RL circuit practice

[Tony Stewart]

jjw although , can you confirm your L value.
https://goo.gl/wYguDC
I calculate LC of 10uH 10nF to match my sim output of 712kHz

and reduced linear amp gain to 0.9 to prevent saturation near ground.

 

[Tony Stewart]

I simulated this in LTSpice, L=10uF C1=C2=10nF using 74HCU04 as inverter, frequency ~ 225kHz and the second harmonic measured from input pin was about 0.3% and from output pin about 3%

I see your problem , no big deal, you said 10uF but meant 100uH which yields 225kHz

 

[jjw]

I see your problem , no big deal, you said 10uF but meant 100uH which yields 225kHz

Yes, 100uH, my typo

 

[Ratchit]

You incorrectly identified the classic design as an electronic multi-vibrator

I said it looked like a multi-vibrator. I never insisted it absolutely was a multi-vibrator. I later corrected myself and called it a switching oscillator which both oscillators are. But that begs the question. How can you say the caps are in series when I showed that they are energized in parallel by the gate, and only connected to one point (ground)? Why are not the caps separated by a inductor in a power supply considered in series also.

The sine wave is very clean but the Simulation time sampling gives quantization effects which can be reduced by reducing the sample time.

in Options > other options but then unlike digital scope this has less memory. For better slew rate , I edited the gate to be faster below.
View attachment 96490
I changed one cap to pullup and one to ground to speed up the startup DC bias.. Same f results as Vcc to ground ~0 Ohms impedance

Good to know it has a nice sine wave.

Ratch

 

[Tony Stewart]

Dear Pedantic;
The correct analysis for the Kirchoff KCL equations for loop current are that the two caps are in series which together form a parallel LC resonant circuit. ( high impedance)
This also correctly predicts the frequency exactly as shown in both the Falstad and Vspice sim.

This has nothing to do with power supply decoupling caps.
Also a "switching oscillator" is a self-invented term, not universal as I described.

I cannot explain further now. HNY.
Do not doubt me, but try to understand.

 

[Ratchit]

Dear Pedantic;
The correct analysis for the Kirchoff KCL equations for loop current are that the two caps are in series which together form a parallel LC resonant circuit. ( high impedance)

They can also be analyzed by the alternate and equally valid node method, where the topology indicates the caps are in parallel. It is all a matter of perspective.

This also correctly predicts the frequency exactly as shown in both the Falstad and Vspice sim.

As it should if the correct parameters are submitted. But since you don't know how F or V calculate the circuit, that factoid does not support your claim.

This has nothing to do with power supply decoupling caps.

Whoever said it did? Certainly not I.

Also a "switching oscillator" is a self-invented term, not universal as I described.

If it describes the operation correctly, then go with it.

I cannot explain further now. HNY.
Do not doubt me, but try to understand.

HYN to you also. I do understand what you are saying. I just don't agree with everything you aver, and your assumption that your way is the only way. In other words, I believe there are alternate ways of looking at and explaining things. Doubt is an essential ingredient in discourse.

Ratch

 

[Tony Stewart]

If you wish to continue this thread in private with someone else , in private conversation, please do so. I have no interest.

 

[Tony Stewart]

This LC parallel resonant oscillator uses the PI (shaped) filter (as you know) with an equiv value of two series caps with Ceq= 1/2C when both C's are equal. The phase is inverted 180deg twice ( once by 2nd order filter and once by inverting amplifier (CMOS logic)) making a positive feedback loop also with excess gain to square up sine feedback input.

All Logic IC's have analog properties in this discussion, so we do NOT refer to them as "switches" as you have done, which might be relevant for a more logical discussion on logic.

Your mention of this as a Switched Oscillator is illogical. It reminds me of an old TV guy who once had 500m$ in public stock on his company Del Secur based on his "analogic" understanding of electronics and 2D fingerprints. When I reviewed his design and theory, I found it fundamentally flawed. Within a year his company stock was worthless as stock traders discovered the same thing that his pretense was incorrect and he had no un-hackable security solution. there is no such thing as "analogic" as he used it.

Why argue against fundamental oscillator theory when you know you are wrong by deflecting the issue with irrelevant info . this is fallacious.
e.g. in 106 you said
[QUOTE="Ratchit ] I don't think it is either a series or parallel resonant circuit. It looks more like an electronic multi-vibrator to me, which is not a linear oscillator.
In some equipment, like power supplies, there are many capacitors connected to ground. You would not say those capacitors are in series, would you? [/QUOTE]

You should know better to be contrary to any text book and wiki page. ( and never back your math with proof or any examples in this case as I have done countless times also confirmed by others.)

Your repeated contradictions e.g. of my true statements without your proof only reflects poorly on you.

If you wish to continue this thread in private with someone else ,
I have no interest.

 

[Ratchit]

This LC parallel resonant oscillator uses the PI (shaped) filter (as you know) with an equiv value of two series caps with Ceq= 1/2C when both C's are equal. The phase is inverted 180deg twice ( once by 2nd order filter and once by inverting amplifier (CMOS logic)) making a positive feedback loop also with excess gain to square up sine feedback input..

Describing how the circuit works does not prove that the capacitors are in series. In a different topology, using different equations, the capacitors can be thought of a being in parallel. For instance, in a "T" circuit of cap----inductor----cap of a power supply, the caps are thought of as being in parallel even though they are connected together at ground.

All Logic IC's have analog properties in this discussion, so we do NOT refer to them as "switches" as you have done, which might be relevant for a more logical discussion on logic.

Not physical switches, no. But the buildup of voltage of the opposite polarity at the input of the inverter causes the inverter to switch and introduces nonlinearity into the circuit.

Your mention of this as a Switched Oscillator is illogical.

It is descriptive.

It reminds me of an old TV guy who once had 500m$ in public stock on his company Del Secur based on his "analogic" understanding of electronics and 2D fingerprints. When I reviewed his design and theory, I found it fundamentally flawed. Within a year his company stock was worthless as stock traders discovered the same thing that his pretense was incorrect and he had no un-hackable security solution. there is no such thing as "analogic" as he used it.

That is arguing by analogy, which is a fallacious tactic. It does not prove I am wrong.

**broken link removed**

Why argue against fundamental oscillator theory when you know your are wrong by deflecting the issue with irrelevant info . this is fallacious.

Who is doing that? Certainly not I.

e.g. in 106 you said
/QUOTE] I don't think it is either a series or parallel resonant circuit. It looks more like an electronic multi-vibrator to me, which is not a linear oscillator.
In some equipment, like power supplies, there are many capacitors connected to ground. You would not say those capacitors are in series, would you?

Which I corrected in posts #112 and iterated that correction in post#116. Did you forget about that?

You should know better to be contrary to any text book and wiki page.

I am not in disagreement with those materials with respect to how the circuit works. In fact, you have not cited any references to that particular circuit. The question is whether the caps are in series or parallel. Expounding on how the circuit works does not resolve the issue.

( and never back your math with proof or any examples in this case as I have done countless times also confirmed by others.)

No you haven't. You have not submitted any math showing how the circuit works, only a descriptive. Others have only showed that the circuit does work. That was never in doubt. What math you did submit pertained to measuring L by the OP. If I have time and inclination, maybe I will do some math.

Your repeated contradictions e.g. of my true statements without your proof only reflects poorly on you.

Not so. No one should swallow anything that comes along.

If you wish to continue this thread in private with someone else , in private conversation, please do so. I have no interest.

To each his own.

Ratch

 

[Tony Stewart]

Show your calculated or tested results with backup simulation not rhetoric or prove #128 is wrong frequency or L or C values. then you have substance, as I do, not baseless contradictions.

 

[Ratchit]

Show your calculated or tested results with backup simulation not rhetoric

Yes, I will if I do it.

Ratch

 

[Tony Stewart]

For now it is just contradictory rhetoric if you dont

 

[Ratchit]

Show your calculated or tested results with backup simulation not rhetoric or prove #128 is wrong frequency or L or C values. then you have substance, as I do, not baseless contradictions.

First of all, I never said that your simulation gave the wrong frequency. You keep saying things that I never did. I challenge you to find where I said the frequency was wrong. I did say those caps were not in series like you insisted they were. I did do what you never did, calculate the formula for resonance of this circuit. You never showed any math for this circuit even though you said you did. You can see the results below. The calculated frequency from the formula is quite close to the simulation.

Ratch

 

[Tony Stewart]

When you suggest 2 parallel caps rather than the correct perspective of 2 in series you imply results must be wrong.
Your calculations include choke DCR term C2R2 where R2 =15, which can obviously be neglected from the beginning using R1=10k in this LC high Q impedance.

your results reduce to

which is identical to mine ( ah the one with the series cap formula ) which you denied several times

Actually your results now prove I was correct all along. Thank you.

 

[Ratchit]

When you suggest 2 parallel caps rather than the correct perspective of 2 in series you imply results must be wrong.

Not the results, the wrong perspective. If the caps were in series, then you could combine them into one cap. You can't, so they are not.

Your calculations include choke DCR term C2R2 where R2 =15, which can obviously be neglected from the beginning using R1=10k in this LC high Q impedance.

The calculated frequencies for choke resistance values of 0,15,73 ohms are 225079, 225163, and 225518.

your results reduce to

which is identical to mine ( ah the one with the series cap formula ) which you denied several times.
Actually your results now prove I was correct all along. Thank you.

Actually my results do nothing of the kind. Where would you put the combined series cap Cs? Those caps are distributed in parallel and cannot be thought of as being in series.

Ratch

 

[jjw]

Ratchit. If R1 ( I assume it is 10k ) is zero, the circuit still works in LTSpice, but your formula gives infinite frequency.

 

[Tony Stewart]

Hopeless,

You have the wrong perspective in every statement in previous comment.`

Parallel Caps have an equivalent value of C = C1+C2

Series caps have an equivalent value where you agreed with my formula and results.

nitpicking wont help your case with insignificant figures in the 4th digit.

I'm sure others would agree.

 

[Tony Stewart]

Ratchit. If R1 ( I assume it is 10k ) is zero, the circuit still works in LTSpice, but your formula gives infinite frequency.

A choke with DCR=0 also implies infinite current into a Cap with ESR=0 with a driver with ESR=RdsOn=0 which is impossible.

Since this design is fairly intolerance to R2 for low DCR values but in order to maintain a high Q, R1 must also be high then the DCR is easily neglected and R1 has insignificant effect as the R1 terms cancel out. as I indicated .


It is the ESR of the driver that makes your Simulator and actual hardware work if R1=0, but this is not recommended design.

 

[Ratchit]

Ratchit. If R1 ( I assume it is 10k ) is zero, the circuit still works in LTSpice, but your formula gives infinite frequency.

No problem, the formula is correct. The simulator probably includes the output resistance of the inverter, whereas I do not. Therefore, even if you set R1=0 in the simulator, the output resistance is still there. Find the resonant frequency with R1 = 0, and you can calculate what value the simulator uses for the output resistance of the inverter. If I set R1=2 in the formula, I get 490548 hz for the frequency.

Ratch

 

[Ratchit]

Hopeless,

You have the wrong perspective in every statement in previous comment.`

Parallel Caps have an equivalent value of C = C1+C2

Series caps have an equivalent value where you agreed with my formula and results.

As I said last time, you cannot implement a single combined Cs. Therefore, the circuit does not contain any caps in series.

nitpicking wont help your case with insignificant figures in the 4th digit.

I did not give the three frequencies related to the coil resistance to prove anything. I just wanted to show the sensitivity of the frequency to the coil resistance.

I'm sure others would agree.

I don't think so.

Another thing, you said in post #128 that the filter network changes the phase by 180°. I believe that is wrong, and I should have caught it earlier. A passive network can only change the phase a maximum of plus or minus 90°. Since the circuit is at resonance, there is no orthogonal component to the input of the inverter. In other words, the phase change is zero from the output of the inverter back to its input during resonance. In fact, I theorize that if you connected a straight wire from the output of a inverter to its input, it would oscillate with its frequency depending on the slew rate of the inverter and other internal parameters.

Ratch