Art of Electronics - Study support

[Wimberleytech]

Hi Wimberleytech,

I am not sure if i understand your response correctly. What i am trying to understand is why the voltage on the wire between inductor and resistor is zero.

It is because the resistance of the inductor is zero ohms.  Thus even with 1mA flowing through it (dc) there will be no voltage drop.

An ideal inductor only opposes current flow if the current is changing.

Does that help?

[Wimberleytech]

Wimberleytech,

Please disregard my last response. I understand what you mean now.

if di/dt = 0  then voltage across inductor is becoming zero.  Thank You.

oops I already answered...looks like you got it now!! Great!

[iMo]

FYI.

[alex_palvai]

Thanks Imo,. Your circuit is depicting clearly initial resistance and change in voltage for few nano sec's (innitial resistance of inductor to change in current ) .   

Thinking loud:
And i guess what else is interesting is voltages comes back to zero in a nano second and current stays there at 10 mA ( I am guessing that's a charge stored in magnetic field of L (incorrect  I realized its a current passing through inductor) ) .. And as input voltage drops from 10 to 0v , Inductor is starting to sense the change the current and starts resisting the change which creates voltage spike . ( Is it appropriate to assuming inductor is current sensing device as a general idea ? - I guess so that's appropriate , i answered my own question  )

Actually One more thing . I see the current across resistor is going negative -10 mA as source voltage is 10 V (I am still trying to find why this is happening ) .

Attaching my LT spice file.

[alex_palvai]

I am trying to build circuit 1.52 A  in LTSpice. Having trouble with switch. Is there a way i can configure this switch to switch between two circuits with 50% duty cycle?

[Wimberleytech]

I am trying to build circuit 1.52 A  in LTSpice. Having trouble with switch. Is there a way i can configure this switch to switch between two circuits with 50% duty cycle?

Yes, use two switches  driven 180° out of phase of each other.  What out for a glitch at each transition.  I hacked one together and it seems to work.

[Wimberleytech]

For some reason "quote" is not working for me...

"When current is reading 0 mA , why is voltage on node l1 reading (1.9950) volts instead of zero."

Because there is an "off" resistance of your switch.  So, the initial operating point is calculated with a resistance connected to the 2V source.

So then, why isn't the initial voltage 2V?  It is simply due to the simulator convergence accuracy.  When the iteration got to 1.9950 volts, the tolerances in the simulator were met so it stopped.

Change your off resistance to something lower, like 10Meg and see how the results change.

[alex_palvai]

I am trying to build circuit 1.52 A  in LTSpice. Having trouble with switch. Is there a way i can configure this switch to switch between two circuits with 50% duty cycle?

Yes, use two switches  driven 180° out of phase of each other.  What out for a glitch at each transition.  I hacked one together and it seems to work.

I successfully created two switches which switch between voltage source and ground at 50% duty cycle.
I am stuck on this circuit for a week clueless of whats going wrong .  As per the attached instructions in page "Inductor Voltage Diverer.PNG" Circuit Fig (1.52 A) must Vout have 1/2 input voltage and (Fig. 1.52B )must have Vout equal to 2 times Vin.

I am experimenting on first circuit now "1.52A" and expecting to get Vout be 1/2 Vin without a success.

Loosing the sense of volt-second balance rule, I started to go trial and error method for different combinations .  Hope someone helps me with a fix and help me understand how to bring the voltage and current into steady state .

Details of trials :
Ind :.1 u    Cap: .1u  -- Vout is square wave with voltage spikes at switch open and close peaks at 4V and -2V reduces back to 2 and 0.
Ind : 1      Cap: .1u  -- Vout rises endlessly 0 to 20V ++
Switched source voltage from 2v DC to pulse -  Ended up with very wierd result Vout (unexplainable shapes) in nano volts.
Disconnected second switch to ground with pulse voltage input  ---- Vout is sign wave between 2 and -2 volts .
Switch back from pulse input to DC input of 2 V - Vout is abnormal signwave peak to peak from ( 2.00006 v to 1.999993 v )

[Wimberleytech]

I am trying to build circuit 1.52 A  in LTSpice. Having trouble with switch. Is there a way i can configure this switch to switch between two circuits with 50% duty cycle?

Yes, use two switches  driven 180° out of phase of each other.  What out for a glitch at each transition.  I hacked one together and it seems to work.

I successfully created two switches which switch between voltage source and ground at 50% duty cycle.
I am stuck on this circuit for a week clueless of whats going wrong .  As per the attached instructions in page "Inductor Voltage Diverer.PNG" Circuit Fig (1.52 A) must Vout have 1/2 input voltage and (Fig. 1.52B )must have Vout equal to 2 times Vin.

I am experimenting on first circuit now "1.52A" and expecting to get Vout be 1/2 Vin without a success.

Loosing the sense of volt-second balance rule, I started to go trial and error method for different combinations .  Hope someone helps me with a fix and help me understand how to bring the voltage and current into steady state .

Details of trials :
Ind :.1 u    Cap: .1u  -- Vout is square wave with voltage spikes at switch open and close peaks at 4V and -2V reduces back to 2 and 0.
Ind : 1      Cap: .1u  -- Vout rises endlessly 0 to 20V ++
Switched source voltage from 2v DC to pulse -  Ended up with very wierd result Vout (unexplainable shapes) in nano volts.
Disconnected second switch to ground with pulse voltage input  ---- Vout is sign wave between 2 and -2 volts .
Switch back from pulse input to DC input of 2 V - Vout is abnormal signwave peak to peak from ( 2.00006 v to 1.999993 v )

You are clocking at a very low frequency compared to the filtering from the LC.  L and C need to increase in value.  To avoid ringing, you need to add some resistance to the inductor model as that would be real-world anyway (I put 10 ohms, but a more realistic value is closer to 1 ohm).  You will need to run more cycles of your pulse, and run the simulation out longer.  See attachments

[alex_palvai]

I am trying to build circuit 1.52 A  in LTSpice. Having trouble with switch. Is there a way i can configure this switch to switch between two circuits with 50% duty cycle?

Yes, use two switches  driven 180° out of phase of each other.  What out for a glitch at each transition.  I hacked one together and it seems to work.

I successfully created two switches which switch between voltage source and ground at 50% duty cycle.
I am stuck on this circuit for a week clueless of whats going wrong .  As per the attached instructions in page "Inductor Voltage Diverer.PNG" Circuit Fig (1.52 A) must Vout have 1/2 input voltage and (Fig. 1.52B )must have Vout equal to 2 times Vin.

I am experimenting on first circuit now "1.52A" and expecting to get Vout be 1/2 Vin without a success.

Loosing the sense of volt-second balance rule, I started to go trial and error method for different combinations .  Hope someone helps me with a fix and help me understand how to bring the voltage and current into steady state .

Details of trials :
Ind :.1 u    Cap: .1u  -- Vout is square wave with voltage spikes at switch open and close peaks at 4V and -2V reduces back to 2 and 0.
Ind : 1      Cap: .1u  -- Vout rises endlessly 0 to 20V ++
Switched source voltage from 2v DC to pulse -  Ended up with very wierd result Vout (unexplainable shapes) in nano volts.
Disconnected second switch to ground with pulse voltage input  ---- Vout is sign wave between 2 and -2 volts .
Switch back from pulse input to DC input of 2 V - Vout is abnormal signwave peak to peak from ( 2.00006 v to 1.999993 v )

You are clocking at a very low frequency compared to the filtering from the LC.  L and C need to increase in value.  To avoid ringing, you need to add some resistance to the inductor model as that would be real-world anyway (I put 10 ohms, but a more realistic value is closer to 1 ohm).  You will need to run more cycles of your pulse, and run the simulation out longer.  See attachments

Good Morning Wim,

Hope you are doing well.

Thanks a lot for showing me that this actually works. I am sure your years of expertise played nailing the issue very quickly. For someone who just started learning like me , what would be the key elements to consider and indicators to realize the right frequency and time for a specific configuration of capacitor and inductor combination .

[Wimberleytech]

Good Morning Wim,

Hope you are doing well.

Thanks a lot for showing me that this actually works. I am sure your years of expertise played nailing the issue very quickly. For someone who just started learning like me , what would be the key elements to consider and indicators to realize the right frequency and time for a specific configuration of capacitor and inductor combination .

Glad to help.  Think of the L and C as an LC filter that filters a square wave input.

[alex_palvai]

I am reading about Center-tapped full-wave rectifier explanation in AoE section 1.6.4 (B) - Attached references .

I find it really hard to construct/visualize what he is trying to explain in this section .

Here's what i struggle with ; typing this paragraph and the questions in parenthesis . kindly please someone  guide me through this .

Center-tapped full-wave rectifier :

The circuit in Fig 1.63 ( attached ) is called center-tapped full-wave rectifier . The output voltage is half what you get if you use bridge rectifier . It is not most efficient circuit in terms of transformer design, because each half of the secondary is used only half the time ( ME:clear to point ) .

To develop some intuition on this subtle point ,  consider two different configurations that produce same rectified DC output voltage ( ME:clear to point , he'z going to compare two circuits )
cont..
(a) circuit in Fig 1.63( attached) and
(b) the same transformer (( ME: which same transformer  ) , this time secondary cut at the center tap and rewired with two halves in parallel ( ME: i am thrown off already at this point ,  where did you cut and how did you join  ) cont.. the resultant combined secondary winding connected to full wave bridge ( ME: What is that resultant circuit , how does that look which is connected to full wave bridge  ) .

[Wimberleytech]

 (( ME: which same transformer  ) , this time secondary cut at the center tap and rewired with two halves in parallel ( ME: i am thrown off already at this point ,  where did you cut and how did you join  ) cont.. the resultant combined secondary winding connected to full wave bridge ( ME: What is that resultant circuit , how does that look which is connected to full wave bridge  ) .

You have reason to be confused.  I think this is a poorly-written section.  They should have drawn both configurations and given a little better explanation.  I will spend a little time and try to draw something up that makes a little more sense.  Stay tuned.

[alex_palvai]

For those like me "Starters struggling with shortcut's in Art of electronics" , What's helping me other two references I have

I also have Grob's Basic electronics by Mitchel E Shultz - 10th edition --- Provides a great break down explanation progressively.

My other reference is "Practical electronics for Inventors - —Paul Scherz and Simon Monk - 4th Edition .

Some topics are well written by one writer than others . 

Although Expensive , Grob's Basic electronics by Mitchel E Shultz is probably great reference (well written and easier to understand )

[alex_palvai]

To the OP: Life is short, so skip over analog and jump straight into digital and MCU's.  Digital and microcontrollers are both easier and more fun.  Buy specialized chips to do whatever you need done in analog rather than needing to build it from scratch.  Chips come with datasheets, and it's largely plug-and-play.  Then, as time permits, and as needed, selectively backfill your knowledge of analog to accomplish whatever you're needing that you can't achieve using only digital or specialized chips.

Also, if need be: with a rough schematic and a SPICE simulator, for simple circuits you can often what-if your way to an accurate analog solution.

This way you learn only what you need and you won't get bored/stuck/discouraged on the parts of analog that you may never use anyway.

I'm not dissing the more methodical school taught EE approach, but if you don't have time for that, these are some short-cuts where you can probably solve whatever problems you need solved and get workable solutions sooner rather than later!  If you're doing hobbyist stuff, you don't need to sweat for days to arrive at the most minimal, cheapest way to do things, as you would if you were making a product that would be sold in the thousands or millions.

Thanks for feedback, it motivates me more towards analog whenever someone says that, reasons below:
I played with(arduino , rasberry pi) for years before i jumped into Analog :-) (they are only good for starter's  ,  i cannot imagine to create a product of my own and do a mass production with Adriuno / rasberry pi . ) , they are simply not ideal.  My perspective, plug and plays are not challenging & lack in power and limited- not customization(don't make me contended or give me any sense of "Yes i know electronics" ) . 

[alex_palvai]

Wanted to share with you all , i ordered  my first oscilloscope(Rigol DS1054Z 4-channel 50 Mhz) and function generator(Koolertron Upgraded DDS Signal Generator Counter) and excited to explore those two .

back to question:
I tried quiet sometime trying to decode the values "3/8 and 8/3" in Irms solving section of attached jpg. Can someone help me understand how he got those two values based on figure 3N.30( attached ) .

I just learnt few concepts yesterday about phase angle, phasor drawing and tried to apply what i learnt to decode but still couldn't .

On a high level i understand what instructor is trying to do  ( calculate rms current ) based on heating in order to pick the right capacity of transformer for unregulated power supply.

[Wimberleytech]

back to question:
I tried quiet sometime trying to decode the values "3/8 and 8/3" in Irms solving section of attached jpg. Can someone help me understand how he got those two values based on figure 3N.30( attached ) .

I believe the 3/8 is derived empirically from the timing diagram.  Nothing magic about that number.  The 8/3 number comes from 3/8 * 8/3 = 1.
The duty cycle for the current delivered during recharge is factored by 3/8 so whatever current from the transformer is multiplied by 3/8.  In this case a full 1A must be delivered to the load, so more current must be supplied...thus 8/3 * 3/8.

Does that help?

[alex_palvai]

back to question:
I tried quiet sometime trying to decode the values "3/8 and 8/3" in Irms solving section of attached jpg. Can someone help me understand how he got those two values based on figure 3N.30( attached ) .

I believe the 3/8 is derived empirically from the timing diagram.  Nothing magic about that number.  The 8/3 number comes from 3/8 * 8/3 = 1.
The duty cycle for the current delivered during recharge is factored by 3/8 so whatever current from the transformer is multiplied by 3/8.  In this case a full 1A must be delivered to the load, so more current must be supplied...thus 8/3 * 3/8.

Does that help?

Thank's a  ton for giving me the logic Wim. Your the best teacher.  I read through that paragraph at least 20 times and it didn't strike me . It makes sense, and i could relate that explanation to other worked example(3W) in AoE Lab course. That fraction is actually a percentage of current against total cycle .  I think where he is getting at with all this is estimate your best and take a transformer with a current rating 2 or 3 Amp's bigger than your recorded estimate.

I wonder if this is AoE specific method for calculating transformer current rating or industry standard practice . I am asking this because, i couldn't find this formula or type of explanation in any other books i was referring or anywhere on internet.

[Wimberleytech]

You might get some additional insight by downloading the National Semiconductor Power Supply Handbook
https://ia800703.us.archive.org/0/items/NationalSemiconductorVoltageRegulatorHandbook1980/NationalSemiconductorVoltageRegulatorHandbook1980.pdf

[Simon_RL]

There is no silver bullet, magic pill or the like.  You need the basics to advance.  If you just want to touch the surface as a hobbyist, there are many cookbook style texts available.  Math is essential, and there are a multitude of texts available with applications of calculus and trigonometry to electronics.  Get the specialized books geared to electronics.  My feeling is that if you master the math, the rest is a knock.  JMO.

Totally agree @GerryR, as a newbie I had all these great vision of learning electronics and making students, man was I delusional . I very quickly learned that I needed the maths skills as well. At 53 I soon realised that I don’t remember much math from uni and have spent more time brushing up on algebra, trig and calculus over the last 4 months than actual electronics.

[alex_palvai]

There is no silver bullet, magic pill or the like.  You need the basics to advance.  If you just want to touch the surface as a hobbyist, there are many cookbook style texts available.  Math is essential, and there are a multitude of texts available with applications of calculus and trigonometry to electronics.  Get the specialized books geared to electronics.  My feeling is that if you master the math, the rest is a knock.  JMO.

Totally agree @GerryR, as a newbie I had all these great vision of learning electronics and making students, man was I delusional . I very quickly learned that I needed the maths skills as well. At 53 I soon realised that I don’t remember much math from uni and have spent more time brushing up on algebra, trig and calculus over the last 4 months than actual electronics.

Thanks for sharing that story Simon_RL, Part of your story matches mine, I've taken GerryR's comments seriously and learnt math .In order to remove fear of Math , i did a course of "Algebra", "PreCalc", "CalC1" and "Calc2" in thattutorguy.com. It took me a month+ to recap ,  but i thankfully i am much comfortable now seeing differential equations , integration and other things than I was before .

[alex_palvai]

I am testing signal rectification circuit Fig 1.70 Chap 1 , attached.

Per what i learnt so far , Diode must block a reverse voltage & current .  This is conflicting with what i am seeing in Vout ( has negative spikes) unfiltered by diode. I am wondering if this issue is specific to LtSpice.

[rstofer]

I have no idea what is causing the negative spike but I will say that there is a reverse voltage characteristic to the diode.

You have chosen very fast edges - dV/dt approaches infinity.  Therefore, it seems reasonable to assume that you can create a reverse voltage far in excess of the diode's rating.

See what happens if you slow down the edges.  Or add some series resistance to the voltage source.  Try a 1us edge.

Expand the scale so you can see what is happening at both spikes.

A lot of people hammer away on LTspice and there doesn't seem to be a crowd yelling about the diode models.  There is also an LTspice group

https://groups.io/g/LTspice/topics

There used to be very active on Groups.Yahoo.Com but that went away.  Bummer...

[rstofer]

See about 1/2 way down here:

https://www.electronics-tutorials.ws/diode/diode_3.html

The idea that a diode blocks reverse current flow is wrong for low levels of current.

[alex_palvai]

I have no idea what is causing the negative spike but I will say that there is a reverse voltage characteristic to the diode.

You have chosen very fast edges - dV/dt approaches infinity.  Therefore, it seems reasonable to assume that you can create a reverse voltage far in excess of the diode's rating.

See what happens if you slow down the edges.  Or add some series resistance to the voltage source.  Try a 1us edge.

Expand the scale so you can see what is happening at both spikes.

A lot of people hammer away on LTspice and there doesn't seem to be a crowd yelling about the diode models.  There is also an LTspice group

https://groups.io/g/LTspice/topics

There used to be very active on Groups.Yahoo.Com but that went away.  Bummer...

Hi Rstofer,

Good to see your response again. I tried your suggestions 1. added source resistance 2. slowed down edges (set trise and tfall to 1n) ... And tried changing diodes to different model , also tried increasing capacitor size to increase currents . It remained same .

I received my new oscilloscope but not my function generator yet, assuming that's required to generate signals and test this practically . I strongly feel this particular module in LTspice is buggy.

I read the suggestions , Its mentioned a small leakage in (uA) is possible through the junction , In my case the leakage is few 12 mA (not small )

From "https://www.electronics-tutorials.ws/diode/diode_3.html" (thanks for sharing) :

"This condition represents a high resistance value to the PN junction and practically zero current flows through the junction diode with an increase in bias voltage. However, a very small leakage current does flow through the junction which can be measured in micro-amperes, ( μA ).