ISC-027BW missing the target temperature and go over

[matteo.sorba]

Hello,
I tried to follow the instructions:

• I initially turned on the bbq;
• I inserted the ISC-027BW module (without any air leakage).
• I set the vent-outs to half open;
• I set a target temperature of the bbq to 107°C;
• I set a demo time of 2h;
• I used 4 other test probes as differential temperature detection;

Despite these settings, the end result was that the bbq reached 141°C, exceeding the target temperature by 34°C.

As the screenshoot testifies, with 3°C degrees missing from the target temperature of 107°C the fan should have been off a while ago, but instead it was still at 15%.

All probes tell confirms that the temperature was 140°C (± 2°C) exceeding the target of 107°C.

What am I doing wrong?
How can the fan trend be adjusted to meet the thermal hysteresis cycle of my bbq?

Test environment data:

• Kamado Monolith classic (46cm grill diameter)
• Dry run test (no food)
• Iphone 12 pro
• Main Module V2.5.2
• MCU module V1.0.3

[matteo.sorba]

On this screenshoot you will see the entire graph:

[matteo.sorba]

And this is the exact moment when the regulation plan breaks down: the fan turning off too late and too little to succeed in slowing down temperature growth.

[Tania]

Hello,
According to the information you provided, ISC-027BW is the latest version. Its algorithm is optimized.
Please understand that in actual use, the charcoal of the grill is a variable. The ISC-027BW adjusts the fan speed according to the temperature change. Since it is not electronically controlled, the more/less charcoal may cause a slight deviation in temperature.
When the target temperature is 3°C away, based on the ISC-027BW database, if the fan speed becomes 0%, some ovens cannot heat up to the target temperature. 15% fan speed is very small. If it causes the temperature to overshoot, it may be that the charcoal fire is relatively strong. Maybe you can lower the target temperature by 10~15°C? Will it be improved?

[matteo.sorba]

Hi Tania, thank you so much for your reply.
I lit only a small amount of charcoal and kept the lid closed for the duration of the ignition.

Let me say that I would not want to try to guess a target temperature other than what it should be: it would mean that the PID would not work if it detected the temperature of 107° and its target was 80/90°C.
I have purchased this device to not go crazy with temperature regulation so…

Many people have the problem of temperature control could you know which library you are using?
Take a look on this discussion about temp overshooting:
https://groups.google.com/g/diy-pid-control/c/XxE8jJP_Hzc?pli=1

Each device has its specific characteristic thermal hysteresis curve and by itself the algorithm does not seem to work.
Would it be possible to provide an update where to adjust these parameters?
Or would it be possible to adjust fan speeds?

I saw that in the settings of the other model, the ISC-007BW, you can specify the type and size of the BBQ. I guess this affects the parameters of KP, KI and KD in some way.

In any case there is official documentation on:

• how much to keep the vent-out open (the top one, the more air comes out the faster the airflow will be);
• knowing what the timer functionality is for? How does it affect the control parameters?
• best-practices, can the device be used only when the ignition is complete or can it also be used to bring the BBQ up to temperature?

Thank you so much for your help.

[matteo.sorba]

Hi Tanya, any info about?

[Tania]

Hi,
I contacted the engineers, and they reviewed the screenshots of the fan changes and replied that the fan output changes were reasonable. Perhaps you could run the oven through several cycles and see if the temperature overshoot returns to the target temperature with less oscillation over time, potentially reducing or converging the overshoot amplitude?
The ISC-027BW automatically adjusts the fan speed based on the temperature, so there is no need to select an oven model.

1. Regarding the oven’s top vent, it needs to be open, but there are no fixed data since different ovens have different insulation characteristics.
2. The timer is used to trigger an alarm when the set time is over and does not affect temperature control.
3. Both scenarios you mentioned can be used.

[matteo.sorba]

Hi Tania,
let me tell you what I found out from my tests these last days and if you can share this information of mine with the engineering team for everyone’s benefit.

In a generic kamado – mine is a Monolith classic – because of its efficiency (large thermal inertia, low dispersion, direct air flow, more uniform heat) combustion is more efficient resulting in slower airflow than in a traditional bullet smoker or offset bbq.

The algorithm can hardly get the device’s thermal hysteresis right the first time: the stage at which it would go into correction, stopping ventilation at zero, after going above the target temperature is now too late.

In fact, a kamado, with its efficient combustion and great insulation, would maintain that temperature even with combustion off for more than 1h under optimal conditions.
Too much time to then manage the adjustment, even the risk is to turn off the combustion altogether, then making the fan intervention futile when the algorithm realizes that the device is now below the target temperature.

So understanding this, I ran a test trying to make combustion as inefficient as possible by almost completely shutting down the vent-out and igniting a small amount of charcoal, so that I had to rely entirely on the controller’s ventilation to ensure combustion.

And so the magic happened: the controller was able to not overshoot the target temperature too much and keep the range within a reasonable range and in a reasonable time.

So I invite you to reconsider your responses to those who complain of the same problem:

• Yes, charcoal counts among the factors, especially in the low&slow (107°C as the target temperature) little of it should be turned on and in one spot only;

• Yes, vent-out is very important, it is not true that it doesn’t matter, especially in those high efficiency devices like kamado;

• Yes, instructing the algorithm with the device type (even if only in broad strokes, just as you do on the ISC-007BW) would help the device to readjust how to intervene on ventilation;

I remain available for further testing; I would be very happy if you would consider my feedback.

Attached are curves demonstrating the success of my test starting at 110°C and after at 170°C.

Matteo

Member of the Devils BBQ team - Meet us at the European BBQ world championships in Stuttgart on July, 27-28, 2024

[Tania]

Your test data is exceptionally valuable and serves as an important reference. I deeply appreciate you sharing this crucial information . I will promptly send it to engineering team, I believe it will greatly assist in improving the product.