P2 Edge - Thermal image

I received my P2 Edge module & a JM P2 Edge Module Breadboard yesterday. I decided to fire it up today, to see if it all works. I ordered a new prop plug with it but it wasn't in the package. Luckily my old prop plug still works (the cable is bad on it) the P2 module & Breadboard are both excellent. Great designs.

I used a Flir One Pro camera - Android C connector version.

The second picture is the HDMI Spiral program running on the P2 Edge module and displayed on my HDMI monitor.

So far everything is working great!

Comments

  • Bob, I will have Chantal send you a new Prop Plug. Sorry it wasn't in the package. Did it indicate we were out of stock, or anything?

    I am really curious to see a thermal image of the back side of the P2 Edge. The ground plane is split, so I expect there is going to be a hot portion and a cool portion. On the next revision, the ground plane is solid, so the heat should be very slowly wicking away from the center.
  • nice photos Bob
  • re: Did it indicate we were out of stock, or anything?

    Thanks, Chip. I sent sales an email yesterday, I'm sure they will look after it. The invoice did not indicate that it was out of stock. I will use a glue gun on the old Prop Plug to enhance its longevity, for now. It's only the outer casing that is frayed. I added a picture of it :)


    I added a picture of the P2 Edge from the back. The one thing I thought of since the first image was taken is that I did not time how long the chip was powered on before I took the image. I waited for 5 minutes, of warm-up time before I snapped this image.
    6000 x 8000 - 9M
    1080 x 1440 - 1M
  • Thanks, Bob. The gap in the back ground plane is not very evident. Maybe it doesn't really matter.
  • evanhevanh Posts: 10,230
    edited 2020-11-23 (5:26 PM)
    Bob,
    Run this attached program. It's a benchmark I made a while back for testing power consumption. I've got it preset for the hungriest program but at 100 MHz so it shouldn't overload the power supply.

    EDIT: Forgot it was still forced to RCFAST. Corrected. :)
  • @evanh
    Bob,
    Run this attached program. It's a benchmark I made a while back for testing power consumption. I've got it preset for the hungriest program but at 100 MHz so it shouldn't overload the power supply.

    Ok, thanks, I'll try it :)
  • I ran evanh's benchmark program and captured some new images. the program seemed to work great. Great job evanh :)

    Front View - P2 Edge Module - I ran the program for 30 minutes before capturing the image.

    P2_30M_test_FrontView.jpg

    Back View - P2 Edge Module - I ran the program for 30 minutes before capturing the image.

    P2_30M_test_BackView.jpg

    Oscilloscope Output on Pin 16

    ScopeOutput.png

    1080 x 1440 - 1M
    1080 x 1440 - 1M
    979 x 575 - 994K
  • @ Tubular
    nice photos Bob


    Thanks! I'm using a brand new camera and haven't had time to even read the instructions yet. However, it does look good to me too. I wonder what options I have to enhance it a bit? LOL
  • Looks pretty useful how it is. Interesting that there's a 10 degree diference between the front (chip top) and back, I wouldn't have guessed that.

    I have a flir C3 that I'll try doing the same thing on. We did take some MicroPython current measurements the other day, it'd be good to tie that back to temperature rise

    One thing I noticed with the C3 is that if you take the photo kind of upside down, it narrows the distance between the optical and infrared cameras, so you don't get as much 'offset' in the image. Well, it does when taking photos on a bit of an angle, rather than directly overhead. I don't know whether your camera would do the same.

    It won't be too long before we have the melexis grid (90640/641?) and hope to get this running with the P2, and Rayman has the Flir Lepton talking already
  • evanhevanh Posts: 10,230
    edited 2020-11-24 (3:01 AM)
    Bob,
    Feel free to notch the clock frequency up further for higher temperatures. At that 100 MHz it will be drawing around 650 mA on Vdd, which translates to 1.17 Watts. 180 MHz should push past 2.0 Watts.

    All you need do is edit the XMUL = 10 constant on line 110, and change it to 18.

  • Oh, good you noticed the frequency output pin P16. It's sysclock/1000, so the 100 kHz means all is correct. I needed that monitoring when pushing the 360 MHz testing I did. It didn't take long before internal heating would limit the PLL frequency and reduce the speed of the prop2.

  • The thermal picture from the back of the board does show the split in the ground plane, it seems. The next version will have a continuous ground plane.
Sign In or Register to comment.