What is this

what is this thing I found in my newly acquired arduino shit?

Attached: urphey.jpg (2268x4032, 1.23M)

Other urls found in this thread:

youtube.com/watch?v=sI5Ftm1-jik
twitter.com/NSFWRedditVideo

Attached: 20180528_162229.jpg (4032x2268, 2.35M)

Bridge rectifier

looks like an AC converter or something?

Attached: 20180528_162233.jpg (4032x2268, 2.29M)

Looks like it’s missimg heatsink

For the size id say it's a Full Bridge Rectifier.

neat. I makes DC from 100V 4A. I wonder what the last person was using it for. I got all this stuff second hand with no documentation

Rectifier(I think). Converts AC to DC. Try to find some markings on it to know for sure.

As long as it wasn't this guy youtube.com/watch?v=sI5Ftm1-jik you're OK.

It-Its a comb for thinning hayer.

Attached: images.jpg (214x235, 9K)

air conditioning module

It's a bridge rectifier. Sometimes they come in packages like that instead of using discrete diodes. There are also ones in this package as well, mostly for higher power applications where there is likely to be high forward current through the diodes when they're conducting.

Attached: 25574.jpg (400x400, 21K)

FULL

It doesn't make DC, it rectifies AC. There's an important difference here. Rectified AC doesn't go negative, but it's still very ripple-y. The ripple frequency is twice the input though which makes filtering easier.

To make decent DC from rectified AC you need a capacitor to match the load.

full bridge rectum fryer, looks to be good for around a hundred volts of angry pixies

Rectified AC is still technically considered DC. Direct current is simply any current whose polarity doesn't change over time. There is no requirement that it maintain a constant value. I will grant you when people say DC they generally imply a flat voltage value whose value does not change outside of some very small ripple. The proper term for rectified un-filtered AC is pulsed DC.

bridge rectifier. its just 4 diodes in a specific arrangement, and converts ac to dc. however, your dc output will still have a wave form. you may also want a smoothing cap, and/or some sort of voltage regulator

>I will grant you when people say DC they generally imply a flat voltage value whose value does not change outside of some very small ripple.
but thats not true for anything outside of batteries. most DC power supplies are witched mode, which have an even more horrendous wave form then rectified AC, albeit it at hopefully a much higher frequency

*switched mode

and when I say its not true for anything outside of batteries, that is an exaggeration. There are *some* linear power supplies out there. but since they are heavy, expensive, usually ifefficient, they are rarely used outside of specialty cases

Switched mode power supplies short circuit an inductor or transformer briefly and then immediately break the circuit (usually a mosfet doing the switching) which creates a high voltage spike on the inductor/xfmr since the current in an inductor cannot instantly change. This spike goes through the output diode and charges a capacitor. The whole flyback effect IS pulsed DC. The output is heavily filtered in nearly all cases so you get clean DC on the output.

I've designed conventional boost as well as flyback converters before so I'm somewhat familiar with it.

Attached: flyback.jpg (1867x929, 501K)

What's the difference between a transformer and a rectifier?

They have nothing in common

I am sorry but I have never grasped how a switched mode power supply could produce a clean output. Maybe you could help me understand

Say you want to produce 6V from 12V. That would be a 50% duty cycle. How do you produce a clean 6V output?

He said a rectifier does AC > DC and a transformer does AC > DC

while a inverter does it in reverse DC > AC
and a buck converter steps up DC
Where am I in the wrong?

a transformer converts one AC voltage to another AC voltage
(example: 120V AC to 12V AC)

A rectifier converts AC to DC
(say 12V AC to 12V DC)
but it is a very ripply DC. You will usually want a smoothing capacitor

Not that user but just add a lot of capacitors until you have the desired ripple

a transformer does not convert AC to DC
It converts one AC voltage to a different AC voltage

A transformer converts one AC voltage to another AC voltage. So for instance it can take 120VAC and step it down to 12VAC. The additional power becomes current capacity.

A rectifier takes AC and converts to DC
An inverter takes DC and converts it to AC
A transformer converts between AC voltages

Transformer steps voltage up or down (it also steps current up or down inversely to voltage). In power applications this is the primary purpose. Transformers are also used for impedance transformation but that's kinda beyond the scope of anything in this thread. Finally transformers can provide electrical isolation, at least if they are properly designed. Oh, and transformers are often used as common mode and differential mode chokes too.

Rectifier simply rectifies AC to pulsed DC.

In a power supply you will usually have a fuse or inrush limiting resistor on the hot or neutral side followed by the transformer followed by the rectifier, followed by the filter capacitor, followed by regulation circuitry like a linear regulator or zener diode or whatever. In offline SMPS you will have the bridge rectifier first then a filter cap, then the transformer as the transformer needs a DC supply which will be switched at high frequency via a FET.

Ah.
So it's like a step down converter for AC essentially?

But then wouldnt it just be the full voltage?

Seriously my current understanding is that switched mode power supplies output a square wave form but most devices just dont care, especially if it is at a high enough frequency

to be clear, inverters usually do not produce a very clean AC waveform.

They can go in any direction. So you can have a transformer that convets from 12V AC to 120V AC. Transformers are very efficient at changing AC voltage

If you give me a few minutes I can build up that circuit and show you output waveforms from my scope, maybe you'll get a better idea.

Gotcha.

Thanks for the informative reponses Jow Forumsents.

That would be awesome

>to be clear, inverters usually do not produce a very clean AC waveform.
Yeah they're usually square waves which work for simple appliances and devices made of passive elements but sensitive electronics might not function properly on them.

There's the setup, don't mind the messy bench

Attached: IMG_20180528_202937.jpg (3024x4032, 2.39M)

Input and output voltages. 12V in 180V out.

Attached: IMG_20180528_202953.jpg (3024x4032, 2.56M)

BRIDGE

are you going to show it in switched mode on the scope?

Waveforms. Yellow is the PWM waveform on the gate of the MOSFET. Purple is the waveform on the output diode's anode on the secondary side. Blue is the filtered DC. It's not totally clean since the 47u filter cap I have is probably not big enough but I don't have many large capacitors with high enough voltage ratings so we'll make due. The spikes on the output are mitigated somewhat with a snubber on the input consisting of a TVS diode + 47 ohm resistor.

Attached: SDS00002.png (800x480, 25K)

Where do you find these for cheap, pawn shops?

ebay mostly

Any schemes to find them cheaper on ebay? Do you just bid on stuff and walk away or search new listings, all i see is 30 year old shit for $400

Not who youre asking but I got a bench power supply from eBay thats from the 70s and the specs put modern power supplies to shame. 36V 6A and incredibly low ripple (it actually has an internal variac, a giant 50V smoothing cap, and linear regs)

I got it for like $75.

Nothing terribly helpful. Searching by instrument manufacturer can help sometimes, at least it'll weed out some cheap kits (like those sjitty $20 function gen kits). Otherwise just comb through listings until you find something you're willing to pay for. Often times stuff that's listed as for parts/not working is in fact working it's just the seller doesn't know how to use it. But it's a bit of a gamble though. Just keep coming back and searching every now and again.

I scored a $450 Agilent vector network analyzer and $300 HP spectrum analyzer that way. Those deals were insane. I didn't get quite as good deals on the vintage tek and HP scopes I have but I got each for under $150 and a tek 2225 for $50. PSUs I've never paid over $60 or so. Sig gens the small 2MHz ones are usually under $50 and my better 20MHz wavetek sig gen was more like $90. Multimeters are all over the place. idk what the Fluke 8840A was but it was under $100 I believe, might be wrong though. I have two Fluke 8375A with nixie displays one of which was over $100 and the other was broken so it was like $60. My Siglent scope I paid whatever the retail price for the SDS 1104X-E is. I don't remember off hand. $400ish?

thanks for the info niggas, i'll keep a search notification up

Attached: 1525366806031.png (380x349, 52K)

what's this thing?

Attached: thing1.jpg (2268x4032, 1.36M)

what's this?

Attached: thing2.jpg (2268x4032, 1.36M)

Light dependent resistor (LDR)

Probably a transistor but not necessarily. Google the part number.

I lied, I believe its a hall effect sensor. I got this by searching 3144 to-92.

>electronics engineers
>niggas
user...
We're fucking white males.

>transformers provide electrical isolation
>requires a fuse or regulating resistor before it
Explain in more detail?

>followed by regulation circuitry like a linear regulator or zener diode
Do "regulation circuits" just prevent backflow? Is that their main intended purpose?

photoresistorTM

>square waves which work for simple appliances and devices made of passive elements

Could you give a couple examples of passive appliances?
Is there a niche market for square wave electronics for off the grid living or mobile homes?

Not an EE, someone correct this if wrong.

A transformer is two coils where the current is induced in the output coil through fucking magnetic force and shiet. You can still have reasons to limit the input coil current (especially in case of a short), hence the fuse.

An LED is passive. Switching it on and off will just dim it. However, that would fry your processor.

That is a full bridge rectifier. It's basically 4 Schottky diodes in a shape like pic related.

Attached: 557px-Diode_bridge_alt_1.svg.png (557x232, 6K)

The fuse and in-rush limiting resistor aren't there to protect you (from electrical shock), they are there to protect the circuitry and to protect you (from a fire). If there's a fault and something goes short circuit and starts pulling a lot more current than it should the fuse will preferentially blow up before all the expensive electronic bits overheat and melt or blow up. The fuse also protects from shorts that could lead to overheating causing a fire. In-rush limiting resistors are a bit different in that they help to dampen current spikes when the device initially powers on as that can happen in some devices.

Regulation circuits simply spit out a regulated voltage. The output of a power supply may vary over a small voltage range depending as the load changes. This is not desirable as it can fuck with digital stuff and screws with biasing for amplifiers and things like that so you regulate the voltage. Zener diodes clamp the voltage to whatever their zener voltage is. If your input voltage is 9VDC a 5.1V Zener diode will clamp it to 5.1VDC. Linear regulators do much the same thing but better, they have a feedback loop with a precision voltage reference, error amplifier, etc all which controls a pass-transistor which will vary it's collector emitter voltage in order to maintain a constant output voltage.

ok homo

Real switching power supplies use output feedback to drive the switching process and maintain the desired voltage. The duty cycle is modified as the load changes and the filter just keeps the ripple smoother than it would otherwise be.

If you want exceptionally clean output then a typical setup would use a switching supply to feed a linear supply at just over the linear supply's minimum voltage. That way you can get near-switching efficiency with linear stability.

okay thanks.
What about passive electronic appliances?

Can a DC toaster run on a square wave for instance?

Depends on which part. If it has a micro, that will get fucked up. However the toaster itself is basically a big resistor, so it doesn't matter what you feed it, it simply coverts amps into heat.

Okay so electrical isolation in the transformer sense is a physical isolation because of the magnetic field.
Is that isolation mainly intended to protect the user frock shock and also prevent electrical static build up?

Also does that magnetic field only transfer in one direction?

Passive components tend to be two legged components that work by changing the impedance in some way. Passive components are resistors(like resistive heating elements in ovens/ranges/toasters), capacitors (most kinds of single phase AC motors have a start/run capacitor), or inductors(any kind of motor such as in fans or compressors) for their primary functionality. Many kinds of kitchen appliances make use of them. They pretty much don't care about the shape of the waveform.

Toaster would probably work on DC but DC voltage is not directly comparable to AC. AC is measured as an average, AC tops out at 33% higher the average. If you ran a toaster on DC it wouldn't cook the same way.

Thanks for all the help professor.

Yes. The output of the transformer is not referenced to ground so, assuming the secondary has high voltage on it, so long as you do not connect yourself between each side of the secondary winding you will not get a shock. Pic related. Touching 2 to 0 is safe, touching 3 to 0 is safe, touching between 2 and 3 is bad. The transformer doesn't do anything regarding electrostatic buildup and the magnetic field does change direction in a transformer. See magnetic hysteresis diagrams.

Attached: xfmr.png (420x233, 4K)

Very interesting.
Electronics always pulls me deeper with it's near mystical properties.

If you like mysticism delve into the depths of RF and microwave electronics. Most of the methods of analysis for low frequency circuits start to fall apart when you get into the GHz range. It may as well be it's own separate field of engineering.

Attached: IMG_20161030_234451.jpg (1944x2592, 1.75M)

>Also does that magnetic field only transfer in one direction?
I believe the transformer can only induce current in one direction at a time. The transformer essentially becomes a magnet with the north and south poles oriented with the primary and secondary sides of the transformer (at that specific moment, AC constantly flips north and south). My understanding is if you induce a large enough current in the secondary side the secondary will begin to flip the magnetic field in the transformer instead and the secondary side will induce current back into the primary side which will backfeed into the mains and probably cause an electrical fire. If the two sides are inducing an equal amount of power I think the transformer would be unable to magnetize and instead would heat up and eventually burn up.

The primary and secondary sides of a transformer are arbitrary. We assign the names of the sides based on what side we're using to induce power (primary) and what side is receiving power (secondary).

You make him sad.

Attached: Science-Behind-Railgun-ElectroBOOM-Style.jpg (633x355, 43K)

RECTIFAYA

what is that

Radar detector guts. There's gonna be an LNA and some filters in there and probably some sort of mixer. It's designed to pick up police radar guns which can operate in a few different frequency ranges 10.5GHz (X-band largely obsolete), 24.15GHz (K-band), 33.8GHz 34.7GHz or 35.5GHz (Ka-band)

Rekt

>Schottky diodes
No

>It may as well be it's own separate field of engineering.
It IS it's own field of engineering, at least it's distinct at my university here.

thanks for the info fellas. Anybody live in Reno? Wanna buy this 1981 Osborne 1? HMU

Attached: 20180423_211400.jpg (4128x2322, 3.5M)

I'll buy that pussy.

The absolute state of modern electronics hobbyists

You're fucking a white male?
Are you a white female at least?

> questioning gender on a trap board

Attached: image.jpg (500x500, 162K)

I love you guys.

its a JUUL

dc to ac?

Thats a haisenburg compensator

A bridge rectifier in a package. You still need capacitors, probably.

God bless you anons

You'd get some serious hum out of it. Might damage the element over time from the vibration.

Attached: 1527160081207.gif (575x420, 507K)

The output voltage is never 'exposed' to the switching voltage, instead the output voltage comes from the inductor which doesn't change voltage immediately upon switching. In a buck converter (step-down) when the switch is closed or supply is connected, the current charges the inductors magnetic field which causes a voltage drop over the inductor compared to supply, once the switch is open and supply is disconnected the load is supplied power from the inductor. Switched mode supplies do not simply average out the voltage with a duty cycle as that would make them PWMs instead and unsuitable for anything but simple loads.

Attached: sw psu.jpg (598x494, 92K)