PopularFX
Home Help Search
Advanced search 
Login Register
Welcome,Guest. Please login or register.
2025-04-01, 06:55:26
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 2 3 4 5 [6] 7 8 9
Author Topic: Tetra Replication  (Read 16403 times)

Group: Professor
Hero Member
*****

Posts: 3670
D-Flip Flop (~50Mhz max clock, can be driven by 2.5V 50 ohm input voltage, looks perfect):
Isn't that one faster ?
https://www.onsemi.com/pdf/datasheet/74vhc74-d.pdf
   
Group: Moderator
Sr. Member
*****

Posts: 308
Isn't that one faster ?
https://www.onsemi.com/pdf/datasheet/74vhc74-d.pdf

Do you think the schmitt trigger functionality of the gates I listed is necessary? If not, I can switch series to 74vhc, which does seem much faster. It can handle 2.5V input too, which is super nice. I used to use 74ls alot, and those were finicky at 2.5V.
   
Group: Moderator
Sr. Member
*****

Posts: 308
Candidate pulse timing/pulse width control circuit sim.
   
Group: Moderator
Sr. Member
*****

Posts: 308
One thing i need help with is how to ensure each coil has the same exact pulse delay to it.

I.e. i need to eliminate minor differences in propagation delay from each driver and the wires leading to each coil.

Maybe overkill but I like to be thorough. Maybe ensure matched delays to the output of the drivers is good enough, but how to compare all three phases when they naturally have a phase delay? Maybe swap out each driver and compare with some kind of reference delay?

The transmission line from the timing circuit to each driver was going to be my overall propagation delay adjustment, just need a procedure to ensure even delays everywhere.
   

Group: Professor
Hero Member
*****

Posts: 3670
Candidate pulse timing/pulse width control circuit sim.
Why all these inverters/buffers and OR/NOR gates when the DS1023 input is triggered by the rising edge of the input waveform in the PWM mode ?
Also, I hope you are getting the DS1023-25 and not the DS1023-50, DS1023-100, etc...
   
Group: Moderator
Sr. Member
*****

Posts: 308
Why all these inverters/buffers and OR/NOR gates when the DS1023 input is triggered by the rising edge of the input waveform in the PWM mode ?
Also, I hope you are getting the DS1023-25 and not the DS1023-50, DS1023-100, etc...

The longer the delay is after the inverter, the longer the pulse width. It's purely for pulse width control. I've used 5 inverters in a chain before and the pulse width is the total propagation delay.

The NAND gate pulses briefly due to the difference in propagation from the undelayed and the delayed path. The NOR gate is designed to drive 50 ohm lines.
   

Group: Professor
Hero Member
*****

Posts: 3670
The longer the delay is after the inverter, the longer the pulse width.
I don't understand.
The width of the output pulse is determined by the value programmed into the DS1023 - not by the length of the pulse applied to the DS1023's input.
After the rising edge at the IN pin (e.g. from ph1), the PWM pin outputs the programmed pulse width + tREF.  The OUT pin does not have to be used.



   
Group: Moderator
Sr. Member
*****

Posts: 308
Right, I see. The ds1023 can just replace my Jerry rigged monostable circuit. Oversight on my part.
   

Group: Professor
Hero Member
*****

Posts: 3670
One thing i need help with is how to ensure each coil has the same exact pulse delay to it.

If you do this, you will not have that asymmetric NOR gate delay. Here is the Falstad sim.



Note: The RC circuit just presets the flip-flops upon the power up using the Preset and Clear pins of the 74VHC74.

WARNING:  The 74VHC74 has negated Preset and Clear pins but the Falstad simulator does not !
If you want it to work in Falstad then swap the resistor and capacitor (i.e.: cap to +5V, resistor to GND) and don't invert the S and R pins (because that feature is buggy).
   
Group: Moderator
Sr. Member
*****

Posts: 308
Great improvement, you have been a great help throughout this process. I'll continue forward with this approach.
   

Group: Professor
Hero Member
*****

Posts: 3670
Instead of the Resistor and Capacitor you can use one of these 3-pin ICs to reset that circuit on power-up ...and other circuits, too.
   
Group: Moderator
Sr. Member
*****

Posts: 308
Resistor and cap seems simple enough for a test circuit
   
Group: Moderator
Sr. Member
*****

Posts: 308
Some interesting thing I watched this morning.
   

Group: Professor
Hero Member
*****

Posts: 3670
Also, it should be possible to use a shift register in 1 chip instead of the D flip-flops in 2-chips in order to shrink and simplify the PCB and reduce the BOM. 
The circuit remains the same - only the number of chips goes down.
« Last Edit: 2025-02-25, 17:08:11 by verpies »
   
Group: Moderator
Sr. Member
*****

Posts: 308
I never figured out a reliable start mechanism but your circuit solved that for me. Shift regs can clock extremely fast.
   
Group: Moderator
Sr. Member
*****

Posts: 308
Calculations from spherics for proper coil distances and frequency peaks. Will compare to actual results.
   

Group: Professor
Hero Member
*****

Posts: 3670
...how to compare all three phases when they naturally have a phase delay?
Sum all 3 signals together and and trigger the scope on the rising edge with the timebase set in such manner that fewer than 3 pulses per sweep are captured.
Enable persistence.

Any phase jitter will manifest itself as edge thickening or doubled edges of the pulses.
If your scope has built-in automatic TIE or cycle-to-cycle phase jitter measurements - they will work, too.  See this video.
   
Group: Moderator
Sr. Member
*****

Posts: 308
Waiting for board to arrive. Going to set up thermal camera to monitor chip temperatures and see how the board spreads heat.

Need to finish timing board and have it produced.

In the meantime assembling and testing each driver board is next.
   

Group: Professor
Hero Member
*****

Posts: 3670
Need to finish timing board and have it produced.
Perhaps you could leave some unpopulated pads for one of these 3-pin ICs in SOT23 package to reset the timing circuit on power-up instead of the Resistor and Capacitor, ...even if you do not plan to use it for the tests now.

The problem with the RC solution is that it does not work when the power supply rails are not completely discharged to ~0V on power-down. 
If you always power-up from 0V then - no problem.
   
Group: Moderator
Sr. Member
*****

Posts: 308
Sounds good, I might just use it to start.

An idea I had for the future is testing output coil designs. It's a toroid, but I want to test different winding pitches. Looking up formulas to calculate slots for the windings in the output toroid for a given pitch/toroid diameter. Going to be really hard to test most configurations blind.

So one test I want to do is try to detect the shape of the field (it should corkscrew in toroid shape at certain frequencies). Maybe a very small coil on an insulated rod, have a servo move it around and map out field strengths at various points.

If the device is working it should induce a current in the probe.
   

Group: Professor
Hero Member
*****

Posts: 3670
Well, according to this we would be getting into coil design and field geometry, now.  A very different topic.
What are these coils supposed to generate ?  i.e. magnetic flux or something else ?

For starters, I have not seen any consideration of the inter-turn capacitance and inter-layer capacitance.  At high frequencies these capacitances dominate.

   
Group: Moderator
Sr. Member
*****

Posts: 308
Here is the relevant information.
   

Group: Professor
Hero Member
*****

Posts: 3670
What are these coils supposed to generate ?  i.e. magnetic flux or something else ?

Quote from: Spherics
The key piece of knowledge concerning manipulating the ether is it is not done using magnetic fields.

So that sentence answers my question negatively -  these coils are NOT supposed to generate magnetic flux.

However, the passage below implies a contradiction, i.e. that the coils generate a magnetic field. Concretely a rotating magnetic field.

Quote from: Spherics
You need only supply correctly phased DC offset square waves of approximately 300V (levels of 0V and 300V not -150V to 150V) to succeed in creating a rotation magnetic field which in reality is vortexing ether.

Elsewhere, the text mentions North and South poles of the coils.  Such poles constitute features of magnetic fields of solenoidal coils, which suggests that the purpose of these coils is to generate magnetic fields.

Also, the passage above states that square waves are applied to these coils, not rectangular pulses like the ones generated by the DS1023 and depicted in the "missing attachment".

If the coil's purpose is not to generate the magnetic field then I do not know how to optimize them, because "optimize for what ?"



   
Group: Moderator
Sr. Member
*****

Posts: 308
The goal is create the so called "comp" field, which at certain frequencies has effects identical to a magnetic field, and since the coils are pulsed in sequence, it causes the apparently artificial magnetic  field to rotate, best as I can tell.

Pulse width is an unknown factor and wasn't quantified properly. Unfortunately.
   

Group: Professor
Hero Member
*****

Posts: 3670
If the coil's purpose is not to generate the magnetic field then I do not know how to optimize them, because "optimize for what ?"

Optimization of coils's construction for the generation of magnetic fields is done by minimizing the inter-turn capacitance and inter-layer capacitance and minimizing resistance while maximizing the number of turns.  Transmission line effects also play a role.
Basket weaves are examples of such optimizations. 

Quote from: Spherics
Your aim with wire choice is to use the highest gauge for maximum resistance per foot and to use as much of it as possible within the space available for the coil.
The text above suggests winding thousands of turns with a very fine wire.  Such construction produces coils with HUGE inter-turn capacitance and inter-layer capacitance as well as a large resistance.  The opposite of what would be needed to generate magnetic fields efficiently.

When you apply a high frequency HV pulse to such coils, high electric current flows through these capacitances - not in the copper wire !
At high frequencies these capacitances dominate and the coil behaves more like a capacitor than a coil.

This begs a question -  why bother with a coil at all ?  Why not build a good capacitor instead ?

Quote from: Spherics
It is done by applying the highest voltage tension possible across the largest ->mass<- of copper or other non-magnetic material, with the ->lowest non-zero current<- possible. For the tension to have the required effect on the copper you need only the smallest amount of current.

The text above begs another question.  What limits the current flowing in/through the coil ?
If it were not for the inter-turn capacitance and inter-layer capacitance then the coil construction described in the previous Spheric's quote would limit this current naturally all by itself.
This happens because a large amount of turns means a large inductance and a large inductive reactance that does not allow a high current to build up during the short voltage pulse.  Also, the large resistance of the wire limits the current regardless of the pulse's duration.

However, the large inter-turn and inter-layer capacitances of such coil exist and throw a monkey wrench into these 2 current-limitation mechanisms, because these capacitances provide a shortcut for the current of these high frequency pulses, ...albeit not for DC.
   
Pages: 1 2 3 4 5 [6] 7 8 9
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2025-04-01, 06:55:26
Loading...