PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
2024-03-29, 07:03:30
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 ... 5 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 34 35 ... 39
Author Topic: Itsu's workbench / placeholder.  (Read 105220 times)

Group: Moderator
Hero Member
*****

Posts: 4056

Here the currents (green) through D1 and D2 during start of oscillations (as seen by the yellow trace  being the "switch" voltage):

Be aware they have different amplitude settings.


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

Posts: 1092
Hi Itsu, 

Thanks for doing the measurements. It is good the measured base and collector current waveforms look pretty close to the simulated ones. Yes, please try to expand the base current to see it better, at what time it raises above zero where the negative collector current starts.

Regarding the current via the diodes: after the switch-on transients seen on the left, they maintain their periodic nature seen on the right side, is that right?

IF yes, then this is greatly different from the simulated diode currents because the latter ones die out in the simulation after roughly 100 us later the start-up, albeit the oscillator remains working for the rest of the simulation time. 

Gyula
   

Group: Moderator
Hero Member
*****

Posts: 4056
Gyula,


yes they maintain their periodic nature seen on the right side.

Be aware i updated the collector/base screenshot post with a 2th screenshot in different amplitudes taken with 2 current probes.

Here again the col / base current, again with 1 probe (AC/DC) but with different amplitude settings

The other current probe used in the before "dual probe measurement" is an AC only probe which misses some resolution

Itsu
« Last Edit: 2020-04-28, 21:28:47 by Itsu »
   
Group: Tech Wizard
Hero Member
*****

Posts: 1092
Itsu, thank you. The updated base current waveform is still much different from the earlier (i.e. from the single current probe) measured one or from the simulated one. Could you recheck tomorrow so that you use the current probe applied earlier at say the collector to check the base current and vice versa? i.e. replace the two probes with each other. 

And maybe the differential voltage measurement across resistor R1 could be rechecked again and also expand it from the earlier 50 mA to say 10mA sensitivity. 

Thanks again and good night,
Gyula
   
Full Member
***

Posts: 114


Buy me a drink
Itsu ,

Thanks for the shots, and good night to you and 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: 4056

Gyula,

the first screenshot in my post #474 and the one in post #477 are taken with the same procedure using my AC/DC current probe A6302 only.

That first screenhot kept the same amplitude setting to show the amplitude differences between collector and base currents, while the one on post #477 shows the same but with adjusted amplitude setting
for the base, so in principle they should be the same.
(it could be i screwed up on the rms value readout as i need to set it manually based on my current probe controller setting).


The 2th screenshot in post #474 i made using my other, AC only, current probe P6021 for the collector together with the earlier used AC/DC A6302 for the base while triggering on the oscillator voltage signal.

This AC only current probe seems to miss some sensitivity or resolution as there is some difference compared to the white traces made by the AC/DC probe (causing flatlining inbetween).
Measuring the low base current with it gives to worse results due to it being so low in current.

I did not use the differential voltage probe method across R1 yet, but can do that tonight.

Sorry for the confusion.

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

Posts: 1092
Okay Itsu, thanks, no problems.  Will be back later tonight.

Gyula
   

Group: Moderator
Hero Member
*****

Posts: 4056

I made myself a 13mH CMC and put it in the circuit as L2/L3.
L1 was already 2.7mH which is close to the measured 3mH, only the DC resistance is only 1.6 Ohm, so i added a 1 Ohm resistor in series.

For the rest i used caps as measured by Nelson.
I tried a similar bridge as Nelson showed in some video's, but this made little or no change compared to my UF4007 bridge.

Only difference now is the MJE13009 transistor and 7 Ohm bulbs.

Oscillation frequency is 13Khz (lower then the 19Khz from Nelson his circuit) and input current is 58mA which is higher then Nelson (33mA).

Output voltage unloaded is 278V and when shorted it pulls about 13mA (barely glowing the output bulb) with a lowered input current of 31mA.

So the effect showed by Nelson where the output bulb is about twice as bright (60mA??) as the input bulb (24mA) when shorting the output is not there.

Video here:  https://www.youtube.com/watch?v=XPE0eXqxysc&feature=youtu.be


Hopefully the  MJE18008 will make the change.


Itsu

   
Full Member
***

Posts: 114


Buy me a drink
I made myself a 13mH CMC and put it in the circuit as L2/L3.
L1 was already 2.7mH which is close to the measured 3mH, only the DC resistance is only 1.6 Ohm, so i added a 1 Ohm resistor in series.

For the rest i used caps as measured by Nelson.
I tried a similar bridge as Nelson showed in some video's, but this made little or no change compared to my UF4007 bridge.

Only difference now is the MJE13009 transistor and 7 Ohm bulbs.

Oscillation frequency is 13Khz (lower then the 19Khz from Nelson his circuit) and input current is 58mA which is higher then Nelson (33mA).

Output voltage unloaded is 278V and when shorted it pulls about 13mA (barely glowing the output bulb) with a lowered input current of 31mA.

So the effect showed by Nelson where the output bulb is about twice as bright (60mA??) as the input bulb (24mA) when shorting the output is not there.

Video here:  https://www.youtube.com/watch?v=XPE0eXqxysc&feature=youtu.be


Hopefully the  MJE18008 will make the change.


Itsu

Itsu,
I want to thank you for the video, as well as your dedication to this topic.
After watching your last video, I’ll notice the way you created the L2-L3 coil.
I could comment after an analysis, very superficial, that with this coil it will never work, with the values of the current components.
If you look closely at the original coil (I send photos attached) in addition to the magnetic wire being much thinner, I would like you to take into account the position as the original L1-L3 coils are positioned.
Both are superimposed, vertically as if they were two flat coils, with the difference of having a core crossing the two coils.
Could that be the reason? ;)
Of course, it is just my opinion, but in any case, I am available to send the original circuit to you if necessary.
Once again, my thanks for your help in this matter.

PS -i forget put link to images :)        https://photos.app.goo.gl/WRpwLy9uMLZUCExD9


---------------------------
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: 4056

Hi Nelson,

well, i did use some similar like yours CMC's which are "vertically as if they were two flat coils", but they are all to high in inductance (lowest is 22mH each).

So i tried to "whip up" a CMC that's 13mH, but indeed not "vertically as if they were two flat coils".

On the other hand, i am not sure that would make much difference, as its the inductance / DCR value's that are most important imo, but you could be right that this coil would never work.

So i was looking also to this CMC, see picure below, which is a 13mH CMC like yours, so perhaps i can order one.


Please keep your "working" unit safe for now as the main component (MJE18008) is still missing here.

Regards Itsu
   

Group: Moderator
Hero Member
*****

Posts: 4056
I redid the collector / base current measurements in this new circuit, see screenshot 1.

I used the P6021 current probe (AC) for the collector current (purple)
I used the differential voltage probe method for the base current (red) and
i used the A6302 current probe (AC/DC) also for the base current (green).

I found out that the magnetic top L1 has much influence on the signal of the A6302 probe, so i had to extend the loop to get the probe out of the way of this L1, so base current is different from yesteday,
but the differential probe methode and the A6302 probe sigs now match.

We loose some resolution on the differential probe methode due to the low signals and math.

2th screenshot is the same, but collector (purple) and base (green) currents put on same line.

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

Posts: 1092
...
I left the last shots for today . My question is as follows:
The current in E2 (Follow the scheme) has a negative peak of 168.39mA at 0,05ms with a peak voltage of 8.0092V at N004 giving a 1.3307W .

The current in input 9.8mA at V1 and 24V at V1, giving a 224.49mW of power consumed .

My question is as follows:
Why does LTspice show a lower consumption value in V1 at the entrance, compared to the value displayed at the exit at E2, even if it is in a small fraction of time?

Is LTspice supposed to present this type of values, being a reference software in electronic simulation?
...

Hi Nelson,

I return to your questions you asked in connection with the ultimo-funcional.asc circuit file and sreenshot Math of input Vs Output.png and I included your shot below from your above post for reference only.

I repeated the simulation (note that I also used r=2 Ohm for L1 as you did back then) and plotted the same input and output currents and power levels you did, see 2nd attachment.
I drew a vertical yellow line at 35.5 us where the highest output power (1.3 W, red curve) manifested and where the current via E2 bulb was the highest too (-165.5 mA, dark blue curve). Here the simulator gave 210 mW (light blue curve) input power, input current taken from 24V voltage source was 8.66 mA (green curve) at the moment of 35.5 us.

These values are very close to your values, the small differences should have come from the reading accuracy of the moving cursor over the curves i.e. our own hand accuracy.
And you considered the time at 0.05 ms i.e. at 50 us but in your plot the maximum peak current in E2 is at less than 40 us time as I see it. Well this difference is no problem for me, and this does not make your questions invalid in any way.

It is also ok that you chose time range between 20 us and 260 us to display the input and output currents and power levels to illustrate your questions (I refer to your plot file again: Math of input Vs Output.png).

Now on your 1st question: LTspice showed lower input power taken from V1 voltage source than the output power in E2 bulb was because at that particular moment (at your 50 us or at my 35.5 us) those peak power levels must have been ruling in the simulated circuit.

On your 2nd question: LTspice and other Spice based circuit simulators are supposed to present correct results provided the component models in them are close to the electrical parameters of the real world components, this depends also on the software user what he or she enters.

To further answer your 1st question, we need to evaluate the total input power the circuit received between time range from 0 to 20 us.  Notice that there is an initial condition: 8V is pumped out from C4 which kicks the oscillator in and let me show you what power and current levels are present in the same circuit within the 0 to 20 us time duration, see 3rd attachment.

The red curve shows L1 coil current, the pink curve shows voltage across L1 coil, the yellow curve shows the instanteneous peak power in L1 and the green curve shows the DC input current taken from V1 voltage source. 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.

Notice that this circuit stops in the simulator at around 197.7 ms if run as long as say 200 ms. I attached the current and power level curves for the time range between 196 and 200 ms. The output power (red curve) in E2 bulb converges from 3.6 mW average power to zero, while the input power converges from -198 mW average power to also zero, both levels are for the 196 to 200 ms time range (the power levels are received by Ctrl + left mouse clicks of course). 

General notice if I may: in circuits that include capacitors and coils that oscillate at a certain (mostly at resonant) frequency, the instanteneous peak power levels may exceed the instanteneous peak input power levels taken from a voltage or current source. 

Maybe I answered your 2 questions, if not please ask.

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

Posts: 1092
Hi Itsu,

Many thanks for doing the measurements.  Will digest and comment them tomorrow.

Greetings
Gyula
   
Full Member
***

Posts: 114


Buy me a drink
Hi Nelson,

well, i did use some similar like yours CMC's which are "vertically as if they were two flat coils", but they are all to high in inductance (lowest is 22mH each).

So i tried to "whip up" a CMC that's 13mH, but indeed not "vertically as if they were two flat coils".

On the other hand, i am not sure that would make much difference, as its the inductance / DCR value's that are most important imo, but you could be right that this coil would never work.

So i was looking also to this CMC, see picure below, which is a 13mH CMC like yours, so perhaps i can order one.


Please keep your "working" unit safe for now as the main component (MJE18008) is still missing here.

Regards Itsu


Itsu ,
I honestly think that the configuration of the coil, such as the thickness of the wire, and the position of the orientation of the coils will make difference in the operation of oscillator .  In my perspective even the gap between turns, completely changes the coil's inductance. I think the coupling between the L2 and L3 is much different  from original coil configuration .
I will keep this unit free , no problem , but I wanted you to understand, that I am 100% available and ready to send you the original circuit .
About the SMC choke you  attach , seems to have some type of gap in the middle of winded coils ?? I'm not sure .  This coil was removed from old Sony TV that i savaged some components , i really don't know their reference rssssssss .  it is a pity .... Let's see how it goes when you have MJE10008;).

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: Tech Wizard
Hero Member
*****

Posts: 1092
Itsu,  if I see Nelson's choke transformer correctly in his photos, it consists of a C and I core pieces,  and the I core on which the bobbin is with the coils closes magnetically the C core. There is usually a small air gap left between the two core pieces.

Here are similar constructions:

https://www.digikey.com/product-detail/en/kemet/SS21V-080136/399-10682-ND/4290748 

https://www.digikey.com/product-detail/en/kemet/SS30V-R200132/399-10600-ND/4290666 

https://www.digikey.com/product-detail/en/kemet/SSR10HS-10135/399-19019-ND/9664316 

https://www.digikey.com/product-detail/en/kemet/SSR10VS-10135/399-19041-ND/9664338 

Perhaps Nelson wish to comment these.

Gyula
   
Full Member
***

Posts: 114


Buy me a drink
Hi Nelson,

I return to your questions you asked in connection with the ultimo-funcional.asc circuit file and sreenshot Math of input Vs Output.png and I included your shot below from your above post for reference only.

I repeated the simulation (note that I also used r=2 Ohm for L1 as you did back then) and plotted the same input and output currents and power levels you did, see 2nd attachment.
I drew a vertical yellow line at 35.5 us where the highest output power (1.3 W, red curve) manifested and where the current via E2 bulb was the highest too (-165.5 mA, dark blue curve). Here the simulator gave 210 mW (light blue curve) input power, input current taken from 24V voltage source was 8.66 mA (green curve) at the moment of 35.5 us.

These values are very close to your values, the small differences should have come from the reading accuracy of the moving cursor over the curves i.e. our own hand accuracy.
And you considered the time at 0.05 ms i.e. at 50 us but in your plot the maximum peak current in E2 is at less than 40 us time as I see it. Well this difference is no problem for me, and this does not make your questions invalid in any way.

It is also ok that you chose time range between 20 us and 260 us to display the input and output currents and power levels to illustrate your questions (I refer to your plot file again: Math of input Vs Output.png).

Now on your 1st question: LTspice showed lower input power taken from V1 voltage source than the output power in E2 bulb was because at that particular moment (at your 50 us or at my 35.5 us) those peak power levels must have been ruling in the simulated circuit.

On your 2nd question: LTspice and other Spice based circuit simulators are supposed to present correct results provided the component models in them are close to the electrical parameters of the real world components, this depends also on the software user what he or she enters.

To further answer your 1st question, we need to evaluate the total input power the circuit received between time range from 0 to 20 us.  Notice that there is an initial condition: 8V is pumped out from C4 which kicks the oscillator in and let me show you what power and current levels are present in the same circuit within the 0 to 20 us time duration, see 3rd attachment.

The red curve shows L1 coil current, the pink curve shows voltage across L1 coil, the yellow curve shows the instanteneous peak power in L1 and the green curve shows the DC input current taken from V1 voltage source. 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.

Notice that this circuit stops in the simulator at around 197.7 ms if run as long as say 200 ms. I attached the current and power level curves for the time range between 196 and 200 ms. The output power (red curve) in E2 bulb converges from 3.6 mW average power to zero, while the input power converges from -198 mW average power to also zero, both levels are for the 196 to 200 ms time range (the power levels are received by Ctrl + left mouse clicks of course). 

General notice if I may: in circuits that include capacitors and coils that oscillate at a certain (mostly at resonant) frequency, the instanteneous peak power levels may exceed the instanteneous peak input power levels taken from a voltage or current source. 

Maybe I answered your 2 questions, if not please ask.

Gyula

Gyula ,
Many thanks by your answer and analyze to my questions .
My questions were asked because I am a beginner at LTspice, and I am not familiar with circuit simulations and theory, I would say that I am more of a bench person, if I make myself understood.
However, without taking value from the LTspice tool, and as I am an open-minded person, receptive to new working methods, I wanted to take the opportunity to get started.
I will read your answers carefully, and do a personal analysis on Ltspice as well, However, it would not be honest on my part, to confess, that I think there will always be profound differences between a simulation and real life, even more when we walk on paths, still little explored.

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

Have a good night

PS- You may not have noticed, but in the post Reply #460 i publish shots with the resistance value already corrected to 2.6 ohms in L1 .
« Last Edit: 2020-04-29, 22:49:18 by nelsonrochaa »


---------------------------
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: Tech Wizard
Hero Member
*****

Posts: 1092

....
PS- You may not have noticed, but in the post Reply #460 i publish shots with the resistance value already corrected to 2.6 ohms in L1 .

Yes Nelson, I noticed and knew it. The reason I used the 2 Ohm for L1 was to have very close current and power levels for your quoted simulation which also used 2 Ohm due to the use of comma instead of the decimal dot.

Will continue tomorrow, have a good night too.

Gyula
   

Group: Moderator
Hero Member
*****

Posts: 4056

Finally i got my MJE18008g transistors.

I went back to the 22mH CMC for L2/L3 and adjusted pot R3 to have 31mA input current unloaded (96 Ohm).

Unloaded output is about 250V, but when shorting it we drop input current to 21mA, but the output current is around 10mA and not able to light up the output bulb.

Oscillation frequency unloaded is 15.7Khz, loaded 12Khz.

Collector (purple) and base (green) currents as shown below, 1st screenshot unloaded, 2th loaded.

So still no effect noted as shown by Nelson (brighter output bulb then input bulb when shortend).

Video here: https://www.youtube.com/watch?v=71h3J5FkBVY&feature=youtu.be

Regards Itsu
   

Group: Professor
Hero Member
*****

Posts: 3354
On the other hand, i am not sure that would make much difference, as its the inductance / DCR value's that are most important imo,
Intra and Interwinding capacitance can matter for the Electronic Engineering aspect of the circuit (the thing the LTspice is concerned with).

IMO the placement of the windings, their spacing and shape, as well as the core type and shape do not matter from EE's point of view.
However from the physics point of view, they can matter very much because of field geometry and core material's effects.
   

Group: Moderator
Hero Member
*****

Posts: 4056

Hmmmm,  thanks,  so like the CMC used in Nelson his circuit can create things like f.i. NMR / NAR which influences
his circuit to show the effect which won't be there when there is no NMR / NAR (my CMC or LTspice).

Itsu
   
Sr. Member
****

Posts: 281
Intra and Interwinding capacitance can matter for the Electronic Engineering aspect of the circuit (the thing the LTspice is concerned with).

IMO the placement of the windings, their spacing and shape, as well as the core type and shape do not matter from EE's point of view.
However from the physics point of view, they can matter very much because of field geometry and core material's effects.

Good day All:

Following and concurring with Verpies train of thought above, I will also add that the composition of physical build, ie; spaceing(mentioned above) and the number of parallel strands that form a winding (think litz) and the method used to interweave those strands (ie; braided/intertwined/twisted vs parallel) also can have profound effects on *k* factor, bandwidth, T_rise, T_fall, etc.

take care, peace
lost_bo
   
Full Member
***

Posts: 114


Buy me a drink
Finally i got my MJE18008g transistors.

I went back to the 22mH CMC for L2/L3 and adjusted pot R3 to have 31mA input current unloaded (96 Ohm).

Unloaded output is about 250V, but when shorting it we drop input current to 21mA, but the output current is around 10mA and not able to light up the output bulb.

Oscillation frequency unloaded is 15.7Khz, loaded 12Khz.

Collector (purple) and base (green) currents as shown below, 1st screenshot unloaded, 2th loaded.

So still no effect noted as shown by Nelson (brighter output bulb then input bulb when shortend).

Video here: https://www.youtube.com/watch?v=71h3J5FkBVY&feature=youtu.be

Regards Itsu


Itsu ,
Thank you very much for your video again.
I am sorry that the MJE18008g transistor has not given better results.
Perhaps the choice of the coil may be making a difference, because even in the current one, it seems to me that there are clear differences.
I have some identical ones that you are using, and if you notice there is a gap in each winding L2-L3, which is seen from the bottom shows a cross winding in each part of the coil, as well as the magnetic wire thickness.
It is the only reason that occurs to me now. I am sure that with patience and perseverance, we will be able to reach the main objective of reproducing the original circuit.
My thanks for making your time available on this topic.




---------------------------
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
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 .



---------------------------
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: Tech Wizard
Hero Member
*****

Posts: 1092
Hi Itsu,

I have been curious on the collector and base current waveforms in a working oscillator like this and I also wished to compare them to the simulated waveforms. The collector current waveforms more or less comparable but the base current waveforms are not so much and of course we have to rely on the measured waveforms. And what rather differs in the simulation is the currents through D1 and D2 diodes you measure in your circuit.

Thanks you very much again for your kind efforts. It is very good you are dedicated and I will be available further on if I can be of any help. It is unfortunate that the new transistor has not produced closer results to that of Nelson's.  Do you think one of the the choke coils I gave link to Digikey above could work better ? 

Gyula
   

Group: Moderator
Hero Member
*****

Posts: 4056

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
   
Pages: 1 ... 5 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 34 35 ... 39
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2024-03-29, 07:03:30