FlirAI,
I can provide some comments on the Microbolometers as there are data-sheets already in the public domain.
It is no secret that modern microbolometers can output much higher frame rates than normally seen on cameras. 120 frames per second has already been seen from a ULIS Microbolometer running in a fire fighting camera.
When considering a microbolometer we need to remember its architecture when considering its capabilities.
In précis, it has a matrix of pixels that are read out in rows and columns in a serial or parallel reading mode.
The pixels are only biased for the read cycle to avoid self heating issues. The Read-Out IC carries out the reading of the rows and columns plus, in some Microbolometers, it handles timing functions and analogue to Digital conversion of the signal coming from the read-out matrix. The ROIC may be quite specialist in its content, but it is not the limiting factor in maximum frame rate, nor is it something that would lead to performance binning of finished modules. Self heating can be an issue with the ROIC but is not a great issue at 30fps or 60fps.
Where binning is concerned, I can advise that such is not applied to the FLIR cameras. All of their Microbolometers have to achieve a minimum standard of functionality to be used in a camera. This usually relates to the number of functional pixels and not read out IC performance. FLIR offer a 10 year warranty on their Microbolometers and have no interest in sub spec units ending up in their cameras. One of the challenges in Microbolometer production is achieving an industry standard acceptable percentage of functioning pixels and, where dead pixels are concerned, a maximum number of pixels in failure in a particular defined area. It is the creation of the Microbolometer pixels that is challenging and not really the ROIC that resides behind the die.
The read out limitation of a Microbolometer is, in reality, set by the physics of the pixels and associated read-out capacitors. It is well known that many thermal cameras that are limited to less than 9fps actually output 30fps or 60fps to the video processing stages. The exact nature of how the <9fps limit is imposed is what I cannot comment on. There are several ways to achieve it in an effective manner. Some are better for the cameras end user experience than others.
So now you know, the Ex series and Exx series both output high frame rate data from the Microbolometer. If you look at Mike’s Teardown and investigation videos you will see him discovering this with an oscilloscope. That does not help you with circumventing the frame rate limitation though. You could build a new video processing stage to create a high frame rate camera but you would basically be rebuilding the camera to do so.
This is where my commentary must stop though
All the best
Fraser
Home
Help
Search
Login
Register
