Weird coil configuration, does it classify as parallel LC circuit?

[wasedadoc]

The two ends of each coil is connected to the other two end of the other coil. That is, both coils are shorted together.

The signal generator input is at the top of the shorted coil configuration and the bottom of the two coils is the oscilloscope probe. Both signal generator GNDs and oscilloscope GNDs are connected together and not touching the shorted coils and then also connected to Earth GND.

Also, the oscilloscope is 10X attenuation so the output voltage is 10 x 26.4 Volts.

Like pulling teeth!

Finally you have revealed that you have a series LC circuit.  The series C is the input capacitance of the scope probe.  Yes, in that case you can expect the probe to experience a higher voltage than the generator is producing.  But maybe not 264 Volts.  Almost all scopes have a menu to select 'x1', 'x10' and maybe others. If that is set to 'x10' the scope has already done the calculation and the displayed 26.4 is correct and should not be multipled by 10 again by you.

[Dejan567]

I will have to correct the point about 10x and my oscilloscope. As a sanity check,  when I connect my 10 Vpp input signal to my probe I see 10 Vpp at 1x. When I switch to 10x, the Vpp displayed drops down to ~ 1V. So I still need to account for this in my measurements manually. Perhaps there is something in my oscilloscope that can be adjusted in its settings to account for this itself, but the measurement I showed did not do that so there is a 20x voltage multiplication.

Also, from Chat - GPT:

"However, it's important to note that this voltage multiplication occurs across the individual components (inductor and capacitor) and not across the entire circuit. The total voltage across the series LC circuit will still be equal to the applied voltage."

Can you clarify as I am measuring across the entire series LC circuit?

[wasedadoc]

I will have to correct the point about 10x and my oscilloscope. As a sanity check,  when I connect my 10 Vpp input signal to my probe I see 10 Vpp at 1x. When I switch to 10x, the Vpp displayed drops down to ~ 1V. So I still need to account for this in my measurements manually. Perhaps there is something in my oscilloscope that can be adjusted in its settings to account for this itself, but the measurement I showed did not do that so there is a 20x voltage multiplication.

Also, from Chat - GPT:

"However, it's important to note that this voltage multiplication occurs across the individual components (inductor and capacitor) and not across the entire circuit. The total voltage across the series LC circuit will still be equal to the applied voltage."

Can you clarify as I am measuring across the entire series LC circuit?

1.  The x10 probe attenuates the signal. The 10V p-p at the probe tip becomes 1v p-p at the scope's BNC socket.  Look in the scope's manual and I expect you will see there is some setting to inform the scope whether you have switched the probe to x1 or x10.

2.  You are not measuring across the entire series LC circuit.  The generator is across the entire series LC circuit.  The scope is not. Your measuring apparatus is in series with the inductance.  The input capacitance of the probe and scope is the series capacitance.  It is the voltage across that only capacitance, not across the inductance and capacitance that the scope is displaying.

[Dejan567]

Yes I am aware what the x10 probe is doing.

Thanks for clarifying the input capacitance of the oscilloscope probe, that is helpful.

What would be the formula to calculate V_out based on the series LC configuration and a given V_in? Assuming of course there is non-zero R.

[fourfathom]

What would be the formula to calculate V_out based on the series LC configuration and a given V_in? Assuming of course there is non-zero R.

This is apparently not a series LC, but a series L into a shunt C, which happens to be an impedance-matching circuit.  With the proper component values and frequency, the voltage step-up can be as much as SQRT(Rsource / Rload).  With a 50 Ohm source, and a 10 Meg load (the oscilloscope input R) this will be 447 x the input voltage.  With real-world inductor loss the step-up will be significantly less.

In this case we can probably ignore the inductor self-resonant parallel capacitance.  The typical 'scope input capacitance is about 15 pF when using a 10x probe.  I don't know what your inductance is.

I can't go beyond this without pure guesswork.  What are you really trying to do, and what are you attempting to measure or accomplish?  Why two parallel inductors?  What frequency range?

And your first diagram showed a parallel resonant circuit, not what you later described.  If this stuff is new to you that's nothing to be ashamed of, but without more information it's going to be hard to give you useful advice.

[Dejan567]

Well everything you measure voltage across with an oscilloscope is considered a shunt, isn't it?

This is not new to me, was just trying to understand how I was getting voltage multiplication with this setup. I don't think we can ignore the self-resonance of the coils since there is a definite resonance point. If we were to ignore that then we wouldn't expect a voltage magnification spike for a range of input frequencies.

I'd rather not get into specifics of the application at the moment.

[wasedadoc]

With the circuit topology you actually have, the self resonance of the coil with its own parallel capacitance makes it higher impedance which reduces the voltage displayed on the scope.

[Dejan567]

So seems like there are opposing forces? Some factors are reducing the output voltage others are multiplying it?

Just trying to become less confused by estimating what the output voltage would be for a parallel LC circuit followed by a shunt C. It seems like there are alot of conflicting factors that could lead to a range of effects depending on exact values. I don't have the L or C values of the coils/scope but the resonance point is around 700 kHz.

[Sensorcat]

1. The point internal to the generator prior to its 50 Ohm output is not accessible.  For the physical connections the composite coil has two points, the generator has two and the scope has two.

2.  The schematic cannot match what the OP claims his words and photo show.  The generator says 10 Volts peak to peak.   Even if that is referring to a 50 Ohm load it means 20 Volts peak to peak inside the generator.  Yet the scope is saying the probe is experiencing 26.4 Volts peak to peak.

• That's what I explained above. However, for the behaviour of the circuit the accessibility is unimportant. There is a voltage drop across the internal resistor depending on the load, so it does not matter it you prefer to see only two nodes, it's three. And because the node at which the oscilloscope probe connects and the signal source are not the same, your claim that the voltages must be the same is wrong. It is staring to get funny, because in Reply #18 you insisted that the two voltages must be the same, and now you argue against me with the measurement of the OP that they are different.
• It was not my intent to show something that matches the - still obscure - configuration the OP has, it was my intent to demonstrate that your claim from Reply #18 is wrong. Do you know that most generators have a setting for 50 ohm and Hi-Z output that does not change the output circuit, but only the amplitude display?

[Sensorcat]

So seems like there are opposing forces? Some factors are reducing the output voltage others are multiplying it?

Just trying to become less confused by estimating what the output voltage would be for a parallel LC circuit followed by a shunt C. It seems like there are alot of conflicting factors that could lead to a range of effects depending on exact values. I don't have the L or C values of the coils/scope but the resonance point is around 700 kHz.

It is still unclear what you actually have, partly because you did not show everything and partly because you do not use standard nomenclature. For instance, the two coils are not shorted, they are in parallel. May sound like hair-splitting, but there's a potential problem when you are creative with your own words and we don't realize it. No offence!

May I suggest that you disconnect the coils and measure the unloaded output of the generator with your ocilloscope in an otherwise unchanged configuration? This will make some points clear, for instance how the instrument settings affect your result. There is this output impedance setting I mentioned above, and I wonder if your scope acconts for the 10x probe or not. Do you perhaps have 10/1x switchable probes?

[Dejan567]

Yes probes are switchable 1x - 10x. That means with my current arrangement as I mentioned before the 10 Vpp with the probe shows ~ 1V.

Two coils being shorted together at both ends means they are in parallel connection so not sure what the confusion or difference is there.

[PCB.Wiz]

So seems like there are opposing forces? Some factors are reducing the output voltage others are multiplying it?

Just trying to become less confused by estimating what the output voltage would be for a parallel LC circuit followed by a shunt C. It seems like there are alot of conflicting factors that could lead to a range of effects depending on exact values. I don't have the L or C values of the coils/scope but the resonance point is around 700 kHz.

Of course there are, which is why in a circuit with this many unknowns, you are best to do exactly what you have done :

ie Measure it, ideally under conditions as close to your desired operating, as possible.

A signal generator (+ series RS) is ok for quickly finding any peak, but you should use your intended driving circuit when chasing details like operating Q.
 

[Sensorcat]

Yes probes are switchable 1x - 10x. That means with my current arrangement as I mentioned before the 10 Vpp with the probe shows ~ 1V.

Did you actually test that? Since all this is really looking weird, it's time to question everything. Checking the measurement without the coil is one way to do this. So you have, with the generator set to 10Vpp, and probes set to 10x, the scope set at 1x probe attenuation, 1Vpp measurement with coil disconnected and 26.4Vpp with coil connected, nothing else changed?

Next test would be to work without the earth GND connection, as it is not needed for this measurement, but sometimes causes problems.

Also, you could measure and plot the peak-peak voltage over frequency.

Two coils being shorted together at both ends means they are in parallel connection so not sure what the confusion or difference is there.

The point is that usually no one in EE would call the parallel coils shorted, Knowing that you do, means that there could be misunderstanding lurking somewhere in this thread, unnoticed.

[Dejan567]

Did you actually test that? Since all this is really looking weird, it's time to question everything. Checking the measurement without the coil is one way to do this. So you have, with the generator set to 10Vpp, and probes set to 10x, the scope set at 1x probe attenuation, 1Vpp measurement with coil disconnected and 26.4Vpp with coil connected, nothing else changed?

Yes, I tested it exactly as you stated and have mentioned this at least twice.

Next test would be to work without the earth GND connection, as it is not needed for this measurement, but sometimes causes problems.

I tested with and without earth GND and it is just more stable with earth GND connection.

The point is that usually no one in EE would call the parallel coils shorted, Knowing that you do, means that there could be misunderstanding lurking somewhere in this thread, unnoticed.

I have a background in EE, with my usage of phrase was trying to emphasize the connections with the signal generator. Perhaps better use of words could have been given, sure. Again, no difference in what I said and the alternative.

[fourfathom]

Well everything you measure voltage across with an oscilloscope is considered a shunt, isn't it?

This is not new to me, was just trying to understand how I was getting voltage multiplication with this setup. I don't think we can ignore the self-resonance of the coils since there is a definite resonance point. If we were to ignore that then we wouldn't expect a voltage magnification spike for a range of input frequencies.

I'd rather not get into specifics of the application at the moment.

With the series L and shunt C you would absolutely expect a "voltage magnification spike", that's the whole point of the LC impedance-matching network.  With your test circuit, as I understand it, the self-resonance will result in a voltage dip (null) at the SRF (Self Resonant Frequency).  There are ways to characterize inductor L, R, and SRF -- this has been discussed on this forum, and it's not very difficult to get fairly accurate numbers.

If you can't tell us about your application, can you at least give us a proper description of your test circuit?

FWIW, here's a simulation of an LC network (ideal) that matches a 50 Ohm source to a 10 Meg / 15 pF scope probe.  The generator delivers 1V to a 50 Ohm load, and the matching network boosts the voltage to over 400V at the load.  I assume that your circuit won't have that inductor value, but the principle should be obvious.
 

[wasedadoc]

1. The point internal to the generator prior to its 50 Ohm output is not accessible.  For the physical connections the composite coil has two points, the generator has two and the scope has two.

2.  The schematic cannot match what the OP claims his words and photo show.  The generator says 10 Volts peak to peak.   Even if that is referring to a 50 Ohm load it means 20 Volts peak to peak inside the generator.  Yet the scope is saying the probe is experiencing 26.4 Volts peak to peak.

• That's what I explained above. However, for the behaviour of the circuit the accessibility is unimportant. There is a voltage drop across the internal resistor depending on the load, so it does not matter it you prefer to see only two nodes, it's three. And because the node at which the oscilloscope probe connects and the signal source are not the same, your claim that the voltages must be the same is wrong. It is staring to get funny, because in Reply #18 you insisted that the two voltages must be the same, and now you argue against me with the measurement of the OP that they are different.
• It was not my intent to show something that matches the - still obscure - configuration the OP has, it was my intent to demonstrate that your claim from Reply #18 is wrong. Do you know that most generators have a setting for 50 ohm and Hi-Z output that does not change the output circuit, but only the amplitude display?

What I wrote applied to the information available to me and everyone else except the OP at the time. Namely that the sig gen was across the coil. That was stated in words and shown in the diagram in Reply #4. (There is only one way to connect the scope and retain the sig gen directly across the coil.) Now that the OP has revealed that was untrue, the configuration is very different and any previous post by myself or anyone else may be inaccurate. Your own Reply #21 models the same incorrect configuration of both sig gen output and measurement being across the coil.

I am well aware that many generators have a fixed output impedance (typically 50 Ohms) and the different settings really mean the external load. My own sig gen does that. However the non-zero output impedance of the generator was irrelevant at that stage of the discussion. Whether comparing to the voltage on either side of that impedance, the voltage on a scope in the configuration that the OP led everyone to believe was in use could never exhibit multiplication. (As was the intent of your Reply #21).

[Njk]

There is something magic in these inductors and resonances.

In 90's I was working for a small company and once upon a time a chess master came to our office. He used to travel around the country with the purpose to perform simultaneous game show. He was thinking about how to improve the show. So he asked if we could design a hi-tech chess board with a computer interface. After some thinking, I came to the conclusion that the only reliable way would be to use the RFID method. Then I did some research and found a nice RFID tag from Temic, in the form of tiny glass tube. That's easy to plant that tag in a chessman. But I was not sure about actual availability, lead time and cost for that part and finally decided to design everything from scratch as that likely will be faster. And I succeeded. A working prototype was built in two months (see the attached images).

Two surprises were encountered in the process. In RFID, the energy is wirelessly transferred between two resonant tanks. That's obvious, the Q-factor in the transmitter tank depends on energy consumption by the receiver tank. The more energy is consumed, the less the Q factor is (and therefore, the less is the voltage/current amplitude across the tx resonant circuit). But it not always so simple. In reality, when the rx tank is almost short circuited (max. current consumption), it effectively becomes aperiodic and disturbs the tx tank less. That results in increased AC voltage across the capacitor of the tx tank, not decreased.
Soon after I'd found a better job and left the company. Much later I learned that the guy who was assigned to that project had a problem to reproduce my prototype. I used a high frequency space-grade capacitors (from my stockpile), while he tried to use cheap random caps, arguing that the design is actually not of HF (100 kHz). But that's actually not a surprise that the components quality matters.

To the OP: The person who's most capable to solve your project problems is you. Everyone else is less aware of your particularities and therefore is much less capable. If you can't sort this out then nobody can. It's as simple as that.