PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
2022-05-26, 05:27:21
News: Forum TIP:
The SHOUT BOX deletes messages after 3 hours. It is NOT meant to have lengthy conversations in. Use the Chat feature instead.

Pages: 1 ... 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 [33] 34 35 36 37 38
Author Topic: Itsu's workbench / placeholder.  (Read 62134 times)

Group: Moderator
Hero Member
*****

Posts: 3433
Yes, without increased output voltage to push the current into the inductor very quickly you are limited to low frequencies for the BH curve.
Because you took the hard way.

The easy way is to buy several pre-made isolated ±15V DC-DC converters to power the gates and use two isolated gate drivers (or digital isolators) for the high side MOSFETs. 
They are available now from Mouser, Digikey, RSonline, etc... but were not 8 years ago.

 O0  I have some RK-0515S DC2DC converters and some IL610 isolators, but i see that those are old and there are new ones integrated together like the ADUM5241ARZ

Itsu


   

Group: Professor
Hero Member
*****

Posts: 2746
... and some IL610 isolators, but i see that those are old and there are new ones integrated together like the ADUM5241ARZ
These ADUM... digital isolators have 70ns propagation delay time.  Depending what your target frequency is, these might not be fast enough.
There are faster ones like the ADN4654BRWZ with 4ns propagation delay ...or even 2ns from Skyworks Solutions, Inc.

There are also gate drivers already integrated with isolators like the UCC21540DWK.
   

Group: Moderator
Hero Member
*****

Posts: 3433
YT has set your video referred there to "private"

Hmmm,  thanks, i notice YT has made some changes lately which i was not aware of (small print?) like ads on some of them and the change to private.

I have put that video on hidden again.

Itsu
   

Group: Moderator
Hero Member
*****

Posts: 3433
These ADUM... digital isolators have 70ns propagation delay time.  Depending what your target frequency is, these might not be fast enough.
There are faster ones like the ADN4654BRWZ with 4ns propagation delay ...or even 2ns from Skyworks Solutions, Inc.

There are also gate drivers already integrated with isolators like the UCC21540DWK.

Thanks for the info, i see many of those chips are not available due to the chip shortage.

Perhaps its faster to look for a completed product on Ebay or Ali.

Itsu
   

Group: Professor
Hero Member
*****

Posts: 2746
Thanks for the info, i see many of those chips are not available due to the chip shortage.
Some are on Mouser some are on www.rs-online.com and some on Digikey and some on Farnell.com ...and if you don't mind CN then on lcsc.com

Perhaps its faster to look for a completed product on Ebay or Ali.
Let me know if you find something.
I don't think there is much consumer demand for MHz digital amplifiers with isolated inputs (so the FG is safe). 
Only EEs and scientists want them, because they do not contain an integrated power supply and music would sound horrible* if amplified by these amps.

* Unless PWM was added at their input and LC filter was attached to their output.
   

Group: Moderator
Hero Member
*****

Posts: 3433

I saw this experiment (see diagram) from member Jagau on aboveunity.com here:  https://www.aboveunity.com/thread/capacitor-recharging/

It shows what happens at the instance a charged capacitor (disconnected from its source (SW1 open)) is connected (SW2 closed) to an inductor.

Somehow i would have expected that the cap would not be negatively charged due to the diode, but it does.

Below screenshot shows:

blue:  the voltage across the cap (10uF)
yellow: the voltage across the coil (8.4mH)
green: current between C1 and SW2 (closed)

Traces have a small offset from zero for better visibility.

   
Itsu
   

Group: Professor
Hero Member
*****

Posts: 2746
It is important to remember that a diode blocks reverse current - not reverse voltage.
   

Group: Tinkerer
Hero Member
*****

Posts: 3055
Actually a beautiful experiment.

It demonstrates nicely Capacitive Discharge and Inductive Charge and Discharge leading to Capacitive Re-Charge with Polarity Change.

A form of Resonant Charging but it would be better if the Capacitor was Non-Polarized.

This behavior is the "secret" to the Switching Converter which Changes Polarity of the Input Current when the Output is directed to an external Load.


---------------------------
For there is nothing hidden that will not be disclosed, and nothing concealed that will not be known or brought out into the open.
   

Group: Moderator
Hero Member
*****

Posts: 3433
It is important to remember that a diode blocks reverse current - not reverse voltage.

Indeed, and this is a nice example of that IMO.

Itsu
   
Group: Experimentalist
Hero Member
*****

Posts: 918
I saw this experiment (see diagram) from member Jagau on aboveunity.com here:  https://www.aboveunity.com/thread/capacitor-recharging/

It shows what happens at the instance a charged capacitor (disconnected from its source (SW1 open)) is connected (SW2 closed) to an inductor.

Somehow i would have expected that the cap would not be negatively charged due to the diode, but it does.
...

Yes, it is surprising. So I did the ltspice simulation, and it is confirmed.


---------------------------
"Open your mind, but not like a trash bin"
   

Group: Moderator
Hero Member
*****

Posts: 3433

Thanks F6FLT,   i was thinking about doing a simulation, but as the circuit was so simple i went for the real one.

Glad you did the simulation and that it confirmed it  O0

Itsu
   

Group: Moderator
Hero Member
*****

Posts: 3433

For Jagau on his remark to me in the above mentioned thread (https://www.aboveunity.com/thread/capacitor-recharging/):

Quote
P.S to Itsu, If you do not want to see the parasitic resonance at the end of the waveform, place your scope probe on
the anode of the diode, you will have a cleaner waveform.

Thanks Jagau, but i do want to see/show this "parasitic resonance" on the inductor (yellow trace).
And, i did have the (blue) probe on the "anode of the diode" (SW2 closed) which indeed shows a cleaner waveform (and thus the lack of "parasitic resonance")  O0


Itsu
   
Newbie
*

Posts: 2
you seem to me a good experimenter.
Tell me what type of current probe do you use?
thank you
   

Group: Moderator
Hero Member
*****

Posts: 3433

Hi Spark2,

i use a A6302 / AM 503B current probe / controller combo (Tektronix).

Itsu
   

Group: Moderator
Hero Member
*****

Posts: 3433

Without the diode the Jagau circuit and signals show as below.

Massive ringing on the resonance frequency of the cap (10uF) and inductor (8.4mH) on 549Hz.
   

Sr. Member
****

Posts: 335
Enjoy your trek through life but leave no tracks
It is important to remember that a diode blocks reverse current - not reverse voltage.
Yes it does! if its depletion layer becomes reverse biased it will be cut off, most e prom programmers and micro controller programmers use that principal in programming by switching the higher programming voltage on and off, over the supply voltage ie 3volt or 5volt ect.

Sil
   

Group: Professor
Hero Member
*****

Posts: 2746
Yes it does! if its depletion layer becomes reverse biased it will be cut off,
"Cut off" what ?
   

Sr. Member
****

Posts: 335
Enjoy your trek through life but leave no tracks
"Cut off" what ?
Reverse biased, turned off none conducting, rather like a veractor tuning diode.
But be aware if the depletion layer is stretched beyond the device breakdown limits it's likley the device
would be destroyed and become a S/C hazard.

So if it has say 20 volts reverse bias you would need to over come that voltage plus it's
depletion layer bias voltage to get it to pass your 20 volts or conduct your current through it.

Sil
« Last Edit: 2022-02-23, 06:20:40 by AlienGrey »
   

Group: Professor
Hero Member
*****

Posts: 2746
That is not an answer to my question.
You are talking about reverse breakdown voltage, which does not mean that diode conducts or blocks that voltage.
It only means that the diode conducts current above that voltage.  Conduction of voltage is a contradiction in terms.
   

Sr. Member
****

Posts: 335
Enjoy your trek through life but leave no tracks
That is not an answer to my question.
You are talking about reverse breakdown voltage, which does not mean that diode conducts or blocks that voltage.
It only means that the diode conducts current above that voltage.  Conduction of voltage is a contradiction in terms.
No that's not what i am saying at all. your getting confused here, no matter such is life.

What i am saying is unless your voltage is in the forward direction to over come the diodes depletion layer you wont
forward pass any current through the diode. Reverse breakdown voltage doesn't come into it here.

Sil
   

Group: Professor
Hero Member
*****

Posts: 2746
What i am saying is unless your voltage is in the forward direction to over come the diodes depletion layer you wont
forward pass any current through the diode. Reverse breakdown voltage doesn't come into it here.
The existence of the forward voltage knee still does not mean that the diode blocks voltage.

https://www.electronics-tutorials.ws/wp-content/uploads/2018/05/articles-iv2.gif
Itsu's workbench / placeholder.
   

Sr. Member
****

Posts: 335
Enjoy your trek through life but leave no tracks
The existence of the forward voltage knee still does not mean that the diode blocks voltage.

https://www.electronics-tutorials.ws/wp-content/uploads/2018/05/articles-iv2.gif
Itsu's workbench / placeholder.

Verpies , thanks for the graph, I think your referring to over voltage breakdown which can have destructive consequences.

However regarding Itsu's circuit diagram above It reminds me of a tunnel diode circuit with the tunnel diode in series with a tuning coil just like that circuit, however all the devices i ever experimented with had to biased and passably fed into driver, output transistor, however  some very fast switching can be achieved with such a device

If you wanted to experiment that type of device there was a lot of Russian devices of the market some time back possibly ex military.
Regards Sil
   
Group: Experimentalist
Hero Member
*****

Posts: 918
I may be a bit off topic... I took a look at aboveunity.com. Jagau shows us things that may seem surprising but remain conventional, the models giving the same results as his observations.
Moreover, one must be careful with his assertions, for example at the end of this page https://www.aboveunity.com/thread/capacitor-recharging/, he says "the oscillation that we see at the end is caused by the spread capacitance of the inductors" whereas a model of the device shows the same oscillation without any spread capacitance for the inductor, the oscillation comes from the classical LC circuit with discrete elements.
It is a pity that he romanticizes his speech with unrealistic hypotheses, but it is still interesting because his basic setups remind us of fundamental things that we ignore or forget. Wanting to model the device that we see here in this video of Chris pointed out by Jagau (these 2 people seem close to each other in their way of thinking, maybe the same?): https://youtu.be/-IE_UZtKr-I?t=2311, I didn't find anything particular, but by replacing the diodes by capacitors, I found a behavior that I had forgotten, that is that you can obtain a resonance by using a fictitious inductance.

In the first diagram (see attached file), each of the 2 inductances L1 and L2, strongly coupled to each other (coefficient 0.9), is coupled with a coefficient 0.5 to the inductance L0 connected to the generator. We see in the AC analysis the resonance around 13.5 KHz.

If, without changing anything in the setup, we create an imbalance of the coefficients of mutual inductance by taking 0.45 for L1/L0 while we keep 0.5 for L2/L0, we see that we obtain a second resonance.
The resonance around 13 KHz of each circuit is maintained, but there is in addition a resonance linked to the common circuit L1+L2, around 51 KHz. In fact L1 and L2 being in opposition, it is as if we had a lower inductance (up to theoretically zero, which means that in the 1st case, we do not see the second resonance). This is an old technique used for example for antenna tuning (see http://w5jgv.com/11.7uHy_Delta_Variometer/).

We therefore understand that it becomes very complex to analyze setups like that of Kapanadze, where the degree of coupling of each coil to each other will generate multiple resonance frequencies that are difficult to control in practice. But on the other hand, we see that with these coupling coefficients, we have a large degree of freedom to vary inductors, for example to produce a parametric device with a mechanical or permeability vibration which would vary the coefficient of mutual inductance, idea on which I am at the moment.


---------------------------
"Open your mind, but not like a trash bin"
   

Group: Moderator
Hero Member
*****

Posts: 3433
F6FLT, not much "off topic" items here, within limits of course.

Concerning  Chris and Jagau being the same, i don't think so.

Anyway,   when you mention:   "whereas a model of the device shows the same oscillation without any spread capacitance for the inductor" you point to your simulation in above post #809 here: https://www.overunityresearch.com/index.php?topic=3691.msg97054#msg97054.

So do you mean that the ringing at the end of your simulation would be the result of the L1 (10mH) and C1 (10uF) which according to this calculator: http://www.1728.org/resfreq.htm would be around 503Hz?

But if closely looking at your ringing frequency (and when i replicate your sim), i see a ringing frequency of ~215KHz.

This ringing frequency of 215KHz with an inductor of 10mH points to a capacitance of ~55pF.

Why is the ringing frequency so different as the LC in the circuit?
Could it be that the sim includes the diode and/or probe capacitance?

   
Group: Experimentalist
Hero Member
*****

Posts: 918
Good point, Itsu! In the simulation, the oscillation cannot come from the intrinsic capacitance of the coil because by default C=0 and I did not change this value.

But you are right, the frequency is much higher than what we would expect with 10µF, I should have seen it.  >:(

I thought it could be a question of the diode's capacitance because I used a real diode (1N5179). So I just redid the simulation with an ideal diode: no oscillation (see attached picture).

I added a 70 pF capacitor (value found in a datasheet of the 1N5819) in parallel on the perfect diode, and bingo, the oscillations came back!

But removing the capacitor from the perfect diode and putting it in parallel with the coil, the oscillation also occurs. It is therefore the set of parasitic capacitances around the inductance that contribute to the oscillation.

So I was wrong to attribute the oscillation to the pure LC circuit and Jagau is right but not 100% because of the forgotten capacitance of the diode which is also the cause of the oscillation.


...Concerning  Chris and Jagau being the same, i don't think so.

I think you're right, and that it was Chris' sometimes unserious comments in his video that made me think that Jagau, who I thought was the same person, was not completely reliable. So don't be influenced by appearances...   :(


---------------------------
"Open your mind, but not like a trash bin"
   
Pages: 1 ... 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 [33] 34 35 36 37 38
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2022-05-26, 05:27:21