, the pulses to the TRIAC seem to change so the left half of the pulses are 0V DC with the right half remaining the same (I assume the frequency has changed but I didn't bother changing the scale). As I increase the voltage, the waveform turns off. Decrease the voltage below 600 degrees F, and the waveform returns.

Technically I guess using the iron in this state wouldn't hurt anything, but it doesn't really assure me the iron has reached its set point without the blinking green LED.

Unfortunately I don't have a conclusion. The board has power coming in, the 5V regulator is putting out a clean 5V DC, the TRIAC does heat the iron (when the iron is plugged in), the op-amp seems to have the correct gain, and, from Andy Watson posted, the PIC doesn't rely on any external factors that would cause the code to mess up.

I've done factory resets without seeing any changes. The dial needs to be set to 600 degrees F at turn on, so that's the mid point. I usually set the voltage on pin 6 to 2.5V which is the mid point, pull pin 4 to ground, power the unit, and remove ground from pin 4.

The only thing left is assuming something funky is going on with the transformer causing the frequency to deviate enough to cause the PIC to mess up, the 220pF on pin 5 is bad, or the thing is just hopeless (which I refuse to believe since the circuit isn't complicated); unless someone has a better theory.

[bostonman]

I wanted to reach out one last time about this thread.

Unfortunately I think a decision needs to be made at my end on whether to junk this solder station, or toss it on the shelf to never address it again.

The only odd measurement I got was measuring from the gate of the TRIAC to MT1 having only 75 ohms - I posted my question in another thread and referenced this thread, but someone stated that seems normal.

The voltages all measure correctly throughout the circuits. I injected my own voltage into the op-amp to simulate a thermocouple and the output op-amp voltages matched the circuit simulator. The soldering iron is getting hot indicating the TRIAC circuit is working (I was hoping that 75 ohms was an indication that maybe the iron was fluctuating too much and that's why the 'ready' LED wasn't blinking).

At this point, everything points to the PIC as the culprit, however, the code was evaluated by Ian.m and Andy Watson to be solid. Plus I bought a new PIC and burnt the code to it (the modified code that allowed the flags or whatever it was to work on a blank PIC).

If anyone has some last minute ideas, I'd appreciate hearing them as I've been grasping at straws for the last several times I've tinkered with this iron.

[Ian.M]

Unless its in *VERY* poor physical condition, I wouldn't junk it.  Worst case, if you don't have the space to keep it, someone will want it for spares, though personally I'd chuck it on the shelf till I could get around to either replacing the whole temperature controller board, or writing my own temperature controller code in C to run on an upgraded but pinout-compatible PIC12F (reprogrammable) chip, or if you are feeling lazy, on an Arduino Pro Micro or similar mounted on 3M VHB tape and patch-wired over to the PIC footprint. 

[bostonman]

I don't disagree with you. Maybe my choice of words were a bit excessive, however, I dislike keeping stuff on my shelf that doesn't work, but, also, periodically it would force me to address the issue again.

This is eating away at me because the circuits are quite basic and straight forward (ignoring the deeper workings of the TRIAC circuit). My go to is the PIC being the culprit, but the code has been checked.

As for programming my own PIC, I don't know anything about code, and will need my iron soon. So far I've got away with a cheap Radio Shack, but I need a better one.

Thinking out loud, the whole top of the schematic isn't in question because it's producing 5V DC. The bottom left isn't in question because the op-amp seems to be producing the correct voltages for the corresponding input voltages. The TRIAC seems to be getting the proper signals and it's turning on indicating the whole top right of the circuit is good. The bottom right circuit is straight forward, and the LEDs work (they just don't blink ready).

I did question the RC coming from the secondary to the PIC (I don't have availability of the schematic at the moment). I posted a picture of the waveform earlier.

Otherwise, the only thing I can point fingers to is something funky going on in the TRIAC circuit causing the PIC to think the temp isn't stable enough to blink ready, or something in the code has been overlooked and got corrupt.
 

[Ian.M]

As I said *way* up-thread, there is a small but real possibility that the original PIC's OTP program EPROM is corrupt.  All it would take would be one bit loosing charge (i.e. going from 0 to 1) to change an instruction to a different one, and if the bit-flip changes the destination of a branch or the 'sense' of a test or compare, so closely studying the code^* doesn't show an obvious nonsensical instruction sequence, it would be near impossible to determine the program is corrupt, without a known good 'clean' program dump to compare with.

At this point, it would be very helpful if anyone else with a WES51 in full working order, and a programmer that can handle PIC12CE673 chips could dump their station's PIC and post the resulting IntelHEX file.   Unfortunately  Microchip's ICSP programming tools for modern FLASH PICs cant handle their old PIC12/16/18C parts, so unless you've got a legacy Microchip programmer for their old EPROM PICs or a universal programmer that supports the above part, you wont be able to help.

* Previously, I studied approx. 20% of the code to the level of detail required.

[bostonman]

You're correct, you did say this.

Hard to believe that the PIC could get corrupt so minor that it affects one tiny portion of its function.

I'm not saying you're wrong, I'm just stating the irony of it.

Somewhere above I posted a picture of the secondary sine wave and the pulse on the RC that goes into the PIC. I questioned why the pulse is higher in voltage than the sine wave wondering if that could be some indication of a problem.

Personally I feel the only things it could be: the RC is an issue (which I doubt) as I mentioned in the previous paragraph, the TRIAC measuring 75 ohms from gate to MT1 (although someone in the other thread claims to have measured 75 ohms confirming mine is good), or the PIC.

As for keeping this board as parts, it's somewhat hacked due to using a junk iron. I know what needs to be fixed, so for me it's an easy repair, but the diode at the TRIAC gate is removed, capacitor at the gate, and the TRIAC itself, along with other stuff that can be touched up.

[bostonman]

Although not cost effective, I bit the bullet and purchased a used and (hopefully) working WES51 according to the description for $50 including shipping.

My plan is to remove the PIC and extract the code so I can not only hopefully repair mine, but also provide the code to everyone else. I believe Ian.M fixed the code on my original so it can be burned to a blank PIC, so can you perform the same magic when I upload the new code?

I don't plan to do anything fancy with this PIC until the code is extracted and safe to burn to a blank PIC, so I'm not going to measure anything, compare the original code to the code on the PIC, just get the code extracted first.

Unfortunately I'm into this solder station for probably more than I should. The new iron was around $50, plus the cost of a replacement station, and a few components that I replaced and will replace. I'm hoping in the end I fix mine, and have a spare solder station too.