Science/Engineering/Technical forums

Merry Christmas to all! Brisket and all the trimming are here at my house in Texas. If anyone wishes to come, be here an hour ago.

Same to all - sorry I was so wrapped up in meditation I didn't post myself. But hey!
There's a party on-demand here most all the time - but like Donovan says, you gotta show up!

Here's a timing jitter on the pi data aq project. Good enough for what I want - worst ones are in the 600us range, and usually much better.
(I'll be posting plans and code when they're baked) Right now I'm working up a little radio transmitter and rcvr to put out in the mailbox.
Not to overthink it - it's just an oscillator @ 20mhz or so, and a receiver that can pull a logic level. The mailbox is well outside wifi range, and that would actually have been harder to do - I'd need some kind of power latch if the mailman didn't keep the box open long enough to ... This is simple, and lower frequencies reach better. Just a magnetic switch on the mailbox door.
LoRa would have worked, with the power latch and a ton more money - in that case, range overkill, I'll save those rather expensive proto boards (arm based like a teensy) for something that needs kilometer(s) range. KISS!

Since it's line of sight, I could probably just use a bright led and photodetector sort of thing, but then the mail truck is in the way when I care...(it's on the other side of the road or I'd just run a wire).

I thought of putting something on my mailbox to tell me when it's been opened, but I just check it every evening. Maybe your box is father away than mine (about 1000 feet or so) and it's more of a hassle for you to manually check it.

More like I'm impatient. Often the mail-person is bringing new toys...whenever they come, which is all over the place. And of course, I want those (or any food that comes) pretty much right away - like the old saying - if I wanted it tomorrow, I'd order it tomorrow(!). But yes, there's a hill in the way, which is sometimes snow covered, I have some emphysema (recovering, but), and it seems silly to start up the tractor for a short run - and dangerous in snow, to check a few times so I'm not wasting an afternoon waiting for things that are already in the box (and then waiting for them to equalize temperature again). Since I don't go out daily...

But also, fun, tech challenge, etc. Transmitter is right now one xistor power oscillator, xtal controlled - keep it simple, magnetic switch on the box door, 9v battery. The receiver should be fun, though.
At that end I can afford a few mA, but again, want to keep it simple. I might do a direct conversion rig with an xtal pulled off a few hz and look for a beat or something, to trigger a gpio on a nearby "lan of things" raspberry pi....dunno. First I have to make it work in testing and get a decent false alarm rate or at least know what it is (Assuming I can reset the detector count via the pi's web interface). I don't think an SDR or even a superhet is justified on the KISS philosophy for this range...So, it's an aesthetic exercise in minimalism. (old man justifies silly hobby/distraction)

Something to do while the other stuff comes together, which is looking good, mostly. The new little grid is probably headed for the big fusor where it looks very good for the normal fusor mode - once I use it to help test and rough calibrate the neutron sensors and so on, I may dupe the ion source with is outperforming expectations, but might need to work a little harder on the thermals for one meant to really sock them out there. Since all that is in another building that takes a lot of time and some effort to heat...I'm doing that part of the work on good weather days. If my hypothesis on what will happen with moving ions (or a semi polarized soup) around with some RF is right, I'll want a grid around 2" diameter for this - so the transit times inside and outside the grid are about the same. I'll have to make that, the constraints will be a little different, and I'm hoping PeterS will help me with a little modelling for field shapes....when I get there.

Meanwhile, some of the new stuff I'm doing on the fusor-2 pi control looks like it will benefit #1 when backported to there - and I'm putting some fairly major skull sweat into that development. This time I'm planning on automating the gas controls in a fairly flexible way - the smaller new setup needs it more badly, bit it'll benefit both. Most of the software work I can do here on the comfy and warm couch. I'm actually talking to the tinker bench 20' away via remote desktop so I don't have to walk back and forth all that much...Just to look at a scope now and then, or re-lash some test stimulus.
Coding and testing on a big-screen TV with voice recog to look up error messages and the like...no point setting up the super luxury and then not using it! Pretty sweet setup.

Happy new year Doug and all that frequent here

Back to you.

So, I've been "chinesed" on my little mailbox project. I ran across this: https://www.mpja.com/315MHz-Remote-Cont ... /31960+MP/
A transmitter and reciever for ~ $3... for 315 mhz (more or less).
OK, it was likely near-useless for the original purpose, so surplussed, but it can easily be bent to my desires for the mailbox indicator.
It's a Pierce oscillator using a SAW device, one transistor, on-off modulation via another transistor. Simple, almost the same thing I'd already proto'd up, but different frequency, fine.
The receiver is a preamp, a superregen detector, and two sections of an opamp used as gain and thresholding. It toggles randomly in the absence of signal (and also sometimes when signal is there).
No DC restoration, so it'd take some fancy error-detect munging to use for what it was meant for.
No biggie....
I modified the receiver to sit at 0 output with no signal.
I made a little multivibrator to modulate the xmitter with at around 650hz (what I got with standard round number parts values, like 10k, 100k, .01 uf).
I took advantage of the characteristic of the lm358 opamp being better at pulling up than down - it's as good as a diode for this, and simply put
a 100k/.2uf lowpass on the output, feeding a comparator (another lm358) that drives one of those little beepers...and we're done, no need to have a computer watch this, it's been running
a few days with near zero or zero false alarms, and has considerably more range than needed. I'd guess half a click. At 100yds it's near-overload on my SDR dongle.
The plan is to repackage this stuff off protoboards, and put the xmitter in the mailbox, run off a 9v battery and a burglar alarm magnetic switch to go when the mailbox door opens and beep me here inside.

Moderately elegant I suppose, I'd hacve a little more satisfaction had I made the receiver myself, and it'd work better too - but it's that enemy of the perfect - good enough.
I'll do a proper writeup once I get things packaged...with schematics. It's really simple... I'm gonna be real glad when winter abates...between the short (no power) days, cold (takes hours to heat the shop with a few trips to add wood to the fire)...other winter infrastructure, it's been silly busy but not with physics. So I haven't been posting, my bad...The good news is my health is getting better and letting me catch up and even improve things despite weather and "life".

At least this was fun, and gave me an excuse for the RTL-SDR radio dongle I bought for myself, which is a cool toy. Not noise-theoretic or anything, but it's amazing for something gumstick sized.
And it let me measure that this xmitter was really 315.065 mhz...nice debugging tool. It will have some uses on the fusor/ion trap stuff...not to mention just being entertaining.

Well, now I have someone of a reason to implement something like this. I put a deck box out by the road for parcel deliveries. It's working. But I do have to remember to check it if a package is due to arrive, least a "porch pirate" get it.

I'll try and get those details up here soonest. I need to put in such a box too, my poor mail-people have been hiking down here to the living quarters with lots of heavy stuff lately ... I'd go out to meet them if i knew they were here, which this will let me know about.

Another friend wants this too. While simple once built, there's a couple oddball tricks to what I did...worth documenting. It should be easy to add a lfipflop latch and a led - or something similar - to the receiver side of things here. Since I'm almost always here - I don't need those - I check the mail when coming home from errands anyway. Having "no bills" means I don't check it much otherwise. But then Amazon (and Banggood, and ...) changed everything...

It's up there now, under RF. http://www.coultersmithing.com/forums/v ... 64&start=0

If a clean desk represents an empty mind, well...good things to come? The data aq and control for the new ion-trap recirculation tests is coming along apace. Raspi 4, 8 15bit ch of a/d sampled at 10hz, 2 12 bit DAC channels, and control outputs for gas control valving. There will be an arduino nano clone or two added to be fast frequency and rad counters, and some more a/d and digital io stuff. All that is going into a nice new rack panel to be installed up at the new rig. I'm just doing the coding and prototyping for now, since the weather in here is salubrious compared to the shop. I'm hoping to get automated gas control really hammered down so as to add it to the other fusor setup.

This design looks like it will (finally) be modular and general enough that people might want to duplicate it. I'll be putting the software up on my github repos for that.

The idea is to stuff the acquired data into an SQL database, with a design such that what each channel is, and some code to translate the raw data into real units will also be in a table in the database that can be edited to accomdate whatever lashup. This will be controlled by a runtime GUI program like the one I use now for Fusor1. There will be some nice plotting code that can take the results of a run and plot it, using the schema data and code to translate channels. There will be both 2d and "4d" versions of that, and both will be able to plot from the database in real time during a run.
Given how this kind of thing works, the database server can either be on the pi, or on some other machine on the lan, as can the other applications.

I have a lot of this now, but it's special purpose dedicated to Fusor1 with some things hard coded that don't generalize well.
Now that I'm doing another one...after it's time to build a railroad, it can be good to build an elegant one.

It's a moderately big deal to get something close to hard real time out of a Raspberry pi - in Fusor1, anything like that had to be in an arduino the pi could talk to now and then via USB serial. This new version will allow, but not require that extra complexity. In the older model, it wasn't so realistic to get things like automatic and interacting gas and power supply control, because things were on different CPU's and the complexity and time lag started to become unreasonable, especially if some fairly "smart" algo would be needed for those.