PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
2020-07-11, 09:08:19
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 2 3 4 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
Author Topic: Itsu's workbench / placeholder.  (Read 21539 times)

Group: Moderator
Hero Member
*****

Posts: 2232

Steve,   thanks.



Ag,

i don't think the MJE18008 transistor will change much, so something else could be not correct yet.

Its not that simple to measure "the back end of the output wave compaired to the osc wave" as the backend is "floating" compared to the input.

I did scoped the backend seperately and try to influence the sigs by adjusting the severall caps to get a nice resonance sine wave f.i.

Not sure yet what "the key" is there.

Itsu
   

Newbie
*

Posts: 23
Enjoy your trek through life but leave no tracks
 quick video here
https://youtu.be/J4e6gd7rIdw
AG

if I connect the multi lamp direct and get it to the same brightness it's not far off the same brightness  it's probably about 98% efficient
 ;D

« Last Edit: 2020-04-15, 18:34:41 by AlienGrey »
   
Group: Restricted
Hero Member
*

Posts: 688
love it - dogs , kids , noise, domestic strife and you still do  -- onward onward - you put me to shame


---------------------------
How many more to be .threatened, abused murdered, Their research in the hands of evil corporations intent on total control ?
http://dnp.s3.amazonaws.com/b/b9/suppressed.pdf
whilst we know little .. friends remember,
In the kingdom of the blind, the one-eyed man is king.
D. Erasmus
   

Group: Moderator
Hero Member
*****

Posts: 2232

Yes thanks AG,

could be this haystack approach is benificial for this type of circuit (spontane oscillations).




Finally got some oscillations in the sim like i have in the real circuit, see picture.

Had to keep it simple, so removed the swich and used the .IC attribute for C4 (ic=1.4).
Then had to lower the C1 value to 5nF to get a similar waveform on the emitter as in real life

Problem is that i had to remove the coupling to the output stage (L2 / L3).
Probably the load from that part is to heavy right now.

Frequency is close, 9Khz while the real circuit is at 12Khz.

Itsu
   

Group: Moderator
Hero Member
*****

Posts: 2232

I understand from Nelson that the effect we are after is the "negative resistance effect".

I did some work on negative resistance here:
https://www.overunityresearch.com/index.php?topic=2.msg36020#msg36020

and i think the 2n2222 transistor is one of the transistors mosly used due to its specifications.


So i think we need to find the correct setting for the used MJE18008 transistor to get it
oscillating in this negative resistance region.

I hope Gyula would be able to shed his knowledgeable light over this.

Regards Itsu
   
Newbie
*

Posts: 21
I understand from Nelson that the effect we are after is the "negative resistance effect".

I did some work on negative resistance here:
https://www.overunityresearch.com/index.php?topic=2.msg36020#msg36020

and i think the 2n2222 transistor is one of the transistors mosly used due to its specifications.


So i think we need to find the correct setting for the used MJE18008 transistor to get it
oscillating in this negative resistance region.


  Itsu:  I would hope that Nelson can also give a hand here.
  You mentioned "negative resistance region", are you basically saying that at that certain "region", the resistance in the circuit is reduced, compared to other "regions"? But, not that there is no resistance, at all? Just less resistance, and therefore higher output, and higher gain. Right? Just trying to rap this around my head. Yet, unless there is an output section in any circuit, there is no usefulness to it. Or, is that part up coming...
 

   

Group: Moderator
Hero Member
*****

Posts: 2232

Hi Nick,

please see the link i provided above from JLN labs:
http://bingofuel.online.fr/cnr/negosc.htm

It will explain some.


Itsu
   
Newbie
*

Posts: 21
   Itsu:
   Thanks, that was a more than just a handful of info.
   Isn't this negative resistance idea, something similar to what happens at resonance? Less resistance, thus higher output?
   Have you tried different frequencies, on the input from your SG?
   Perhaps the 3 or more of those series connected transistors work similar to Dr. Stiffler's diode loop idea? Not just allowing for higher input, but allowing a much higher output. Unfortunately on my tests, at the expense of higher draw from the input. Higher efficiency, but not free lunch, yet. Still the idea is for any circuit to be able to self run, itself.  Are we getting any closer...
   
   
Group: Tech Wizard
Hero Member
*****

Posts: 596
Hi Itsu,

I think when we build oscillators we need to differentiate active devices with negative resistance regions in their V-I characteristics (Esaki (tunnel) diode, Lambda diode, reverse biased p-n junctions) from the 'normal' active devices (transistor, FET) operated in a positive feedback circuit which is able to create negative resistance just by the positive feedback i.e. when there is no positive feedback then there is no negative resistance appearing between two pins of an active device like a bipolar transistor or a field effect transistor.

I think Nelson's oscillator is much closer to a positively feedbacked oscillator circuit (the transistor does not receive reverse bias to be able to exhibit negative resistance).

However, this oscillator needs a 'kick' to start, the switch when pushed provides the initial DC (positive for npn) bias for the base-emitter and oscillation could start if two further conditions are met.
One is that there should be a positive AC feedback maintained (this is likely provided by C3 in Nelson's circuit) and the other condition is the transistor should have an (initial DC) voltage gain > 1. This latter seems to be insured by the 6 Ohm bulb (when cold) in the collector and the 2.6 Ohm coil in the emitter (6 / 2.6 > 1). The bulb resistance then increases (hot resistance) as the circuit start working as you know this well.
So I think resistor R1 (i.e. the bulb) in the simulator should be inreased to say 47 Ohm to increase gain.

This is I could comment at the moment.  It would be good to know the purpose of this oscillator circuit and what is meant on "negative resistance effect" ?

Gyula
   

Group: Moderator
Hero Member
*****

Posts: 2232
   Itsu:
   Thanks, that was a more than just a handful of info.
   Isn't this negative resistance idea, something similar to what happens at resonance? Less resistance, thus higher output?
   Have you tried different frequencies, on the input from your SG?
   Perhaps the 3 or more of those series connected transistors work similar to Dr. Stiffler's diode loop idea? Not just allowing for higher input, but allowing a much higher output. Unfortunately on my tests, at the expense of higher draw from the input. Higher efficiency, but not free lunch, yet. Still the idea is for any circuit to be able to self run, itself.  Are we getting any closer...
 

Nick,

negative resistance is to be believed by many the opposite of loosing power namely: generating power.

So a negative resistor could/should be supplying power instead of consuming it.

The trick here is to use that small region of this negative resistance to indeed generate
power, preferable more then that it costs to create it.

It is in that sense comparable (but not similar) as resonance as there also the trick is to
use that small region of resonance to our advantage without disturbing it.

Nelson never claimed this specific circuit produces OU or so, but it makes sense for me to look
into it.

Itsu
   

Group: Moderator
Hero Member
*****

Posts: 2232
Hi Itsu,

I think when we build oscillators we need to differentiate active devices with negative resistance regions in their V-I characteristics (Esaki (tunnel) diode, Lambda diode, reverse biased p-n junctions) from the 'normal' active devices (transistor, FET) operated in a positive feedback circuit which is able to create negative resistance just by the positive feedback i.e. when there is no positive feedback then there is no negative resistance appearing between two pins of an active device like a bipolar transistor or a field effect transistor.

I think Nelson's oscillator is much closer to a positively feedbacked oscillator circuit (the transistor does not receive reverse bias to be able to exhibit negative resistance).

However, this oscillator needs a 'kick' to start, the switch when pushed provides the initial DC (positive for npn) bias for the base-emitter and oscillation could start if two further conditions are met.
One is that there should be a positive AC feedback maintained (this is likely provided by C3 in Nelson's circuit) and the other condition is the transistor should have an (initial DC) voltage gain > 1. This latter seems to be insured by the 6 Ohm bulb (when cold) in the collector and the 2.6 Ohm coil in the emitter (6 / 2.6 > 1). The bulb resistance then increases (hot resistance) as the circuit start working as you know this well.
So I think resistor R1 (i.e. the bulb) in the simulator should be inreased to say 47 Ohm to increase gain.

This is I could comment at the moment.  It would be good to know the purpose of this oscillator circuit and what is meant on "negative resistance effect" ?

Gyula

Hi Gyula,

i was hoping you would chime in,  thanks.

Well the negative resistance idea comes in mind due to fact that the current under certain conditions
in Nelson his setup (see his 3 video's) on the output is higher as the input, measured / shown
by the intensity of the bulbs in both the input and output.

I have not detected this effect yet, but have not installed the correct transistor (MJE18008).

Untill i have one i am optimizing my oscillation setup by trying to match the same conditions
(30mA input current) and the correct series resonance frequency on the coils.

My setup pulls between 70 and 100mA (MJE13009) at 24V and the output current is around 12mA max.
with about 18V across some leds as load.

Itsu
   

Group: Moderator
Hero Member
*****

Posts: 2232

Gyula,

some further comments, C3 (Nelson)= C1 in my diagram has SOME influence on the oscillating
frequency; at 100n its about 12Khz, at 5n its 14Khz, so a limited difference and not what
i would expect in a series circuit made up by C1 and L1, so there must be another frequency
determining component there.

Changing R1 in the sim from 6 to 48 or even 0 Ohm makes no difference in oscillations or frequency.


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

Posts: 596
Dear Itsu, 

IMHO, when negative resistance manifests in the V - I characteristic curve of a two terminal device by biasing it within a certain voltage - current range from an outside source like a battery,  (see Naudin http://bingofuel.online.fr/cnr/images/negosndr.gif ) then this device cannot "produce" higher output power than what its DC input bias provides. This is valid for both kinds of devices I mentioned in my previous post  i.e. for 2 terminal devices like tunnel and Lambda diodes, reverse biased pn junctions etc and for 3 terminal devices like bipolar and field effect transistors used in circuits where positive feedback is intentionally created by appropiate components to create negative resistance between two terminals of the 3 terminal device.

Regarding the brightness of the input and output bulbs:  In Nelson's videos, if the input bulb is the one labeled as E1 in his original schematic, then its brightness refers to how much power the bulb consumes, in series with the circuit, nothing else. This bulb acts in the circuit as the upper member of a voltage divider and the lower member of this divider is the rest of the circuit, ok? Power levels consumed are different for the two because the voltage levels are different across them while the current is the same.

I understand that the output current (38.8 mA) is higher than the input current (which is between 20 and 30 mA) as measured in video 2 for instance. The input voltage is 24 V as shown but I do not know the output voltage level across the bulb E2 if that is assigned for the output load.
Nelson mentioned in video 1 the bulbs are 12 V 1W rated bulbs with 6 Ohm cold resistance. This means that almost the full 24 V feeds the circuit (E1 bulb in series with the input remains dark because it has no enough voltage remaining from the rest of the circuit) and E2 output bulb may get roughly 8-10 V output voltage to have the brightness shown (it is a 12 V bulb).

I understand that you wish to see similar behaviour in your replicated circuit (for instance see the difference in the two currents).
It is possible that with MJE18008 transistor you get closer to those values. With oscillators it is a bit difficult to deal with to make them operate similarly.  Sorry that changing R1 in the simulation circuit did not have any effect (what I expected).

Regarding C3 in Nelson's circuit (C1 in yours) I still think it establishes positive feedback from collector to emitter (but only in case R1, the bulb in Nelson circuit has at least 6 Ohm or higher value) and as a feedback capacitor it surely influences operating frequency too but in a smaller extent than a real tuning capacitor would do say in parallel with the coil L1.

If you vary C4 10 nF in Nelson circuit, does it change frequency better than C3 (i.e. your C1) ? Try to vary it if you have not done it yet.

Regarding the roughly 3 times as high input current, does the 300 Ohm resistor in series with the base in Nelson circuit (you show a variable potmeter in the sim) influences input current much? If you put say a 470 Ohm potmeter in your real circuit, would input current change by varying the pot?

What is not ok in the sim is you had to remove the coupling factor between L2-L3: can power be transformed to the output with zero coupling? Or how the heavy load of R2, 10 Ohm (if you meant that) could appear for the oscillator with zero coupling for L2-L3? 

Regards
Gyula
   
Newbie
*

Posts: 21
   Guys:
   I've tried to look for Nelson's 3 videos about this device, but, it seams that he no longer has a channel up on YouTube.
   It also seams to me, that we are being guided by the bulb brightness, on those videos. Which may certainly relate to what Gyula just mentioned above. As well as the no load tests, performed so far.
   Yet, Nelson's later videos do show his device self running, with a feed back circuit attached. So I suppose that that is the purpose and idea of any lessons that may be learned from the negative resistance tests. Which he wanted us to see.
But, when we will get the full info of that self running device, which is what I've been waiting for.  Keep waiting Nick...
   Nelson, common guy, drop another nugget, or two. Please.
             
   
   

Group: Moderator
Hero Member
*****

Posts: 2232
Dear Itsu, 

IMHO, when negative resistance manifests in the V - I characteristic curve of a two terminal device by biasing it within a certain voltage - current range from an outside source like a battery,  (see Naudin http://bingofuel.online.fr/cnr/images/negosndr.gif ) then this device cannot "produce" higher output power than what its DC input bias provides. This is valid for both kinds of devices I mentioned in my previous post  i.e. for 2 terminal devices like tunnel and Lambda diodes, reverse biased pn junctions etc and for 3 terminal devices like bipolar and field effect transistors used in circuits where positive feedback is intentionally created by appropiate components to create negative resistance between two terminals of the 3 terminal device.

Regarding the brightness of the input and output bulbs:  In Nelson's videos, if the input bulb is the one labeled as E1 in his original schematic, then its brightness refers to how much power the bulb consumes, in series with the circuit, nothing else. This bulb acts in the circuit as the upper member of a voltage divider and the lower member of this divider is the rest of the circuit, ok? Power levels consumed are different for the two because the voltage levels are different across them while the current is the same.

I understand that the output current (38.8 mA) is higher than the input current (which is between 20 and 30 mA) as measured in video 2 for instance. The input voltage is 24 V as shown but I do not know the output voltage level across the bulb E2 if that is assigned for the output load.
Nelson mentioned in video 1 the bulbs are 12 V 1W rated bulbs with 6 Ohm cold resistance. This means that almost the full 24 V feeds the circuit (E1 bulb in series with the input remains dark because it has no enough voltage remaining from the rest of the circuit) and E2 output bulb may get roughly 8-10 V output voltage to have the brightness shown (it is a 12 V bulb).

I understand that you wish to see similar behaviour in your replicated circuit (for instance see the difference in the two currents).
It is possible that with MJE18008 transistor you get closer to those values. With oscillators it is a bit difficult to deal with to make them operate similarly.  Sorry that changing R1 in the simulation circuit did not have any effect (what I expected).

Regarding C3 in Nelson's circuit (C1 in yours) I still think it establishes positive feedback from collector to emitter (but only in case R1, the bulb in Nelson circuit has at least 6 Ohm or higher value) and as a feedback capacitor it surely influences operating frequency too but in a smaller extent than a real tuning capacitor would do say in parallel with the coil L1.

If you vary C4 10 nF in Nelson circuit, does it change frequency better than C3 (i.e. your C1) ? Try to vary it if you have not done it yet.

Regarding the roughly 3 times as high input current, does the 300 Ohm resistor in series with the base in Nelson circuit (you show a variable potmeter in the sim) influences input current much? If you put say a 470 Ohm potmeter in your real circuit, would input current change by varying the pot?

What is not ok in the sim is you had to remove the coupling factor between L2-L3: can power be transformed to the output with zero coupling? Or how the heavy load of R2, 10 Ohm (if you meant that) could appear for the oscillator with zero coupling for L2-L3? 

Regards
Gyula


Gyula,

an excelent post, you worded it basically as my thought about this, thanks.

The whole circuit "seems" to act like there is negative resistance somewhere based on the brightness of the bulbs only.
It could be the transistor (unlikely) or something else in the setup (delay line?).
But also Naudin stated that his circuit is NO OU.

The key effect seems to be the less brightness in the input bulb compared to the output bulb (when shorting the output) AND the dimming of the input bulb when shorting the output.

Neither of them i am able to confirm which is my goal here and then take some proper measurements.

C4 10nF has not much influence on the frequency either, even less then C1 (mine).

The pot (500 Ohm) has almost no influence on the input current, only at one end (the low Ohm end) it rapitly drops the input current until it stops oscillating.

In the sim when removing the coupling factor (or zeroing it) there is oscillation, but no sigs on the output.
I tried severall things like an array of coupling value's and all kind of loads, but it won't start oscillating then, could be i did not find the correct value's yet as my real life circuit proves it can oscillate.

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

Posts: 596
Hi Itsu,

Okay on the behaviour of the 500 Ohm potmeter,  at its low Ohm end it shunts the base-emitter bias voltage too low and the transistor fails operating then.  Thanks also for reflecting back on C4.

I think the negative resistance may manifest between two terminals (likely they are the collector and the emitter) of the transistor, due to the positive feedback caused mainly by C3 in Nelson's schematic and it influences the operating frequency too.

So there surely is a negative resistance occuring in the circuit which changes its resistance value dynamically as the load dictates within certain ranges. Neither Nelson nor Naudin claims OU with these circuits, I understand.

Because we do not know the inductance values in Nelson's circuit, the possibility of resonating his choke coil (T1, pins 2 and 3) with the 10 nF C5 capacitor at the oscillator operating frequency could be attempted.

Gyula

PS :

For those interested in the behaviour of negative resistance devices in oscillators, here is a good presentation on two terminal negative resistance devices whereby the device itself can be built from a P and an N channel JFET combination or from an N channel and a PNP transistor combination. see here:
 http://www.zen22142.zen.co.uk/Theory/neg_resistance/negres.htm    
I mean for instance Figure 5 where the 1 kOhm resistor could be replaced by a parallel LC circuit of quasi any L and C choices and one gets a simple and working oscillator.  And the change in  input current could be studied when the LC circuit is loaded, indicating the dynamic change of the negative resistance as the impedance of the LC circuit changes (within certain range).  Of course, Figure 8 shows an oscillator when the 1 kOhm regeneration potmeter is omitted or turned towards zero Ohm.
 
   

Group: Moderator
Hero Member
*****

Posts: 2232

Interesting link Gyula, thanks.


I reverse connected L2 compared to L3 to see if that makes any difference, but it does not.
Makes sense to me as the L3 output part is floating compared to the L2 input part.

As the oscillation frequency of the circuit (with or without L2/L3) is about 12Khz, i tried to make a series resonance circuit with L2/L3.

Both are 37mH, so to resonate at 12Khz, there must be a 4.7nF cap in series.
So i changed C5 and C6 (my diagram) to 4.7nf, but no improvement is seen nor the effect that the output bulb is brighter then the input bulb when shorting the output.

Guess i have to wait for my MJE18008 transitor to show up for any further testing.

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

Posts: 596
Itsu,  you wrote C6 (as per your diagram) which  is here I suppose:
https://www.overunityresearch.com/index.php?topic=3691.msg81058#msg81058
  and I see C6 in parallel with one of the full wave bridge diodes. It cannot change frequency at that place. Or you made a typo and meant another capacitor?
 
In fact I notice this parallel capacitor as C7, 50 nF in Nelson's schematic too, you included here:
https://www.overunityresearch.com/index.php?topic=3691.msg80986#msg80986
 I would like to understand what advantage is received from this capacitor?
I am aware of such capacitors in parallel with the individual diodes of full wave diode bridges used at the input AC side of switch mode power supplies but all the four diodes have identical value capacitors connected in parallel with them, not with a single diode out of the 4 only.
The reason for using the 4 caps across the 4 diodes of the bridge is to help reduce the quick switching signal products coming from the switch mode circuit (or from the 4 diodes themselves) and going towards the AC input, the 4 capacitors shunt most of these unwanted signals by forming a 'capacitor bridge'. 

Regarding the value of C5 in your schematic (in series with L2), I think several values would need to be tested because if we consider the possible other capacitors that may influence the value of C5 in the circuit,  we can see the top pin of C5 is connected to the top pin of C1, the bottom pin of C1 is connected to the emitter which is connected to C4 and C4 is then connected to the bottom of L2, ok? So the actual "tuning" capacitor is not only C5 but influenced by C1 and C4 in series with it (even if C4 is shunted by diode D1).

This may mean you would need to test several cap values for C5 in the some ten nF range if C1 and C5 are to be left unchanged for a while.

Gyula



   

Group: Moderator
Hero Member
*****

Posts: 2232
Gyula,

sorry about the confusion on the components in the original circuit from Nelson and my redraw, i should of have payed more attention and should have made them the same notation.

Talking my redraw notation now:
I regarded C6 the series cap of L3 and in fact it does influence the output signal of L3.
But the signal there is a mix of square wave, sine wave and high frequency ringing sigs, so hard to see if anything changes for the good.

If this C6 is really the series cap meant also by Nelson i do not know.
Perhaps he included it for a different purpose.
I did add extra caps in both series and parallel with L3, but no nice sine wave like signal (resonance) is seen up till now.

I also used many different caps for the major C's (C1, C4, C5 and C6), but the effect as described by Nelson in his video's did not manifest.

I will try some further tonight.

EDIT,   according to Nelson, my C6 is meant to dump HF noise from L3, so  not as series cap and it should be 4nF, so not 50.

Regards Itsu
« Last Edit: 2020-04-18, 21:35:08 by Itsu »
   

Group: Moderator
Hero Member
*****

Posts: 2232
I did a quick characterisation of my used 6V bulbs to indicate the resistance and voltage drop.

I can now see at what current there is a specific resistance and/or voltage drop.



I eddited the start of the data by starting at 0.1V


Itsu
« Last Edit: 2020-04-19, 18:19:39 by Itsu »
   
Group: Tech Wizard
Hero Member
*****

Posts: 596
Hi Itsu,

Okay on returning to the role of your C6. Notice though that in Nelson's original schematic C6 is 200 nF and is across the output of the diode bridge. And he labeled the "50 nF" capacitor as C7 which is in parallel with one of the diodes in the diode bridge, and this is your C6, right?

 No problem, just to clarify so that everybody following this thread should understand what labels we refer to.

It is good you got acquainted with the nonlinear V-I curve of your bulbs, will surely help.

Gyula
   

Group: Moderator
Hero Member
*****

Posts: 2232
OK,  see below the latest diagram using MY notations.
C6 is 4nF (corrected by Nelson) and is to "dump HF noise", so not to get series resonance at the output stage.

Other adjustments are the 13 Ohm (cold) bulbs R1 and R2

Important components are C1, C4 (oscillations), C5 (resonance), L1 (2.6 Ohm!) and L2/L3.

SIM still not running with coupling >0.


Itsu
   

Group: Moderator
Hero Member
*****

Posts: 2232
Finally got the sim doing something at the output stage.

Had to add a ground sign on the floating output part (below C8).
Doing the same on the real circuit makes no difference, so that point could be connected to the 24V return line.

But still only coupling factor 0.1 works to get some signals from the output stage.
Also C8 cap needs to be low, so i toke 6u instead of the 460u.

Dark blue is across L3,
green is across C7 and
light blue is across R4.

Itsu
   

Group: Moderator
Hero Member
*****

Posts: 2232
I put 2 of my 7 Ohm (cold) bulbs in series at the input to mimic the 13 Ohm (cold) bulb of
Nelson, but the input current dropped from 77 to 65 mA only.

Also the effect of decreased current while shorting the output was not seen.

A 2n2222 transistor (used in negative resistance oscillators) was not able to start oscillating
in my circuit, so i run out of options untill the MJE18008 arrives.


Itsu 
   

Group: Moderator
Hero Member
*****

Posts: 2232

I found in my junkbox some chokes similar to Nelson's L1 which measure 2.7mH / 1.6 Ohm and 9.8mH / 5 Ohm.

The latter one (9.8mH) drops the input current to about 9mA (oscillations around 4KHz) and first one (2.7mH), drops the input current to 24mA (oscillations around 7Khz).

I did some tests with this last one and it seems the input current decreases when loading the output cap, but still almost no output current (1-2 mA range).

Varying C1 from 100n to 3n makes the output voltage rise to over 250V, so had to watch for overloading the caps.

Anyway, a step closer to replication, but not the whole effect.

Itsu 
   
Pages: 1 2 3 4 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
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2020-07-11, 09:08:19