OverUnity Research

Here i will attempt a replication of Tiger's Blocking Generator.

Itsu

i quickly put something together (yoke i had, MBR4045 dual diode, MJE13009, 600nF base cap, 2nF L2 cap, 1M base resistor, schottky diode and the coils like mentioned etc.) and it seem to be working right away when powered by 6 to 12V.

I do see that after removing power, it keeps on running for half a minute or so with a similar signal as in the video, then drops voltage to 0.5V and
then keeps on oscillating with some pulses for some additional 5 to 10 minutes (i have a 1000uF 50V electrolytic cap at the input).

I did reverse L1 leads as when in the circuit diagram.

But it does work as a blocking generator (joule thief) as expected, so some tuning and different components could further improve on it.

Itsu

Quote
Hello Itsu,

I looked up some Data:

For KT805 there are A, B and C-Types so there might be a difference

https://alltransistors.com/search.php?search=KT805


The Diode you choosed is a general rectifier, they are slow in the reverse recovery-time, but above all they have a
high instantaneous reverse current of 1 mA which gives proof of beeing slow,so you loose energy, should be in the range of 50 to 100 myko-amps.

I have found that it depends on the application- circuit which choice you make : either general recovery time ( myko-seconds) and low Forward-Voltage-drop or higher Forward-Voltage-drop >1 V and fast recovery time in the range of 20 to 50 nsec.

I had a flyback-recovery-circuit for a transformer-primary ( pulsed and the switched off)  and tested fast Diodes ( 50 ns , Vf 1,1 V) versus an AA143 Germanium Diode and the AA143 outperformed the differend fast diodes I had available.

Concerning the inductance:
pay attention to the reversal of magnetism losses but I do not have much experience in selecting the right core. His monitor-chocke-core might be low on these losses ( depends on the airgap and material)

Its a good start though, thank you Vasik  O0

Also look up the end paragraphs of my post #35 above about capture of flyback-pulses which might be of importance.


Mike

Hi Mike,  thanks for the reply.

I put something together without taking notice to much of what i got, so any NPN transistor.
I have some KT805AM on order and will start using them to get as close to the circuit as possible.

I have some Germanium Diodes which i could try instead of the Schottky's.

I have my doubts about the way the coils suppose to be wound, i think i have the L3/L4 wrong (like a push pull now), will try different methods.

Presently i use a TV yoke i had with a 0.1mm Mylar insulation inbetween the gaps.

Looking into your post #35 later today, and i also will make some screenshots of the signals (talking Collector signal yesterday) then.


Itsu

Here another document from Tiger, I think it is relevant to the topic

Thanks Vasik, that document could come in handy.


Mike, good information on paralleling a MLMC (MultiLayer Ceramic Capacitor) to the electrolytic ones to overcome the inertness to fast pulses.



Here a video on how my present setup functions.

I measure the collector of the transistor and the time it takes to show these signals after removing the 12V battery from the circuit.

The fist 20 seconds show a similar signal i think as shown by Tiger in his video.

Then the voltage suddenly drops from 12V to 2v or so and the pulse train becomes differently for some 6 minutes.

Video here:  https://youtu.be/cFdOazjelQs

Regards Itsu 

Hi Itsu
your video is in private navigation
right click on utube link and open
thank you also to Vasik for his many pdf, very interesting
well done Itsu thank you
Spark2

Quote from: Itsu on 2021.09.17, 20:53:06
Thanks Vasik, that document could come in handy.
Mike, good information on paralleling a MLMC (MultiLayer Ceramic Capacitor) to the electrolytic ones to overcome the inertness to fast pulses.
Here a video on how my present setup functions.
I measure the collector of the transistor and the time it takes to show these signals after removing the 12V battery from the circuit.
The fist 20 seconds show a similar signal i think as shown by Tiger in his video.
Then the voltage suddenly drops from 12V to 2v or so and the pulse train becomes differently for some 6 minutes.
Video here:  https://youtu.be/cFdOazjelQs

Regards Itsu

Nice work Itsu  O0
You can try add more capacitors in parallel (with smaller capacitance) after rectifier like we see in the video, this will reduce ESR and may improve circuit performance.

Regards,
Vasik

Just a historical note: this setup was quite popular before "Tesla coil rush" started
Many people experimented with it and had interesting results.
More similar schematics in the attachment.

-Vasik

QuoteHi Itsu
your video is in private navigation
right click on utube link and open
thank you also to Vasik for his many pdf, very interesting
well done Itsu thank you
Spark2

Hi Spark2,

my video's suppose to be "hidden", so everyone which has the link should be able to open/see it.
Checking on this latest one, its also "hidden" and works for me when clicked on.

Are you sure its says "private" to you?

QuoteNice work Itsu  O0
You can try add more capacitors in parallel (with smaller capacitance) after rectifier like we see in the video, this will reduce ESR and may improve circuit performance.

Regards,
Vasik

Vasik,


thanks for the info, i will add some MLCC's parallel to the rectifier cap and experiment with different transistors.


Those new diagrams remind me of the "joule thief" days where there were many threads all with different joule thief (blocking oscillator) circuits all able to light up leds (some hundreds) and with the intention to let is selfrun.

Itsu

Itsu and others can some one try  a torroide with a air gap in the loop, called the magnetic diode!
it's supposed to cancel #~~~Lenz law.

Why don't I try it ? I dont have a grinder  and I have no gash cores to experiment with and i dont play with let is,  :D
I mean is it worth a go or not ?

Regards Sil

Quote from: Itsu on 2021.09.17, 20:53:06
Thanks Vasik, that document could come in handy.


Mike, good information on paralleling a MLMC (MultiLayer Ceramic Capacitor) to the electrolytic ones to overcome the inertness to fast pulses.



Here a video on how my present setup functions.

I measure the collector of the transistor and the time it takes to show these signals after removing the 12V battery from the circuit.

The fist 20 seconds show a similar signal i think as shown by Tiger in his video.

Then the voltage suddenly drops from 12V to 2v or so and the pulse train becomes differently for some 6 minutes.

Video here:  https://youtu.be/cFdOazjelQs

Regards Itsu

Itsu,

How does the circuit behave with the single diode omitted from the circuit (clip lead across it, etc)?

PW

Quote from: Itsu on 2021.09.18, 07:33:04

Hi Spark2,

my video's suppose to be "hidden", so everyone which has the link should be able to open/see it.
Checking on this latest one, its also "hidden" and works for me when clicked on.

Are you sure its says "private" to you?


   

















Vasik,


thanks for the info, i will add some MLCC's parallel to the rectifier cap and experiment with different transistors.


Those new diagrams remind me of the "joule thief" days where there were many threads all with different joule thief (blocking oscillator) circuits all able to light up leds (some hundreds) and with the intention to let is selfrun.

Itsu

   Guys:  As far as I know, Tiger never had any OU devices, nor one that would self run. Even after all these years.
   Yes, like in the joule thief days. But, Joule Thiefs never could self run.  Wishful thinking at best...

   This current circuit is just running off of the input cap, (rectifier cap).  Take out the cap, or just replace it with a lower value, and you'll see that it's only there to fool you, without having any load on. What's the point?

   I once asked Nelson Rocha, if he actually had what appears to be a self runner.  He said no, it's actually just recycling the stored energy in the caps and such, for a while. But, it ran for much longer than one would expect, which, made me think...

   Just trying to let you know what I think, if it matters to anyone.  As I did spend some time working on this type of thing, but, to no avail.
  You can't get something from nothing. A joule thief type of circuit is not an energy conversion device, nor an open circuit, it may be more efficient, but that's it. Perhaps that's why we have wasted so much time, barking up the wrong tree. Just saying...

    NickZ

   P.S.  Itsu your last link to the video is opening up just fine for me on YT.
   

Quote from: picowatt on 2021.09.18, 14:55:07
Itsu,

How does the circuit behave with the single diode omitted from the circuit (clip lead across it, etc)?

PW

PW

I added a 0.1uF MLCC across the rectifier cap (1000uF / 50V).

Still using the MJE13009 transistor and signals as shown yesterday seems the same.

With the single diode omitted from the circuit (clip lead across it) the first stage (20s at 11V with single pulse plus ringing) is omitted as it directly seems to go to this tapering of pulse train signal at 2V decreasing to zero.

But there does seem to be a difference.


Itsu

Changing the transistor has a big impact on the signals and time the oscillations keep on going as expected.

A 2SC4632 was not changing a lot, a MJE3055 increased the time of oscillations well over 8 minutes and a present BF871 is still running after 13 minutes (stopped after 14 minutes) producing only the signal shown in the 1st 20s of the MJE13009 earlier (peak plus ringing).


But as Nick mentioned, they all slowly stop oscillating, but its fun to try to extend this dying out as long as possible.

Itsu

With this BF871 running at start, there is about 1V on the rectified side (anode) of the single schottky diode and a (decreasing) 11V on the plus (kathode) side.

This of course needs to be reversed for any selfrunning to occur.

Shorting out this diode only increases the loading from the input and shortening the oscillation times me thinks.

Itsu

Looking at the below diagram, i see the ringing frequency should be around 30 to 40KHz.

L2 together with C1 is the frequency determinating LC.

With C1 = 1.5nF, and a ringing frequency of 35KHz, then the L2 inductance (40 turns) should be 13.7mH.

My L2 (and L1) are 40 turns and measure 450uH, so my resonance frequency with C1 = 1.5nF is way higher 193KHz.

To get to 35KHz, i have to increase C1 to be 45nF

Where does this difference come from, the ferrite core used?

Itsu

Itsu,

do you have Germanium Diodes like AA143 available ?

Then replace all diodes with this type. AA143 has a peak reverse Voltage of 20 V so this will work.

I did some tests with energy-recovery from collapse-spike lately and these germanium-type outperformed all.
Much of the energy dissipates because of the forward-voltage-drop. Even power-schottky diodes never go below 0.7 Volt,
Some of the AA143 even go below 0.3 V

I also think that it would be worth while to use a germanium-Type transitor if you have saved some from the transistor-radios of the eighties.

Now here you can see one of the main influences concerning the ringing-frequency. He leaves the base open so you can see it clearly. The beat-frequency is just another story

https://www.youtube.com/watch?v=Mz4VYcuXL5E (https://www.youtube.com/watch?v=Mz4VYcuXL5E)

I documented this in 2011 and we are lucky it's still there.

There are more influencing factors which I forgot ( have to revisit my memory-palace :) ), but base-collector-capacitance strongly determines the ringing and its obvious why. Base-collector capacitance together with the collector coil is a LC-series
configuration. Also battery-capacitance must play a role in this setup.

Mike

Quote from: Itsu on 2021.09.20, 20:21:07
Looking at the below diagram, i see the ringing frequency should be around 30 to 40KHz.

L2 together with C1 is the frequency determinating LC.

With C1 = 1.5nF, and a ringing frequency of 35KHz, then the L2 inductance (40 turns) should be 13.7mH.

My L2 (and L1) are 40 turns and measure 450uH, so my resonance frequency with C1 = 1.5nF is way higher 193KHz.

To get to 35KHz, i have to increase C1 to be 45nF

Where does this difference come from, the ferrite core used?

Itsu

Itsu,

Do you have a gap in the core ?
If not, then it is probably different ferrite.

-Vasik

Kator,

i used some 1N34 Germanium diodes for both the single schottky diode as for the dual schottky's rectifier, but with the BF871 transistor which worked the best up till now (16 minutes) i have comparable running time, so no improvement using the Germanium diodes.

I have some AD161 transistors which i tried, but with the germanium diodes in the circuit, it did work, but only for 1.5 minute.

So although the various components do make some difference (mostly the transistor), i think the main activator is something else, probably the used ferrite in combination with a certain frequency.


Vasik,

i have 0.1mm mylar in between the both halves of the yoke.



Itsu   

Quote from: Itsu on 2021.09.21, 20:14:25
Vasik,

i have 0.1mm mylar in between the both halves of the yoke.

Itsu,

you will get significantly higher inductance if you remove/reduce gap between halves of the core.

-Vasik

Itsu it wants to be like a 'C' not 2 half sections O0

Sil

I have removed the (mylar) gaps and rewound the L1 and L2 coils to be 40 turns each around one whole circumference of the yoke measuring both 562uH @ 10KHz.

The L3 and L4 coils are 13 turns each and both measuring 66uH @ 10Khz.

Still with the 3 germanium diodes, the BF871 and C1 being 2nF it still runs for 16 minutes.

Below screenshot shows the following signals:

Yellow:          collector voltage (AC coupled!).
purple / blue:  Anodes L3 and L4 diodes.
green:           L2 (collector) current.



The delta 145.8KHz ringing frequency in the upper right box of the screenshot very closely match the calculated (http://www.1728.org/resfreq.htm) parallel resonance frequency (150KHz) of the used 562uH @ 2nF.



So, gap or no gap, L3/L4 more or less turns (10 v 13), germanium or schottky diodes, the thing oscillates for a similar time, so i conclude that those components are hardly influencing its behavior.

Itsu

I received my KT805AM transistors and used it in the above mentioned circuit.

The circuit was running for 14:30 minutes, so better then most transistors used, but not better as the BF871.


I will try to sweep the L2 coil to see if there is a sweet spot the ferrite responds to and use that frequency to get L2 and C1 in resonance.

Itsu

A 10s sweep from 1KHz to 300KHz using a single turn around the yoke does not show any specific peaks in this range when scoping across C1 / L2.

So finding any sweet-spot will be on trail and error base i think.

Itsu

Quote from: Mike on 2021.09.17, 08:15:17
For KT805 there are A, B and C-Types so there might be a difference
...
Mike

That is a good point even if the difference is only the package and not the silicon structure inside, the parasitics can be different between the transistor versions.

Quote from: Itsu on 2021.09.17, 20:53:06
The fist 20 seconds show a similar signal i think as shown by Tiger in his video.
Then the voltage suddenly drops from 12V to 2v or so and the pulse train becomes differently for some 6 minutes.

Video here:  https://youtu.be/cFdOazjelQs

It looks like in Tiger's video, the system never switches to the 2V operating mode but instead something weird happens at 3:10 (https://youtu.be/bcftGrBEaL0?t=184).

If I were to guess what is going on, I'd venture that the ferrite responds in an unusual manner when the circuit attempts to enter the 2V operating mode ...and reenergizes the system, effectively keeping it permanently above the border of the 2V operating mode. 
This is probably ferrite specific behavior (including its remanent magnetization) which is also dependent on internal and external/ambient magnetic flux densities (or their histories) that the ferrite was subjected to.

One other clue about matching his operating mode to yours more precisely is his >2Hz pulse repetition frequency (PRF) versus yours at 1.8Hz. 
I suggest that you measure the time between the pulses that his wireless mic picks up at 1:35 (https://youtu.be/bcftGrBEaL0?t=95) to obtain a more precise PRF and attempt to match yours to his by varying the capacitances, windings, air gap and the elephant in the room.

Yes, i can see the decrease in amplitude in my setup, but this "turnover point" where it bounces back to an increase is missing here.

According to an audio recording of the clicks, they are ~495ms apart (10.434 - 10.929s), see red circle in the below picture pointing to PRF of ~2.02Hz.


I will see if i can match that (not sure what it is now without the mylar gaps in between the yoke halfs).


Itsu

Quote from: Itsu on 2021.09.25, 08:38:26
Yes, i can see the decrease in amplitude in my setup, but this "turnover point" where it bounces back to an increase is missing here.

Well, at least you know what to look for.

Quote from: Itsu on 2021.09.25, 08:38:26
According to an audio recording of the clicks, they are ~495ms apart (10.434 - 10.929s), see red circle in the below picture pointing to PRF of ~2.02Hz.

Excellent. This is exactly what I had in mind.

Quote from: Itsu on 2021.09.25, 08:38:26
I will see if i can match that (not sure what it is now without the mylar gaps in between the yoke halfs).

I would expect the capacitance, resistance and inductance at the base of the BJT to have the most influence on the PRF. Could also be the turn ratio.

Some initial tests shows that the PRF is changing in time.

Right after start (12V) its around 4Hz, after a few minutes (5) it passes through 2Hz, stays several minutes (5) around 1.8Hz, then ends around 0.8Hz.

Will do some more tests later today.

Itsu

Quote from: Itsu on 2021.09.25, 15:44:35
Some initial tests shows that the PRF is changing in time.

Does the same input cap voltage always result in the same PRF ?

Quote from: Itsu on 2021.09.25, 15:44:35
Right after start (12V) its around 4Hz, after a few minutes (5) it passes through 2Hz, stays several minutes (5) around 1.8Hz, then ends around 0.8Hz.

Is the transition from the 12V to the 2V regime abrupt ?

Quote from: verpies on 2021.09.25, 16:44:01
Does the same input cap voltage always result in the same PRF ?

Yes it does look like it, when inputting 6V i always have a certain PRF, same with 8V, 10V etc., see below picture 2 graph of the Inp Voltage v PRF.

Quote
Is the transition from the 12V to the 2V regime abrupt ?

Well, this transition was only there with certain earlier transistors AND with a Lower C1 (like 2nF).
Now that i am using a C1 of 48nF (to lower the ringing frequency to 30KHz) and the BF871, BF869 and now the KT805AM transistors, i get longer runtimes and do not have this transition, they have the ringing signal all the way to the end.


Below picture 1 shows the PRF and Input Cap Voltage versus runtime (14 minutes).

I have 48nF as C1, the 3 Germanium diodes, the KT805AM transistor and a 600nF cap as L1 to ground cap.

Itsu

Nice analysis!

The PRF vs. voltage relationship is linear from 12-6 Volts and between 1-3 Volts the PRF is constant.

Generally, it seems that this circuit auto-sweeps the ferrite until it hits some special combination of the PRF and a much higher frequency component of the ringdown.

This clever single transistor circuit performs a dual parameter frequency sweep, that otherwise would require your FG + PC for SCPI control.

Itsu, you've mentioned that the different Transistors that you've used have
resulted in differing run-times of the unpowered circuit.

Have you determined what specifically about the various Transistors is the
reason for the differing run-times or "transition" capability?


Attached is the Data Sheet for the STN851-A which has some
very interesting characteristics.

Quote from: verpies on 2021.09.25, 22:46:21
Nice analysis!

The PRF vs. voltage relationship is linear from 12-6 Volts and between 1-3 Volts the PRF is constant.

Generally, it seems that this circuit auto-sweeps the ferrite until it hits some special combination of the PRF and a much higher frequency component of the ringdown.

This clever single transistor circuit performs a dual parameter frequency sweep, that otherwise would require your FG + PC for SCPI control.

Yes, but note that it takes only seconds to drop the voltage from 12 to 10 etc., so the first part of that "Inp Cap Voltage v Time" graph is more curved then shown.


Itsu

Quote from: muDped on 2021.09.26, 00:45:38
Itsu, you've mentioned that the different Transistors that you've used have
resulted in differing run-times of the unpowered circuit.

Have you determined what specifically about the various Transistors is the
reason for the differing run-times or "transition" capability?


Attached is the Data Sheet for the STN851-A which has some
very interesting characteristics.

muDped,

no i have not looked at the difference in used transistors, but Kator also mentioned earlier that some specific parameters are important.

The "transition" is mostly gone at a lower ringing frequency (30KHz) which i use now.

Thanks for the PDF on that transistor, any hints on what is so interesting with it?

Itsu

Hi Itsu,

Accidentally I found this picture.
It says that choke with inductance 10x of L3+L4 need to be added to the schematic.
Perhaps you would like to try this modification.

Regards,
Vasik

PS It says that C3 is 6x 2200uF 6V

Quote from: Itsu on 2021.09.26, 08:54:05
Yes, but note that it takes only seconds to drop the voltage from 12 to 10 etc., so the first part of that "Inp Cap Voltage v Time" graph is more curved then shown.

Acknowledged but this time does not change the shape of the  "PRF" vs. "Inp Cap Voltage" graph, ...which is timeless.

Quote from: Vasik041 on 2021.09.26, 12:20:17
Hi Itsu,

Accidentally I found this picture.
It says that choke with inductance 10x of L3+L4 need to be added to the schematic.
Perhaps you would like to try this modification.

Regards,
Vasik

PS It says that C3 is 6x 2200uF 6V

Vasik,

my L3 + L4 = 132uH, so that choke needs to be 1.32mH.

C3 would then be 6 x 2200uF = 13200uF (now having 1000uF).


I found a choke measuring 1.5mH which i will use and some 10000uF / 6.3V caps which i will use one for C3.


Itsu

This mod significantly extended the run time as expected from 14 min. to presently 40 min and still running.

I will make some more measurements later today, but the 1.8Hz PRF is there most of the time.

It does take somewhat longer to "load" to 12V from the battery.

Itsu 

800hz bursts filed with 30-40 kHz oscillations

Quote from: Itsu on 2021.09.26, 16:41:17

Quote from: Itsu on 2021.09.26, 15:43:35
C3 would then be 6 x 2200uF = 13200uF (now having 1000uF).

This mod significantly extended the run time as expected from 14 min. to presently 40 min and still running.

C3 is over 13x larger now, so that is not surprising.
However, the amplitude increase which is demonstrated in the video at 3:10 (https://youtu.be/bcftGrBEaL0?t=184) is what we really want to observe.

Quote from: Vasik041 on 2021.09.26, 12:20:17
Accidentally I found this picture.
It says that choke with inductance 10x of L3+L4 need to be added to the schematic.

Also, two additional annotations on this schematic state:

• L2 40 turns of double wire.
• L4 bifilar coil, the midpoint is its beginning and the end. (Verpies: The screenshot is cropped, so the entire one probably states: "L3+L4...")

...and I can't decipher the meaning of "800" in this one:

• f= 800 (30-40 kHz)

P.S.
For those who are cyrillically impaired, the word "Шотки" means "Schottky".

Quote from: Itsu on 2021.09.26, 16:41:17
This mod significantly extended the run time as expected from 14 min. to presently 40 min and still running.

I will make some more measurements later today, but the 1.8Hz PRF is there most of the time.

It does take somewhat longer to "load" to 12V from the battery.

Itsu

I rember that there were an opinion that gain in this circuit comes not from core, but from capacitors (e.g. due to dielectric absorbtion).
This might be interesting research topic.

-Vasik

Quote from: Vasik041 on 2021.09.26, 17:05:57
I rember that there were an opinion that gain in this circuit comes not from core, but from capacitors (e.g. due to dielectric absorption)

Then we should be seeing coreless circuits exhibiting the same behavior.

Dielectric absorption is a real effect that manifests itself as a slow voltage increase across a discharged capacitor AFTER THE LOAD ACROSS THAT CAPACITOR IS REMOVED.  This effect can be easily observed with a high-Z voltmeter.

@Itsu
To test for this possibility, the "PRF vs. Voltage" graph of this circuit would have to exhibit directional hysteresis. 
In other words: a different shape when the voltage is increasing compared to when the voltage is decreasing.

Quote from: Itsu
Thanks for the PDF on that transistor, any hints on what is so interesting with it?

What I find interesting about this new type of Transistor is its small size, its Collector Current Rating and its Low Voltage Switching Efficiency.

The Transistor is engineered to have High Current Gain and a Very Low Collector to Emitter Voltage Drop at Saturation.

Quote from: verpies on 2021.09.26, 16:46:20
Also, two additional annotations on this schematic state:

• L2 40 turns of double wire.
• L4 bifilar coil, the midpoint is its beginning and the end. (Verpies: The screenshot is cropped, so the entire one probably states: "L3+L4...")

...and I can't decipher the meaning of "800" in this one:

• f= 800 (30-40 kHz)

P.S.
For those who are cyrillically impaired, the word "Шотки" means "Schottky".

Concerning these annotations, i think they are the same as mentioned before in the below diagram from vasik:

L1 и L2 40 витков в два провода
L3 и L4 бифилляр средняя точка начало и конец

L1 and L2 40 turns in two wires
L3 and L4 bifillary middle point start and end     (google translate).



I was asking about L3 and L4 before as i do not see how they can be bifilar, so i have them "in series" now, with the midpoint the ending of L3 and
the beginning of L4 all CW.

But looking at the diagram from verpies, i can imagine how to wind them bifilar with the midpoint the ending of L3 and the beginning of L4.
I will have to rewind L3 and L4 to follow that.

Itsu

Quote from: Itsu on 2021.09.26, 18:52:22
Concerning these annotations, i think they are the same as mentioned before in the below diagram from vasik:
L1 и L2 40 витков в два провода
L3 и L4 бифилляр средняя точка начало и конец

Yes and notice that the cropped screenshot did not mention L1 consisting of 40 turns, too.
This would change my translations to L1 & L2 being be wound with two wires, 40 turns each.

Quote from: Itsu on 2021.09.26, 18:52:22
I was asking about L3 and L4 before as i do not see how they can be bifilar, so i have them "in series" now, with the midpoint the ending of L3 and the beginning of L4 all CW.

But looking at the diagram from verpies, i can imagine how to wind them bifilar with the midpoint the ending of L3 and the beginning of L4.

That's how I understand that annotation about winding the L3 & L4:
(https://www.overunityresearch.com/index.php?action=dlattach;topic=4181.0;attach=41570)

OK, i will rewind L3 and L4 that way tomorrow.


This evening, with the above modifications, it ran for 1 Hour 19 Minutes, starting at 12V, ending with 0.62V and with a PRF around 1.82Hz for ¾ of that time.

Voltage steadily decreasing over time, so no increase noticed.


Tomorrow i will show the Excel graphs.


Itsu

To prove/disprove the Dielectric Soak/Absorbtion hypothesis, force the supply voltage to slowly increase over time and note whether the shape of the "PRF vs. Voltage" graph is the same as when this voltage is decreasing.

One important thing: please note the capacitors voltage.
It's 6v but capacitors charged to 12v.
Over voltage cause  non-linear behavior in capacitors, similar to core saturation.
So when voltage decrease to nominal, capacity increases.
This could be source of "stabilization" effect when voltage decreases.

-Vasik

Quote from: Vasik041 on 2021.09.27, 04:58:01
It's 6v but capacitors charged to 12v.

Overvolting electrolytic caps outgasses their electrolyte and damages them permanently even if their dielectric is not punctured.

QuoteTo prove/disprove the Dielectric Soak/Absorbtion hypothesis, force the supply voltage to slowly increase over time and note whether the shape of the "PRF vs. Voltage" graph is the same as when this voltage is decreasing.

I am using a 12V battery as input now, but will see what a variable PS does as i manually increase the voltage.


Concerning the 6V cap as C3, here the 12V from the battery / PS does not "reach" it, doing a quick test i see it reaches 3.7V or so (depending how long i hold the 12V on the input) then decreasing.
Could be the L3 / L4 coils are not according to the specs (not bifilar), so will see if that increases after changing them.

I do have another 9 of those 10000uF / 6.3V caps in series (3) and parallel (3), so will have 10000uF @ 18.9V.


Itsu

Quote from: Itsu on 2021.09.26, 20:24:13
OK, i will rewind L3 and L4 that way tomorrow.


This evening, with the above modifications, it ran for 1 Hour 19 Minutes, starting at 12V, ending with 0.62V and with a PRF around 1.82Hz for ¾ of that time.

Voltage steadily decreasing over time, so no increase noticed.


Tomorrow i will show the Excel graphs.


Itsu

Excel data / graph from this test run yesterday:


Itsu

I increased the supply voltage from 2V to 4V in 0.1V steps, but the PRF stayed the same value as noted down yesterday (xls graph's above).


Itsu

Quote from: Itsu on 2021.09.27, 14:06:52
I increased the supply voltage from 2V to 4V in 0.1V steps, but the PRF stayed the same value as noted down yesterday (xls graph's above).

I was looking for a difference between two graphs like in the attachment below (i.e. hysteresis), that would have to be present for the amplitude to bounce up like in the Tiger's video.  The dielectric soak cannot be responsible for the amplitude bouncing up like that when such hysteresis is not present.  Without the hysteresis it can be only responsible for stabilization without bouncing up.

This reasoning hinges on the assumption the the peak amplitude of the ringdown is proportional to PRF.
To be certain, you'd have to directly measure the peak amplitude of the ringdown vs. supply voltage (in both directions).

The red graph is for the PRF when the voltage is decreasing (retyped from your latest spreadsheet, arrows hand drawn)
The blue graph is for the PRF when the voltage is increasing (all hand drawn - fabricated).

Ok, i think i understand, my below graph shows the 2V to 4V part of the data from the excel sheet from yesterday which is with a decreasing supply voltage.
This morning i used my variable PS to measure the PRF with an increasing supply voltage from 2V to 4V in 0.1V steps and the result was that the PRF data was the same as from yesterday.

So the graph lines (decrease and increase) would be the same also, so no hysteresis, see the graph below.

I can do that again (decrease and later-on a manual increase) for the whole 12V range, but i think again the PRF data would be the same.


In the mean time i have removed my L3 and L4 series coils and replaced them by L3 and L4 bifilar coils (18 turns each).
A run time test is running now (34 minutes and going).


What i did notice just now is that the supply voltage (12V momentarily from a battery) drops very quickly (seconds) to 5 / 6V then slower to end at 0.6V or so, but the ringing amplitude increases the first seconds.

I will try to capture that on video later today.

Itsu

Quote from: Itsu on 2021.09.27, 15:26:03
I can do that again (decrease and later-on a manual increase) for the whole 12V range, but i think again the PRF data would be the same.

I think we already have this data but you could do the Ringdown ampliude vs. Supply voltage measurement (in both directions) which constitutes a more direct proof of this hysteresis.

The below video shows the start of a cycle by touching the input cap with a 12V battery.

It shows the input cap voltage in blue and the collector signal (ringing) in yellow.

The blue input cap voltage can be seen dropping from 12V to a few volts in seconds, but it also shows the ringing amplitude signal increasing in those first few seconds, then stabilizing.


As you also can see in the below video is that capturing the ringing amplitude in data is not so easy as the PRF of 2Hz or so probably is to slow,
so getting the "Ringdown amplitude vs. Supply voltage measurement (in both directions)" will be hard.
Putting the scope in "normal mode" helps, so i will try to capture that data.

Video here: https://youtu.be/N41Z_BdcDq4

Itsu

Quote from: Itsu on 2021.09.27, 19:24:22
so getting the "Ringdown amplitude vs. Supply voltage measurement (in both directions)" will be hard.

I agree. You'd have to force the supply voltage with a variable PS in both directions and keep it there for the amplitude measurement.

When it comes to the "switching" action of the Transistor in this particular circuit, which is preferred:  Soft or Hard Switching?

Should the efficiency of the switching device be as high as possible or does it matter?

Have any switching characteristics been identified which may enhance circuit performance?

Is the switching Pulse Width an important consideration?

Quote from: muDped on 2021.09.27, 22:47:01
When it comes to the "switching" action of the Transistor in this particular circuit, which is preferred:  Soft or Hard Switching?

I think Hard Switching because a half-open transistor with current flowing through it dissipates a lot of energy as heat. Fully closed or fully open transistors dissipate the least of energy.  Thus half-open durations should be minimized.

Also, since the transistor feeds an inductance and current in an inductor cannot change quickly by definition, then pulses with short risetimes are smoothed out by the inductor and don't cause losses.  Short falltimes will generate high reverse voltages when no path for the current in the inductor is provided when the transistor turns off quickly and that can lead to the loss of energy which was stored in the inductor. A capacitor connected in parallel with the inductor can recycle this energy.

The change yesterday from L3 and L4 in series (13 turns each) to L3 and L4 bifilar (18 turns each) has resulted in a decrease in run time from 1 Hour 19 minutes to 1 Hour 6 minutes.

Itsu

With different number of turns it is not an apples-to-apples comparison.

Below graph shows the supply voltage versus ring-down voltage for both decrease (blue) as increase (orange) of the supply voltage.

They are very closely matched.

The supply voltage of 12V will cause the circuit to start oscillating, so the 15.1Vpp ring down voltage mentioned could be off.

Itsu

Looks monotonic. I expected some extrema based on your video.
Hysteresis looks minimal.

I don't think this data can explain the bounce back shown at 3:10 (https://youtu.be/bcftGrBEaL0?t=184).

Is Smudge here?

You've tried a lot of variations with some very interesting results.

I wonder, for the DC Feedback portion of the circuit;  have you considered configuring it as a Voltage Doubler to more efficiently extract energy from the oscillations?

such as this configuration:

(https://www.electrical4u.com/wp-content/uploads/What-is-Voltage-Doubler.png?ezimgfmt=rs%3Adevice%2Frscb37-1)

Would it make any difference at all?

muDped,

i was wondering about "the DC Feedback portion of the circuit" myself too.

It seems to me that the relation L2 to L3/L4 should be equal at least (now 40t v 10t) to have enough energy to keep on going.
I already increased L3/L4 to 18 turns, but that seemed to have not enough influence.

Your Voltage Doubler could be an option, but at first i would like to stay as close to the original circuit as possible.

Thnaks,   itsu 

Quote from: muDped on 2021.09.30, 02:41:44
I wonder, for the DC Feedback portion of the circuit;  have you considered configuring it as a Voltage Doubler to more efficiently extract energy from the oscillations?

I don't think so because that would entail capacitor-to-capacitor energy transfers which are inherently inefficient.

Capacitor-to-inductor (and vice-versa) energy transfers are very efficient and IMO that is the principle of the improvement suggested here (https://www.overunityresearch.com/index.php?topic=4181.msg94780#msg94780).

Also, that circuit probably has currents with two frequency components circulating through it (just like that other device (https://www.overunityresearch.com/index.php?topic=3926.msg93335#msg93335)).
I'd venture that the energy contained in only one of these components is the result of the gain mechanism and should be looped back to the input of the device. This would mean that a frequency-selective feedback circuit could be of benefit, too (as described in the 3^rd paragraph here (https://www.overunityresearch.com/index.php?topic=3926.msg93335#msg93335) and here (https://www.overunityresearch.com/index.php?topic=3926.msg93337#msg93337)).

Quote from: Itsu on 2021.09.30, 08:45:11
...but at first i would like to stay as close to the original circuit as possible.

So you should not omit the large area aluminum heatsinks, which can act as HF capacitor plates even if they are thermally superfluous.

Quote from: verpies on 2021.09.30, 12:16:26
So you should not omit the large area aluminum heatsinks, which can act as HF capacitor plates even if they are thermally superfluous.

Right, was thinking about putting back the schottky diodes and attach heatsinks to them and the transistor.

Itsu

Verpies,

I agree with your prognostication (https://www.overunityresearch.com/index.php?topic=4181.msg94847#msg94847) that it may not work as hoped.

I was principally wondering if anyone had given the concept any thought.

Often we are surprised by how things actually perform when put to the test.

I made some changes these last days like:

increase the choke from 1.5mH to 9mH
increase C3 from 10000uF to 15000uF to 20000uF.
decrease Cap in series with L1 from 600nF to 450nF
increase ringing frequency from 29KHz to 35KHz by decreasing C1
changed the rectifying germanium diodes to a dual schottky diode and put it on a heatsink.
installed a heatsink on the KT805 transistor.

These changes where done one at a time after which the circuit was monitored for a while.

But is amazing to see how little effect those changes have on the basic operation of this circuit.
Only the run time is influenced by the value of C3, and changing the ringing frequency makes little difference while the PRF stays around 2Hz.

Itsu

Looks like you have precisely replicated Tiger's circuit (including the winding layout).

At this point, the only parameter that I can suggest further experimentation with is the ringing frequency by varying the C1 and/or the air gap between ferrite's halves.

Yes, it seems C1 (ringing frequency) is a parameter which can be tweaked to sync f.i. with the ferrite resonance frequency



Concerning the below diagram, when putting 12V on the input cap C2 (right), it takes some time for this 12V to "reach" output cap C3.
When tapping only, it will not load C3 more then a few volts.
It will take some 10s or so for C3 to reach 6V.

This loading of C3 is done via inductive coupling from L2 to L3/L4 as the blocking diode in the plus lead prevents direct loading of C3.

So to see in Tigers video his voltage rise / stabilizing to 6/7V with only a short connection of the battery on C2 (is there a C2 there?) raises some eyebrows here.

A quick way to load C3 to above 6V is to connect the battery directly across C3 and let it run from there.


Itsu
   

I know you want to keep to the Tiger circuit, but why has he put the choke where it is.

Please see the attached

Regards

Mike 8)

Quote from: Centraflow on 2021.10.01, 17:33:59
I know you want to keep to the Tiger circuit, but why has he put the choke where it is.

This is kind'a elementary.
Tiger has put a choke there because cap-to-cap energy transfers are inherently inefficient.

For example, if you take a charged capacitor C₁ and close the switch S₁ in Fig.2 then the capacitor C₂^* will become charged to a higher voltage than in the circuit shown in Fig.1, which does not have the choke L₁.

The same is true when a pulsed voltage source is substituted for C₁ and S₁.


* ^{Initially discharged.}

Quote from: Itsu on 2021.10.01, 14:36:32
This loading of C3 is done via inductive coupling from L2 to L3/L4 as the blocking diode in the plus lead prevents direct loading of C3.
...
A quick way to load C3 to above 6V is to connect the battery directly across C3 and let it run from there.

I agree

Quote from: Itsu on 2021.10.01, 14:36:32
So to see in Tigers video his voltage rise / stabilizing to 6/7V with only a short connection of the battery on C2 (is there a C2 there?) raises some eyebrows here.

Yes, I can't see exactly what component he is charging with that battery, either.

Centralflow,

no, he placed the choke in the path to Ground-connection of L3/L4 rectifier,  thus avoiding higher frequencies shorting out to the ground...I assume

@Itsu: A year ago I did some intensive testing with Kacher-circuits. What you can do is to vary in small steps the resistor
           between collector and base of the transistor. For example the more you will decrease the value from 30 Kohm -> 28
           KOhm etc. I would expect an increase in PRF and change in Voltage of the ringdown and frequency. If you do this you
           may have to reduce the voltage input to C2 as the ON-time of the transistor increases in relation to the negative current
           from L1 to the base. The relation between these two competing currents is responsible for the PRF until it finallly is
           oscillating continually.
           Also the quality of C1 will play an important part in the Q of the parallell resonance L2/C1. I mostly use high-voltage
           ceramic capacitors from my "old-radio"-collection

Mike

Quote from: verpies on 2021.10.01, 21:48:13
This is kind'a elementary.
Tiger has put a choke there because cap-to-cap energy transfers are inherently inefficient.

For example, if you take a charged capacitor C₁ and close the switch S₁ in Fig.2 then the capacitor C₂^* will become charged to a higher voltage than in the circuit shown in Fig.1, which does not have the choke L₁.

The same is true when a pulsed voltage source is substituted for C₁ and S₁.


* ^{Initially discharged.}

Exactly, you are not teaching me anything I don't know. Look at what I have drawn and what Tiger has drawn, they are not the same, what I have drawn is tipical of a resonant parallel  converter.

Regards

Mike  8)

Quote from: Centraflow on 2021.10.02, 07:26:44
Look at what I have drawn and what Tiger has drawn, they are not the same, what I have drawn is tipical of a resonant parallel  converter.

Indeed the positions of the chokes differ but according to Kirchhoff's laws the positions depicted in Fig.3 and Fig.4 are equivalent.

Quote from: Kator01 on 2021.10.02, 00:50:30
Centralflow,

no, he placed the choke in the path to Ground-connection of L3/L4 rectifier,  thus avoiding higher frequencies shorting out to the ground...I assume

@Itsu: A year ago I did some intensive testing with Kacher-circuits. What you can do is to vary in small steps the resistor
           between collector and base of the transistor. For example the more you will decrease the value from 30 Kohm -> 28
           KOhm etc. I would expect an increase in PRF and change in Voltage of the ringdown and frequency. If you do this you
           may have to reduce the voltage input to C2 as the ON-time of the transistor increases in relation to the negative current
           from L1 to the base. The relation between these two competing currents is responsible for the PRF until it finallly is
           oscillating continually.
           Also the quality of C1 will play an important part in the Q of the parallell resonance L2/C1. I mostly use high-voltage
           ceramic capacitors from my "old-radio"-collection

Mike

Mike,

i am using a 1M resistor between base and collector right now as that suppose to "work better" then a 30K resistor.
But yes, i could replace it with a 1M or 50K trmmerpot and vary its value to see how it influence the circuit behaviour.

I tried severall types and value's of capacitors for C1 to change the ringing frequency, but will try to find and use some
high-voltage ceramic capacitors.  O0

Itsu

Quote from: verpies on 2021.10.02, 07:44:49
Indeed the positions of the chokes differ but according to Kirchhoff's laws the positions depicted in Fig.3 and Fig.4 are equivalent.

https://www.youtube.com/watch?v=UNmEayHrCJg&list=PL9_sR6Qqqcyn51rN4BaPhjckWWoMPvHJg&index=5

Regards

Mike 8)

Quote from: Centraflow on 2021.10.02, 09:43:36
https://www.youtube.com/watch?v=UNmEayHrCJg&list=PL9_sR6Qqqcyn51rN4BaPhjckWWoMPvHJg&index=5

...and how does that rebute my observation that circuits in Fig.3 (https://www.overunityresearch.com/index.php?action=dlattach;topic=4181.0;attach=41609)  & Fig.4 (https://www.overunityresearch.com/index.php?action=dlattach;topic=4181.0;attach=41609) are electrically equivalent ?

I did not rebut what you said, only it is not kirchoffs law.

Regards

Mike  8)

My observation was according to the Kirchhoffs's current law, because the algebraic sum of currents is commutative and thus the order of current-carrying components does not matter in a series circuit.
This is why it does not matter whether the choke is placed before or after the winding+rectifier block.

After changing the fixed 1M resistor for a 1M trimpot, i was able to vary the PRF with this pot.

In the below PDF, somewhere below the diagram there is this statement:
"The blocking frequency is about 800 hertz, the ringing frequency of the circuit is about 30 - 40kHz."

Does this mean that the PRF should be around 800Hz (mine presently is 2Hz also measured from audio from the Tiger video)?

If i tune the trimpot so the PRF reach 800Hz, the circuit starts to squeel, the voltage drops and the ringing almost disappears to only 1 peak (the trimpot reads 5K).
So there probably is more load on the transistor like Kator mentioned earlier (ON-time of the transistor increases).

Therefor the 800Hz as a PRF looks not realistic to me.

Itsu

Quote from: Itsu on 2021.10.02, 20:27:40
"The blocking frequency is about 800 hertz, the ringing frequency of the circuit is about 30 - 40kHz."
Does this mean that the PRF should be around 800Hz...

It seems that way based solely on the meaning of these words.

However a little further the PDF states:

Quote from: 001-lab.comI clarify the data: I measured the distance between pulses - 6 milliseconds,...

Which calculates to 167Hz PRF.


...but don't these 001-lab.com forum posts refer to a different circuit ?

Itsu,

what you describe is exactly what I was anticipating. The squealing is very well known among HAM-guys and old radio-engineers...: I had a good friend way back in the 80s who teached me my first lessons in oscillators. When we were building some high-frequency circuits ( 40 Megacycles) then at times there was this squealing and he said "oh no its peeping"
( in german : er piept ) meaning the 40 Megacycles were interrupted at audio-PRF, so the blocking frequency you entered at 800 Hz is in the audio-band which for some reason is so strong it can be heard.

I had one simple Kacher with a PNP-Transistor and a small 1200 V
cold-cathode-transformer with just one Base-Resistor of 1000 Ohm ( no Z-Diodes whatsoever). The high-voltage inductance
was connected to a 20 pf-capacitor ( ceramic, old radio) thus forming a Series-LC-tank with high Q
It runs on very low Voltage in pure sinus-oscillation. The blocking was not present as the Collector-current was dominating.
See Kacher-basisc-PNP.jpg

Another one does not have a base-Resistor at all but in the power-input-line.
See PNP-old-russian-Kacher.jpg.
This one I did not test.

I personally regard these basic circuits a good lesson to learn.

The russians have a good history with peeping-signals...or better damped waves
If smudge is around he certainly will remember as I do this:

https://www.youtube.com/watch?v=lfnfNe31fmY (https://www.youtube.com/watch?v=lfnfNe31fmY)

It had a tube-oscillator with just 1 Watt !! and was received around the globe at these hights !! ??...unbelievable.

https://en.wikipedia.org/wiki/Sputnik_1#cite_note-NKA-36 (https://en.wikipedia.org/wiki/Sputnik_1#cite_note-NKA-36)

The transmitter scheme :
https://www.radiomuseum.org/forumdata/users/15793/Sputnik1_xmtr.jpg (https://www.radiomuseum.org/forumdata/users/15793/Sputnik1_xmtr.jpg)

Sorry guys I dont like to side-track you,  but I think its important because we have to deal here with damped waves and they are not liked much by professionals.

I have to look up my archive...had recently found a picture of the signal....cant find it right now.

Mike

Quote from: verpies on 2021.10.02, 20:31:09
It seems that way based solely on the meaning of these words.

However a little further the PDF states:Which calculates to 167Hz PRF.


...but don't these 001-lab.com forum posts refer to a different circuit ?

Yes, the PDF jumps to another webpage i understand which is not available anymore.

But this statement: "The blocking frequency is about 800 hertz, the ringing frequency of the circuit is about 30 - 40kHz." is also on the diagram in that PDF and is what i am following.



Itsu

Quote from: Kator01 on 2021.10.03, 00:50:32
Itsu,

what you describe is exactly what I was anticipating. The squealing is very well known among HAM-guys and old radio-engineers...: I had a good friend way back in the 80s who teached me my first lessons in oscillators. When we were building some high-frequency circuits ( 40 Megacycles) then at times there was this squealing and he said "oh no its peeping"
( in german : er piept ) meaning the 40 Megacycles were interrupted at audio-PRF, so the blocking frequency you entered at 800 Hz is in the audio-band which for some reason is so strong it can be heard.

I had one simple Kacher with a PNP-Transistor and a small 1200 V
cold-cathode-transformer with just one Base-Resistor of 1000 Ohm ( no Z-Diodes whatsoever). The high-voltage inductance
was connected to a 20 pf-capacitor ( ceramic, old radio) thus forming a Series-LC-tank with high Q
It runs on very low Voltage in pure sinus-oscillation. The blocking was not present as the Collector-current was dominating.
See Kacher-basisc-PNP.jpg

Another one does not have a base-Resistor at all but in the power-input-line.
See PNP-old-russian-Kacher.jpg.
This one I did not test.

I personally regard these basic circuits a good lesson to learn.

The russians have a good history with peeping-signals...or better damped waves
If smudge is around he certainly will remember as I do this:

https://www.youtube.com/watch?v=lfnfNe31fmY (https://www.youtube.com/watch?v=lfnfNe31fmY)

It had a tube-oscillator with just 1 Watt !! and was received around the globe at these hights !! ??...unbelievable.

https://en.wikipedia.org/wiki/Sputnik_1#cite_note-NKA-36 (https://en.wikipedia.org/wiki/Sputnik_1#cite_note-NKA-36)

The transmitter scheme :
https://www.radiomuseum.org/forumdata/users/15793/Sputnik1_xmtr.jpg (https://www.radiomuseum.org/forumdata/users/15793/Sputnik1_xmtr.jpg)

Sorry guys I dont like to side-track you,  but I think its important because we have to deal here with damped waves and they are not liked much by professionals.

I have to look up my archive...had recently found a picture of the signal....cant find it right now.

Mike

Kator,

yes, the squealing is interesting as i wonder what it is that produces it.

But the 800Hz PRF is way to high in this configuration as it drains the stored energy in the caps very quickly.

Perhaps its meant to be 800mHz.

Thanks for the links and info, might come in handy.

Itsu

Itsu,

the 001- lab site is still in the archive..but only this page #58

https://web.archive.org/web/20110425160412/www.001-lab.com/001lab/index.php?topic=1056.2850 (https://web.archive.org/web/20110425160412/www.001-lab.com/001lab/index.php?topic=1056.2850)

...however no schematics

@vasik: can you help find out more ?

Also I found the original russian pic of the sputnik1 signal here:

http://sputnik.rusarchives.ru/dokumenty/forma-radiosignalov-isz-1 (http://sputnik.rusarchives.ru/dokumenty/forma-radiosignalov-isz-1)

It consisted of two alternating damped frequencies about 20 Megacycles and 40 Megacycles. However this is just a theoretical
signal-form. In reality - as one wave stops it fades in the manner we see in tigers presentation...so this exponential fading interferes with the start-sequence of following higher or lower puls...and so on.

Why did they do this ? What did they know then in 1957 ?  Just ONE watt from this obit reaching  almost all parts of the earth.


Mike

Kator,

thanks for the link, it does contain some information f.i. about the right most capacitors (C3).
Google Chrome does a good job translating.

Itsu

Itsu,

the problem is that when I try to go to previous pages < 58 or forward > 58 the archive returns a notice
that this page has not been archived. That was the reason I called out to Vasik if he can help with this because
I have the impression that sometimes the return of information might be restricted to the russian web
Seems we are stuck here for now.

One remark however I found on page 58 was that this recovering action of the damped wave at the beginning might have its cause in a base to emitter Zehner-performance. If this is the case than the chance to find a matching transistor is almost impossible. For example with negistor-circuits using different transistors I found that the same type of a transistor - however from a different manufacturer or a different batch makes on a negistor the other not.

Also I wonder what will happen if you use rf-Litz-wires ( 25 or more stands)  for L1 / L2 because these thick speaker-wires have low real resistance however  certainly a good portion of capacitance-coupling which might be detrimental to the performance, also the skin-effect might be negative.

Anyway, Itsu, you have done a very good job O0

Mike

Thanks Mike,

i did notice the problems with other pages of the archived website, besides page 58 there are few others that worked (are archived) like 1 and 62 or so, but no additional info was found there.

Anyway, i never noticed the "increase effect" like seen in Tiger his video, so probably there are some parameters which are not matched yet.

I do have some litz wire i could use, not sure if it worth the effort.

Thanks for your insights so far.

Itsu

Quote from: Kator01 on 2021.10.04, 00:21:58
Itsu,

the 001- lab site is still in the archive..but only this page #58

https://web.archive.org/web/20110425160412/www.001-lab.com/001lab/index.php?topic=1056.2850 (https://web.archive.org/web/20110425160412/www.001-lab.com/001lab/index.php?topic=1056.2850)

...however no schematics

@vasik: can you help find out more ?

Also I found the original russian pic of the sputnik1 signal here:

http://sputnik.rusarchives.ru/dokumenty/forma-radiosignalov-isz-1 (http://sputnik.rusarchives.ru/dokumenty/forma-radiosignalov-isz-1)

It consisted of two alternating damped frequencies about 20 Megacycles and 40 Megacycles. However this is just a theoretical
signal-form. In reality - as one wave stops it fades in the manner we see in tigers presentation...so this exponential fading interferes with the start-sequence of following higher or lower puls...and so on.

Why did they do this ? What did they know then in 1957 ?  Just ONE watt from this obit reaching  almost all parts of the earth.


Mike

I guess this is offtopic.

There is quite detailed explanation what and why it was done

https://translate.google.com/?sl=ru&tl=en&text=http%3A%2F%2Fwww.radioscanner.ru%2Finfo%2Farticle593%2F&op=translate

Even design documentation

http://russianspacesystems.ru/wp-content/uploads/2017/10/Otchet-o-razrabotke-bortovoy-radiostancii-pervogo-sputnika.pdf

Regards,
Vasik

I used to be very active on all satellites at one time, including Russian

http://www.arrl.org/news/russian-dosaaf-85-rs-44-amateur-radio-satellite-transponder-now-active

Regards

Mike 8)

Itsu,

can you please give data on your last, or best performing Tiger-device based on the scheme here in the
attachment. Just windings, capacitor and semiconductors and runtime ( as a final spec for other replicators ) ?
I certainly will give it a try...will take some time however.

Mike

Ok, will do some test runs.

But the run times greatly vary with the time you connect the 12V across C2 (it will load C3 through induction which determines the running time)
Longest times are when loading C3 directly with the 12V.


Itsu

   Yet, there will be a maximum run time, once the caps are all full.
   
   I still wonder how Nelson is able to get his rig self running for quite a while, even with a load on, with just a single one second tap on the input.
It doesn't make sense to me, nor add up as we know it. He said that it's just recycling the juice, but how can it be doing that with a load on.

   NickZ

Kator,

L1 and L2 are 39 turns 944uH @ 10KHz each (speaker wire, 1 circumference of the yoke)
L3 and L4 are 13 turns each bifilar connected (134uH each, total 537uH).
C1 is  13nF
C2 is  5000uF @ 12.6V
C3 is 15000uF @ 12.6V
C in series with L1 is 450nF
Choke is 9mH
Transistor is KT805AM
Single diode is 1N5819 (Schottky)
L3/L4 diodes is dual schottky MBR4045PT
Base / Collector resistor is 1M trimpot set at max.
PRF is 2.7s (after some time).
Ringdown frequency is 43KHz
Run time 5H 11Minutes (with C3 loaded to 12.3V) 

I tried the yoke without an air gap, with a 0.1mm air gap from mylar tape and this last run with 0.09mm printer paper.

I never saw that effect of increased ringing signal stabilizing around 6V as in Tiger his video.

Regards itsu

Itsu,

thank you for the data.
One thing catched my eye...see attached detail from your video.

This chocke is wound from thin rf-litzwire ( air-coil with an tuning ferrite) and certainly is from an LC-tuning-part of an old radio:
can you please measure the DC-resistance of this coil ?

I think that the dc-resistance is too high and responsible for additional thermal losses in the rectifier-section.

Mike

Good catch Kator,


yes, it came from a 1950's TV set, one of my first dismantle projects to get some parts.

I measured it to be 7.5 Ohm (1.5mH).

But i was looking for a bigger one later, so found a more modern 9mH choke which i was / am using in the last setup's, but it also measures 6 Ohm.

Itsu

Itsu,

yeah, 6 Ohm...too much. I have to look at my little stock to measure some with ferrite core.

Here they are:

The blue one is 2 x 3.3 mH, you  mainly find them in the 230V-Input-section of power-supplies.
Pretty good low resistance
The big one I bought for some other purpose than filter.

I apologize, but the orientation of the pics  looked ok before upload, but changed in the upload...some hidden information >:(

Mike

Quote from: Kator01 on 2021.10.08, 22:33:22
I apologize, but the orientation of the pics  looked ok before upload, but changed in the upload...some hidden information >:(

You can use this online tool (https://coding.tools/exif-remover) to remove the EXIF information from your photos.
...or another exif stripping tool that you install on your computer.

Below is a video of a very similar device to the one discussed in this thread.
https://youtu.be/9kTBbVyLzf0

How long will a 2200μF cap keep this LED lit ?

Verpies,

thank you for this hint., very helpful. On my old computer ( XP) which I do not use any longer for browsing the web.  I have a picture-viewer where you can see and modify the exif data.

Yes, G.Sav's Cicuit is a kacher in the blocking-mode, very interesting, however no info about this big core, He says it's from ebay :(
But I think the big coils in old computer-power-Supply might do the job also.

Did you read the comments ?


Mike

I did now.

Hello all,

I was doing research on a topic about the so called Romanov nod.
There is a collection of vids I found. The first one caught my attentions because of the circuit I saw on the initial still-picture

https://www.youtube.com/watch?v=3G42P-Hwz6s&list=PLkNjPIYiMgr9gy5mud0ZQRiQHrK_Y2SZ5 (https://www.youtube.com/watch?v=3G42P-Hwz6s&list=PLkNjPIYiMgr9gy5mud0ZQRiQHrK_Y2SZ5)

So it seems, Tiger has used this setup and changed it.

This guy Romanov is referred to by some other guys and it seems to me that he is some crucial figure these Kacher-based Long-Runners,  so I posted some of my findings in Vasiks thread about "different things"

Hope Vasik can help us with translation

@Itsu what do you think ?

Mike

Hi Kator,

thanks for the links here and in the "About different things" thread.

I know the "romancorp" guys have good video's etc., but like you i have problems understanding what is said, even with (or because?) google translate.

It looks similar as the Tiger setup, but with an extra (buffer?) part to the right.     Not sure if they use a yoke for the transformer.


Itsu   

Quote from: Kator01 on 2021.10.30, 23:16:16
Hello all,

I was doing research on a topic about the so called Romanov nod.
There is a collection of vids I found. The first one caught my attentions because of the circuit I saw on the initial still-picture

https://www.youtube.com/watch?v=3G42P-Hwz6s&list=PLkNjPIYiMgr9gy5mud0ZQRiQHrK_Y2SZ5 (https://www.youtube.com/watch?v=3G42P-Hwz6s&list=PLkNjPIYiMgr9gy5mud0ZQRiQHrK_Y2SZ5)

So it seems, Tiger has used this setup and changed it.

This guy Romanov is referred to by some other guys and it seems to me that he is some crucial figure these Kacher-based Long-Runners,  so I posted some of my findings in Vasiks thread about "different things"

Hope Vasik can help us with translation

@Itsu what do you think ?

Mike

As I wrote before, Romanov is very interesting case. He is running private courses and teach
people electronics and how to build OU devices probably for more than 10 years now. Many
persons who you see in Russian internet showing something OU are his students. He also discuss
many different OU setups and it seems that many of devices we see now are originates from Romanov.
Taking into account his military background and specialization in electronics (!) raise lots of disturbing questions (at least for me).

There is some useful information in his videos, but it buried under tons of fancy bla-bla-bla. I do not consider watching and even translating this "bla-bla" this as reasonable way to spend time. Sorry guys.

Regards,
Vasik

PS Romanov's kacher takes energy from radio waves as crystal radio. If you interested in such device I can share few much more reasonable devices.

  Vasik:
  As far as I know, Romanov has never shown a self runner, nor OU.
Same with Tiger. Tiger has not lied, but it seams to me that Romanov wants people to think that he knows more than he does.
However if you find some information showing a self runner, and not just bla bla bla, please let us know.

   NickZ

Hello Vasik

Quote
There is some useful information in his videos, but it buried under tons of fancy bla-bla-bla. I do not consider watching and even translating this "bla-bla" this as reasonable way to spend time. Sorry guys.

Regards,
Vasik

Understood...I fully agree. The problem is that I did not know anything about this guy. Also I noticed something hard to prove
concerning the quality of translation with google.
When I was - after a pause of 3 years - resuming my research in 2016 I was amazed about an increase of translation-quality and it was obvious that AI was optimized after these years.

Now it is my suspicion that AI has improved in selective downgrading the translation quality depending on the language, a new
stage of censoring.

i.e english to german ...very good
     russian to english or german ....very bad.

That is the reason why i asked for another alternative translation machine ( of course deepL - Translator but its not for free )

@vasik, no I do not think, crystal radio fits in here but its just my opinion.

Mike

Kator01

QuoteThis guy Romanov is referred to by some other guys and it seems to me that he is some crucial figure these Kacher-based Long-Runners,  so I posted some of my findings in Vasiks thread about "different things"

That's an interesting phrase, "long runners", with a neat history.

Many use batteries as a source however they have many drawbacks including trying to measure energy. If we try to use capacitors which are superior we run into issues with the input voltage dropping and effecting the switching/process. Then many creative FE inventors started to realize we could use an input capacitor and an output capacitor and loop the output back into the input... a circle with no beginning or end.

Once we start looping using capacitors energy calculations are super easy and based on the voltage, Energy = 1/2 C V^2. At this point we simply monitor the input/output voltage showing us energy and also the system run time. I like using these methods because it gives us quick feedback whether were moving in the right direction. Not wasting time with complex measurements/calculations or wondering what's actually going on. In this respect many people are fooling themselves pretending there's a gain using ambiguous or complex measurements.

I don't like wasting my time, I want to know exactly what's happening and whether there is a real gain or not every step of the way.

Regards
AC

Quote from: Kator01 on 2021.11.03, 21:20:19
Hello Vasik

Understood...I fully agree. The problem is that I did not know anything about this guy. Also I noticed something hard to prove
concerning the quality of translation with google.
When I was - after a pause of 3 years - resuming my research in 2016 I was amazed about an increase of translation-quality and it was obvious that AI was optimized after these years.

Now it is my suspicion that AI has improved in selective downgrading the translation quality depending on the language, a new
stage of censoring.

i.e english to german ...very good
     russian to english or german ....very bad.

That is the reason why i asked for another alternative translation machine ( of course deepL - Translator but its not for free )

@vasik, no I do not think, crystal radio fits in here but its just my opinion.

Mike

Mike,

there are some things you need keep in mind about translation:
google works quite ok nowadays, but if original text/speech is not clear... you get fuzzy translated text.
It depends on culture of speech of the person. Also Russian is way more complex than English, and also spoken language could be quite confusing, it developed during many years of oppression. Every word separately can be clear, but not whole meaning :)

You know old saying "soup from one can can't have different taste" ? Why would suddenly Russian troll disclose you some working FE device ?  ;D

Regards,
Vasik

Vasik,

QuoteYou know old saying "soup from one can can't have different taste" ? Why would suddenly Russian troll disclose you some working FE device ?  ;D

Haha..agreed. But that would presume that I have the ability similar to an X-ray like troll-sensor ;D

So I will learn in time...of course your help is very much appreciated

Mike