Electrical Engineering

Asmyldof

00:47

@efox29 case you get in: MOre blue-sky-ing can be done here as well

Guys... It's totally normal to still be at the office at 2:30 AM, right?

Right?

The Photon

02:19

@Asmyldof wrong

Asmyldof

02:52

:'(

Only just arrived home.

Nearly dawn

But! My protocase design is complete and the control PCB has been designed

No idea what I'm going to do the rest of the week though

The Photon

@Asmyldof Is that even legal where you are?

Asmyldof

No-ish

In the sense that, if I write down on a form that I worked 18 hours without clearance from a Health and Safety officer, my boss has to pay a fine

But if I write that I have worked 12 hours, and then write 8 tomorrow and 8 the day after, but do 6 hours of work for my freelance company on thursday while at the office, everyone happy

:-)

We have strict rules about working more than X per day and Y per week I think, but my payroll work is only 4 days, so I can "go over" by a full day's worth. For my freelance work the only rules is that I have to pay all my taxes and communal health premiums and try not to die of a heart attack.

And in a lot of places there's a lot of freedom anyway

The Photon

@Asmyldof You need to appoint yourself health & safety officer so you can approve your excess hours.

Asmyldof

No I don't, because I'm a freelancer

I also don't need them to check my sick days, because I can freely decide whether I pay myself or not for my sick days

The Photon

@Asmyldof Are you incorporated?

Asmyldof

03:01

My employer has to pay me after 2 days, but they are insured for that, so they have a H&S officer who then checks up on me, and after X weeks they want you to go to their doctor to check up and work on getting better as soon as

Nope

Else I would have to pay myself wages

If I were the director of my corporation I would have to keep all the money cleanly separated and pay myself an explicable wage

Even if I'm also the one and only engineer and valued at entry-level, if my company does well (for whatever reason), I'd still have to pay myself a director's wage, in stead of an entry-level engineer's wage

The Photon

@Asmyldof But (at least here) you could pay yourself minimum wage and issue yourself a dividend whenever you wanted.

Asmyldof

If my company makes $1000000/yr the tax agency will not accept me paying myself $2200/month, because that is not in line with my performance as director of a $1000000/yr company

They immediately flag that as tax evasion schemes

The Photon

@Asmyldof If you make $100000000 per year, I don't feel sorry for you.

Shalvenay

@Asmyldof -- wow. they flag that before they look at where the rest of the money's headed?

Asmyldof

6 zeroes would be possible if I went double-full time in my own company and didn't mind rotting away my soul

Shalvenay

03:08

I mean...calling Warren Buffet a tax evader is just plain silly -- he's the type to wonder why he pays less in income taxes than his secretary

Asmyldof

If I have a company that makes $1000000 and my only employee is me... whee exactly do you propose that money going?

where*

The Photon

@Asmyldof "expenses"

Shalvenay

yeah -- expenses!

The Photon

Like the important conference your engineers have to attend in Burmuda.

Or Bali or Hawaii

Asmyldof

I'd have to continuously fly around the world to make that happen. Expenses need receipts, or they aren't expenses

The Photon

03:10

@Asmyldof Rough life, that would be.

Shalvenay

there's also your business needing a rainy day fund

Asmyldof

Fact is, if your company invoices $1200000 and pays $200000 and you declare that as you should and you then pay yourself $50000. There's only thing that will happen: immediate audit

If you don't declare the income, that's another thing, it;s 90% computers.

But at some point you will be audited and they will then also check the books of your customers if they even smell the slightest hint of bad

And I'd have to avoid doing anything through a bank, which is a hassle...

If I'd get that much in cash I'd probably actually be travelling all the time

Shalvenay

@Asmyldof -- the key is they can look all they want and they'd find nothing wrong :P all the money would be sitting in the business' cash pile

The Photon

@Asmyldof I'm willing to help you out. If you have $1000000 lying around that you need to make disappear, I am available as an engineering consultant.

Asmyldof

And then go bankrupt, because after 6 months I'd have wasted everything and also done no work :-)

@Shalvenay No, because I cannot have the money if I don't declare it. Or they will conclude tax evation

Over here they do actually tax people with money. The more for it

;-)

@ThePhoton I'm going to keep that in mind

The Photon

03:16

I even accept payment in Euros, guilder, or whatever.

Asmyldof

@ThePhoton I actually should go to a cool conference again soon, been way too long

((I only get to declare 75-ish % though, by the way))

The Photon

@Asmyldof I think your first expense should be to hire a decent accountant.

Asmyldof

Nope.

I have a very good accountant

He wrote all my bookkeeping software when his EE/SW/FW work wasn't coming

Asmyldof

03:37

Nap time!

Ah, I wish I could see why voted to close

who*

pilot

17:48

Is wireless sensor network a saturated field for research?

akaltar

19:07

Hello! Could anybody recommend me a simple stepper motor driver that uses only 2 pins for control( step + direction )? I couldn't find any....

Asmyldof

BALLS!

I am only allowed 30000 characters in an answer!

:'(

@akaltar Allegro Microsystems makes those

Have microstepping even, allowing you to pulse 1/16th of a step

akaltar

19:23

They still use 4 pins minimum as I looked at the smallest one, but even that wasn't available in my region (Well... Aliexpress has it for a 1 month delivery...)

Could I use a L293D H-bridge + buffer to achieve the same?(or similar)

I am trying to control 8 stepper motors from my Raspberry Pi using demultiplexers + some controller

Asmyldof

19:43

What do you mean with the 4 pins?

I have used allegros with just 2 pins in test setups

(I'm also finishing up a monster answer on a seemingly simple question, so my attention is a bit divided)

akaltar

I am a newbie, so I might be misinterpreting the datasheets(well, for sure now), still it seems a waste to use so many pins...

Nick Alexeev

@Asmyldof What were you going to write about at that kind of length?

Asmyldof

If you explain to me how you came to 4 pins, by also linking the datasheet and virtually pointing to what you saw and thought about that, I might be able to help you.

@NickAlexeev I'm afraid to tell you, or you'll delete it

:-P

Nick Alexeev

@Asmyldof Would you think I'd delete it as off-topic, or low quality, or abuse, or copyright infringement, or ... ?

akaltar

I searched on their site for the lowest pin count controller and found this: allegromicro.com/en/Products/Motor-Driver-And-Interface-ICs/… The others which used 2 pins for control seemed to have other pins that were necessary.

Asmyldof

19:56

@NickAlexeev Not really, to be honest. Someone asked about "how to find values" in a single transistor design with way too ambitious intentions (hard on/off), so I am calculating the specifics, but explaining each step.

I have added nice little tea breaks for the reader though

@akaltar You'd have made me very happy by linking the datasheet, their website is all garbled on my screen (probably the hack-a-thon of Windows/Curl I have going on which I am taking a 5 hour break from) so I had to google. Got it now though. Hang on

Yup, that's pretty much one that needs 4 pins.

Because it doesn't do anything smart

Also, it's a thermal-pad-ed QFN. Are you sure you want to do that at the Arduino stage?

They are very pissy to solder by hand if you haven't practised a couple of times before.

The chip you chose, though, in your enthusiasm, is one that only does the H-bridge-ing

akaltar

I can manage SMD-s, but I prefer SOP or similar design.

Asmyldof

So you have to give signals yourself to control the coils. They have chips that do that for you

It's not just SMD. It's a QFN with a thermal pad that needs to be cooled, or your chip will melt

akaltar

I understand, and Its probably not what I want.

Asmyldof

cooled thermal-pad = hard to get hot enough with normal tools

bit of a chicken-egg kind of thing...

Anyway

akaltar

At the moment I am considering using a simple H-bridge(L293D) and demultiplexers + shift registers.

Asmyldof

20:02

I'll start virtualbox to browser you a better chip, gimme a minute

akaltar

Thanks.

Asmyldof

Holy crap. Even un-garbled their website is made of the poops

maybe 2 or 3 minutes

akaltar

And also most of their datasheet links are broken.....

Asmyldof

Ah, for f.... A PDF downloaded selection guide? Seriously? Is this 1992?

Well... stepper motor driving is very 90's, so I suppose that means they're serious about their core business

Ah! Tablez!

Tell me, what are your parameters in the sense of supply voltage and currents and all these cute numbery things?

@akaltar If it's in the 100's of mA region I have exceptionally good news for you

They have a SOIC for that

akaltar

5V supply. Annnd.. not really sure about amperage.

Asmyldof

20:08

Still pissy because of the thermal pad, but the pad is larger and further away from the pins which makes it a tiny bit easier

Give me what you have about the motors then

1. what will they move/rotate and how, and which motors are they (preferably directly a datasheet link)

and 2. see the second part of 1, because while typing that I forgot halfway that I was doing numbers

(use @), I'm back to my answer for a little bit in betweem

akaltar

@Asmyldof Thanks for looking for specifics, but the situations is, that we have about 5-10 different motors(to be reused). After reading a bit I have found out that stepper motors are operated not by constant voltage but constant current, if that means that it will work, just be less powerful, then 100mA by all means is enough. We would be happiest with the simplest solution requiring the least external components.

Asmyldof

20:23

@akaltar Then you should look at A3967 (there are other companies that make similar ones and it is possible, or maybe likely, that a few of them still have SIP-packages, that you can cool with a normal heat-sink like a longer, bigger version of a TO220 transistor with more legs) for a SOIC24 with thermal PAD, maximum current 750mA, goes from 4.5V to 30V (so you can tweak with the voltage)

You have to be aware though, that neither constant voltage or constant current really applies

They are a combination of a resistive and inductive load, and if you are using them at their most efficient, their current in the coils is constantly increasing and decreasing

Often more complex drivers use higher voltages than the defined motor voltages to achieve these waveforms quicker

But, if you use an allegro chip, all you have to take away from that is: Set the maximum current as you think is safe, hook it up, pulse it.

If it can't go fast enough, you can increase the voltage to the allegro a few volts as an experiment, if it's not strong enough, maybe tweak the current limit a little.

akaltar

I have 5 volts from battery, can't change that...

Asmyldof

That said, I'll browse to the chip mentioned, A3967, with you to help you figure it out

akaltar

But thanks, I am sure this will help me a great lot.

Asmyldof

Well, you can change any voltage into any other, it's just harder.

akaltar

Yeah, I know... just built a multiplier.

Asmyldof

20:28

But, it's not for me to design your system, it's just for me to tell you what I know, or presume to know, and let you figure out what to use and how

A booster can make 5.5V, or 6V, or 7V, or 50V from your 5V, though the latter would not be wise if you then drive 5V rated steppers with that: lot of watste

waste*

Anyway, hang on, gooling...

googling*

allegromicro.com/~/media/Files/Datasheets/A3967-Datasheet.ashx

(will download and open)

And I see right away that this is exactly the type I meant

But I'm cheating, I've used them before

akaltar

I have this open ofcourse already.

Your help is very great.

Still it seems that this is an overkill for my simple setup.

Asmyldof

I'm just procractinating from my work ;-)

Probably is, but a chip like this is very easy to use

And again, there's other companies who may have something more appropriate

akaltar

I want to learn to properly control a stepper motor.

So a simple H-bridge will be most likely what I stick to...

Shift registers + 2 demultiplexers. (1 to 8 )

Asmyldof

Ah, you can look at the pages 9 through 13 for some cool images

It might be faster in a few senses to use a buffer chip for your signals, I presume that's your plan, or use an I/O expander from Microchip on your I2C

That said, if you need to control a lot of motors reliably and/or well timed on a small arduino, using multiplexing tricks may not be the best way.

akaltar

I plan this: Raspberry GPIO -> Optocoupler -> demultiplexer -> shift buffer -> H-bridge.

I don't need to be reliable

Asmyldof

20:36

It depends on the target of your project really. If you're just having fun and seeing how far you get I can only support you trying everything yourself

akaltar

Its for moving a camera, one axis at a time.

Asmyldof

Fair enough

akaltar

Yeah I just want to know if its possible at all.. I have way too many times wasted components by using them wrongly.

(3 years at electrical engineering school didn't help me much)

Asmyldof

I'm thinking in the AdaFruit sphere of hobby people there must be a couple of good tutorials on driving steppers with arduinos with simple components. Step 1 is doing that with your arduino without any multiplexing tricks

akaltar

But your help has teached me much!

Asmyldof

20:38

Jumping in at the deep end will have you drown for sure, start small

akaltar

I am designing a PCB and want it to be usable in the future.

Asmyldof

make one motor move one camera-weighted thing

akaltar

Yes, of course

Thats the plan

Just want to design the pcb to be expandable.

Asmyldof

If you are unsure about what you need and making a PCB is a lot of work, you are best of not doing that right away

Either you make a PCB design now that will have 1000% of the parts you actually need later, or 10% of them

akaltar

We already have DC motor control with pwm, this would be the next step.

Asmyldof

20:40

If you just use a protoboard or a breadboard for most of it, and if there's a cooling pad solder a big 4mm2 wire 50cm long onto it, you can actually experiment with it and see what might go wrong

Yes, but stepper motors are a very different beast to DC motors and PWM

akaltar

We prefer to order components as a bulk, both because a week minimum delivery time and prices...

We do prototype them first.

Still your tips are very welcome.

But now I am going.

Thanks for all the info.

Its remarkably hard to get such detailed information when you are hoarding it from the web.

Everyone expects you to know the basics which you don't....

Now I'll go.

Have a nice day

Bye.

Asmyldof

@akaltar Good luck. Enjoy, and I'm happy to help if I can

(and have time to waste I don't feel like putting into actual work)

Branden Boucher

20:57

Hello all

PlasmaHH

gaaaaaahrgh, these scope probes suck

Asmyldof

21:33

@PlasmaHH yup, they do

@BrandenBoucher Hai

PlasmaHH

@Asmyldof give me better ones, now!

Asmyldof

Eh

No

PlasmaHH

and while you are at it, tell me what I am doing wrong here...

Asmyldof

Learn to use the crap before you get the shit.

You're measuring crap with an oscilloscope?

PlasmaHH

no, its just that this crap is falling apart when I use it...

Asmyldof

21:37

So you are measuring crap

PlasmaHH

of course, I hacked it together ^^

well, the other part is crap from ebay ^^

Asmyldof

And you're still asking me what you're doing wrong?

I will trade you an old HP for a hack to be allowed to post a 32000 character answer, without linking off-site

Oh, no never mind, found a paragraph I can scrap

Well, edit

Asmyldof

21:58

aaaaaaah

10 more characters to make go away

As a first ever I have summarised a whole paragraph that had some actual, though be it weak, ties to electronics into the sentence "Because pink and purple are the only real colours in the world!"

I'll make me laugh and that's good enough for it to be small on the edge of a schematic

The Photon

@Asmyldof You might want to consider getting more sleep.

Anonymous

good day everybody

Asmyldof

@ThePhoton I'm pretty sure sleep-deprivation was invented in the 8th century BC by very early Romans to keep their slaves in check. As I do not believe in Romans (it was all an alien conspiracy), I don't believe in sleep-deprivation. So, I don't see the point really

@PatoSáinz hello

Anonymous

is anybody familiar with frequency doublers?

Asmyldof

@ThePhoton Although, I would have put all the maths in schematics from the start had I known about the 30000 character limit for answers.

@PatoSáinz depends on what kind

Anonymous

22:09

@Asmyldof I have to build something that takes 50 Hz (mains here is fine) and outputs 100 Hz of AC

Anonymous

and there must be more than one way to do it, but I wonder what's the simplest way

Anonymous

I've been googling extensively

Asmyldof

Oh, that's not too difficult

Anonymous

yea, but I've never done AC stuff before

Asmyldof

couple diodes, capacitor, resistor, transistor and some lower AC voltage. Or lower DC voltage and another diode and possibly an optocoupler

How many of what kind of depends on what for exactly

Anonymous

22:11

and I've been thinking a couple of ways to do it but most of them just end up as pulses with no alternating current

Asmyldof

Oh, hold on, I didn't read that part

Anonymous

for now the best way I found is

Anonymous

Anonymous

but I have no LC meter

Anonymous

@Asmyldof yea, what's hard is to end up with AC

Anonymous

22:13

of course I could set up a diode bridge and get 2f but there wouldn't be any ac there

Asmyldof

What for? might be the best question at this point

I thought you meant a 50% square at 100Hz or some such

The Photon

@PatoSáinz How much power?

Anonymous

@Asmyldof education, amusement

Anonymous

@ThePhoton enough to drive a little motor

Anonymous

it's a little coil that has a little ink hammer

Anonymous

22:15

so currently the hammer goes down with f = 50 Hz

Anonymous

I want it to be 2f

Anonymous

I also found

Anonymous

but I have no LC meter so that'd be hard. also I don't know if my NPNs can switch that fast (of course, 50 Hz isn't fast but it's something)

Asmyldof

Any normal NPN can handle into kHz

Most will do that trick up to the given 1MHz

Anonymous

22:18

oh wow I thought they weren't that handy

Asmyldof

But, that trick doesn't work very well if you take out power

Another but: your coil is an inductore

-/- e

Anonymous

oh of course I know it is but do you mean that I should account for it in my LC circuit?

Asmyldof

So, you can actually do that exact trick with your coil as the inductance. It will not be as perfect or easy with that coil as with a real inductor

But it will be easier to use the coil directly in tricks like that, than make a tuned circuit and then connect another L

Anonymous

I have no LC meter tho... making tuned circuits is hard

Asmyldof

And what's great is, you may not even need an LC(R) meter

That last one is a bit down to the values.

Anonymous

22:20

@Asmyldof are those random turns*diameter formulas actually useful?

Asmyldof

Eh, those formulas aren't really random, but in your case, probably no

What I mean is quite different, fun and playful

Anonymous

@Asmyldof I've found endless variations on them

Anonymous

@Asmyldof what do you mean haha?

Asmyldof

That's because they depend on the core material and such

Which is also why it's pointless for you: your core is icky and it moves

Anonymous

hmm

Asmyldof

22:24

What kinds of things that measure do you have?

We are going to MacGuyver you along

Anonymous

a cheapo multimeter

Anonymous

a makeshift CRT TV oscilloscope (so 59.9Hz horizontal refresh) that will probably burn if I feed it anything over 6V to the deflection coil I guess

Asmyldof

Ah, this is good news-ish

Anonymous

ish

Asmyldof

Truly good news would be "I have an 10kHz+ scope with 1s phosphor"

But, anyway

digressing

Any acceptable DC voltage sources?

Such as there are, let's say batteries

Anonymous

22:28

hobbyist electronics are expensive. Also, yes, batteries

Anonymous

I used to have a diy bench psu but I threw it away

Asmyldof

Hobbyist electronics are expensive because they are sold at hobby stores

Hobby stores buy them slightly too expensive in 100 ~ 1000 quantities and then mark them up 500%

Okay, what else do you have handy in the way of resistors and/or transistors and such?

Anonymous

yea but even used instruments are expensive. What do I have handy? Well varied resistors (I guess I can always just add them in series to get what we need: and transistors I have some NPNs from power supplies... also I remember having some nice opamps-in-a-chip and diode bridges-in-a-chip

Asmyldof

Okay, now it's up to me to find out if I'm allowed to use circuit lab without an account or without EE.SE

Anonymous

Also, some assorted caps

Asmyldof

22:34

Becuase images = easier

Anonymous

@Asmyldof thanks a lot

Asmyldof

For instruments you should look at universities and companies that do EE, they throw away crap every year, 5 years and 10 years, depending on the type of crap

Some scopes might have a uni-lifetime of up to 25 years, but those scopes that do are of the sort that are still perfectly fine (though less accurate by now) 30years after they were bought

They won't have memory, of FFT, or whatnots, much like your DIY TV one,, but they'll be 10MHz+ with fancy dials and knobs and settings that your DIY one doesn't

Anonymous

I know someone with an old 15Mhz phywe with a blown fuse, might try there

Asmyldof

Anyway, stop distracting me, I need to help you!

Anonymous

ok hahaha

Asmyldof

22:37

Oh!

Hang on

What's the specs of the coil you want to pulse?

aproximately

Anonymous

@Asmyldof oh god I wouldn't know... I don't have it with me right now... I guess I can give a rough estimate of its length and diameter

Asmyldof

I sort of meant, what voltage do you have it running on, what currents (even just a maximum) do you expect it uses

Anonymous

ah

Asmyldof

stuff like that

ball-cakes! CL needs an account

Toilet-break-ponder-time

Anonymous

well it's 220VAC @ 50Hz... I have not hooked an ammeter to it so I wouldn't know how much it drew but it wasn't some heavy duty motor... it was actually kind of artesanal and it just moved a little stick with ink on one end

Asmyldof

22:45

Ah. High voltage

Hohum

Then you do need to look up the transistors you want to use

Many stop at about 60V

Or actually, they start there. Start always being on and getting hot

And of course: be careful with trickery

That does pose a little problem in my plan though

Different plan

What kinds of capacitors do you have?

@PatoSáinz Oh, sorry, I'll grab your attention.

Anonymous

@Asmyldof electrolytic mostly

Anonymous

@Asmyldof oh god i forgot about that little detail... I guess I can just use a transformer?

Asmyldof

In principle, but if it is made for 220V direct connection, you'll be hard pressed to get the required current through it at a lower voltage

Anonymous

@Asmyldof it's just hacked to work that way

Asmyldof

And with hard-pressed I mean it'll have to be much slower most likely

Anonymous

22:51

do you think it really makes a difference?

Asmyldof

What do you mean?

how is it "hacked"

Anonymous

i mean that right now it's just a coil that someone put in front of and old cable that someone connected to mains

Asmyldof

You do have a current measurement by the way, needed to point that out: I = V / R

Anonymous

I'm not sure it was product of "design"

Anonymous

@Asmyldof I wish I had hooked up an ammeter... I can do that tomorrow

Asmyldof

22:52

Ah, well, still, if it works without evaporating, it's likely not going to be very enthusiastic about 40V

Ah you do have the meter, but didn't use it, fair enough

Anonymous

@Asmyldof most of my NPNs are big and come from power supplies, I guess they'll be fine... I'll find the datasheet

Asmyldof

How much V's can your best caps take?

Ehhh, power supply tors are often 2N-thingermajigs in metal cans for the linear secondary control, so it's unlikely they go over 100V

And that's break down, not "smart to use in a resonant circuit"

Anonymous

thanks for you help... should we pause this and resume tomorrow when I have photos and more concrete numbers? also: some ceramic were kilovolt-rated... I have some 50V electrolytics... I can't remember off the top of my head but I'm sure I have a couple of 300V-rated

Asmyldof

If you have capacitors that can take 400VAC or 250VDC you can use those to estimate the L

Ah, there you go

I'm not sure I'll be on much tomorrow

Too much procrastination done today

Did a lot of work yesterday, so now I made promises to other people I'd have docs done and such

:-D

Buttt

I can give you

1728.org/resfreq.htm

and for theory:

LC circuit

An LC circuit, also called a resonant circuit, tank circuit, or tuned circuit, is an electric circuit consisting of an inductor, represented by the letter L, and a capacitor, represented by the letter C, connected together. The circuit can act as an electrical resonator, an electrical analogue of a tuning fork, storing energy oscillating at the circuit's resonant frequency. LC circuits are used either for generating signals at a particular frequency, or picking out a signal at a particular frequency from a more complex signal. They are key components in many electronic devices, particularly radio...

Anonymous

@Asmyldof hahaha it's alright I'm not that hard-pressed to finish this now... I guess we can get much more done if I have more information and stop being vague

Anonymous

22:58

@Asmyldof this one seems especially useful

Anonymous

thanks!

Asmyldof

poo! why doesn't it just link wikipedia

Anyway, wiki LC resonance

for theory

If you have an AC current and AC voltage measuring device you can get L by knowing C

Anonymous

@Asmyldof when would you use a parallel LC circuit vs. a series one?

Asmyldof

It's not accurate or the best way

Anonymous

I have AC current and have my multimeter that can measure RMS

Transcript for