I still have perhaps 5000 diodes. They aren't worth much, maybe a penny or two, but at least I won't ever have to buy one again.
Sounds like a nice hand-made diode ROM is in your future.
Some of the things i did like to pull out are power semiconductors on heatsinks (having heatsinks with already threaded holes including screws, nylon washers and silpads included is really useful), rectifiers, relays, ferrite ring cores, IEC mains plugs (Especially ones with filters), large film caps, switches, motors... Another thing i liked to keep are PC ATX power supply cables since i could easily clip them off and throw them in a box for when i need some short colorful wire to wire some things together.
Speaking of ferrites and other magnetics -- try to note where they came from, or what they were doing:
- Common mode chokes are typically at the mains input, are 1:1 ratio, and have high permeability and low saturation. They make okay transformers, though the leakage inductance can be difficult.
- Small power transformers are likely flyback supplies, often with a single-transistor (blocking oscillator) circuit, or a TNYxxx or other off-line regulator. The ratio and inductance will be suitable for whatever it was: the frequency is typically fixed and modestly low (~100kHz), the primary will be suitable for 120 and/or 240V operation, and the secondary will be suitable for whatever the output was. There's usually an aux winding on the primary side as well, which will be good for say 6-15V.
- Alternately, they might be driver transformers, the tall ones from ATX supplies being a classic example. These are mostly used with BJT based inverters. These have complex windings (a feedback winding is included, which was coupling load current into the transformer -- positive feedback drives the BJTs), and aren't good for much besides what they did; you could maybe use them for small (< 10W?) flyback or forward converters, especially in push-pull primary and/or secondary configuration. The tapped windings give some opportunities for ratios (1:1, 2:1, 1:2, 1:+/-1, and a few oddball ratios when including the short feedback winding). The one isolated secondary (it's for the low side BJT by itself) also allows dual isolated outputs (with more limited options on ratios).
Also kind of in this category, the triple winding toroid in CFLs. Same idea, BJT drive with current feedback. Probably high mu core.
-Bigger power transformers follow the same scheme, but expect even more custom pinouts and windups, and pay attention to the configuration. You'll see 1-switch and 2-switch flyback and forward, half bridge forward and resonant, and sometimes odd things like self-resonant (usually half bridge) with saturable control (Sony used a ton of these in the 80s-00s, particularly with Trinitrons?). Expect oddball combo secondaries, like the ATX's +/- 5 and 12V windings. Or a whole bunch of outputs for monitors (typically something like 5V, +/-20V, 100V, and probably a few more).
- Filter chokes, toroids, ferrite, the works: usually identifiable by being on the primary side (active PFC) or secondary side (choke-input rectifier, forward converter), in addition to a power transformer in the middle. If ferrite, these will be gapped to give usefully low inductance and high saturation current. If powdered iron, these will be... well, usually they're the ugly yellow-white or green-blue (#26 or #52) types, very lossy and suitable for low ripple fractions (and hence, average current mode control, or even more ugh, voltage mode control) at low frequencies (< 50kHz say). Other colors, and solid colors, are usually okay, but solid colors you can't really tell without taking measurements, there's no consistency on it. (Ferrites come in painted/coated styles, too!)
That's what they were, by the way -- typical ATX supply with a TL494 or KA7500 or whatever in it, voltage mode control. Depends upon output capacitor ESR for compensation. This is why you don't want to repair them with super-low ESR caps, and why they go nuts as their caps dry out and ESR rises.
You'll also see toroids in car power amps, usually a ferrite toroid for the charge pump. (It's mostly wired like a forward converter, but almost always lacks the output choke -- it's more of a charge pump, with the primary side switches dumping charge directly into the output caps, limited only by transformer leakage inductance, which is relatively generous at least. It's no wonder they are so prone to blowing up!)
Ferrite beads, chips -- pretty generic. You'd have to measure Z @ F to see what they're good for. May be high or low frequency material; beads of course you can count turns (usually just one, the wire straight through), chips are multilayer so who knows. Multi-hole beads are usually good for quite some current, and high impedance (the complex magnetic path between all the holes allows it to saturate much more gently than an ordinary bead, which often saturates at less than an ampere).
(Not that beads and chips are worth anything, and you can buy a sack of a hundred, or a thousand, of known type, for the labor cost of sorting them.)
Although I take that back, there is this somewhat oddity:
https://www.hitachimetals.com/materials-products/amorphous-nanocrystalline/surge-absorbers.phpparticularly the last couple pages. Saturable cores in ferrite-bead styles; hard to identify though, needs a pulse test pretty much.
Saturable cores in general aren't that uncommon, despite how boutique they may seem -- it depends, but millions upon millions have been used, in ATX power supplies. Usually a stripwound core in a plastic box, with a few turns of heavy wire around it. This was used to regulate the 3.3V supply (using the power of only a TL431, and the property of the saturable reactor / magamp), which is constructed as halfway between a "2.5V" supply and a "5V" supply -- it was derived from the 5V taps on the power transformer, with one always on, and the other conditionally on, gated by the magamp. This way it could be varied between 2.5-5V (give or take), 3.3V being the set point.
Other stuff; you'll find lots of neat but obscure power electronics in plasma TVs, mostly obscure because they were either: house marked, or poorly documented or custom-run parts (transistors, diodes, arrays), or custom arrays/drivers (ASICs, or probably hybrids/multi-die devices too). I've seen some transistors that were completely glued to heatsinks, no screw or clamp, no way to service them separately.
The beefiest PDP parts are a few IGBTs for sustain current, typically rated 300-360V, and anywhere from 30A continuous to 600A+ pulsed. As far as reuse, these aren't quite as efficient as MOSFETs of comparable ratings, but you might make a cromulent power supply with them, or a modest sized Tesla coil, say.
LCD TVs, monitors -- if the backlight is CCFL, can always take that out, separate it from the LCD panel itself, and use it as a light box or lamp. The supply is usually a push-pull BJT oscillator (somewhere between Royer and Baxandall, depending on how much supply inductance they provided, and exactly how the magnetics are designed). The output is high voltage and high impedance, suitable for igniting and operating, well, CCFL lamps, and maybe some other shorter discharge tubes (don't expect them to ignite regular length neon lights, but short ones, or stacks of regular fluorescent bulbs, sure). If LED, same idea, probably with a constant current buck driver, using a regulator or controller. If saving for illumination purposes, consider tracing for enable and brightness signals; hopefully it's analog and you can just slap on a switch and pot. Worst case, it's digital (SPI or I2C?), which may be documented in the datasheet, or if you can't find a datasheet (or it's all in Chinese or whatever), who cares, it was junk anyway!
Tim
Home
Help
Search
Login
Register
