The advice I wish I knew is to not buy an inventory of parts. I bought a bunch of resistors, capacitors, LEDs, etc., with no specific plan. Most of it sits unused.
There's no point in keeping an inventory because almost every project will require a parts order anyway. Just order what you need per-project.
There's no point in keeping an inventory because almost every project will require a parts order anyway. Just order what you need per-project.
That assumes that everything one does is a "project". I don't know about anybody else, but sometimes I just sit down at my lab bench intent to tinker or experiment, or do some ad-hoc repair on something from around the house. And I've found having a modest stash of components on hand is very useful for those activities.
It's a balancing act, of course. There's a limit to how much one would want/need to spend on stocking components, but I posit that the limit should probably be "greater than 0". I can't tell you what the upper limit is though. My component stockpile has been building for many years, and there's no way I could even tell anyone how much is invested in it.
All of that said the one other thing I've often heard quoted, and tend to agree with, is this sentiment:
"When you order components for a project, always order more than you need, and put the unused ones in your stash". The thinking being, something along the lines of "If you needed this component once, you'll probably wind up using it again."
Of course it is, again, a judgment call. Order a few extra $0.50 comparators? Sure. Ordering a few extra of some multi hundred dollar FPGA or ADC or something? Yeah, no so much. :-)
If you mostly do digital, I'd say just get the E6 resistors and capacitors, common connectors and switches, and but there's as needed for projects.
Unfortunately multi packs usually have the whole E12 series full of odd sizes, like as if I'm going to need a 68k resistor rather than 100k, in a situation other than a project I'm ordering stuff for anyway, but having the extra selection isn't the worst thing.
E12 is about being +/- 5% of your target resistance using only a singular resistor.
A couple of notes:
1. E12 kits plays well with +/-5% tolerance. You're covering the entire spectrum of resistances possible with your kit, within the error-specs of your resistor.
2. Trimmer potentiometers (cheap pots designed for only ~200 cycles of use or less) are the key component needed for most analog / accurate circuits.
-------
You can usually design a circuit with +/- 5% accuracy (nominal), and then have one or two key trimmer-pots tune to 0.1% accuracy at the end. Ex: You might have one trimmer-pot control an a voltage offset (ex: from an instrumentation amplifier), and a second pot control the scale / multiplier (ex: op-amp multiplier and/or resistor divider).
You may have 20 to 30 resistors all across your analog circuit, all within +/- 5% tolerances for various reasons. But these two control points at the very end can "trim" away your error and reach 0.1%, or better, accuracy, in practice.
But you have to be "close enough", and have "linear enough" behavior. E12 (aka: resistors all within 10% of each other) seems to be enough for most such projects.
---------
Digital circuits on the other hand, don't really care too much about this stuff. So you end up with a bunch of 100nF capacitors with 1k resistors just all over the project. That's fine too. You should build a stockpile of the components you plan to use.
E12 resistor kits + a set of trimmer pots are for analog circuits. At least if you want 0.1% or 0.01% tolerances (3 or 4 digits).
Yeah, I didn't really mean "there's no point in keeping an inventory". I totally agree about buying extras and building inventory over time. But that's all you need to do, IMO.
For the beginner you can apply this principle to your first few breadboard projects and you're off to the races with some power-of-10 resistors, one color of LED, a few tactile switches, etc.
The main point is that you shouldn't try to anticipate what parts you will need more than a couple weeks into the future.
The main point is that you shouldn't try to anticipate what parts you will need more than a couple weeks into the future.
No doubt. You could easily wind up spending a TON of money if you do that! Not to say that I've never bought something on totally speculative basis, of course... :-)
Last year I started with a small ESP32/WS2812B project, after that was finished I wanted to get an understanding of how all this electronics stuff works and also get an understanding for embedded programming. Now I'm sitting here with lots of electronics components, some PiPicos, ESP32 boards and my newest addition, some STM32 Nucleos. :) I think I should really sell some of this stuff and only concentrate on one project at a time.
There's some truth to this, but on the other hand I find it cost effective and useful in the long run to build up inventory by cheap kits.
For example, if I need specific resistor (and don't have any high precision/tolerance/current requirements — i.e. 90% of hobby projects) I could buy online:
- 10 of them for maybe £3
- 100 of them for around £6
- a kit containing 1000 resistors of 25 different values for <£12
Not only is the last option exponentially better value, but it means that I almost certainly don't need to buy anything the next 10 or more times I need a resistor. There are a good number common components, fittings, etc. that I've found this works for.
If you're doing specific high frequency, RF, hifi, etc. stuff, this approach might not work so well, but for other hobby electronics, messing around with microcontrollers and the like it's served me well.
Especially don't buy: old-school digital IC assortments (if they still market those) and electrolytic capacitors.
If you are using substantial discrete digital logic in 2023, we can be friends if we're not already, but you shouldn't be giving people advice on stocking their workshop.
Electrolytic capacitors have a shelf life, and you'd be amazed how time flies. I've thrown out old electrolytics (>10 years) at more than one employer now.
mostly true, but always infuriating when you need a specific resistor value and don't have it. I like having a resistor kit around, but it doesn't need to be a big sorted set of drawers, can be a box of bags from digikey
Ehh, if you're not doing analogue stuff how specific do you really need? And even if you do you can combine a few to get closer to what you need. Like if you have 220/390R & 1k & 10k & 1M on hand, are you really ever going to be that annoyed for breadboarding purposes that you don't have 680R or whatever?
I think that I could go for an entire decade not needing any passives other than powers of ten except for power supply work when you need an exact ratio for feedback.
But analog tinkerers do things that are very different from what I do!
Don’t buy resistors on aliexpress. They tend to be steel legged and go rusty invisibly leading to all sorts of weird problems. Also some of the tend to be mislabelled.
Spend on some decent ones. Tayda sell decent quality Royal Ohm ones for not much money.
This depends - my projects are usually small and spontaneous, my attention span is short, and not having parts to at least start means the project is dead right there.
I do most of my stuff on microcontroller boards though, and I keep a small stock (maybe $30 worth of ESP32, ESP8266, and something-that-can-pretend-to-be-a-USB-keyboard dev boards).
Cat5 (6,7,etc) cuts make for some very good breadboard cables; they're just the right size, colored, stiff, and much much better quality than the specific ones sold online. Unless one needs more colors or different sizes, they're just perfect.
The ready made ones aren't single core and not ideal for breadboarding, still they're handy for other uses; A while ago I was about to throw away a long (15 meters or so) cat5 one that I swapped after it had been chewed by a dog, but rather cut the damaged part, stripped away the external insulation and saved each wire, and man, they're really good, from the copper used to the consistency of their insulation.
I try to solder as little as possible and I find having a large selection of task specific breakout boards, power management, motor controllers, etc. and various wires and connectors to be essential. It's much easier to experiment if I can do it adhoc when inspiration hits and then put in a parts order when I know what the missing pieces are.
There's no point in keeping an inventory because almost every project will require a parts order anyway. Just order what you need per-project.