PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
OverUnity Research > Benches > AlienGrey > Dally, Shark & Ruslan workbench
2024-11-26, 07:01:24
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 ... 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 [26] 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 ... 101
Author Topic: Dally, Shark & Ruslan workbench  (Read 309764 times)
Topic Options img
img  Itsu
Re: Dally, Shark & Ruslan workbench « Reply #625 on: 2021-05-25, 08:29:26 »

Group: Elite Experimentalist
Hero Member
*****

Posts: 4159
Quote from: verpies on 2021-05-24, 22:15:56
...or increase the load on the secondaries so the magnetic flux energy, which is stored in the core, has other places to go rather than breaking down the D-S junction of the MOSFET, after the gate goes low.

Good tip, i can try to increase the load from the now 40W bulb to 1KW of bulbs.

Itsu
   
img  Itsu
Re: Dally, Shark & Ruslan workbench « Reply #626 on: 2021-05-25, 08:36:21 »

Group: Elite Experimentalist
Hero Member
*****

Posts: 4159
Quote from: partzman on 2021-05-24, 22:32:09
Yes, this will work if the coupling is tight between the secondary and primaries.  However, my guess is that the coupling with the yolk core is not that great so even with the secondary shorted, there would probably still be evidence of avalanche.

I'm not that familiar with the overall operation so I'll ask, is the output of the squarewave of the yolk transformer converted to sine somewhere before being used?

Pm

PM, 

yes,   the 5 turn secondary drives a series resonance LC circuit running at 122KHz (5th harmonic) and the 18 turn secondary forms a parallel LC with some caps running at resonance on 24KHz driving the output.

So we have sine wave signals on both secondaries.

Itsu
   
img  Itsu
Re: Dally, Shark & Ruslan workbench « Reply #627 on: 2021-05-25, 08:47:32 »

Group: Elite Experimentalist
Hero Member
*****

Posts: 4159
Quote from: Vasik041 on 2021-05-25, 05:45:59
Itsu,

What is voltage on inductor ?
My understanding is that it should be around 200v
and same should be output of gradient coil.
If voltage too high you might need remove some turns from yoke secondary windings.

I would suggest tuning inductor and gradient coil separately.
When you come close to 24.4Khz you should see HF harmonics, this will help you with precise tuning (look for highest HF signal).

Vasik


Vasik,

in post #589   (https://www.overunityresearch.com/index.php?topic=3926.msg92214#msg92214) i measured with 2 probes (differential method) the voltage across the inductor (only connected), and it showed a sine wave of 163Vpp.

This points to the problems we have in measureing the voltage on the inductor and grenade as both are floating compared to ground, so care should be taken to measure these voltages.

Any idea how Stalker measured those voltages?



Yes, i will try to match the inductor series LC to resonante at 122Khz (now 113KHz) by lowering the series caps value.
Then tune the grenade capacitance to resonate on the base 24KHz.

Itsu
   
img  Vasik041
Re: Dally, Shark & Ruslan workbench « Reply #628 on: 2021-05-25, 09:10:39 »
Group: Guest
Quote from: Itsu on 2021-05-25, 08:47:32
Any idea how Stalker measured those voltages?
Itsu,

163V I think is ok, probably you need unwind few turns from 18 turns coil.
I think voltages just for reference, you can check it with some digital meter (which can handle 100Khz frequency).
As I understand it is more important to look on the currents.
Stalker use current transformers for that.

BTW you can try connect bigger load, it also could drop power consumption.

Vasik

PS there are same profile but higher heatsinks, you try those to reduce heat issue
   
img  Vasik041
Re: Dally, Shark & Ruslan workbench « Reply #629 on: 2021-05-25, 09:28:53 »
Group: Guest
Itsu,

Could you please confirm

Quote from: Itsu on 2021-05-25, 08:36:21
and the 18 turn secondary forms a parallel LC with some caps running at resonance on 24KHz driving the output.

Why it is parallel ?
My understanding of schematic is that it all connected in series (18 turns, gradient coil, capacitor across output)

Vasik
   
img  Itsu
Re: Dally, Shark & Ruslan workbench « Reply #630 on: 2021-05-25, 10:36:46 »

Group: Elite Experimentalist
Hero Member
*****

Posts: 4159
Vasik

Yeah, sorry about that, initially i thought it was a parallel LC circuit, but its not, but somehow it is stuck in my mind

Itsu
   
img  Vasik041
Re: Dally, Shark & Ruslan workbench « Reply #631 on: 2021-05-25, 13:42:36 »
Group: Guest
Quote from: Itsu on 2021-05-25, 10:36:46
Vasik

Yeah, sorry about that, initially i thought it was a parallel LC circuit, but its not, but somehow it is stuck in my mind

Itsu

NP, but did you connect it correctly ?  :)
   
img  partzman
Re: Dally, Shark & Ruslan workbench « Reply #632 on: 2021-05-25, 14:39:52 »
Group: Experimentalist
Hero Member
*****

Posts: 1808
Quote from: verpies on 2021-05-24, 23:14:37
You can estimate the coupling coefficient by the fact that the inductance of his primary is 42μH when the secondaries are open and 4μH when the secondaries are shorted.

Yes, in this case using k = (1-Lpss/Lp)^.5 = .95 where Lp is the primary inductance with secondary open and Lpss is the primary inductance with secondary shorted.  This is higher than I would have anticipated from experience!

Edit:  The leakage inductance referenced from the primary would then be Lpleak = (1-k)*Lp = 2.05uH .

Pm
   
img  NickZ
Re: Dally, Shark & Ruslan workbench « Reply #633 on: 2021-05-25, 15:04:35 »
Group: Guest
Quote from: Itsu on 2021-05-25, 08:23:11
Nick,

those MOSFETs get hot quickly then, like within half a minute they are 70°C and rising, but look at the drain signals, they avalange? now twice/cycle.
Need to tune (frequency / resonance / duty cycle / number of turns) to see if i can improve.

A perfect match between the MOSFETs will be very hard to get i think.

Itsu


   Ok, thanks.  That is what I thought would happen. As I have the same effect, at or close to resonance. However, I don't recall seeing the avalange type of signal from my device. But, perhaps you are right about that being a needed effect on these builds.
    I can tune to reduce the heating, but it also drops the output. Perhaps a perfect match of the Fets, is not even really needed. But, I always have one fet hotter than the other one.
    I would suggest swapping the fets, and or replacing them with new ones to see if there is any difference. I normally find that there is some difference. As once the fets get too hot and cook, they are useless, even if they still work, just like Verpies says.
   
img  NickZ
Re: Dally, Shark & Ruslan workbench « Reply #634 on: 2021-05-25, 15:22:46 »
Group: Guest
Quote from: Vasik041 on 2021-05-25, 13:42:36
NP, but did you connect it correctly ?  :)


   Vasik: We have previously heard that the 3 turn coil/inductor circuit, is a series type of circuit. And that the grenade output coil is on a parallel type of circuit. This is the first time that I've heard someone mentioning what seams right to me, as well. That the output grenade coil circuit is also a series circuit. But, does it really matter, in this case?
   I don't understand why the inductor needs to be tuned so high, at 122KHz? Grenade coil at 24KHz, I can accept, but the inductor circuit running that high? And why the 5th harmonic? Is that the magic harmonic frequency?
Any way, we'll see what happens there. 

   itsu:  Please show your device lighting your big bulbs, or at least some bulbs over 300 to 500w worth of load. This, once you reach the proper resonance of both circuits, of course. Thanks, and good luck with the heat issue. Remember to keep your finger on the hot fet. My fets can get untouchable in less than a minute. And ready for the trash in less than 5 minutes run time. But, that can also depend on the running frequency, sometimes they can run nice and cool, also.

   BTW: Guys there are some very very inexpensive TL494 PCBs out on the market now, that would work on this device. They do have control able frequency and duty cycle. You'd still need to connect those little boards up the fets, but it would save a lot of time and trouble, and also, the variable DC to DC booster circuit are available to drop or raise the circuit voltages, and are also very cheap as well. But, they can all take some time to get, like sometimes a month, or so.
               
    NickZ
« Last Edit: 2021-05-25, 18:44:48 by NickZ »
   
img  Vasik041
Re: Dally, Shark & Ruslan workbench « Reply #635 on: 2021-05-25, 15:52:54 »
Group: Guest
Quote from: NickZ on 2021-05-25, 15:22:46
   Vasik: We have previously heard that the 3 turn coil/inductor circuit, is a series type of circuit. And that the grenade output coil is on a parallel type of circuit. This is the first time that I've heard someone mentioning what seams right to me, as well. That the output grenade coil circuit is also a series circuit. But, does it really matter, in this case?
   I don't understand why the inductor needs to be tuned so high, at 122KHz? Grenade coil at 24KHz, I can accept, but the inductor circuit running that high? And why the 5th harmonic? Is that the magic harmonic frequency?
Any way, we'll see what happens there. 

Nick,

here are two points as I understand from Sergey's explanations:

- push pull NOT supposed to drive load directly, it is a modulator
- inductor LC tuned to high frequency to allow some "free space" for Tesla coil action (when both MOSFETs are off). 

Vasik

PS If this does not work at all we can try Alexeev's tuning (inductor 24.4KHz, gradient coil 12.2Khz)
   
img  Itsu
Re: Dally, Shark & Ruslan workbench « Reply #636 on: 2021-05-25, 15:54:41 »

Group: Elite Experimentalist
Hero Member
*****

Posts: 4159
Quote from: Vasik041 on 2021-05-25, 13:42:36
NP, but did you connect it correctly ?  :)

Yes,  connection is like in this overview:

Itsu
   
img  Itsu
Re: Dally, Shark & Ruslan workbench « Reply #637 on: 2021-05-25, 16:01:36 »

Group: Elite Experimentalist
Hero Member
*****

Posts: 4159
Quote from: NickZ on 2021-05-25, 15:22:46

   Vasik: We have previously heard that the 3 turn coil/inductor circuit, is a series type of circuit. And that the grenade output coil is on a parallel type of circuit. This is the first time that I've heard someone mentioning what seams right to me, as well. That the output grenade coil circuit is also a series circuit. But, does it really matter, in this case?
   I don't understand why the inductor needs to be tuned so high, at 122KHz? Grenade coil at 24KHz, I can accept, but the inductor circuit running that high? And why the 5th harmonic? Is that the magic harmonic frequency?
Any way, we'll see what happens there. 

   itsu:  Please show your device lighting your big bulbs, or at least some bulbs over 300 to 500w worth of load. This, once you reach the proper resonance of both circuits, or course.
   Thanks, and good luck with the heat issue. Remember to keep your finger on the hot fet. Mine can get untouchable in less than a minute. And ready for the trash in less than 5 minutes run time. But, that can also depend on the running frequency, sometimes they can run nice and cool, also.
               
    NickZ

Nick, 

one 100W bulb can be lit fine (200V), see next post video, but adding more lowers the voltage and the bulbs get dimmer and dimmer.
With 4 x 100W, there is only 52V DC left.

Concerning the avalanche effect, if needed or not, please look at rk_rev2.pdf at page 39  and  this video at the 9:30 minute mark: https://youtu.be/XDXTxPRPxSU?t=577


Itsu
   
img  Itsu
Re: Dally, Shark & Ruslan workbench « Reply #638 on: 2021-05-25, 16:03:05 »

Group: Elite Experimentalist
Hero Member
*****

Posts: 4159

I added a little fan above the 2 MOSFETs to cool them.
I changed the 40W output bulb to a stack (10) of 100W bulbs, but only have 1 such bulb active initially.
I removed the ground connection to be able to measure the Grenade/Inductor voltages (PS used is isolated from ground).
Then tuned the Grenade to max voltage/current which turns out to be around 21Khz, so also a little to low for the target 24.4KHz.

Voltage (yellow) and current (green) signals across the Grenade coil are messy, but kind of sine wave like, see screenshot.
Seems we have roughly 400Vpp @ 17App running in the Grenade circuit.

Input is 24V @ 6.7A and adding more 100W bulbs will lower the input current and also the DC output voltage (with 4 bulbs we are left with 52V dc).
DC output voltage with 1 bulb is 200V.


Video here:  https://youtu.be/JFJTWqAO8PU

Will try to get resonance on the Grenade circuit at 24.4Khz, then tune the inductor circuit to be in resonance at 122KHz......


Itsu
   
img  Vasik041
Re: Dally, Shark & Ruslan workbench « Reply #639 on: 2021-05-25, 16:23:42 »
Group: Guest
Quote from: Itsu on 2021-05-25, 16:03:05
I added a little fan above the 2 MOSFETs to cool them.
I changed the 40W output bulb to a stack (10) of 100W bulbs, but only have 1 such bulb active initially.
I removed the ground connection to be able to measure the Grenade/Inductor voltages (PS used is isolated from ground).
Then tuned the Grenade to max voltage/current which turns out to be around 21Khz, so also a little to low for the target 24.4KHz.
Voltage (yellow) and current (green) signals across the Grenade coil are messy, but kind of sine wave like, see screenshot.
Seems we have roughly 400Vpp @ 17App running in the Grenade circuit.
Input is 24V @ 6.7A and adding more 100W bulbs will lower the input current and also the DC output voltage (with 4 bulbs we are left with 52V dc).
DC output voltage with 1 bulb is 200V.
Video here:  https://youtu.be/JFJTWqAO8PU
Will try to get resonance on the Grenade circuit at 24.4Khz, then tune the inductor circuit to be in resonance at 122KHz......
Itsu

Itsu,

I think you are getting closer  O0

There is a picture in Alexeev's archive, looks similar to yours.

Vasik
   
img  verpies
Re: Dally, Shark & Ruslan workbench « Reply #640 on: 2021-05-25, 19:08:56 »

Group: Professor
Hero Member
*****

Posts: 3498
Quote from: Vasik041 on 2021-05-25, 16:23:42
There is a picture in Alexeev's archive, looks similar to yours.
There is a conspicuous difference, though.

Namely, in Itsu's scopeshots the spikes are more diffuse and they also occur at zero crossings (marked in blue below).
In that Alexeev's scopeshot they occur at half of Itsu's pulse repetition frequency and they DO NOT occur at the zero crossings.

...if they did then this scopeshot would look like this:
« Last Edit: 2021-05-25, 21:21:21 by verpies »
   
img  Vasik041
Re: Dally, Shark & Ruslan workbench « Reply #641 on: 2021-05-25, 19:25:40 »
Group: Guest
Quote from: verpies on 2021-05-25, 19:08:56
There is a conspicuous difference, though.

Well, let see how it looks when tuned  :-\

Vasik
   
img  Vasik041
Re: Dally, Shark & Ruslan workbench « Reply #642 on: 2021-05-25, 19:49:02 »
Group: Guest
Quote from: Vasik041 on 2021-05-25, 15:52:54
- inductor LC tuned to high frequency to allow some "free space" for Tesla coil action (when both MOSFETs are off). 

Here what happen when Tesla pulse get in a wrong place ;)

https://www.youtube.com/watch?v=yypJQKDoYXQ
   
img  Itsu
Re: Dally, Shark & Ruslan workbench « Reply #643 on: 2021-05-25, 21:37:58 »

Group: Elite Experimentalist
Hero Member
*****

Posts: 4159

Vasik,

i was following in rk_rev2.pdf, the part on "Push-pull schematic and tuning" on page 37 using the video there.

The signals shown are very well to be replicated by me (no secondaries connected), but do i understand correctly that at the end (page 43 top) he reduces the duty cycle to slightly above 3% and that this is the best setting for getting "static burst maximal"?

See his last screenshot (1) and mine (2).

Itsu
   
img  Vasik041
Re: Dally, Shark & Ruslan workbench « Reply #644 on: 2021-05-26, 05:47:01 »
Group: Guest
Quote from: Itsu on 2021-05-25, 21:37:58
Vasik,
i was following in rk_rev2.pdf, the part on "Push-pull schematic and tuning" on page 37 using the video there.
The signals shown are very well to be replicated by me (no secondaries connected), but do i understand correctly that at the end (page 43 top) he reduces the duty cycle to slightly above 3% and that this is the best setting for getting "static burst maximal"?

Itsu,

Yes, at the end of the video he reduce duty cycle to 3% to show how scope traces look like
and says that it gives maximum bursts.
I think he means that 3% is enough to get 200v and later MOSFET starts limiting it with internal zener diode.
(IRFP260N 200v 50a)

You can do resonance tuning with small duty cycle like 3 or 10% and then increase power until you get proper power on gradient coil and stay with reasonable power consumption/heating.

Vasik

PS I read some youtube comments about tuning and it could be that gradient coil should be tuned also to 120KHz just to have 30 degrees phase shift to inductor. It's a little unclear, so many different variants :)
   
img  Itsu
Re: Dally, Shark & Ruslan workbench « Reply #645 on: 2021-05-26, 09:31:46 »

Group: Elite Experimentalist
Hero Member
*****

Posts: 4159

Thanks Vasik,

so yet another variable into the mix to test  :D


Itsu
   
img  Itsu
Re: Dally, Shark & Ruslan workbench « Reply #646 on: 2021-05-26, 09:41:48 »

Group: Elite Experimentalist
Hero Member
*****

Posts: 4159
Quote from: verpies on 2021-05-25, 19:08:56
There is a conspicuous difference, though.

Namely, in Itsu's scopeshots the spikes are more diffuse and they also occur at zero crossings (marked in blue below).
In that Alexeev's scopeshot they occur at half of Itsu's pulse repetition frequency and they DO NOT occur at the zero crossings.

...if they did then this scopeshot would look like this:

Yes, and the number of spike(groups) is more / cycle, mine has 5 / cycle, Alexeev's has 3 (or is it 4 versus 2?).

As if the grenade is tuned to half the TL494 frequency.

Itsu
   
img  Itsu
Re: Dally, Shark & Ruslan workbench « Reply #647 on: 2021-05-26, 16:27:49 »

Group: Elite Experimentalist
Hero Member
*****

Posts: 4159
I changed the inductor series capacitors from a measured 27.6nF to 23.1nF and now the Inductor resonance peak (5th harmonic) is at 122KHz
I changed the Grenade series capacitors from a measured 306.7nF to 256nF and now the Grenade resonance peak is at 23.5KHz,       so almost there.

The TL494 frequency was set at 23.5KHz with 44% duty cycle, see purple (gate) trace in the screenshot.

Green is current through Grenade circuit, yellow and blue are the drain signals.

Input was 24V @ 10A, output load was 2x 100W bulbs on 210V DC and the temperature on the MOSFETs was stable around 50°C

Itsu
   
img  Vasik041
Re: Dally, Shark & Ruslan workbench « Reply #648 on: 2021-05-26, 16:33:24 »
Group: Guest
   
img  verpies
Re: Dally, Shark & Ruslan workbench « Reply #649 on: 2021-05-26, 19:32:37 »

Group: Professor
Hero Member
*****

Posts: 3498
Quote from: Vasik041 on 2021-05-23, 11:26:07
Found in comments to this video
https://www.youtube.com/watch?v=XDXTxPRPxSU
Quote
Inductance of one primary 110uH
The inductance of the "ОС-90.38ПЦ12" yoke ferrite from an old Russian 32-inch TV is 792nH/turn2 so 12 turns wound on it would have the inductance of 114µH.*

* When its two halves are well clamped together.
   
Pages: 1 ... 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 [26] 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 ... 101
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2024-11-26, 07:01:24