Philips Air Fryer XXL (HD9654/90) Heating Issue - Seeking Advice

[w0nk1]

I'm facing an issue with my Philips Air Fryer XXL, model HD9654/90. Here's a breakdown of the problem:

Startup:
Upon powering on and setting a temperature (e.g., 180°C for 2 minutes), the display immediately shows 5 dashes.

Fan Activity:
The fan begins to spin, suggesting the unit is trying to start the cooking process. However, there's no heat being produced.

Unresponsiveness:
After a short period, the fan stops, and the unit becomes unresponsive. I have to unplug and replug to reset it, but the heating issue persists.

Thermal Fuses:
I've checked them, and they're intact, showing continuity.

Temperature Sensor:
When unplugged, it reads around 150 kOhm and decreases when exposed to a heat gun. But when connected to the PCB, I measure 220 Ohm at the connector, and this value doesn't change significantly with heat application.

Given these observations, especially the unusual behavior of the temperature sensor when connected to the PCB, does anyone have insights or suggestions on potential causes or solutions?

Any advice would be greatly appreciated. Thanks in advance!

[Swake]

150kOhm is a high value for a NTC sensor but it exists. As you say that resistance is reducing when heat is applied that sensor seems ok. Try searching for the specs of a spare part of that sensor and see if it matches what you measure.

Not heating might be caused by a blown thermal fuse. Frequently found near the heater.

[w0nk1]

Thanks for your answer! I checked both thermal fuses and they are fine. I accidentally touched the heating element during the "on" period and got a little electric shock, so I assume this is not the problem.
According to the official FAQ page:

https://www.philips.sa/en/c-f/XC000006976/the-display-on-my-philips-airfryer-shows-5-or-6-dashes

It is an overheating issue, which obviously makes no sense, but it could explain why it is not doing anything when it thinks it is already overheated during the startup.
I don't understand why the Ohm value of the temperature sensor won't change when it is plugged to the PCB. It is always 220 Ohm, even without the sensor plugged into the connector. Could this be the right behavior?

[Swake]

When unplugged, it reads around 150 kOhm and decreases when exposed to a heat gun.

It is always 220 Ohm, even without the sensor plugged into the connector.

It cannot be two things at the same time.

If the sensor is 150 kOhm on its own and the PCB reads 220 Ohm without the sensor plugged in, then yes it is normal to measure just half an Ohm less than those 220 Ohm when the sensor is plugged in (assuming it is all resistive). This is a parallel connection of resistors. R = R1xR2 / R1+R2

Should the sensor read 150k? It sounds somewhat high, but plausible, plenty of different models out there.
Should the PCB connector measure 220 Ohm? This measurement is not very relevant. Depends on the circuit, can be totally different ones powered. But it is very low compared to the sensor indeed and that doesn't make the measurement any easier for your frier.

Do you have a schematic, or can you trace it somewhat around that connector? Can you post pics of the thing?
Do you have a variable resistor of about 150 kOhm to simulate the sensor; higher resistance is good too?

[w0nk1]

If the Sensor is 150 kOhm on its own and the PCB reads 220 Ohm without the Sensor plugged in, then yes, it is normal to measure just half an Ohm less than those 220 Ohm when the Sensor is plugged in (assuming it is all resistive). This is a parallel connection of resistors. R = R1xR2 / R1+R2

It is always 220 Ohm, with the Sensor plugged or unplugged.
The Sensor alone reads at ~21 °C the mentioned 150 kOhm. I found some replacement parts for a similar model => 200 kOhm NTC temperature probe.

Do you have a schematic, or can you trace it around that connector? Can you post pics of the thing?

Unfortunately, I have no schematics, but I attached a picture of where the temperature probe connector directly goes. All resistors in the image I checked are fine and in range. The caps had no shorts, but I don't know their values. It looks messy because I desoldered them out of the circuit for measurement - sorry for that.

Do you have a variable resistor of about 150 kOhm to simulate the sensor? higher resistance is good too?

I tried a 200k potentiometer, but no difference at all.

[fzabkar]

If the sensor is really reading 150Kohm, and if it is an NTC type, then a reading of 220 ohm would correspond to "very hot". This means that there is something wrong with the 220 ohm reading. Or am I missing something?

[w0nk1]

If the sensor is really reading 150Kohm, and if it is an NTC type, then a reading of 220 ohm would correspond to "very hot". This means that there is something wrong with the 220 ohm reading. Or am I missing something?

I would definitely agree! But I have no clue where it comes from! :-/

[fzabkar]

What is the resistance of C106? If it is 120 ohms, then there is your culprit.

What is the resistance between R107 and the other end of R108?

[w0nk1]

What is the resistance of C106? If it is 120 ohms, then there is your culprit.

It is indeed in between the 115 - 120 Ohm range! (Measured in the circuit)

Your conclusion seems to be correct. ^^

What is the resistance between R107 and the other end of R108?

From the right side of R107 to the upper pad of R108, it is ~220 Ohm.
From the left side of R107 to the upper pad of R108, it is ~124 Ohm.

BTW: The value for the R107 is 101. The red circle in my previous post's image hid the last digit.

[fzabkar]

I think C106 is faulty.

[w0nk1]

I think C106 is faulty.

Do you think it is safe to remove it, or do you have an idea of a replacement value? Thank you so much for you support!

[fzabkar]

I think the capacitor is probably filtering noise. You could at least remove it and confirm whether the low resistance goes away.

[w0nk1]

I think the capacitor is probably filtering noise. You could at least remove it and confirm whether the low resistance goes away.

I desoldered it, but unfortunately, it is still 220 Ohm. This is driving me nuts!

[Swake]

'More' information is needed. Please trace into a schematic the components from the connector up to the PIC.

The picture that you have posted shows via's and traces going of screen and as component numbers are not sequential there is much more on that board.

[w0nk1]

'More' information is needed. Please trace into a schematic the components from the connector up to the PIC.

The picture that you have posted shows via's and traces going of screen and as component numbers are not sequential there is much more on that board.

I tried to come up with a schematic (I am not an expert) and added Front and Back side images of the complete board.
Can I do something else to clarify things?

I am very happy that so many are willing to help me; thanks! ;-)

[Swake]

Did just a quick check, have not much time right now but wanted to share this small finding:

One side of the sensor is on the GND. The red circle you have drawn on the bottom side of R108. It is connected to pin 29 on the PIC, and also connected on the right side of R108 under the letter R there is a via, on the other side of the board it is clearly ground plane.

The other side of the sensor is the red wire on the connector, right?

[fzabkar]

I would think the top side of the voltage divider for the temp sensor would be connected to Vref+ (PIC pin #2) via another resistor. Perhaps Vref+ is shorted to ground???

R108 is a 10 ohm resistor. I would measure the resistance between ground and each end of this resistor. This will tell you which side of the resistor the short is on.

[w0nk1]

One side of the sensor is on the GND. The red circle you have drawn on the bottom side of R108. It is connected to pin 29 on the PIC, and also connected on the right side of R108 under the letter R there is a via, on the other side of the board it is clearly ground plane.

Meh, I have seen, that I got an error in the PIC pins labeling, it was the wrong side. I updated the schematic. The pins are: 33 - 23.
The bottom side of R108 is connected to pin 34 of the PIC and it is indeed connected to GND.

The other side of the sensor is the red wire on the connector, right?

Yep, and the wire beside it, is the other pin of the temp sensor.

// edit: BTW: I added an image of the PIC datasheet and hope it is the right one.

[fzabkar]

I think pin #27 of the PIC is leaking to ground internally.

[w0nk1]

I think pin #27 of the PIC is leaking to ground internally.

This would mean the MCU itself is dead, right?

[fzabkar]

I think pin #27 of the PIC is leaking to ground internally.

This would mean the MCU itself is dead, right?

Yes, I can't see any other possibility.

[w0nk1]

I think pin #27 of the PIC is leaking to ground internally.

This would mean the MCU itself is dead, right?

Yes, I can't see any other possibility.

This is such a pitty!
BTW: I desoldered the PIC and now I measure something around 9 kOhm and it is changing when applying heat to the temperature sensor.

[Swake]

I have no better conclusion unfortunately.
A last chance is to replace the PIC but that would require that you can read it out. I would guess they set the protection bits on this one to make that impossible. Don't know much about PICs, there might be ways to circumvent that.

[w0nk1]

I have no better conclusion unfortunately.
A last chance is to replace the PIC but that would require that you can read it out. I would guess they set the protection bits on this one to make that impossible. Don't know much about PICs, there might be ways to circumvent that.

Okay, thank you guys! Not sure it if makes sense to go into this rabbit hole - but we will see! ;-)

[NiHaoMike]

Controlling an air fryer isn't exactly rocket science, could probably just have the fan be constantly running and install a generic PID controller to regulate the temperature.