PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
2024-03-29, 04:55:26
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 ... 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 [40] 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 ... 101
Author Topic: Dally, Shark & Ruslan workbench  (Read 218436 times)

Group: Restricted
Hero Member
*

Posts: 1460
Enjoy your trek through life but leave no tracks
Pico hi, I more or less asked the active members of the group to do this and got nothing! I'm not one for messing about life is too short
and you guys need to beat rising sea levels before we all self destruct do to stupidity.
mine is built on the old Akula pcb it's a good 5 years old but modified into the Stalker circuit so it should be the same output. PULSE WIDTH.

but we need to know the 'pulse' and the 'width' time and the frame frequency!

Any way I will post what my test rig outputs are ok.

Also noter Delamorto post

give me a couple of hours.
notes on scope shots. Note i'm not using a self tune Tesla coil, my demo had a 35 volt bulb, I dont think i will be using this circuit it's rise and fall transisions are not good enough in my view. it's OK till we get to the Mos Fets then it all falls apart, well that what i have found.
Perhaps using a current sink driver using JT woukld be far faster or a compleat re design. Sorry guy's if you dont like the bottom line but thats what we find.

Regards Sil
   
Sr. Member
****

Posts: 422
PW,

i already have such a screenshot, see:  https://www.overunityresearch.com/index.php?topic=3926.msg91628#msg91628

but i will redo this tonight.

Itsu

Itsu,

Was the scope shot you directed me to above made with R17=3.3K and the recent base signal scope shot made with R17=8.2K?

If so, it appears increasing R17 has slowed the Q6 turn on time, which I suspect is due to R17 having to charge Q4's drain capacitance.

Also of note, if I am following the correct schematic, is that in both scope shots, very soon after Q6 is biased on, it is turned back off, and all the ringing that follows occurs with the Q6 base clamped to ground via Q4 and with Q5 turned off.

An expanded scope view of HO, LO, Q6 base and the collector coil "ringing" would allow us to see how many cycles Q6 is turned on during the initial start of the collector "ringing".

It appears Q6 is being biased on just long enough to start/trigger an LC ring down in the collector coil.

PW

 



   

Group: Elite Experimentalist
Hero Member
*****

Posts: 4056
PW,


you are right, the scope shot i directed you to was still taken with the 3.3K R17 resistor while the recent ones are with 8.2K.

Below screenshots ares with the 8.2K R17,  1st normal,  2nd expanded:

green: collector signal
purple: base signal
yellow: pin 7 IR2113 (LO)
blue: pin 1 IR213 (HO)

Trigger is on rising edge of LO.
12V on the collector   

Let me know if you want additional screenshots.

Itsu
   
Sr. Member
****

Posts: 422
PW,


you are right, the scope shot i directed you to was still taken with the 3.3K R17 resistor while the recent ones are with 8.2K.

Below screenshots ares with the 8.2K R17,  1st normal,  2nd expanded:

green: collector signal
purple: base signal
yellow: pin 7 IR2113 (LO)
blue: pin 1 IR213 (HO)

Trigger is on rising edge of LO.
12V on the collector   

Let me know if you want additional screenshots.

Itsu

Itsu,

Thanks for the scope shots.  Next time you post an expanded view, slow the horizontal down just a bit more so as to include the turn waveforms as well.

Regarding the older scope shot you referenced, have you adjusted RV7 differently (lengthened the pulse) between the old scope shot and your recent posts?

Observations:

Q6 turns on rather slowly due to R15's value and Q5's gate capacitace.  Q6 turns on even slower due to R17 having to charge Q4's drain capacitance.  As Q6 turns on, an oscillation in the collector coil is initiated, the frequency of which, I assume, is determined by the collector coils LC resonance and feedback from the antenna.  Q6 is turned on only very briefly, around a half cycle in the recent scope shots, and perhaps a cycle or two (of collector oscillations) in the older scope shot.  The setting of the pulse width using RV7 determines the number of collector oscillation cycles Q6 is allowed to be on before Q6 is turned off.

The ring down in the older scope shot looks much cleaner than your newer captures.  The newer captures show a clipped signal on the negative going portion of the ringing.  Perhaps RV7 needs to be adjusted so that Q6 is on just a tad longer (for another cycle or two).  Perhaps it is related to R17 having been increased to 8.2K.

I would, however, check for leakage on D1, Q4, and Q5 to ensure they were not damaged in your recent 100V excursion and perhaps the cause of the observed clipping.

I would also consider returning R17 to 3.3K.  For no longer than Q6 is turned on, dissipation at the higher current level should not be an issue.

As Vasik suggested, diodes can be used to protect the Q6 base from overvoltage.  A single diode, cathode to base, anode to ground, and an additional two diodes in a series string, anode to base and cathode to ground, will clamp the base between -.7 and +1.4.

Suggestions:

See if you can get the cleaner ringdown of the older scope shot before considering the use of clamping diodes.  First, check components for leakage/damage.  Second, increase the pulse width slightly using RV7.  Third, return R17 to 3.3K.

The ringdown is much cleaner in the older scope shot, I'd aim for that.

PW 

 
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 4056
PW,

Thanks for your insights.


Quote
Regarding the older scope shot you referenced, have you adjusted RV7 differently (lengthened the pulse) between the old scope shot and your recent posts?

I did change the RV7 to lengthen the base "on" time to get more power into the kacher which worked before changing to the 8.2K R17 resistor.

Quote
The ring down in the older scope shot looks much cleaner than your newer captures.  The newer captures show a clipped signal on the negative going portion of the ringing.  Perhaps RV7 needs to be adjusted so that Q6 is on just a tad longer (for another cycle or two).  Perhaps it is related to R17 having been increased to 8.2K.

The cleaner ring down on the old scopeshot is because that green signal was the signal picked up by the green probe laying near by the kacher secondary coil.
The present green signal is as mentioned the collector signal as that was what you requested i think.


Ok about the protection diodes and checkout of the 3 components, i will take a look at those.

Added screenshot with slowed down horizontal to include turn on waveforms


Itsu
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 4056
In below screenshot i copied the same setup (only difference is the 8.2K R17 instead of the 3.3K) as the "older scope shot".

So i put the frequency from 24.4KHz to 20.16Khz
Adjusted RV7 to have (somewhat) the same "on" time for the base signal
Put the green probe near by the kacher secondary.

What we see is that the base signal is cut off earlier, the output signal of the kacher is way lower (green probe at 1V/div instead of 10V/div).


Itsu
   
Sr. Member
****

Posts: 422
In below screenshot i copied the same setup (only difference is the 8.2K R17 instead of the 3.3K) as the "older scope shot".

So i put the frequency from 24.4KHz to 20.16Khz
Adjusted RV7 to have (somewhat) the same "on" time for the base signal
Put the green probe near by the kacher secondary.

What we see is that the base signal is cut off earlier, the output signal of the kacher is way lower (green probe at 1V/div instead of 10V/div).


Itsu

Itsu,

With R17=8.2K it takes longer to charge the drain capacitance of Q4 to the .6-.7V required to turn on Q6.  Therefore, to have the same amount of Q6 "on time" as the older scope shot with R17=3.3K,  RV7 has to be adjusted for a longer pulse width.  ("On time" meaning the amount of time the base of Q6 is at the required .6-.7 volts required to turn Q6 on, not the actual HO or LO pulse width.)

Is there any guidance from Vasik et. al. regarding how many collector oscillation cycles Q6 should remain on? 

Also, I'm still trying to understand why Stalker felt he needed the three gates worth of propagation delay at HIN.

PW

   

Group: Restricted
Hero Member
*

Posts: 1460
Enjoy your trek through life but leave no tracks
Itsu,

It's just 0.7V as supposed to be.
Those high peaks coming from collector through parasite capacitance.

Vasik

PS if you worry that there could be too much voltage you can put diode from base to ground.
Well none of this adds up to me.

1/. So why use power Mos Fets when some bipolar transisters can do the same job far faster in the VHF range ?
2/. How do we get the device to produce that dancing wave form in Verpies animation a few pages back with the fixed nodes ?
3/. How do we fix the nodes and modulate the rest of the wave Don't we need a special circuit for that phase shift ?

Regards Sil
   

Group: Professor
Hero Member
*****

Posts: 3354
1. So why use power Mos Fets when some bipolar transistors can do the same job far faster in the VHF range ?
I do not understand either, why Stalker simply didn't design something like this if he wanted a gated BJT amplifier:


...a circuit like this could also be built with the IR2110/3 driver.


Well none of this adds up to me.
According to my understanding some of it, too  ...but I do not know everything.
   

Group: Restricted
Hero Member
*

Posts: 1460
Enjoy your trek through life but leave no tracks
I do not understand either, why Stalker simply didn't design something like this if he wanted a gated BJT amplifier:


...a circuit like this could also be built with the IR2110/3 driver.

According to my understanding some of it, too  ...but I do not know everything.
That circuit is DC connected via the driver so it's doubtful the Tesla coil can self oscillate in that environment, a digital count of pulses would need gating with a zero crossing start and finish would need to be used. Complicated.
   
Group: Guest
Itsu,

I made a picture for you. Do you see difference ?

Vasik
   
Group: Guest
Well none of this adds up to me.

1/. So why use power Mos Fets when some bipolar transisters can do the same job far faster in the VHF range ?
2/. How do we get the device to produce that dancing wave form in Verpies animation a few pages back with the fixed nodes ?
3/. How do we fix the nodes and modulate the rest of the wave Don't we need a special circuit for that phase shift ?

Regards Sil

AG,

Answers to these questions were posted here already several times.
I don't see point repeating one more time.
"The more detailed documentation the less probability that somebody will read it" seems to be true  >:-)

If you don't like circuit - do not build it.

Vasik
   

Group: Professor
Hero Member
*****

Posts: 3354
That circuit is DC connected via the driver...
...and the Stalker's circuit is not ?
   

Group: Professor
Hero Member
*****

Posts: 3354
I made a picture for you. Do you see difference ?
In the compared scopeshot, when IR2110.pin1 (red) is high then the amplitude of the HF oscillations at the collector (yellow) is increasing.
In Itsu's scopeshot the HF oscillations (green) are absent when IR2110.pin1 (blue) is high.
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 4056
Itsu,

I made a picture for you. Do you see difference ?

Vasik

Vasik, be aware the my green signal is the collector signal and your yellow signal is the pickup signal from the kacher!

In your puicture i would like to see the base signal so see how that looks like.

 
I see the difference like verpies mentioned, and i too would expect that the green (or yellow in your picture) oscillations start to build up when the below purple (base) signal gets at a certain level.

I do see some change at the green signal at the red line, but its not enough to activate the transistor and cause oscillations to start building up.
Only when the base is clamped to ground by Q4 via IR2113 pin 7 the transistor gets activated all the way causing oscillations to start.


So how comes there is this difference?


Itsu
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 4056
Itsu,

With R17=8.2K it takes longer to charge the drain capacitance of Q4 to the .6-.7V required to turn on Q6.  Therefore, to have the same amount of Q6 "on time" as the older scope shot with R17=3.3K,  RV7 has to be adjusted for a longer pulse width.  ("On time" meaning the amount of time the base of Q6 is at the required .6-.7 volts required to turn Q6 on, not the actual HO or LO pulse width.)

Is there any guidance from Vasik et. al. regarding how many collector oscillation cycles Q6 should remain on? 

Also, I'm still trying to understand why Stalker felt he needed the three gates worth of propagation delay at HIN.

PW


PW,

sorry for the wrong answering sequence, but see my answer to Vasik his post here above, there you see in my screenshot that Q6 is not really turned on by the ".6-.7V required to turn on Q6"

Probably because a transistor is a current device and not a voltage device, so we would need to see the base current.

In the PDF's / video's translated by Vasik there is mentioned somewhere that there should be 5 or so collector pulses, i will try to look that up.

Why there is a  "three gates worth of propagation delay at HIN" i have no idea (could be inuntentional).

Itsu
 
   

Group: Restricted
Hero Member
*

Posts: 1460
Enjoy your trek through life but leave no tracks
AG,

Answers to these questions were posted here already several times.
I don't see point repeating one more time.
"The more detailed documentation the less probability that somebody will read it" seems to be true  >:-)

If you don't like circuit - do not build it.

Vasik
Vasik, Your 'made a picture for you' is interesting to say the least what is or are the width of both neg and pos transisions, i'm working in the dark here
please, you co operation is esential for me to learn, if one doesn't ask questions we dont learn.

regards Sil
   

Group: Restricted
Hero Member
*

Posts: 1460
Enjoy your trek through life but leave no tracks
...and the Stalker's circuit is not ?
Verpies, Yes if the driver is a solid signal held to either one rail of the other then it can not be influenced by deliberate self oscillation one would need to gate it digitally.

This later circuit does the job a bit better. Also the rise and fall transitions appear to be important.

Sil
   
Sr. Member
****

Posts: 422

PW,

sorry for the wrong answering sequence, but see my answer to Vasik his post here above, there you see in my screenshot that Q6 is not really turned on by the ".6-.7V required to turn on Q6"

Of course it is.  In all your scope shots, the base signal ramps up to and is clamped at .6V, which is Q6's Vbe forward voltage.  Once the base signal reaches .6V, current begins to flow between the base and emitter and Q6 is biased on.  Base current from that point onward is determined (mostly) by R17.  Once current flows through the base, Q6 is biased on and oscillation starts.


However, on another note, comparing your waveforms to the ones Vasik just posted makes me question the whether HO and LO signals are connected correctly (i.e., whether they should have been swapped as they were some time back)

PW
   

Group: Restricted
Hero Member
*

Posts: 1460
Enjoy your trek through life but leave no tracks

PW,

sorry for the wrong answering sequence, but see my answer to Vasik his post here above, there you see in my screenshot that Q6 is not really turned on by the ".6-.7V required to turn on Q6"

Probably because a transistor is a current device and not a voltage device, so we would need to see the base current.

In the PDF's / video's translated by Vasik there is mentioned somewhere that there should be 5 or so collector pulses, i will try to look that up.

Why there is a  "three gates worth of propagation delay at HIN" i have no idea (could be unintentional).

Itsu
I think it’s called propagation delay, and I think he needs to set the pulse timing for the Tesla coil to go loose so he says it is more efficient that way in one of his videos
if that’s of any use. So keep up the good work Its we all appreciate your work.

Regards 

Sil

   
Sr. Member
****

Posts: 422
Vasik, be aware the my green signal is the collector signal and your yellow signal is the pickup signal from the kacher!

In your puicture i would like to see the base signal so see how that looks like.

 
I see the difference like verpies mentioned, and i too would expect that the green (or yellow in your picture) oscillations start to build up when the below purple (base) signal gets at a certain level.

I do see some change at the green signal at the red line, but its not enough to activate the transistor and cause oscillations to start building up.
Only when the base is clamped to ground by Q4 via IR2113 pin 7 the transistor gets activated all the way causing oscillations to start.


So how comes there is this difference?


Itsu

Itsu,

Q6 turns on when the base signal flattens out at .6V.  However, as you point out, oscillations do not appear to start when Q6 is turned on, but rather just as Q6 is turned off.

Q6 turns on very slowly (as R17 "slowly" charges Q4's drain capacitance), possibly too slowly to initiate oscillations.  However, once the base signal reaches .6V, current flows in the collector coil.  The rapid turn off of Q6 (by Q4 clamping the base) appears to be shocking/exciting a simple LC ringdown (as opposed to the desired oscillator function?).   

Try going back to R17=3.3K to decrease the Q6 turn on time to see if that make Q6 turn on fast enough to start oscillation.
(which will also increase the base and collector current).

All that said, are we absolutely sure HO and LO are connected correctly?

I have also wondered about the three gates of propagation delay and wonder if they were used to ensure oscillation start up.

PW
   
Sr. Member
****

Posts: 422
Itsu,

If you adjust RV7 to increase the TC pulse width, does that allow sufficient time for Q6 to start oscillating on its own or do oscillations only appear as just a ring down at Q6 turn-off regardless of TC pulse width? 

PW
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 4056

PW,

i will try to answer your posts and do some additional tests later today / evening, but i think this last question could be answered by this post:
https://www.overunityresearch.com/index.php?topic=3926.msg92690#msg92690

Here i have replaced the collector coil by a 470K resistor fed by 24V and showing both the (increased) base signal in red and the collector signal in yellow.

But i will redo that "adjust RV7 to increase the TC pulse width" tonight as i am kind of lost with what i did when and how.

Concerning the correct connection of HO and LO i can only look to Vasik who mentioned earlier that the PCB was wrong in that regard and needed to be corrected.




Itsu
   
Sr. Member
****

Posts: 422
PW,

i will try to answer your posts and do some additional tests later today / evening, but i think this last question could be answered by this post:
https://www.overunityresearch.com/index.php?topic=3926.msg92690#msg92690

Here i have replaced the collector coil by a 470K resistor fed by 24V and showing both the (increased) base signal in red and the collector signal in yellow.

But i will redo that "adjust RV7 to increase the TC pulse width" tonight as i am kind of lost with what i did when and how.

Concerning the correct connection of HO and LO i can only look to Vasik who mentioned earlier that the PCB was wrong in that regard and needed to be corrected.


Q6

Itsu

In the post you referenced, all appears to me to be as expected.  Q6 begins to turn on as Vbe is approached and turns on fully as Vbe is reached and the set current flows through the Q6 base-emitter.

Anyone who has ever designed an oscillator knows that some measures must be taken to ensure oscillation startup.  Q6 is turning on very slowly.  Possibly too slowly to stimulate oscillation start up (without some external stimulus).

Even in your posted scope shots with R17=3.3K, it appears we are only seeing LC ringdown at Q6 turn off (unless increasing TC width allows Q6 to be on long enough to start oscillating).

Try tacking the 3.3K resistor across the 8.2K on the board.  This will increase base current to around 6ma and collector current to 500ma, but at the short on times, dissipation should not be a problem.  It will let us compare what speeding up Q6 a bit more does to the waveforms.

It may be necessary to continue decreasing R17 (speeding up Q6 turn on) until oscillations start on their own.  If that value of R17 is too low (i.e., more base/collector current than desired), R18 can be decreased to shunt additional current away from the base.

I've been studying the schematic with HO and LO connected both ways, and it probably makes more sense the way they are currently connected. 

PW
   
Group: Guest
AG,

Quote
please, you co operation is esential for me to learn, if one doesn't ask questions we don't learn

Ok.
You posted scope traces of IR2110 outputs.
I think they are inverted.
To say for sure I need see TL494 output on same screen.
Kacher enabled when Q4 is off and Q5 is on.

Vasik
   
Pages: 1 ... 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 [40] 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 ... 101
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2024-03-29, 04:55:26