Temperature spikes after FFC

[Tomas K]

Hello,

Lepton 3.5 automatically performs FFC every 3 minutes. When FFC occurs, the temperature jumps approx. 2 °C and it takes some time (30-60 seconds) to stabilize back to the original temperature. The similar effect can be seen on professional termocameras but the temperature stabilizes after a few frames after FFC and not after 1 minute. It is almost impossible to measure the temperature with such behaviour. Disabling FFC is not a solution, because I guess it will degrade the image in time.

This effect can be seen at attached images when camera is facing at (almost) temperature-constant object (not black body).

I'm thinking to create some normalization curve dependent on the time since last FFC using black body and apply it to temperature reading. But I'm afraid that it won't be so easy, because this curve can be different for different temperatures and can have different run even for the same temperature (as seen in FFC2.jpg).

Has anybody already tried to deal with this?

[Lander T'Kindt]

Tomas

Did you execute the measurements in an environment that has a different temperature than the camera? I'll explain what I mean by that: For my experiments, I have to enter a laboratory cold room that is at -5°C. The camera itself has been laying around on my desk for a while before measuring, so we assume that it has a temperature of 20°C. When entering the room, the camera starts measuring immediately but it also gets colder. I noticed that this temperature shock caused big spikes.

If it is not the case, I am very interested as well as to what the source of this problem is. I'm also interested in the code of your FFC generation. It would be rather useful for me as well to have such graphs - would you be willing to share your code of that?

Thanks & good luck!

[Tomas K]

The measurement was done in normal home room environment. I don't know its exact temperature but it is neither cold nor warm. Camera's temperature was stable around 38°C. Our camera is closed in the package, but I tried also unpacked camera (FPA Temp is around 33 °C then) and also after the cold start up (graph attached). It happens in all conditions, just the spikes' "height" is a bit different. The spikes appear after FFC only - regardless the camera's automatic FFC or manual execution via LEP_RunSysFFCNormalization() or LEP_RunRadFFC().

Our software just stores the temperature value for certain temperature-constant object. The graphs have been created in Excel.

Dne pondělí 6. dubna 2020 10:21:37 UTC+2 Lander T'Kindt napsal(a):

[Tomas K]

We were finally able to do Lepton measurement against a black body set to 37°C.

- measured value is higher than the real BB temperature

- measured value seems to be dependent on FPA temperature - the higher the FPA temp is, the slightly higher the measured value is

- 1-1.5°C spikes appear after every FFC

I'm posting the part of the results to see how the Lepton behaves. One is uncovered Lepton camera (window transmission = 1.0), the second one is Lepton covered in plastic box  together with RPi3 (so much higher FPA temp is reached) with Germanium window. Window transmission = 0.89 and window temperature is continously set to Lepton housing temperature (that is get in frame telemetry data) every time it changes more than 0.1°C.

[Zoltan Szalontay]

I experienced exactly the same behaviour with my Lepton 3.5. 

Closed room, stable ambient temp, collecting the temperatures of a certain fixed point that does not change its temperature. Log shows exactly the same spikes of around 1-2 Celsius degrees that go down to normal in cca 30-40 seconds. The spikes happen at the very moment FFC is executed.

This is very disturbing. I do not think disabling FFC would be any better in long terms, let alone that I am developing in Python (flirpy) and it does not support control over USB.

The only thing I can do is to measure a Black Body temp and dynamically correct every single pixel based on the Black Body temperature difference, but that would cause another problem: the distance and the emissivity of my BB (or any known temperature object) and my object to measure are different.

Any idea is highly appreciated. Especially, if it is in Python :-)

[Zoltan Szalontay]

Here's my chart. Time axis is missing, but the spikes are exactly at 3 minute intervals.

I'm posting the part of the results to see how the Lepton behaves. One is uncovered Lepton camera (window transmission = 1.0), the second one is Lepton covered in plastic box  together with RPi3 (so much higher FPA temp is reached) with Germanium window. Window transmission = 0.89 and window temperature is continuously set to Lepton housing temperature (that is get in frame telemetry data) every time it changes more than 0.1°C.

[andrewv...@gmail.com]

This really doesn't help you, but you're definitely not alone with this issue.

My FLIR has spikes like this when FFC kicks in -- fortunately, I'm only monitoring rabbits, so it doesn't really matter to me if the frame is a few degrees hotter for a bit!

[Tomas K]

After a lot of black body measurements at different temperatures in whole Lepton range (-10 - 140°C) and playing in Excel, I found out that it is possible to define "offset" curve and use it to convert the temperatures to correct one. The curve can be either exponential or polynomial of higher (10+) degree with timeSinceFFC as input value. The main problem is that FFC peaks are not always the same - they are similar at same temperature (so you can simply average all FFC intervals) but a lot different for different temperatures so you never get perfect fit in whole temperature range even when using multivariate regression of very high polynomial degree.

Dne neděle 5. dubna 2020 11:38:32 UTC+2 Tomas K napsal(a):

[Fernando Paz]

I too am getting temperature spikes on my PureThermal 2 (Lepton 3.5).

My code is in JavaScript, so I'm streaming it as a webcam. I wonder if there's a way to control the FCC or any other measure to keep the readings consistent.

Any ideas?

[su william]

I want to deal with that, too. I disable FFC. Temperature values can  fluctuate.1 minute interval.

Have you solved the problem?

[su william]

I want to deal with that, too. I disable FFC. Temperature values can  fluctuate.1 minute interval.

Have you solved the problem?

On Sunday, April 5, 2020 at 5:38:32 PM UTC+8 bigm...@centrum.cz wrote:

[Yedidya Harris]

Hi everyone,

I'm encountering a similar issue. Due to ghost images, I perform an FFC before capturing each image. However, I’m noticing that while the plants I’m monitoring are around 25°C, the temperatures reported by the Lepton 3.5 are significantly higher—around 47°C.

Has anyone found a reliable solution to this? For context, I’m using the Cam Mini with a Lepton 3.5 module.

Thanks in advance!