Видео

Just to get started I used the ADI system setup. But it's not required and there is other code you can use, for example there's an Arduino-compatible one and I presume it could be adapted. github.com/jbeale1/DataAcq/blob/master/AD7124-continuous-read.ino

@@dancevideo2 Thank you, I will try it soon

@@dancevideo2 I see that you also used the ADS1256 from TI for another project. Although they are different ADCs, which family do you like better AD or TI? I am looking at the ADS1220 vs the AD7124 in my case. Thank you.

@@ChristianBeltrame Yes, you see this code for example for ADS1256. I never got that particular board to be 100% reliable but it is possible it was just the "waveshare" ADS1256 board or circuit, I never figured out the problem. Basically, it would unpredictably reset after some # of hours of continuous use, and then become unreadable because I wasn't using the default output mode so my code would no longer read it. Or maybe I was doing something out of spec, but I don't know what. github.com/jbeale1/DataAcq/blob/master/ADS1256_read.c

@@dancevideo2 Thank you. Since I already started looking into the AD7124, I may stick with it. I have had bad luck with the ADS1220 boards from Amazon (shorted pins, IC in the wrong position, etc). I got a ADS1263 to test in the TI side until I figure out how to get an ADS1220. Too many things to learn, too little time...

This document shows both a thermocouple and a 4-wire RTD connected to this chip. www.analog.com/media/en/reference-design-documentation/reference-designs/cn0384.pdf

I have the KC761 (no "B") so I can only speak to my unit. On my unit there is a signal on that jack, but first I had to enable it in a menu setting. There is a "Pulse out" menu selection which was OFF by default, but can be connected to either the scintillator or the alpha detector.

Per datasheet it goes up to 19200 sps but note there is a _strong_ tradeoff between sample rate and effective resolution, as described in the datasheet. If you want high sample rates, you may want a different part. www.analog.com/en/products/ad7124-4.html

I didn't take it to a race track, but I don't think there is any practical limit in the module itself. The limit would be in whatever you use to measure or record the output, which is a frequency proportional to speed. If you've got a target moving 1000 mph I would expect it still generates a signal.

@@dancevideo2 I've seen some people suggest there's a 30 MPH limit in the hardware setup, and others suggest there's an 87 MPH limit.

I'd be surprised if that's true. There is one site showing the circuit on this module and basic detector is a single microwave part that I would expect has very wide bandwidth. Normally you need extra parts and extra expense to narrow the bandwidth and this is an absolutely dirt-cheap design. On the other hand I think it was intended for low-speed use like people walking (automatic door opener) so if a simple R-C filter reduces high frequency interference usefully, they might have done it.

@@dancevideo2 Thanks, but you think your setup has no such limits. I want something that can at least handle 60 MPH so I think this should be OK.

@@GeorgeOuWell, now you made me wonder if there's some easy way to measure it directly. The traffic on my road, happily, does not go that fast, and I'm not enthusiastic about hanging out on the side of a freeway, but I will think about what I could try. I guess I could just put it in the car facing forward and drive, and cars coming the opposite way will give me a fairly high relative speed.

TBH I would tend to recommend the Radiacode for most people. The raw count rate on the KC761 is better, and it has a second, diode detector that works for alphas, if you remove the protective label (which makes it very fragile and they suggest you don't) but that's probably the only advantages to most people who aren't trying to build experimental external circuits. The Radiacode software is far more mature and functional; its energy resolution is much better, it's smaller, the battery life is better. I have the Radiacode 102, the somewhat improved 103 model was just released recently.

It's the simplest possible setup: video camera facing the graduated cylinder, no separate external lens, no mirrors. I don't remember, but I may have used a sheet of paper as the background with light shining through it from behind, to avoid glare from specular reflection off the glass cylinder if I had the light in front.

There is nothing complicated. This is a normal piece of chemistry-lab glassware, a plain graduated cylinder filled with water, which I am dropping the glycol into from the top. Light is in fact refracted inside the cylinder, that's why the edges of the cylinder appear dark, where the rays are bent off to the side towards a darker part of the background. If you try the same experiment, you should see the same thing.

Camera works fine for me on the RPi, just like the earlier model Pi cameras. It's the cloudy sky that's the problem, at least if you want to try taking photos of stars...

Camera is Raspberry Pi camera module 3 (NoIR variant). It is an accessory for a Raspberry Pi, just released this year, 2023. Reasonably cheap and not designed for astronomy, but it can see a few stars anyway. www.raspberrypi.com/products/camera-module-3/?variant=camera-module-3-noir

The recording sounds like much less than what I heard. What I heard was the whole house shaking, but that recording was outside and got just the raw pressure wave, not all the shake-rattle-boom effect of it in the house.

Thanks, neat that you saw the flash. It sounds totally underwhelming to me in the recording. Inside it felt way bigger because the pressure wave rattled everything in the house. The recorder is outside; I set it up to answer "Nextdoor: what was that noise?" type questions. The gain is set low so it doesn't saturate even on this kind of sound.

Nextdoor: what was that noise? Haha, a definite growth industry. At least 10 threads in my feed right now.

I have not tested high speeds. This is a super-simple circuit, and I don't think there's much of an output bandwidth limit built into it (or even if there is, changing or removing one capacitor would fix that). It might well be able to detect hypersonic speeds- if your cars go that fast!

It's tricky to see the scale of things at the bottom of the frame but I think they were regular bugs that looked bigger because they were very close, within inches of the camera lens. It was a visible+IR camera but at this time I did not have an IR light, it was just a normal visible white outdoor yard light mounted on the garage that is providing the light.

So far I've only used the Pi Zero for recording the raw video, and I'm running OpenCV on a Pi 4, and this is only a few fps right now BUT everything is hugely inefficient as this was a super-rough, first-day try and the code is doing all kind of unnecessary things. Doing it sensibly, like initial detection on a very low resolution frame, and then the full resolution just in a very small window on each pre-detected dot position, I could imagine it being real time at some resolution even on the Zero. Just not sure what resolution that would be. In my particular application I'm just measuring a creep over hours and days, so I don't need to run the analysis on the Zero at all, let alone in realtime, I was only intending the Zero just for the acquisition.

Here is an example that walks you through the code. ruclips.net/video/8B1LwzkLKXs/видео.html

I think it is pretty good. I have been experimenting with mine. I really hope they have a version some time without the IR-Cut filter.

When you start the www.aposteriori.com.sg/Ev3devSim/index.html page, it comes up with some basic code for a line follower that works for a while, but there's a bug. If you know any python, it is easy to fix. Then you can try playing around with some other changes, that is all I did here. I added some more variables to the print() statement so I can see what's happening. Doing all this on the sim is faster than testing the real robot. Of course eventually you have to use the robot, but here you can try different ideas quickly and see what gets better or worse.

Yep! For me it feels like the cars are driving directly over your head. The mic was not actually in the road, but off to the side about 5 feet from the curb.

Is there a motor operating on this?

The coupler is this thing, which is parametric www.thingiverse.com/thing:81342 I changed some parameters from the defaults I think. It is supposed to be a little bendy to take up shaft misalignment, although that one ended up pretty much rigid. Yes, it is being driven by a gearmotor which is most of the noise you hear. It is out of frame on the left. On the background power supply (a little blurry) you see the current and voltage going to the motor. Looks like 0.12 amps and 2.0 volts.

@@dancevideo2 Just got this printed through a friend. Working to make the worm drive have a spool attched to gear for opening and closing slot blinds string. Thanks! Did you have an intended use for did you make this just for kicks?

@@dancevideo2 Do you have any plans for the wooden parts? Even if you just had a picture of sketch with dimensions would be a great help.

@@thomasmartin8580 Sorry there were never any kind of plans, and I no longer have this thing to measure it. It was thrown together one afternoon from random pieces of wood in my garage, just as a test to see if the gear mechanism would work at all.