Tempterature gauge shoots up, STOP light turns on and then the gauge drops to 0

[Arkis]

Hello,
We have a Peugeot 206, 1.4i 55kW, 2005. Since buying it a few months back, we have one problem with the cooling gauge.
When we drive more than 20 km, there is strange behaviour with the temperature gauge. It starts at 0 and starts to crawl up slowly (way too slow), suddenly it shoots up to the red line and the STOP light turns on. After around 20 seconds, the temperature gauge falls to 0 and the STOP light turns off. It repeats every 20-30 kms while driving. The car never overheats even when idling.

I have done quite a few diagnostics around it already.
1) The OBD scanner showed me the temp of the coolant was -39°C. That's simply wrong data so I changed the temperature sensor. Since then, the OBD scanner is showing good data but the problem remains.

2) The thermostat opens normally at around 70°C (measurement from the OBD scanner, on the dash it barely moved from default position).
I found that there is often a problem with fan relays located next to the fan. There is a single fan which is supposed to be 2 speed. But mine works only at higher speed, which triggers at 99°C. That's the exact point where the dash gauge shoots up and the STOP light illuminates. When the fan cools the engine down a few degrees and the fan turns off, the dash drops to 0 and the STOP light turns off.
I tried swapping the relays but the problem remains.
So I tried to remove the cover from the relay and trigger it manually. It triggers only the high speed, but both relays are good.
Then I tried to measure the voltage on the relay sockets and it matches with each other so there is electricity going to both relays.
Could it be simply a bad fan resistor? From pin 1 to pin 2 it shows 1,2 Ohm and it gradually drops to 0,7 Ohms within around 5 seconds. From pin 1 to 3 there is no resistance measured. From pin 2 to 3 there is no resistance measured. I am not sure of the inner layout of the resistor so I do not know if it is bad. I found on the internet that the resistance should be 0,8 Ohm (+/- 0,1 Ohm). But the resistor wouldn't be responsible for bad gauge reading, would it?

Edit: The coolant was changed this week while doing timing and water pump change. The system is fully bled.
Edit2: When I take the fan resistor out and bridge the socket with a simple wire, the fan works on both relays.

[GiveMeABreak]

Let's have the VIN and the specific fault codes (bear in mind, depending on your OBD scanner, it may not be able to provide detailed constructor codes - only basic codes), so you could be missing fault information.

[Arkis]

Hello, the VIN is VF3**************[VIN obfuscated, can be read by forum staff]. It's 1.4 55kW 8v, with AC, single fan with dual speed control and with single temperature sensor on the thermostat (green sensor).
OBD scanner and Lexia/PP2000 showed no codes (there was only one code related to the battery change).

[GiveMeABreak]

Ok, so no recorded technical issues with the cooling system and data is very limited on these earlier petrol engines. The engine ECU will trigger the fan speed as necessary, but if it's not kicking in at the lower speed, either the data the engine ECU is getting from the other sensors is duff or the fan control unit is shot. Hard to say.

[PaulC5]

The thermostat starting to open at 70°C seems low, I would have thought it ought to be nearer 90°C. If it is faulty then as well as opening low it might also be shutting or not opening fully when it should be full open. I once replaced a thermostat on a Saab and then the temperature would go up and down and it turned out to be a faulty thermostat despite it being new. We had a 2003 C3 with the 1.4i petrol engine and that had 2 temperature sensors. On a couple of occasions in this car in winter on a motorway the temperature jumped to max. I did lots of checks and found nothing wrong. I later replaced the timing belt, water pump, thermostat, anti freeze and it never happened again but I do not know if this fixed anything.

[Arkis]

I still felt bad about the measurement on the fan resistor so I ordered new for comparison. Now I know there should be a connection between pin 1 and 3 on the resistor. The old resistor was unmeasurable. The new resistor showed 1.8 Ohm.

So I changed the resistor and the fan started spinning around 80 °C (data from the OBD scanner, 70 °C on dashboard). I was unable to get to the 99 °C for the second fan stage because the fan was cooling the engine and I am definitely not running the engine on 4000 rpms for 10 minutes to get it to the 99 °C threshold. I will have to do some test runs on the street to see the temps.

But the first speed of the fan is definitely fixed. Now there is only the gauge that bothers me. How is it possible that the OBD scanner and the dashboard say different temps?

[Rp0thejester]

Would the OBD scanner read it at the fan end and the dashboard read it at the other side of the engine?

[Arkis]

There is only one temperature sensor so I assume both (OBD scanner and gauges) have the same data?

[PaulC5]

The dash gauge is probably not an accurate value. It might be a value indicated to make you feel happy rather than an actual one which is used to control anything. You might be able to remove and refit its needle so it gives a value nearer the scanner one. The 2003 C3 we had had a gauge which was 6 lights that lit up with each one being a band of about 10°C and as long as it was on band 3 or 4 I was happy.