PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
2022-01-18, 03:50:48
News: Check out the Benches; a place for people to moderate their own thread and document their builds and data.
If you would like your own Bench, please PM an Admin.
Most Benches are visible only to members.

Pages: 1 ... 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 [21] 22 23 24 25 26 27 28 29 30 31 32 33
Author Topic: Itsu's workbench / placeholder.  (Read 47753 times)
Group: Tech Wizard
Hero Member
*****

Posts: 948
Gyula ,

about :

Note that the input current is very close to zero mA during the full 20 us time duration hence the currents and power levels shown by the curves come mainly from the 8V initial voltage established in C4 by the software.

Respectfully, I will have to disagree with that statement.Let's consider that C4 in its initial state has no voltage present.
Instead, we will replace that state in C4 by zero volts and use a pushbutton, just like on the original circuit.
From this moment on, we would stop considering any external input.
We will consider a pulse of only 1ms, as if the push-butt is pressed.
From that moment on, any input will always be provided by V1 our source in circuit. Do we agree?

Being the time chosen to activate the push button, in 1ms, simulating the transient test, we should not check any activity before that time, only after this same 1ms can we hypothetically begin the analysis.
I will leave all samples for LTspice analysis,
so you can comment on why I disagree with the opinion previously given.
I want to add that the main reason for insisting on this point is that I do not believe in the reliability of LTspice at all, for use in very particular cases.

Being a novice, in Ltspice I would be grateful that yours or someone more experienced could take my doubts.
Thank you one more time .

Hi Nelson,

I am pleased you managed to use the SW switch in the simulator and the oscillator runs without any pre-charged C4, like the initial condition of  5 to 8V established earlier. 

However, what I wrote on the input power for the first 20 us simulation time and then from the 20 us to 260 us etc I meant for that particular simulation and I did not mean it generally.
The close to zero input current taken from V1 was the result of a not yet oscillating circuit in the first 20 us, and from around 20 us and onwards the input current started to increase form 3-4 mA towards the 10 to 15 mA range and finally it stopped at the 198 - 200 ms or so simulation time duration.

Now that the SW switch can mimic the push button you use on the real circuit may open up further test possibilities in the simulator. Of course we cannot fully trust in simulators, I agree and  we have to interpret and evaluate simulation results with caution. Will do so too with this new SW switch variaton.

Gyula
   
Group: Tech Wizard
Hero Member
*****

Posts: 948
....
Regarding your last point  :

General notice if I may: in circuits that include capacitors and coils that oscillate at a certain (mostly at resonant) frequency, the instantaneous peak power levels may exceed the instantaneous peak input power levels taken from a voltage or current source.
My question is as follows:
Where did this apparent current and voltage gain come from? Even if it is at peak level at  nS or mS   time ? :)  I really appreciate your answers, as well as Itsu's. My thanks
...

Hi Nelson,

The instanteneous currents and voltages via or across capacitors and coils cannot really be termed as "gain" because they come from normal operation of such reactive components.

Consider a 10 mH coil with 2 Ohm DC resistance. Suppose we connect a 22 nF capacitor in series with it and at their resonant frequency, 10.73 kHz we drive this series LC circuit from a function generator that has only 0.1 Ohm internal resistance and the output voltage is set to say 1 V.

So the current flowing in this circuit at resonance will be I = 1 V / 2.1 Ohm = 0.476 A.
Now consider this current establishes a magnetic field in this coil, the stored energy in this field would be E = 0.01 H x 0.476 A x 0.476 A / 2 = 0.00113288 J   i.e. 1.13288 mJ.
When the field changes to zero as per the AC input current dictates, the energy in the diminishing field drives a charging current into the 22 nF capacitor. Now we can calculate the voltage this capacitor will be charged up to.

Assuming a nearly lossless capacitor, quasi all the 1.13288 mJ energy would manifest as stored voltage and we can reverse calculate this voltage level:  0.00113288 J = 22 nF x V x V / 2. This gives V = 320.9 V, this will manifest across the 22 nF capacitor periodically at particular moments.
And also this 320.9 V will appear across the coil but just in opposite phase with respect to the capacitor voltage because the discharging capacitor current will build up the magnetic field in the coil and of course the coil will fight against the increase of this current by developing an increasing counter voltage.  Normal induction:  a changing coil current builds up a changing magnetic field and this latter induces voltage across the coil.

Now we can also consider the 0.476 A through the coil causes a voltage drop across the wire resistance this calculates as
2 Ohm x 0.476 A = 0.952 V and yet there is also the 320.9 V across the same coil but not across the 2 Ohm DC resistance but across the coil inductive reactance (the coil is a series RL circuit in this respect).

The inductive reactance is 674.18 Ohm for this coil at 10.73 kHz and the Q quality factor is 674.18 Ohm / 2 Ohm = 337.   And this same 337 is given by the ratio of the voltage across the 2 Ohm to the voltage across the coil  i.e. 320.9V / 0.952V = 337. 

Now what could be said is that the coil in a series resonant LC circuit keeps maintaning a  voltage across itself which is proportional to the voltage drop across the coil wire resistance (neglecting any other series resistance in the circuit).
And the proportionality constant is the Q quality factor, the ratio of the coil inductive reactance to its resistance. 

Regarding a parallel LC circuit at resonance, the input current to the circuit increases Q times inside the LC circuit i.e. if we drive an LC circuit with 10 mA current at the resonant frequency, then inside the LC circuit the current will be 1 A if the loaded Q of this resonant LC circuit would be 100.

Hopefully these 'ramblings' answer your above question. 

Good night,
Gyula
   
Full Member
***

Posts: 114


Buy me a drink
Gyula,   Nelson,


yes, it is kind of disappointing, but not unexpected.

I did some further tests this evening, but none showed any more then about 10mA output current (so no light in the output bulb).

I can mimic the input current going from 30mA unloaded to 21mA when shorting the output, but that is all.

So it could be the CMC (L2/L3) that is contributing something special to the circuit, so yes i could give those Digikey CMC's a go and will order some, but probably its that special one Nelson has that is needed ;) 

Anyway, not done with testing yet, tomorrow is another day.

If you want some more scope shots taken please indicate wich one.

Regards itsu


Hi Itsu ,

I tried to take some pictures to try to illustrate some of the differences that actually seem to exist between the coil of the circuit and the one you use.
I think that these Digikey CMC's coils do not correspond to the circuit coil, they are identical to the ones you used, and the ones I put in the attached photos.
Will this be with a 1: 1 transformer? Who knows ?? ☺ The coil wire in my circuit seems to be very thin compared to the used cmc coils that normally use 0.3 mm thick magnetic wire. I would say it is 0.1 mm, in addition it is covered with wax.

I don't know how I can help with this coil theme ☹ Maybe someone here on the forum, who in the past has worked with Sony TV repair? Maybe it could help with a tip!

Thanks Itsu



---------------------------
Best Rewards
Nelson Rocha

" The goal is not to be successful, the goal is to be valuable.
Once you’re valuable, instead of chasing success,
it will attract itself to you. "
   
Full Member
***

Posts: 114


Buy me a drink
Hi Nelson,

The instanteneous currents and voltages via or across capacitors and coils cannot really be termed as "gain" because they come from normal operation of such reactive components.

Consider a 10 mH coil with 2 Ohm DC resistance. Suppose we connect a 22 nF capacitor in series with it and at their resonant frequency, 10.73 kHz we drive this series LC circuit from a function generator that has only 0.1 Ohm internal resistance and the output voltage is set to say 1 V.

So the current flowing in this circuit at resonance will be I = 1 V / 2.1 Ohm = 0.476 A.
Now consider this current establishes a magnetic field in this coil, the stored energy in this field would be E = 0.01 H x 0.476 A x 0.476 A / 2 = 0.00113288 J   i.e. 1.13288 mJ.
When the field changes to zero as per the AC input current dictates, the energy in the diminishing field drives a charging current into the 22 nF capacitor. Now we can calculate the voltage this capacitor will be charged up to.

Assuming a nearly lossless capacitor, quasi all the 1.13288 mJ energy would manifest as stored voltage and we can reverse calculate this voltage level:  0.00113288 J = 22 nF x V x V / 2. This gives V = 320.9 V, this will manifest across the 22 nF capacitor periodically at particular moments.
And also this 320.9 V will appear across the coil but just in opposite phase with respect to the capacitor voltage because the discharging capacitor current will build up the magnetic field in the coil and of course the coil will fight against the increase of this current by developing an increasing counter voltage.  Normal induction:  a changing coil current builds up a changing magnetic field and this latter induces voltage across the coil.

Now we can also consider the 0.476 A through the coil causes a voltage drop across the wire resistance this calculates as
2 Ohm x 0.476 A = 0.952 V and yet there is also the 320.9 V across the same coil but not across the 2 Ohm DC resistance but across the coil inductive reactance (the coil is a series RL circuit in this respect).

The inductive reactance is 674.18 Ohm for this coil at 10.73 kHz and the Q quality factor is 674.18 Ohm / 2 Ohm = 337.   And this same 337 is given by the ratio of the voltage across the 2 Ohm to the voltage across the coil  i.e. 320.9V / 0.952V = 337. 

Now what could be said is that the coil in a series resonant LC circuit keeps maintaning a  voltage across itself which is proportional to the voltage drop across the coil wire resistance (neglecting any other series resistance in the circuit).
And the proportionality constant is the Q quality factor, the ratio of the coil inductive reactance to its resistance. 

Regarding a parallel LC circuit at resonance, the input current to the circuit increases Q times inside the LC circuit i.e. if we drive an LC circuit with 10 mA current at the resonant frequency, then inside the LC circuit the current will be 1 A if the loaded Q of this resonant LC circuit would be 100.

Hopefully these 'ramblings' answer your above question. 

Good night,
Gyula


Gyula thanks for your elaborate explanation about the operation of resonant circuits in series and in parallel, although it is not unknown to me, but I liked the way you explained the theme itself; It was very clear and succinct.
I could add that  series resonant circuit provides voltage magnification and  parallel resonant circuit provides current magnification.
However, gains in currents and voltages in resonant circuits are not synonymous with producing real work , given their reactive nature. At least that's what the books say . :)

When I ask you, from where apparent current and voltage gain come from, I was referring to the LTSpice circuit simulation, which was the topic we were discussing, given the shots presented,  because in that case LTspice's power calculation apparently presented a gain of instantaneous peak power in W and that is what puzzled me, but given that we already agreed in your last post, that we cannot fully trust in simulators I was completely clarified.
Thanks again for your response and contribution.

Good night Gyula


---------------------------
Best Rewards
Nelson Rocha

" The goal is not to be successful, the goal is to be valuable.
Once you’re valuable, instead of chasing success,
it will attract itself to you. "
   

Group: Moderator
Hero Member
*****

Posts: 3332

Hi Itsu ,

I tried to take some pictures to try to illustrate some of the differences that actually seem to exist between the coil of the circuit and the one you use.
I think that these Digikey CMC's coils do not correspond to the circuit coil, they are identical to the ones you used, and the ones I put in the attached photos.
Will this be with a 1: 1 transformer? Who knows ?? ☺ The coil wire in my circuit seems to be very thin compared to the used cmc coils that normally use 0.3 mm thick magnetic wire. I would say it is 0.1 mm, in addition it is covered with wax.

I don't know how I can help with this coil theme ☹ Maybe someone here on the forum, who in the past has worked with Sony TV repair? Maybe it could help with a tip!

Thanks Itsu

Nelson,

thanks for the pictures, it is a special one, i agree.

I will be looking for such a CMC, up till now my google searches did not yield a match.


Itsu
   
Group: Tech Wizard
Hero Member
*****

Posts: 948
Hi Nelson,

Sony components were given individual Part Numbers if I recall correctly. Your transformer has its Part Number stamped on its top but due to wear and ageing, the numbers may not be fully identified any more. Maybe with a magnifying glass it could be made out?

 It should have these number groups: 1-4XX-XXX-11   

in the bottom line the 8142 (or perhaps the last digit is 8 and not 2), the 81 may mean the year 1981 and the 42 (or 48) may mean the week number it was manufactured in 1981. 
 
Another approach would be to identify the chassis board, type or model number, from which this transformer was scavenged. If such chassis or circuit board is known, then there is a chance to get a service manual on it which includes the needed Part number.

But even if the correct Part number is known, such an old component (at least 35-39 years old or more) would be very hard to find if at all. There is little chance any Sony Spare Parts supplier stocks such old components any more.

There are several Sony Spare Parts suppliers coming up with google search but either the Sony model number or Sony component part number is needed to know in advance.

Ebay has some Sony chokes but none of them is even close to this one. here is such, just for illustration:
https://www.ebay.com/itm/193436933120

In the previous page I also included some Digikey CMC offers that have the C and I cores and 'similar' windings.  I suppose they are not even close to the Sony part you have? 
Here are some more:
https://www.digikey.com/product-detail/en/kemet/SS11VL-R08125/399-10555-ND/4290574

https://www.digikey.com/product-detail/en/kemet/SS26V-150121/399-10583-ND/4290602

https://www.digikey.com/product-detail/en/kemet/SS11H-07120-CH/399-10741-ND/4290807


Gyula
« Last Edit: 2020-05-01, 12:54:09 by gyula »
   
Full Member
***

Posts: 114


Buy me a drink
Hi Nelson,

Sony components were given individual Part Numbers if I recall correctly. Your transformer has its Part Number stamped on its top but due to wear and ageing, the numbers may not be fully identified any more. Maybe with a magnifying glass it could be made out?

 It should have these number groups: 1-4XX-XXX-11   

in the bottom line the 8142 (or perhaps the last digit is 8 and not 2), the 81 may mean the year 1981 and the 42 (or 48) may mean the week number it was manufactured in 1981. 
 
Another approach would be to identify the chassis board, type or model number, from which this transformer was scavenged. If such chassis or circuit board is known, then there is a chance to get a service manual on it which includes the needed Part number.

But even if the correct Part number is known, such an old component (at least 35-39 years old or more) would be very hard to find if at all. There is little chance any Sony Spare Parts supplier stocks such old components any more.

There are several Sony Spare Parts suppliers coming up with google search but either the Sony model number or Sony component part number is needed to know in advance.

Ebay has some Sony chokes but none of them is even close to this one. here is such, just for illustration:
https://www.ebay.com/itm/193436933120

In the previous page I also included some Digikey CMC offers that have the C and I cores and 'similar' windings.  I suppose they are not even close to the Sony part you have? 
Here are some more:
https://www.digikey.com/product-detail/en/kemet/SS11VL-R08125/399-10555-ND/4290574

https://www.digikey.com/product-detail/en/kemet/SS26V-150121/399-10583-ND/4290602

https://www.digikey.com/product-detail/en/kemet/SS11H-07120-CH/399-10741-ND/4290807


Gyula
Hi Gyula  , yes is almost impossible identify the coil , but i have news about that coil. i will publish more info very soon . I'm organize some things but later I will update with data .
See you later ;)



---------------------------
Best Rewards
Nelson Rocha

" The goal is not to be successful, the goal is to be valuable.
Once you’re valuable, instead of chasing success,
it will attract itself to you. "
   
Group: Experimentalist
Hero Member
*****

Posts: 1397
All,

I see comments regarding accuracy simulators in general and in particular LtSpice which is currently being used here for analysis.  LtSpice is a very powerful simulator and when setup properly, is extremely accurate.  Keep in mind that a simulation is no more accurate than the models used and sometimes the circuits themselves.  If a simulation does not match a bench circuit say within 5% or so, the sim models are usually to blame.  There is one exception to this and that is if there is external input to the bench circuit from electrostatic, electromagnetic, or aetheric means that is not accounted for in the simulation.

Even non-linear components can be correctly simulated although the degree of difficulty in modeling becomes significantly greater.  Many times we assume a transformer or coil is linear when in fact it not and this in itself can account for significant errors.

regards,
Pm 
   
Group: Experimentalist
Hero Member
*****

Posts: 1397
Hi Gyula  , yes is almost impossible identify the coil , but i have news about that coil. i will publish more info very soon . I'm organize some things but later I will update with data .
See you later ;)

The special or unique CMC that you've shown can be profiled in the following manner so it can be replicated or simulated by others with reasonable accuracy.

First, measure the inductance of each winding which in this case they will be close to the same.  This can be done at a low level with an inductance meter or with a high level from a switching circuit that will allow one to measure L = E*dt/di .  This last measurement will also allow one to check to see if the core is not saturating at the required operating mmf or coil current.

Next, take inductance measurements of the two coils connected in series aiding L+ and series bucking L-.  The L- or bucking measurement will be lower than the L+ or aiding.  Then using the formula M = ((L+)-(L-))/4 we now have the mutual inductance.

Using the mutual inductance we can now calculate the k or coupling factor with k = M/(Lp*Ls)^.5 but since Lp = Ls we can use k = M/Lp .

With this info, the transformer can now be built or simulated with the additional info on dc resistance.  There will a slight inter-turn capacitance and very little inter-winding capacitance due to the split bobbin design so both can be ignored without much error..

As a check, we can calculate the leakage inductance with Lpleak = (1-k)*Lp and the primary inductance with the secondary shorted with Lpss = (1-K^2)*Lp .

I might add that most CMCs have a slight gap in the ferrite cores to help prevent saturation at higher current levels.

Regards,
Pm
   

Group: Moderator
Hero Member
*****

Posts: 3332
Thanks PM,

i have taken one of my CMC's and measured:

Lp = 19mH
Ls = 19mH
L+ = 74mH
L- = 190uH

thus:

M = 73.81/4 = 18.45
K = 18.45/19 = 0.97

Lpleak = 0.57mH
Lpss = 1.1mH  (measuring Lp with Ls shorted gives 190uH, so this does not match the calculated value).

Itsu
   
Full Member
***

Posts: 114


Buy me a drink
After some investigation what I was able to ascertain, it is apparently a type of isolation transformer . Although the references are not the same, but I think it must be similar.
I leave a summary of the information I was able to collect. In the photos, some apparently identical coils are highlighted.

The insulating converter transformer PIT is equipped with an EE-type core comprising two E-type cores CR1, CR2 of ferrite material which are assembled so that the magnetic legs thereof are confronted to each other, and the primary winding N1 and the secondary winding N2 are wound around the center magnetic leg of the EE-type core while they are separated from each other by using a dividing bobbin B. Further, the center magnetic leg is designed to have a gap G therein, thereby achieving loose coupling based on a required coupling coefficient.
The gap G can be formed by making the center magnetic leg of each of the E-type cores CR1, CR2 shorter than the two outer magnetic legs thereof. Further, the coupling coefficient k is set to about 0.7 to 0.8 so that loose coupling can be attained, and thus it is harder to achieve the saturation state.

 Transformer PIT can be implemented in any combination case where the polarities of the primary winding N1 and the secondary winding N2 are in additive polarity relationship or subtractive polarity relationship and the winding directions thereof are the same (coaxial) or opposite to each other.





---------------------------
Best Rewards
Nelson Rocha

" The goal is not to be successful, the goal is to be valuable.
Once you’re valuable, instead of chasing success,
it will attract itself to you. "
   
Group: Experimentalist
Hero Member
*****

Posts: 1397
Thanks PM,

i have taken one of my CMC's and measured:

Lp = 19mH
Ls = 19mH
L+ = 74mH
L- = 190uH

thus:

M = 73.81/4 = 18.45
K = 18.45/19 = 0.97

Lpleak = 0.57mH
Lpss = 1.1mH  (measuring Lp with Ls shorted gives 190uH, so this does not match the calculated value).

Itsu

Itsu,

None of these calculations take into account the dc resistance of the windings and this is mostly affect the calculations for Lpleak and and particularly Lpss.  The best measurement method for these is to use resonance to arrive at the values. 

For example and you probably already know this, Lpleak can be determined by placing a known C across the secondary, measure the resonance on the primary and calculate Lpleak from the results.

Likewise Lpss can be determined by placing a known C in series with the primary with the secondary shorted, measure the resonance of the primary and calculate Lpss from the results.

Regards,
Pm

Edit: Sometime I would like to modify the equations to include the winding resistances!
   
Full Member
***

Posts: 114


Buy me a drink
All,

I see comments regarding accuracy simulators in general and in particular LtSpice which is currently being used here for analysis.  LtSpice is a very powerful simulator and when setup properly, is extremely accurate.  Keep in mind that a simulation is no more accurate than the models used and sometimes the circuits themselves.  If a simulation does not match a bench circuit say within 5% or so, the sim models are usually to blame.  There is one exception to this and that is if there is external input to the bench circuit from electrostatic, electromagnetic, or aetheric means that is not accounted for in the simulation.

Even non-linear components can be correctly simulated although the degree of difficulty in modeling becomes significantly greater.  Many times we assume a transformer or coil is linear when in fact it not and this in itself can account for significant errors.

regards,
Pm

Hi Partzman ,

Thanks for your opinion, as I said I am not a regular user of LTspice, having started it a few days ago to try follow the simulations made by Itsu and Gyula., in the context of this circuit that we are trying to understand.
I would like to take the liberty of asking you to do your simulation and interpretation based on the model I made available, at
https://www.overunityresearch.com/index.php?action=dlattach;topic=3691.0;attach=34859.
Given that you are a regular user, perhaps you can make some productive considerations regarding some values obtained during the transient simulation.

Grateful for your opinion thank you







---------------------------
Best Rewards
Nelson Rocha

" The goal is not to be successful, the goal is to be valuable.
Once you’re valuable, instead of chasing success,
it will attract itself to you. "
   
Full Member
***

Posts: 114


Buy me a drink
The special or unique CMC that you've shown can be profiled in the following manner so it can be replicated or simulated by others with reasonable accuracy.

First, measure the inductance of each winding which in this case they will be close to the same.  This can be done at a low level with an inductance meter or with a high level from a switching circuit that will allow one to measure L = E*dt/di .  This last measurement will also allow one to check to see if the core is not saturating at the required operating mmf or coil current.

Next, take inductance measurements of the two coils connected in series aiding L+ and series bucking L-.  The L- or bucking measurement will be lower than the L+ or aiding.  Then using the formula M = ((L+)-(L-))/4 we now have the mutual inductance.

Using the mutual inductance we can now calculate the k or coupling factor with k = M/(Lp*Ls)^.5 but since Lp = Ls we can use k = M/Lp .

With this info, the transformer can now be built or simulated with the additional info on dc resistance.  There will a slight inter-turn capacitance and very little inter-winding capacitance due to the split bobbin design so both can be ignored without much error..

As a check, we can calculate the leakage inductance with Lpleak = (1-k)*Lp and the primary inductance with the secondary shorted with Lpss = (1-K^2)*Lp .

I might add that most CMCs have a slight gap in the ferrite cores to help prevent saturation at higher current levels.

Regards,
Pm

Hi Partzman,

Thanks for the information about the calculation, but for the moment all the data I could provide has already provided some posts ago :
Inductance of each coil and its resistance. I think it will be a good starting point, given that initially, I only provided ohmic resistance for each coil.
Given the lack of equipment, I won't be able to do the test you mentioned,
I will have to borrow the LCR measurement instrument again from a friend so that I can do it.
But good tip! Thanks



---------------------------
Best Rewards
Nelson Rocha

" The goal is not to be successful, the goal is to be valuable.
Once you’re valuable, instead of chasing success,
it will attract itself to you. "
   

Group: Moderator
Hero Member
*****

Posts: 3332
Hi Partzman ,

Thanks for your opinion, as I said I am not a regular user of LTspice, having started it a few days ago to try follow the simulations made by Itsu and Gyula., in the context of this circuit that we are trying to understand.
I would like to take the liberty of asking you to do your simulation and interpretation based on the model I made available, at
https://www.overunityresearch.com/index.php?action=dlattach;topic=3691.0;attach=34859.
Given that you are a regular user, perhaps you can make some productive considerations regarding some values obtained during the transient simulation.

Grateful for your opinion thank you

Hi Nelson,

does your ultimo-functional2.asc runs continuously?   

When running here, it stops around 400ms.

Itsu
   

Group: Professor
Hero Member
*****

Posts: 2677
I see comments regarding accuracy simulators in general and in particular LtSpice which is currently being used here for analysis.  LtSpice is a very powerful simulator and when setup properly, is extremely accurate. 
Especially in the electronic domain.

If a simulation does not match a bench circuit say within 5% or so, the sim models are usually to blame.  There is one exception to this and that is if there is external input to the bench circuit from electrostatic, electromagnetic, or aetheric means that is not accounted for in the simulation.
I would even generalize it further and say that LTspice is not a general physics simulator.

For example make an LC tank circuit with a coil wound on an EE core (or pot core)  but bind the two halves of the core loosely so they can clap mechanically.
When the frequency of the current in the LC tank approaches the mechanical resonance of the core halves, then the current/voltage waveforms measured by the scope will be very different than the waveforms simulated by the electronic simulator.

Even non-linear components can be correctly simulated although the degree of difficulty in modeling becomes significantly greater.  Many times we assume a transformer or coil is linear when in fact it not and this in itself can account for significant errors.
Yes but there is a limit what LTspice can simulate. For example, in the example above with the clapping core, it could not simulate the mechanical or acoustic sounds from the core - the ones that are not functions of the instantaneous current flowing through the coil. Even if these functions are non-linear.
« Last Edit: 2020-05-01, 22:02:20 by verpies »
   
Group: Tech Wizard
Hero Member
*****

Posts: 948
Itsu, yes I noticed it too. 

One more thing: if I plot the current through the SW switch, there is a 13 Amper "glitch" at the moment the switch turns on at 1.05 ms moment and the 24 V input voltage appears on the common connection point of SW, C4, R1 and Emitter.   Strange for sure. The highest current via the SW should be not higher than (24V-VBE)  / (300+48 Ohm) = 67 mA  (VBE=0.7V)

Gyula
   

Group: Moderator
Hero Member
*****

Posts: 3332
After some investigation what I was able to ascertain, it is apparently a type of isolation transformer . Although the references are not the same, but I think it must be similar.
I leave a summary of the information I was able to collect. In the photos, some apparently identical coils are highlighted.

The insulating converter transformer PIT is equipped with an EE-type core comprising two E-type cores CR1, CR2 of ferrite material which are assembled so that the magnetic legs thereof are confronted to each other, and the primary winding N1 and the secondary winding N2 are wound around the center magnetic leg of the EE-type core while they are separated from each other by using a dividing bobbin B. Further, the center magnetic leg is designed to have a gap G therein, thereby achieving loose coupling based on a required coupling coefficient.
The gap G can be formed by making the center magnetic leg of each of the E-type cores CR1, CR2 shorter than the two outer magnetic legs thereof. Further, the coupling coefficient k is set to about 0.7 to 0.8 so that loose coupling can be attained, and thus it is harder to achieve the saturation state.

 Transformer PIT can be implemented in any combination case where the polarities of the primary winding N1 and the secondary winding N2 are in additive polarity relationship or subtractive polarity relationship and the winding directions thereof are the same (coaxial) or opposite to each other.



Thanks for the info Nelson,  but it seems to me that your CMC does not have 2 E-type cores, but more like Gyula mentioned a C and I type core.

Anyway, they could  work similar.


I have some (large) E-type cores and a former, so would be able to make one.
Can you determine if your coils are in the same winding direction or opposite?

Itsu

   
Full Member
***

Posts: 114


Buy me a drink
Hi Nelson,

does your ultimo-functional2.asc runs continuously?   

When running here, it stops around 400ms.

Itsu

Hi Itsu,
I just tested 500 ms and it stopped at 400ms.
It was curious to note that there is a negative curve since the beginning of the simulation, being possible to verify that IV1 is approaching the positive value until it stops at zero.

About the wiring of the coils :
Seems to me that coils wired in opposite directions , like L1 >=< L2 .



---------------------------
Best Rewards
Nelson Rocha

" The goal is not to be successful, the goal is to be valuable.
Once you’re valuable, instead of chasing success,
it will attract itself to you. "
   
Full Member
***

Posts: 114


Buy me a drink
Itsu, yes I noticed it too. 

One more thing: if I plot the current through the SW switch, there is a 13 Amper "glitch" at the moment the switch turns on at 1.05 ms moment and the 24 V input voltage appears on the common connection point of SW, C4, R1 and Emitter.   Strange for sure. The highest current via the SW should be not higher than (24V-VBE)  / (300+48 Ohm) = 67 mA  (VBE=0.7V)

Gyula


Hi gyula ,

I had also seen this unrealistic peak current value,  but the SW is dependent from V2 to open , however SW is supposed to remain isolated from V2.
If you measure the voltage peak in Vn007 in 10ms he achieve something like -360V .  lol

PS- maybe LTspice consider a fast electric burst transients in th switch open ??????


---------------------------
Best Rewards
Nelson Rocha

" The goal is not to be successful, the goal is to be valuable.
Once you’re valuable, instead of chasing success,
it will attract itself to you. "
   

Group: Moderator
Hero Member
*****

Posts: 3332
Hi Itsu,
I just tested 500 ms and it stopped at 400ms.
It was curious to note that there is a negative curve since the beginning of the simulation, being possible to verify that IV1 is approaching the positive value until it stops at zero.

About the wiring of the coils :
Seems to me that coils wired in opposite directions , like L1 >=< L2 .


Thanks for the wiring info on your CMC.

Looking closer to my CMC's (like your 1.jpg picture in post #502), it seems to me that these ARE made
of 2 E-type cores (in one piece) and there probably is a gap in the center where the coils are wound
over it, AND mine are wound in opposite direction.

So the CMC's i have do look like your PIT transformer, only the wire thickness could be the difference.

Itsu

   
   
Group: Tech Wizard
Hero Member
*****

Posts: 948
Hi Itsu,

The first 4 pictures are on Digikey or are on similar to Digikey CMCs and the last 3 pictures are that of Nelson's.
so if you mean this https://www.overunityresearch.com/index.php?action=dlattach;topic=3691.0;attach=34880 it is from Digikey.

I attached Nelson's photo on his transformer, no ferrite core bridge is seen at its bottom part, the C core is on the top part and the I core is in the middle, inside the bobbin.

Gyula
   
Group: Tech Wizard
Hero Member
*****

Posts: 948


I found data sheets on Epson Power line chokes, may be useful, includes stray inductances.
   
Group: Experimentalist
Hero Member
*****

Posts: 1397
Nelson,

OK, here is my version of your sim which has a couple of modifications.  Most involve rotation of components so the components conventional current flow is with the circuit current flows, S1 modified, and C7 moved to parallel L3 instead of D7.  Everything else is left the same except the timing of S1 which you can see in the pulse gen V2.

If C7 is left in it's original position, Q1 is starved of base current at about 400ms and the oscillation shuts down.

The first pix is the sim running out to 1 second at which time we reach near complete stabilization.

The next pix is an expanded view of the last 8 cycles of the simulation.

The last pix shows the power levels of E1, E2 and the input power drawn from the supply V1.

I guess what I need to know is, what is the specific function of this circuit?

Regards,
Pm

Edit: Attached the .asc file.
« Last Edit: 2020-05-02, 01:30:43 by partzman »
   
Full Member
***

Posts: 114


Buy me a drink
Nelson,

OK, here is my version of your sim which has a couple of modifications.  Most involve rotation of components so the components conventional current flow is with the circuit current flows, S1 modified, and C7 moved to parallel L3 instead of D7.  Everything else is left the same except the timing of S1 which you can see in the pulse gen V2.

If C7 is left in it's original position, Q1 is starved of base current at about 400ms and the oscillation shuts down.

The first pix is the sim running out to 1 second at which time we reach near complete stabilization.

The next pix is an expanded view of the last 8 cycles of the simulation.

The last pix shows the power levels of E1, E2 and the input power drawn from the supply V1.

I guess what I need to know is, what is the specific function of this circuit?

Regards,
Pm

Edit: Attached the .asc file.






First of all I want to say that I will be as direct as possible in my answer face to your approach in your last posts, where I sincerely  felt some hostility and cynicism by your side just because I expressed my doubts about the reliability of LTspice in certain situations .

When I ask you  some considerations regarding the values obtained during the transient simulation on this specific diagram I was not considering the profound changes you made to the original circuit but try to understand the actual circuit .
The idea would be to justify and interpret the result of the original circuit, but after the changes you made, what can I say? I had the same doubts that I had previously.

This diagram was initially designed by Itsu, and later modified by me to try to recreate a specific real life circuit that we suspect have some sort of negative resistance in their operation, but apparently, you didn't even give the opportunity to check that, face your question at the end of the post,where you ask what is the specific function of this circuit.

Some of the changes you have made, namely in the configuration of V2 pulse time , T Rise and T fall , period  , completely change the behavior of the simulated circuit, but you should know that better than me, because it was you that  conveniently change the values. Period value of 1second ??   SW is supposed to simulate a action of a simple push button , My question is why you made that changes? What is the point of change that values  ?

About C7 capacitor, the one you change from original position to parallel with L3 You justified that if you do not change the position of C7 the oscillation would die at 400ms , but on the real circuit, it doesn't stop …  And one of the main points, would be exactly to understand why in the simulation in LTspice, he stops at 400ms, but you discover  the problem was C7 capacitor...   Should I change the original real circuit according to your simulation to you believe yourself that Ltspice is extremely accurate?  Seem is not, but I understand, to you is now properly configured.

I don't think I have anything else to add to my answer
Many thanks.

Best rewards


---------------------------
Best Rewards
Nelson Rocha

" The goal is not to be successful, the goal is to be valuable.
Once you’re valuable, instead of chasing success,
it will attract itself to you. "
   
Pages: 1 ... 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 [21] 22 23 24 25 26 27 28 29 30 31 32 33
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2022-01-18, 03:50:48