Show HN: Learn how to build electronics with monthly kits

I received the adabox second box just before Christmas. Its a ton of stuff, I feel I got far more than my $60 worth although of course its stuff they selected not stuff I selected. I did not sign up in time to get the first box, so I can't comment on the first box. I think the first box was something with RGB LEDs, looks interesting.

The second box is "make a robot" with a feather which is an adafruit thing which is a custom arduino with on board bluetooth. It looks extremely easy to me, but I am not a noob to either electronics or mechanics. Its a nice little set, everything you need to mess around with a little robot.

Its "make a robot" in the post 1990s sense of a robot is a homemade RC car with homemade UI. Not make a robot in the pre 1990s sense of its possibly a stationary arm, or its autonomous, or it does something other than be a possibly weaponized RC car. So its a rather basic homemade RC car, OK cool.

Needless to say with the holiday obligations I've done nothing with it other than watch the videos and paw it all over and there's no way I'm going to get to mess with it due to obligations etc until well after the new year.

Everything seems very well selected and packed and looks fun. It looks quite expandable. Somewhere on my cluttered workbench I have a nice IMU/magnetometer/accelerometer I2C board and I think it would be fun to bolt on. It is not maxed out and cries out to be messed with and expanded. Its open loop motors, so I'd like to use a simple IMU board to close the loop so a 90 degree turn really is 90 degrees and so on. Also I could use the magnetometer to discipline the straight line such that its really straight not a curve with a large unknown radius. That'll probably be my first experiment on my own after I complete the official published experiments.

My decades of experience with boxes implies that if a company mostly sells "stuff" and occasionally offers a box or gift basket it'll be a great deal with overstocked items marked down as much as 50%, maybe more, but if its a "gift box company" the prices of the components (perhaps a cheese log or a summer sausage) will be marked UP at least 100% to 500%. So because adafruit is mostly a "stuff seller" and not a "box seller" the box price to retail component price ratio is awesome. There must be $120+ of retail price stuff in the $60 adabox. The thimble being a company focused on selling boxes, I would assume the price ratio will be more like the "sausage and cheese holiday gift box" biz where the box is at least 200% of retail component price. I would be happy to be surprised if thimble is in fact a good value, but observations of parallel businesses implies that would be quite an achievement if they pull it off.

I'm okay with that extra 5 bucks for their inventory cost if they allow me to opt out of permanently saving my details on their servers. No matter how many times I unsubscribe from their emails, they always start a new list and I'm strangely on it again.

There are a lot of good book suggestions but I'll toss out an alternative idea of buying "bad" kits. Kits that just take care of one task. Maybe its the RF guts of a 30M ham radio receiver that barely technically works. Now you get to spend engineering effort on the somewhat simpler tasks of making a decent source of regulated noise free power, and build a decent AF amp strong enough to turn the headphones level audio into something driving a speaker using a couple transistors (or maybe an IC). And the input filtering isn't very good so you get to design a nice 10Mhz or so bandpass filter wide enough to let most of the receiver bandwidth thru but block enough interference signals especially AM radio transmitters. And now input signal levels are going to be low enough that a dedicated class A RF amp on the input might be a good idea to boost signals. Next thing you know you're going to be unhappy with the frequency stability of the VFO, so you either rig one of dozens of DDS VFO kits in its place or stabilize the free running VFO in the kit.

Anyway the point is if you can regulate power and build amplifiers and filters but not quite up to the level of building a 30M radio receiver, then improving a bad kit would be a good plan.

Likewise if your "wifi robot" you don't want is close source silo 100% maxed out no expansion possible it is quite useless. But if its got an IO pin or two and you want to learn how to light LEDs using resistors and maybe a transistor or two, you can use the wifi robot to handle the whole "how to turn it on and off" part while you focus on the source vs sink current limits and the excitement of inverting transistor driver circuits and the virtues of putting the LED in the collector or emitter circuit (assuming bipolar transistor) or the joys of trying to bias different types of switching FET transistors, oh its all a barrel of fun but by leveraging the unwanted "wifi robot" you can avoid some hassle.

Also note that the "maker community" of java programmers learning to solder and the ham radio community of EEs learning to program java, never ever cross pollinate. However I live in both camps and the ham radio guys have decades of "learn you some electronics for greater good" general category of book. So if you're not having much luck with the maker style books there are literal decades of the ham radio branch "learn electronics" books. Check the ARRL to start.

That's exactly what Gammel claimed to do in his original videos on this:

https://news.ycombinator.com/item?id=13260343

He pointed out that the classes were usually theory heavy where you do piles of math without building anything. He said the kits normally had you plug stuff up without understanding what's going on. So, he tried to do a course that was incremental and build-oriented like kits but gave you understanding of theory/concepts a bit at a time. I haven't used the service but it was only one that claimed to do this.

Agree. Also not wanting to pick on this particular project but:

I have burned myself - and seen others got burned - too many times on subscriptions.

Feature keys? Thumbs up. Commercial with 15 days trial? Fine. Pay for online storage? Only if you provide a better alternative than Dropbox, Google or AWS. Monthly payment for static apps? Bad. Creating an online experience (syncing in your proprietary "cloud") to justify monthly payments? Yuck.

I want my apps to work without constantly having to pay a protection fee to a vendor.

This of course goes against what every make-a-SAAS-tutorial will tell you but I am speaking here as a consumer :-]