search

Ralim / IronOS

Open Source Soldering Iron firmware
https://ralim.github.io/IronOS/
GNU General Public License v3.0
7.14k stars 712 forks source link

Pinecil V2 has erroneous temp spikes in companion #1915

Open vguttmann opened 3 months ago

vguttmann commented 3 months ago

Describe the bug

When using IronOS companion, sometimes the tip temperature spikes up to 450C for one time point (see screenshot). I'm not sure whether this is an IronOS or a companion bug, therefore I'm opening an issue in both repos Screenshot_20240524-105154

To Reproduce

Details of your device:

Additional context

Other issue: https://github.com/aguilaair/IronOS_Companion/issues/10

vguttmann commented 3 months ago

Thanks to the help of the Companion devs, I was able to figure out that it's indeed an issue with IronOS. Interestingly, the spikes only appear when the iron is actively heating - not when it's in cooldown or idle. Also, the spikes are not consistent in height.

Ralim commented 3 months ago

Hia,

Did you test with the linked PR above? Part of this issue is to do with BLE and the adc not playing nice together (I think). Its normally filtered out on the iron itself, but BLE has no filtering.

This has been a thing around since BLE support was added (roughly, don't have an exact timeline and chat is busy to search back in). If I build with BLE stripped from the build the issues disappear for me here, but would love any other observations. Unsure what the root cause is though, as tracing the raw register read from the ADC does have corrupted data. Lightly suspect its timer + BLE interrupt fighting but unsure.

xam100 commented 3 months ago

Hi I have encountered a similar issue with a similar setup

I wasn't able to reproduce it in the same manner but I did see it happen during normal usage.

I did manage to reproduce temperature spikes by playing (contacting) with the solder tip using tweezers, and that spike did not reflect on the pinecil display. image

I did also notice random temp spikes when the iron is just laying on the table cooling down. image

Ralim commented 3 months ago

@xam100 Did you test with a build from #1916 as mentioned above?

xam100 commented 3 months ago

@Ralim just tried using this build , and yes it seems to be reproduceable.

Firstly a spike appears during the first power on, right after flashing, resetting settings and re-paring. image

Afterwards, the issue comes back occasionally. Strangely, the spikes don't always reflect on the bluetooth web page but seem to always show on the display, i.e. the pinecil display can be seen briefly temp spiking to 300-310 but the graph doesn't always have time to show that spike image

Ralim commented 3 months ago

Thank you for testing. Annoying to see its still occurring for you. I guess I have more work to do šŸ˜¢ . Will try and spend more time on it when I get some more time.

The screen on the Pinecil is showing a very,very lightly smoothed rolling average; where as the BLE will point-in-time sample the unfiltered reading so depending how often the issue arises will vary what you see on each.

Ralim commented 3 months ago

Hia, If you check a chance you can test this firmware image?

Pinecilv2_EN.zip

xam100 commented 3 months ago

Hi, tested 1069e461 04-06-24, three good findings.

Random temp spikes did not show up this time.

Temp spike after pairing is still noticeable and is corelated with high (45W) sustained power output. image

Very similar behavior is observed when re-heating the iron after an idle period (no re-pairing) image

Those spikes max out at around 350C on the pinecil display image

discip commented 3 months ago

@xam100 Sorry, I did not mean to delete your post. šŸ˜Æ

Unfortunately on my device it showed up as if this post was entered twice, so I removed the latter. But for some reason both posts vanished.

Iā€™m so sorry. Please forgive me.

Hope you will find the references you used again.

Ralim commented 3 months ago

Okay this is good. These temperature spikes are likely something else (PID) can you confirm what power source and what tip your using so I can replicate these ?

xam100 commented 3 months ago

pinecil B2 86mm Tip lenovo 65w 20v charger but sometimes only 45w is used image

The post accidentally deleted by @discip was showcasing another temp spike issue resulting from a sudden request of power here are two examples when it happened

https://github.com/Ralim/IronOS/assets/33205747/f1ba5fdf-bc2d-4a6d-8a71-226f71386881

https://github.com/Ralim/IronOS/assets/33205747/7ad8637d-ed79-4f36-a3a5-f7bed7073094

Ralim commented 3 months ago

pinecil B2 86mm Tip lenovo 65w 20v charger but sometimes only 45w is used

Aha, okay I'll have a look at tuning. The wattages displayed are only estimated so take them with a grain of salt.

The system is tuned to overshoot slightly on the readback when you do a large temp change like that but it shouldn't overshoot that much. (Some overshoot is desireable to help bring the outside tip temp up faster since the temp sensor is inside the tip and leads the outside).

Ralim commented 3 months ago

Just to test; could you try the build from here and compare for you: https://github.com/Ralim/IronOS/actions/runs/9426474334?pr=1925

discip commented 1 month ago

@vguttmann @xam100 Is this still an issue after the latest update #1925? If not, please consider closing this issue. šŸ˜Š

thanks in advance