Thread Tools
Dec 08, 2015, 06:25 PM
Registered User
Thread OP
Mini-HowTo

Chickadee Tech DIY Laptimer


Hi all,
I'm excited to announce that I've open sourced all files related to an RSSI based lap timing system. Its still in prototype phase but input is welcome. The idea is to track video signal strength to determine when a quad or other aircraft with 5.8ghz video passes by. This should be a cheaper and easier system than other available timing systems.

Here is a video and a couple pics. More info is available at http://chickadee.tech/lt58/v4 and https://github.com/chickadee-tech/laptimer58

Chickadee Laptimer 5.8 v3 Quick Overview (3 min 21 sec)


~Scott
Sign up now
to remove ads between posts
Dec 08, 2015, 06:29 PM
Registered User
Thread OP
On a separate thread rcdesigns what type of antenna would be used for a small region.

I'm currently just using a rubber ducky antenna so I'd love to hear Shea's ideas on what a better antenna would be. With the rubber ducky antenna in my first initial test I got anywhere between 0.5 - 1 second resolution when comparing against values from 60fps video.
Dec 08, 2015, 06:40 PM
www.laforgefpv.com
crazyshea's Avatar
Quote:
Originally Posted by tannewt
On a separate thread rcdesigns what type of antenna would be used for a small region.

I'm currently just using a rubber ducky antenna so I'd love to hear Shea's ideas on what a better antenna would be. With the rubber ducky antenna in my first initial test I got anywhere between 0.5 - 1 second resolution when comparing against values from 60fps video.
I have a few ideas I can sketch out later tonight. My ideas are based off ground shielding to make the antenna as directional as possible and reduce reflections ( think of a cone with the antenna in the center surrounded by ground). Depending on how well we can get the channel calibration code working will help drastically on the design of the antenna.
Dec 08, 2015, 06:44 PM
Registered User
Thread OP
Quote:
Originally Posted by crazyshea
I have a few ideas I can sketch out later tonight. My ideas are based off ground shielding to make the antenna as directional as possible and reduce reflections ( think of a cone with the antenna in the center surrounded by ground). Depending on how well we can get the channel calibration code working will help drastically on the design of the antenna.
What do you mean by channel calibration code?
Dec 08, 2015, 06:57 PM
Registered User
This looks interesting, show it working please.
Dec 08, 2015, 07:15 PM
Registered User
Thread OP
Quote:
Originally Posted by downloader9
This looks interesting, show it working please.

That's top of my list! I hope to have to going by the end of next week. I'm super busy this week.
Dec 08, 2015, 07:29 PM
www.laforgefpv.com
crazyshea's Avatar
Quote:
Originally Posted by tannewt
What do you mean by channel calibration code?
By channel calibration I was just referring to the RSSI strength and what peak would count as a "pass through the gate" versus noise to ignore while you fly the rest of the race track.

The channel part is so we can monitor multiple racers/channels with this setup.

Channel seek time is about 20ms +5ms to get a stable RSSI reading. This device could be actively looping through the race channels checking if any channels RSSI has reached the peak threshold you could post the time it takes between peaks of that channel.

The peak calibration would be done at the start of each race by walking the craft through the gate.
Dec 08, 2015, 07:52 PM
Durability Tester
This looks Interesting can you post a schematic?
Dec 08, 2015, 08:24 PM
Registered User
artemen's Avatar
I suggest a using an attenuator on RX which would be killing the tops of the signals, at the same time putting an antenna in a vent duct pipe 6-8" in diameter about 15" away from the opening, while grounding the duct.
Dec 08, 2015, 09:26 PM
Registered User
LAFPV's Avatar
posted about this today in the multigp fb group. people are very interested! and now that you're working with shea, you guys will nail it. keep up the good work! let us know how we can help move it along. i would gladly pay $5 for the ios companion app.
Dec 08, 2015, 11:06 PM
Registered User
Android app?
Dec 08, 2015, 11:54 PM
Registered User
Thread OP
Quote:
Originally Posted by crazyshea
By channel calibration I was just referring to the RSSI strength and what peak would count as a "pass through the gate" versus noise to ignore while you fly the rest of the race track.

The channel part is so we can monitor multiple racers/channels with this setup.

Channel seek time is about 20ms +5ms to get a stable RSSI reading. This device could be actively looping through the race channels checking if any channels RSSI has reached the peak threshold you could post the time it takes between peaks of that channel.

The peak calibration would be done at the start of each race by walking the craft through the gate.
Ah, ok. Thats what my code is currently doing. Its waiting 30ms currently.

Quote:
Originally Posted by geofrancis
This looks Interesting can you post a schematic?
Schematic is here which is linked to (but kinda buried) from http://chickadee.tech/lt58/v4/.

Quote:
Originally Posted by LAFPV
posted about this today in the multigp fb group. people are very interested! and now that you're working with shea, you guys will nail it. keep up the good work! let us know how we can help move it along. i would gladly pay $5 for the ios companion app.
Yeah, I'm stoked to get Shea on board. I sent him a dev board today so he can jump onto the project. I'll look on the multigp group too. I was surprised to see so much traffic from Facebook.

Quote:
Originally Posted by Niknjen
Android app?
We'll do that too hopefully. Our plan is to use React Native so we'll be able to share most of the logic.
Dec 09, 2015, 06:02 AM
Registered User
To my mind, before getting too whizzy with the wifi connection to mobile phones, I'd have started by ensuring you can reliably detect the quads as they pass the start gate, without false detecting from another part of the track and doing this with a reasonable amount of accuracy (say 100mS), whilst a couple of quads are flying close together.
I see this as the first and greatest challenge.
It sounds like a tall order to seek through, say 5 channels, taking enough RSSI samples to average and detect a peak in such a short space of time.
Dec 09, 2015, 10:48 AM
Dirt Dobbins
This looks promising! Love the idea of being able to have racers just show up and run... one less thing to worry about!

Does this scan the full frequency range of 5.8? Could it be set so you define freq's to watch and increase accuracy?
Last edited by zanthor; Dec 09, 2015 at 11:32 AM.
Dec 09, 2015, 11:57 AM
www.laforgefpv.com
crazyshea's Avatar
Quote:
Originally Posted by rcdesigns
To my mind, before getting too whizzy with the wifi connection to mobile phones, I'd have started by ensuring you can reliably detect the quads as they pass the start gate, without false detecting from another part of the track and doing this with a reasonable amount of accuracy (say 100mS), whilst a couple of quads are flying close together.
I see this as the first and greatest challenge.
It sounds like a tall order to seek through, say 5 channels, taking enough RSSI samples to average and detect a peak in such a short space of time.

Yes, this is the most important part. However we need to get our debuging invironmwnt set up. This could be the phone app with raw data read out or just a computer reading out the data like his initial python app.

I was also looking at adding an OLED to the esp8266 last night for even more debugging directly from the device.
https://hackaday.io/project/6132-esp8266oled


I think right now Scott's main goal, looking through the code, was to get it testing one to two frequencies. He has it set up to do more but left the frequencies as 0 so they would be skipped.

It is pretty wet where we are but I have a pretty big basement to do some preliminary testing while getting formillar with everything.

Should be a lot of fun and discovery in the next few weeks. I'm pretty excited with the foundation Scott has already set in place for this project.

Nice work Scott and glad I can be a part of this moving forward!
-Shea


Quick Reply
Message:

Thread Tools