There's no function that detects higher/lower temp. After each PWM cycle, the adc reads, and the new values are passed to the PID.
The PID receives new reading, setpoint and gives back the new PWM value. Internally it holds some data based on last values.
PID can cause that and much more with wrong settings, did you figure out how it works? What do proportional, integrator and derivative mean?
Nothing changed on the SW in the dectection/processing. But PID values changed, ADC samples much faster now.
As you said, the delay time changed in last release. If you suspect of too low delay, then why you don't go ahead and try changing the delay yourself?
Is the menu there just for showing? Are you expecting me to try it on a hardware I don't have?
I thought it was a nice idea to port the firmware for other boards, but it has become a nightmare.
When you come here and check for new updates, behind that there are a lot of hours debugging, testing, changing, reverting, searching the source of the problems.
It's already time consuming when I can debug the hardware, but for something that you can't touch, can't see, it's like find a needle in a haystack unless the problem is very clear.
Most developers will directly refuse to develop for something they can't touch of that. And this time-consuming way needs to stop.
Sadly, support for any hardware else than Quicko will be discontinued soon. I will try to find the issues, but I won't promise anything.
I'm gonna set something clear:
- No more stupid suggestion and "maybes" if you don't know how it works.
- When I say no more stupid advices, I mean no more "But JBC does this", "But KSGER blah". Stop losing my time ok?
- There are nice guys that really help and properly report issues. You rock.
- There are others that only cause annoyance with "I want", "It would be better", "Why don't we?". Go to the candy store. This doesn't apply for all suggestions, some have been really good.
Want faster sampling? Because JBC does? Unless you start programming, debugging, measuring waves with the scope and working on the real thing yourself to help in this project, save your ideas.
Timing needs to go back to 10ms+. This is why we choose 20mS delay months ago
- When you measure too quickly, the TC is always hotter than the tip body.
- This is not a problem if there's always some power. Until you reach the setpoint.
- When the power is gone, the TC will cool down to the real tip temperature.
- That takes -You guessed?- around 5-10mS, and if you don't respect the timing, the next adc reading reports 10...20ºC less and the PID goes crazy high.
- The next power stroke reads the extra heat on the TC, not the real temp. PID goes kicks back reducing power. And everything happens again.
- JBC and others use ac, so it's a sine wave, when it reaches zero the power has been decreased being slowly for the last 5mS, so there's no sudden power removal like with PWM. Thats helps a lot.
- And that's why PWM drive requires more time.
- Some tips do much better than others. Some will do great with 200uS delay with some, while others need 20mS (Specially genuine Hakko tips)
- If you want it better, faster -That's why you have a customizable menu-. Tweak yourself, but if it goes wrong don't come complaining. If it works for you, nice, but I'm not going to put risky settings for everyone.