[Solved] Need a little hardware guidance on Servos

Hi @inq,

   The case for and against common grounds and so on, although simple in principle, often represents a difficult problem when designing a new electronic product. Such products are often redesigned several times during the development phase to achieve a reliable product that meets all requirements, including functional performance, unaffected by electrical noise and so on.  'Home' designs do not have to meet all of the criteria a commercial product must meet, but following sound principles improves the chance that the 'home' product will work reliably without redesigns.

It is also quite a difficult subject to cover in a few words .. I am sure Bill could do a great video (it might stretch to a small 'box set' ) about it, with animated diagrams and so on, but it would take a good deal of preparation work.

The basic principle is that current always flows in a complete loop, from one terminal of a power source to the other terminal.  For a simple circuit involving just a battery and a bulb, that is fairly obvious, but for 'real' electronic circuits with multiple power sources and multiple loads, it can quickly become much more complex.

Things tend to 'go wrong' when the currents from two (or more) loop circuits both pass through the same component, such as a wire or a resistor.

If two loop circuits, such as two torches, are only interconnected by one wire, then that interconnecting wire cannot be part of a loop.  The problem can only arise when there are at least two interconnecting wires. Sometimes it is tricky to spot the second connection.

--------------

So assuming you have an Arduino, supplied by a 5V battery. You also have a dedicated motor controller and motor supplied by a 12V battery.

The motor controller has a suitable Input drive pin and associated '0V' pin.

The Arduino has a suitable output 'GPIO' pin and associated '0v' pin.

You connect the Arduino output pin to the motor controller Input pin with one wire, and connect the two '0V' pins together, using a second wire.

With a bit of luck it will work fine ... no problems.... the Arduino commands the controller, and the motor works.

-----------

But now you decide to replace the batteries with power supplies. Things start to become a little murky..

If the power supply output pins are isolated, in particular from mains earth, then it too will probably work.

(You can test for this with a multimeter. Unplug the power supply from the mains. If the supply was powered on recently, leave it for a while to discharge any internal capacitors. Check resistance from Earth pin on the mains plug to each of the power supply output pins in turn. Any indication of a resistance value, including zero, shows connection between the earth and the output pins.)

...........

If the power supplies (internally) both connect one of their output pins, to mains earth, then you have a loop .. and things can start to go wrong.

Perhaps the worst case is one power supply connects mains earth to its positive terminal, the other one connects mains earth to its negative terminal ... the result being one  power supply output is now connected to a short circuit!

But even if this worst case is avoided, small resistances from the wiring, etc. can result in a current around the loop, and strange things begin to happen. Audio equipment users may be familiar with the 'hum' caused by an earth loop ... digital equipment users will not notice this hum, but can suffer problems of their own, including the infuriating once a week crash scenario.

---------------

Thus the rule is to identify each loop, and ensure that there is only one ground connection between any two loops ... including any 'hidden' connections, such as those through mains earth wiring.

If you wish to ensure your 0v lines together to ensure 0V is the same for each circuit then bring one wire from each loop, to a common point. This is known as a star connection.

In addition, if you need to ensure your 0V lines are at the same potential as mains earth, then connect the mains earth wire to that same 'star' point.

-------

Sorry .. this is lots of horrible words ... maybe you can begin to see the pattern.

-------------

As a couple of examples, I have borrowed from Bill's excellent pics

-------------

From his large stepper motor video ... it is easy to imagine the Arduino and the connections to the left hand connector of the controller form one power supply loop, although he hasn't shown the USB plug in the USB socket, providing the 5V supply.

The motor and 24V supply connect to the right hand connector. This is the second power supply loop.

What is not obvious, until you look carefully at the motor controller data sheet, is that the input connections on the left hand connector go to opto isolators.

Thus, there is no electrical connection between the Arduino,etc and the stepper motor,etc, via any of the components and wiring shown on the diagram.

It is still possible for (say) the 0V pin on each of the power supplies to be connected to mains electrical earth, but there is no possibility of this becoming a loop.

An example of this came up on a DBW Forum thread a few months ago ... it was expected that 0V lines from the Arduino had to be connected to 0V lines from the motor controller ... this was an error ... the only connections from the Arduino go to the lefthand connector on the controller.

-----------------

This one is from Bill's Using Servo Motors with the Arduino

As with the previous picture, it is presumed the USB plug is inserted into the Arduino socket and providing 5V supply to the Arduino.

The 5V-6V power supply is providing power for the servo motor, including electronics built into the servo motor casing.

Thus there are two independent power supplies, one in a loop with the Arduino, the other in a loop with the servo.

The black wiring also shows a (T-shaped) 'star' connection point .... a wire from the Arduino 0V line, meets the 5V-6V power supply to servo 0V.

This 0V connection from the Arduino, is also the 0V reference & return line from the Arduino control output pin 9.

As the servo is only low power, the same 0V reference wire is used for both the control and the motor power ... if the servo motor had demanded large currents, these would have been separate.

--

If BOTH the 5V USB power line is connected to mains earth, AND the  5V-6V power supply output are connected to mains earth, then there is now a ground loop, which MIGHT give problems, and is a situation to be avoided.

Ideally, neither the USB 5V supply or the 5V-6V outputs would have any connection to mains earth. If you need to connect the wiring to mains earth, then the correct point would be at the star connection point, shown on the diagram by the meeting point of the three black lines.

------------------------

Best wishes, Dave

@inq 

This is for the Ralf project and it will have at least six servos on it and will be powered by only an AC adapter during use. I think you offered up on another thread some kind of I2C board that can drive up to 16 servos. I need to find that again 

Forgot Bill had this covered.

https://dronebotworkshop.com/servo-motors-with-arduino/

Posted by: @davee

Thus when the 'sandwich is first 'pressed' together, the component is not really a capacitor, since the two plates are not insulated from each other. Thus, there is another manufacturing step ... the component is connected to a DC power supply, and the resulting current causes one of the plates to be chemically changed, probably into aluminium oxide, which is a good insulator. Thus, the current soon reduces to (very near) zero and the component has become a capacitor.

Thank you for taking the time @davee!  I always new it was based on plates and gaps... I remember fondly, about 4 years old, my Dad and I build some kind of radio... that used what I thought was way cool... some fins that rotated in/out of another set of plates and he taught me the fins can never touch each other and so I had to be extra careful.  

And I knew they round can versions were rolled up... but I always thought they were some aluminum alternating with something like mylar, mica or such.  The making oxide step using electricity makes sense from a chemistry standpoint that if you reverse the polarity, you'll be driving the oxide out and returning it to a conductive metal layer, but seems like an extra cost burden.  Surely these days, there are man-made, thin films that can compare to the oxide versions.  Must be a permittivity advantage.

Posted by: @davee

Electrolytic capacitors provide a large capacitance, and hence a large charge storage, for a given component volume. This is useful for handling low frequency demands which take a considerable (more than a microsecond say!) time. If they were competitive runners, these would specialise in marathons!

The ceramic capacitors are better adapted to high speed transitions, with better high frequency response, but the capacitance is too small for the longer time current demands. These are the sprinters!

Ok... this rings a bell also.  Some time ago, I made some home speakers and interested in the acoustics, volumes, ducts, varying the geometry to eliminate standing waves and reinforcement waves... but when it came to the drivers, I just added certain size (very small mostly) across the tweeters to keep out low frequencies... I diverge.

Is it just... that is ceramic synonymous with small and electrolyte with large... and why they are say, "add electrolyte and a ceramic" or are there other properties associated with the two types?  And I've seen where they make it sound like I must use a Tantalum capacitor... like it is some silver bullet to do "something".

Again, thank you for being so thorough.  This I'll retain.  I won't forget about rusting them makes them work! 😋 

VBR,

Inq

P.S. - Were you a teacher of something?

@inq 

I remember fondly, about 4 years old, my Dad and I build some kind of radio... that used what I thought was way cool... some fins that rotated in/out of another set of plates and he taught me the fins can never touch each other and so I had to be extra careful.

That was a variable capacitor connected in parallel with an inductor to tune for a particular radio frequency.

Data was carried by the radio wave by being modulated in different ways, amplitude modulation AM (video in tv), frequency modulation FM (sound in tv) and phase modulation PM (colour in tv). They were the old analogue days before IC's made digital solutions possible.

I have fond memories learning how radio and tv sets worked but I had to do it alone from books as I didn't know anyone that shared the interest in electronics.

Hi @inq,

  Thanks for your reply .. good to see you enjoyed it! As @robotbuilder explains, the variable capacitor with interleaving fins you remember would indeed be part of an inductor/capacitor tuned circuit. When I described the maximum capacitance of 500 pF, that was the example I had in mind ... they were typically 300-500pF with the fins fully enmeshed.

I don't pretend to be up-to-date with all the 'tricks' electrolytic capacitor manufacturers use when they are making them by the billion, but I think the process is at least roughly as I describe ... at least close enough for you to have some idea as to what's going on. The 'supercapacitors' take it a stage further by making the surface 'super rough' at the molecular scale, so that surface area becomes massive, whilst still keeping the insulating layer exceedingly thin.

Tantalum capacitors are not 'magic bullets' ... though if you abuse them they can become miniature 'incendiary bombs'!

As I mentioned, electrolytic capacitors (And indeed all capacitors to a greater or lesser extent), have intrinsic properties which tend to make them 'less good' at their job. They are often expressed as an 'equivalent circuit ... like this one I just found in Wikipedia at https://en.wikipedia.org/wiki/Electrolytic_capacitor    (You might like to read the article .. it is better researched than my note that was just a late night waffle!)

As you can see, the 'perfect capacitor' C has a resistor R-ESR and inductor L-ESL in series with it, plus a second resistor R-leakage in parallel.  (ESR =Equivalent Series Resistance) The series resistor and inductor obstruct the current flows that charge and discharge the capacitor. The parallel resistor allows a small current to flow through the component .. thus if you charge a capacitor 'on the bench', disconnected from other circuitry, then the capacitor will discharge through this resistor. Of course, these resistors and inductor are not actual components in the capacitor ... they are just representations to model how the real capacitor will behave .. typically an electrical simulator like SPICE will be programmed with an equivalent circuit, such as the one shown, and appropriate values for the components inserted so that the simulated circuit behaves in a similar way to the 'real' circuit on the bench.

In addition, if you buy capacitors from one of the 'major' manufacturers, they will list representative values for R-ESR, and sometimes the other resistor and inductor as well, as part of the specification for each individual part number.

Tantalum capacitors, at least at certain times over the last 50 or so years, have offered a smaller component, with lower ESR, albeit at a premium in cost. Hence, at times, they were favoured in military/aerospace applications, with other industries playing the 'if it is good enough for .." premium game ... but at other times, their unsocial habits have caused them to be strongly deprecated/banned.

The present position is probably closer to neutral, until cost is considered, at which point they tend to head towards the bottom and below, of the short list of possible components. At the risk of starting a flame war, I am not aware of an application that must have tantalum, though there will be places where it is a contender. Perhaps the most common 'tricky case' is the input and output capacitors of power supply regulators, especially switch mode type. Most switch mode designs are now based on chips, and the chip manufacturers give guidance which is usually well researched for obvious reasons. I think in most cases, ceramic devices, which can be 10 microFarad or more in small surface mount package will be advised, but take each case on the evidence you can find.

-------

Hope this is of interest ... 

No, I have never made it as a teacher .. most of my life has been nearer the research and development side --- and I have been known to try to explain things to colleagues on 1-to-1 basis over the years.

Best wishes all, Dave

Posted by: @robotbuilder

@inq 

This is for the Ralf project and it will have at least six servos on it and will be powered by only an AC adapter during use. I think you offered up on another thread some kind of I2C board that can drive up to 16 servos. I need to find that again

https://diyi0t.com/servo-motor-tutorial-for-arduino-and-esp8266/

https://nerdiy.de/en/howto-espeasy-pca9685-pwm-port-erweiterung-an-den-esp8266-anschliessen-und-ansteuern/

 

That's the beast.  When you originally posted it, I registered it in my mind, but not a bookmark.  I don't use servos much.  Now doing the Ralf project changed all that.  I can see someone wanting far more than just eyes.

Have you used it?  Am I assuming right it has all the PWM capability to drive the servos and simply send it servo ID and angle you want???  This sounds good in that it actually reduces MPU processing.  Bit-banging PWM isn't easy on an MPU.

VBR,

Inq

Posted by: @robotbuilder

@davee 

I was going to go into the functionality rather than the physical construction of a capacitor but for me I like visual explanations.  I don't know how much @inq knows to begin with.

 

 

Assume an empty vessel.  The derogatory term for an Aerospace Engineer is Air-head.  🤣   My hardware background is:

• Built (I think) Crystal radio with Dad at age 5.  Dad built color TV kit, I learned resistor color-code and picked his resistors for him.
• Require EE 101 - Ohm's law.  

That's about it.  I consider myself a Cookbook hardware guy.  I Google search for the parts I want to use.  "ESP8266 servo potentiometer" and kiss a bunch of toads, let a little smoke out and finally get the princess in the end.

That's why I buy in bulk. 

I like and need the basics... I've always wondered why some had +/- requirements.  Looking at Wikipedia, doesn't even mention that aspect.  With @davee comment... now I know.  Cool stuff. 

VBR,

Inq

Posted by: @davee

If you feel you can clarify things I missed, didn't make clear or even muddled up, please add your own contribution.

Well, I don't think I've ever seen someone elaborate MORE detail on something you have written.  🤣 

Posted by: @davee

Obviously we can only guess what other people know ... I tend to say more than necessary, on the basis that an excess of details is usually better than assuming the recipient knows something, whilst hoping it doesn't get too tedious for the recipient.

I think this is a wise way to proceed even if the pupil is more expert.  Remember... you are not writing just for the one pupil, but for the whole Internet.  In my searches over the years, I get hits to dronebotworkshop all the time.  There will be those just starting out and @davee "War and Peace" will be just right!  As it was for me.

VBR,

Inq

Posted by: @robotbuilder

Forgot Bill had this covered.

https://dronebotworkshop.com/servo-motors-with-arduino/

One of these days... I'm going to be able to search for and watch videos at home.  I promise.  So that board is in that video... good, I'll bookmark it for my next excursion to the library.

Posted by: @davee

As you can see, the 'perfect capacitor' C has a resistor R-ESR and inductor L-ESL in series with it, plus a second resistor R-leakage in parallel.  (ESR =Equivalent Series Resistance)

I'm guessing for government work, its close enough that I just use the C part. 😉 Most of my cookbook hardware projects seem to say use one bigger than X or use a small one like Y.  They seem to be really fuzzy numbers in these Arduino/ESP8266 designs.

Posted by: @davee

typically an electrical simulator like SPICE will be programmed with an equivalent circuit

I've heard of that.  My sister and brother-in-law are EE's... they design CPU's.  I think he had a hand in the still-born joint HP-Intel project that birthed the Itanium series of processors IA-64.  Interesting... that he doesn't mess with Arduino for hobby... he builds and drives race cars for fun.

Posted by: @davee

No, I have never made it as a teacher .. most of my life has been nearer the research and development side --- and I have been known to try to explain things to colleagues on 1-to-1 basis over the years.

Well, you're very good at it.  I too was more research and development (in AE) and was not very patient with peers, but was comfortable explaining to generals, administrators and other over-seers.  Changing careers to software development, where most developers are best served Pizza under the door and receive their work through wires, didn't help my social skills.  Now... I enjoy teaching the next generation.  When I can find one that wants to learn.  

VBR,

Inq

Hi @inq,

  Thanks for the multiple kind comments above... appreciated.

Re:

I'm guessing for government work, its close enough that I just use the C part.

Yes and No ... in many cases, you pick a capacitor according to its capacitance in microFarads, check its voltage rating is comfortably above what it will ever see, and hope, in the knowledge that this works more often than it fails.

However, there are certain 'positions' in a circuit which are more 'demanding'. For capacitors, a few circumstances that immediately come to mind are:

High frequency

 you may know the reactance of an inductor is given by

XL = 2 * Pi * f * L  ...

Given 2 * Pi is (very roughly) equal to 6, it is easy to see that for any given inductance is L, that that the reactance will is linearly proportional to f, the frequency.

So at low frequency, XL of (say) a short length of wire, may be near zero, and happily ignored. But as the frequency the circuit is dealing with increases, so does the reactance.

And whilst we might have an ESP8266/32 GPIO pin pulsing at only few thousand times a second, the same chip of silicon is supporting processor(s) running with (say) 240MHz clock and WiFi at 2.4 GHz ... , the latter about a million times higher frequency than our GPIO pin.

So if our capacitor is acting as decoupler for power to the chip, it has to do a good job for our GPIO pin, the microprocessor running the program and the WiFi functionality. For the higher frequencies, 1-2mm of wire (or PCB track) is a significant inductance, as the reactance can become so high, it is as if the capacitor wasn't there at all.

And the same appies within the capacitor ... if it contains a long length of aluminium foil, that can be an appreciable inductor. Hence, for high frequency work, ceramic (and other) capacitor types with minimal internal inductance are far more suitable than (say) aluminium electrolytic.

High current

As a (perfect) capacitor has an (perfect) insulator between the plates, no current can flow between the plates.

But the charge held on a capacitor is given by the formula

Q = C * V

where Q is charge in Coulombs, C capacitance in Farads, and V is the voltage across the plates in Volts

Hence when the voltage across the plates changes, then charge must flow in (increae in V) or out (decrease in V) of the capacitor.

The movement of charge is more commonly referred to as current! So, if the voltage across a capacitor is changing, there will be corresponding current flows, even though none of the current flows between the plates.

For low power circuits, these current flows are rarely an issue, but in certain positions, such as the input and output of a switch mode (e.g. buck) power supply, they can amount to many amps.

This means that the capacitor needs a small Equivalent Series Resistance, and for the switch modes that work at higher frequencies in particular, a small Equivalent Series Inductance.

Low Leakage

Electrolytic capacitors offer a substantial capacitance in a small volume. But the very thin insulation is not perfect. The resitance between the plates of even a reasonable quality new part may be only tens of MegOhms, so that when a voltage is applied to it, a current of say 1 microAmp (1 uA), could flow.

If this capacitor is a decoupling capacitor in a 1 Amp power supply, 'losing' 1 uA as leakage current is not going to be noticed.

However, if this capacitor was part of a sensitive instrumentation amplifier input circuit, that 1 uA leakage current could be larger than the current of the signal you are trying to measure.

In this case, an alternate type of capacitor, e.g. polyester or polystyrene, would be more suitable, as their leakage resistance can be tens or hundreds of GigaOhms.

And as another example ... several forum contributors have mentioned micropower projects, where a single coin cell is required to power a circuit for years. A leakage current of 1 uA through the power decoupling capacitor could represent a higher average drain than the microprocessor!

-------

I mentioned SPICE simulator ... 30 years ago, this was largely confined to research labs etc. Now several simulators based on the original SPICE code are available for free. Personally, I tend towards LTSpice, now on the Analog Devices website, but there are others including one associated with KiCad, and Tina on the TI website.

Simulators are not magic bullets, but when the circuit has an analogue part, they can often help to give an insight into what will happen ... and even if you put 1000 Amps through a tiny transistor, there is no magic smoke or need for a replacemet part!

They are fairly simple to get started with, but they do have learning curves to get good results, so some patience is required.

-------

Good enough for Government Work

Whilst just outside of my personal experience, I have been reminded that 'Good enough for Government Work' was apparently a phrase used when doing the financial accounting costs ... not to imply the Government was being swindled with outrageously over priced items, etc., but that the calculated cost was so close to the 'precise' cost that any attempt to improve the precision would actually cost more than the difference.

Obviously, some aerospace/military products have to be built to more exacting standards to meet the rigours of their intended use and/or to minimise the consequential cost of a part failing in service, but in general, the more complex model for a capacitor is based more on physics and electronics than the customer.

------

Clearly you have complementary skills to those I perceive for myself. Put simply, Dilbert has more patience and respect for pointy-haired boss than I have had with a few of my 'senior' (but technically challenged) colleagues over the years, and my (luckily) very few and extremely brief encounters with military personnel .... 

I think I have patience with people who are trying to do the right thing ... not with the others...

Luckily, I have also worked some excellent and talented people, so a there has also been a good bit of fun on the way.

Best wishes, Dave

@dronebot-workshop and @codecage -  Might I suggest that you create a section like some forum's have a Wiki, Glossary, etc?  It should be totally controlled/curated by you two.  You could cherry-pick from the best and ask said authors to edit/contribute to you directly for approval before posting.  Stuff like @davee last posts should go in a section that can easily be perused.  Searching over the whole forum for these posts deep in a topic barely related is problematic.  Especially if you don't know the keywords to search.  I'm thinking something with a Contents that we can drill down as deep as we need for the problem at hand... 

• Hardware
  • Resistors
  • Capacitors
    • Capacitance
    • Reactance
    • Resistance
    • Inductance
    • etc
  • Transistors
    • Mosfet - Something I'd really like to learn about that is fully working with 3.3V logic pins of my favorite processors ESP8266.
    • etc
  • etc
• Software
  • Arduino Sketch
  • C++
  • Python
  • etc
• 3D Printing
  • Materials
  • Cura Settings
  • etc

Posted by: @davee

So if our capacitor is acting as decoupler for power to the chip, it has to do a good job for our GPIO pin, the microprocessor running the program and the WiFi functionality. For the higher frequencies, 1-2mm of wire (or PCB track) is a significant inductance, as the reactance can become so high, it is as if the capacitor wasn't there at all.

I'm going to have to go over this several times and lookup strange words. Like I'm looking at Reactance now... but my reluctance is causing resistance. 😉 😋 

Posted by: @davee

Good enough for Government Work

Whilst just outside of my personal experience, I have been reminded that 'Good enough for Government Work' was apparently a phrase used when doing the financial accounting costs

Although intended to be a slur in most situations, I did not find it so in Aerospace design.  I designed carbon/epoxy structures for satellites.  Basically the chassis.  Security was compartmentalized... being, I wasn't allowed to know what connected to "hard points"... just their mechanical properties mass, inertia, etc, how many G's they had to take on launch, how much vibration, how much noise, temperature range, how much contraction/expansion was allowed over that temperature range, etc.  It was not unheard of contracts specifying that up to $250K per ounce be spent in research, analysis and testing to reduce chassis weight.  We used the best material, used grundles of CPU time and many hours of testing and verification to save mere ounces in the chassis so more electronics, optics, fuel could be put on the satellites.  Our margins of safety were very small.  

Posted by: @davee

Luckily, I have also worked some excellent and talented people, so a there has also been a good bit of fun on the way.

I have been very fortunate in my careers with superiors.  All in the AE field were experienced mostly educated to same level, but in different disciplines.  Even the military men sent to oversee us were no slouches.  In software... my early bosses were excellent... the latter ones... not so much.  

VBR,

Inq

@inq 

@dronebot-workshop and @codecage - Might I suggest that you create a section like some forum's have a Wiki, Glossary, etc? It should be totally controlled/curated by you two. You could cherry-pick from the best and ask said authors to edit/contribute to you directly for approval before posting.

They probably don't have time from their own projects for that? Bill has his tutorials on subjects that interest him and are useful to others. There are probably better sources online for teaching basic electronics and 3d printing. Forums are usually used to ask one off questions like "why are two different types of capacitors used" although if you google that question you will get answers.

@robotbuilder I stumbled across a great explainer site a few weeks ago. Didn't book mark it but it was sort of a random find, I was looking for something else. If I run into it again I will let @Inq know about it and I will check my history but it's huge.