« PreviousPages: 1 2 [3]
mit | Posted: 12 May 2025, 01:15 PM |
---|---|
![]() yeah whatever Admin Posts: 601 Joined: 4-May 16 |
Hi Jeff, that's great to hear! Originally I used a Sony RX100 which can do 1000FPS. The rolling shutter has an interesting effect, that makes the last digit look completely static in the viewfinder, but as you tilt the frame up and down the number displayed changes. But it turns out that studios with high-speed cameras are often quite enthusiastic when you approach them and say you've got something interesting to film in slow motion. I got some time on a Phantom Flex 4K, which can do 3000FPS at 1080p, or much faster if we drop the resolution. I'll share some of the footage when I, err, finally publish the clock properly... @SolidElectronics - there's a python script nmea.py in the repo that can be very helpful for debugging. The lookup table overflowing is exactly the kind of problem that means the clock behaves correctly for 50 years, then suddenly updates daylight savings on the wrong day. I get paranoid about these sorts of things. ------------- |
[top] | |
geerlingguy | Posted: 13 May 2025, 09:05 PM |
![]() Jeff Geerling Member Posts: 3 Joined: 12-May 25 |
QUOTE (mit)
I'll share some of the footage when I, err, finally publish the clock properly... Time is relative anyway, right? :) I might have an opportunity to get time on a much higher framerate phantom later this year! Working on that now. I'll be briefly mentioning the clock in a video later this week, but no pressure at all... it happens when it happens, and when you make something as cool as a millisecond-accurate clock, it's kinda timeless anyway. I've been running mine continuously over a week now with various GPS antennas, and the U-blox you're using is nice and selective. Not quite as good as the M9/F9s I'm testing in my Pi servers, but better than the same-generation UBX-M8030 I'm testing in a USB solution. I can get a solid lock from inside my building with almost any of my antennae, whereas I have to place it outside for my USB M8. I also just noticed there's a PPS LED on the back, I hadn't seen that before, but I like that! (And thanks for the PPS test pad, quite useful). ------------- |
[top] | |
naikrovek | Posted: 16 May 2025, 05:04 PM |
![]() Member Posts: 3 Joined: 11-May 21 |
oh geez i can't wait. I will buy it and the nanosecond version mentioned earlier in the thread. i have the skill to maybe create one of these myself, actually. but do i have the focus... no. no i do not. ------------- |
[top] | |
krixt | Posted: 16 May 2025, 07:00 PM |
![]() Member Posts: 1 Joined: 16-May 25 |
I'm here because of @geerlingguy's most recent video (mac gps time) where he briefly showed this clock. Disappointed when I saw the sold out notice, but ray of hope for "new version will be launched soon." Is there a waiting list, or newsletter I can subscribe for updates? Feature request idea: Ribbon cable between "date" and "time" PCBs to allow for vertical positioning, rather than one long horizontal. In my environment, having time on top, and date below would be more aesthetically pleasing. ------------- |
[top] | |
geerlingguy | Posted: 18 May 2025, 12:23 AM |
![]() Jeff Geerling Member Posts: 3 Joined: 12-May 25 |
QUOTE (krixt)
Feature request idea: Ribbon cable between "date" and "time" PCBs to allow for vertical positioning, rather than one long horizontal. In my environment, having time on top, and date below would be more aesthetically pleasing. That's exactly how the MkIV version works (as shown in the video — I have it in the stacked configuration so it fits better in the 16:9 shots); though by default it has the date on top. A firmware change, I'm sure, could make that the opposite. ------------- |
[top] | |
mit | Posted: 20 May 2025, 10:48 AM |
![]() yeah whatever Admin Posts: 601 Joined: 4-May 16 |
QUOTE (geerlingguy)
That's interesting, I wonder if it's to do with the board layout. A lot of the simple breakout boards for GPS modules pay no attention to impedance matching the track between the antenna and the module, because that's very hard to do on a 2-layer board.I've been running mine continuously over a week now with various GPS antennas, and the U-blox you're using is nice and selective. Not quite as good as the M9/F9s I'm testing in my Pi servers, but better than the same-generation UBX-M8030 I'm testing in a USB solution. I can get a solid lock from inside my building with almost any of my antennae, whereas I have to place it outside for my USB M8. The MkIV clock will function as a USB serial device, I did experiment with trying to send PPS over USB, you can send state commands for the secondary signals like DCD, but unfortunately the jitter was several milliseconds, which defeated the purpose. Nothing beats having the PPS pin exposed. In something of a marathon session last week, I wrote up maybe 90% of the documentation, so I'm closing in on being able to release it. Still need to make a video about it (but that can wait). I really ought to set up a mailing list for people waiting to buy one. QUOTE (krixt)
Haha, great minds think alike...Feature request idea: Ribbon cable between "date" and "time" PCBs to allow for vertical positioning, rather than one long horizontal. ------------- |
[top] | |
JulianH | Posted: 24 May 2025, 10:08 AM |
![]() Member Posts: 1 Joined: 24-May 25 |
a mailing list sounds like a great idea. Really looking forward to version IV and thanks for all your work ------------- |
[top] | |
« PreviousPages: 1 2 [3]
Sign in to post a reply.