Staff Assigned: Adam Robichaud Patch notes and latest linux API

Discussion in 'DL Imaging Drivers - Aluma, STC, StarChaser' started by FlavineC, Oct 18, 2023.

  1. FlavineC

    FlavineC Standard User

    Joined:
    Oct 18, 2023
    Messages:
    2
    Hello,
    I work with an AC4040 and our optimal use is with linux, however I noticed differences in performance between my code using the DLAPI on linux, and with using the MaxImLT software on windows.

    My version of the linux DLAPI is the 2.1, whereas I have the latest version of drivers and software on windows.
    I was wondering if you had a page with your patch notes between software and API updates ?
    I also wanted to know if the 2.1 DLAPI is the latest version there is, as the one on the download page seems to still be in 1.4? And if there is a newer one, may I request the latest version of the API for linux?

    The performance issue is related to the camera maintaining its temperature at high FPS using MaxImLT, but not managing to do the same in linux with similar requests.

    I look forward to hearing from you,

    Flavine CABY
     
    FlavineC, Oct 18, 2023
    #1
  2. Colin Haig

    Colin Haig Staff Member

    Joined:
    Oct 27, 2014
    Messages:
    7,962
    Location:
    Earth
    I've asked my colleague to assist.
     
    Colin Haig, Oct 18, 2023
    #2
  3. Adam Robichaud

    Adam Robichaud Lead Developer Staff Member

    Joined:
    Sep 29, 2014
    Messages:
    1,042
    Location:
    Ottawa
    2.1 is the latest version of DLAPI we've ported over to Linux. We have an update in the pipeline, but they're queued behind other critical tasks.

    All temperature regulation for the AC4040 happens in the camera—meaning that unless the setpoint being sent to the camera differs across platforms, or the USB interface is being pinged so frequently that the camera's main loop can't manage the cooler, they should behave the same across all platforms and architectures.

    Can you provide us with the camera's setpoint (preferably the value returned from DLAPI) maximum achieved cooling during normal operation on both OSes?
     
    Adam Robichaud, Oct 18, 2023
    #3
  4. Doug

    Doug Staff Member

    Joined:
    Sep 25, 2014
    Messages:
    10,316
    If you do a very rapid series of short exposures, the temperature of the sensor may rise a bit. Unfortunately that is unavoidable given the CMOS architecture.
     
    Doug, Oct 18, 2023
    #4
  5. FlavineC

    FlavineC Standard User

    Joined:
    Oct 18, 2023
    Messages:
    2
    Thank you for your reply. During testing the camera was set to cool to -10°C, and as the aim is to do a rapid series of short exposure, the cooler power was often close to 100%. Depending on the exposure time and delay between images it doesn't always manage to stay at this temperature.
    What do you consider a rapid serie of short exposure? For example, would 700ms exposure time with 1s delay be considered too close to the camera's limit?

    Currently our system pings the camera for status update at 1Hz, is the frequency for status update on MaxImLT similar?

    We own multiple AC4040 and have noticed that the newer camera are different from the older ones, in that they are heavier and the back of the camera which holds the cooling system is larger. Is there a significant difference of the cooling system between different generations of cameras?
     
    Last edited: Oct 19, 2023
    FlavineC, Oct 19, 2023
    #5
  6. Doug

    Doug Staff Member

    Joined:
    Sep 25, 2014
    Messages:
    10,316
    It really depends on conditions. Temperature regulation is inherently less stable if you are running close to 100%. If you're running at 95%, there's only 5% more power available to lower the temperature, which means it happens slowly - you have much less control authority. A bit more headroom would make a big difference to thermal regulation. If you're running this close to the limit you might want to consider increasing the setpoint or going to the water cooling option.

    As part of our ISO9001 process, we are actively engaged in continuous improvement. Since it was first released, we have made several minor improvements to the Aluma AC4040.

    As you have noted, the most visible change is the heat sink cover. That wasn't done for performance reasons; it was mainly done to help control costs in light of massive inflation in the cost of machined aluminum (over 40% in the last two years). We're trying to hold the line on camera prices as much as possible, and this particular change made a surprisingly large difference to cost.

    There are several internal improvements in the camera - minor design and material changes - that have improved the delta-T performance of newer cameras by perhaps 2-3 degrees C.
     
    Doug, Oct 19, 2023
    #6
  7. Colin Haig

    Colin Haig Staff Member

    Joined:
    Oct 27, 2014
    Messages:
    7,962
    Location:
    Earth
    I think Franck sent some sample images to us this morning.
     
    Colin Haig, Oct 19, 2023
    #7
  8. Colin Haig

    Colin Haig Staff Member

    Joined:
    Oct 27, 2014
    Messages:
    7,962
    Location:
    Earth
    FlavineC - can you provide the camera serial number and firmware version?
    The serial number is on a label on the camera body. It is also available using the Windows-based DL Config Utility or from the DLAPI SDK unsigned int dl::ICamera::Info::firmwareRevision
     
    Colin Haig, Oct 19, 2023
    #8
  9. Colin Haig

    Colin Haig Staff Member

    Joined:
    Oct 27, 2014
    Messages:
    7,962
    Location:
    Earth
    Hi Flavine -
    We checked manufacturing data and believe your camera has revision 13.

    A change was made in v14 to optimize cooler performance and reduce noise. Your camera appears to be having the issue that this revision cures.
    https://diffractionlimited.com/aluma-fpga-firmware/

    The current version is v15. Please upgrade the camera firmware to v15.
    The DL Config Utility on Windows can perform the upgrade.
     
    Colin Haig, Oct 19, 2023
    #9

Share This Page