Scope shots of clean kicks

[bob.rennips]

Hi Eldarion

It would be really good if after you get Earl's synchronous dividers to work and then integrate them into a FPGA. Then with a few external pot & discreet RC the pulses can be generated with minimum parts? Keep us posted. Thanks.

chrisC

Chris,

I was using an FPGA so that if I ever did get something working, the control algorithms could be implemented on the same piece of hardware with minimal effort.  Also, I can generate extremely accurate frequencies with the FPGA, something I am not sure an RC oscillator can do.

Just my $0.02

EDIT: Looking at the coils above, I noticed a sort of spiralling rotation, as each control coil is higher than the previous one, then it resets at the first coil.  I wonder if that would pump "something" through the coil?

Excellent. Look forward to getting your results - good or otherwise. And great to have you on board. Your coils look exactly what I was envisaging. Are the control coils bifilar - I can't tell from the pictures (or my eyesight!) ?

The way you have arranged the coils you have two sequences to try:

1. top - bottom pulsing clockwise.
2. bottom -top pulsing anticlockwise.

If you rearrange the coil you can also try:

1. top - bottom pulsing anticlockwise.
2. bottom - top pulsing clockwise

You might want to try without the output coil to start with and just scope the various coils to see what oscillations you are getting.

Although my timing diagram showed the OFF coinciding with an ON, my instinct says that you need the ON happening just before the OFF occurs. This will set in motion three thing:

1. An expanding magnetic field which attracts the particles from:
2. A collapsing magnetic field.
3. A path from the collapsing field to the expanding field.

In other words as I said in previous post. An expanding magnetic field attracts magnetic particles - via the coil connections - from a collapsing magnetic field.

[c0mster]

OL

[eldarion]

Excellent. Look forward to getting your results - good or otherwise. And great to have you on board. Your coils look exactly what I was envisaging. Are the control coils bifilar - I can't tell from the pictures (or my eyesight!) ?

Thank you for your encouragement!  Yes, the control coils are bifilar (50 turns of #22 ga. speaker wire).  No, it isn't your eyes, it is the lousy camera I used to take that picture, as I couldn't locate the good camera fast enough...

The way you have arranged the coils you have two sequences to try:

1. top - bottom pulsing clockwise.
2. bottom -top pulsing anticlockwise.

If you rearrange the coil you can also try:

1. top - bottom pulsing anticlockwise.
2. bottom - top pulsing clockwise

You might want to try without the output coil to start with and just scope the various coils to see what oscillations you are getting.

That's what I was thinking of doing--or winding, say, 20 turns of the output coil and seeing what shows up and how much power I get out of that small coil.

Although my timing diagram showed the OFF coinciding with an ON, my instinct says that you need the ON happening just before the OFF occurs. This will set in motion three thing:

1. An expanding magnetic field which attracts the particles from:
2. A collapsing magnetic field.
3. A path from the collapsing field to the expanding field.

In other words as I said in previous post. An expanding magnetic field attracts magnetic particles - via the coil connections - from a collapsing magnetic field.

Should be easy enough--I'll just add a variable delay in the pulse synchronization portion of the code.  Gotta love them FPGAs...   Now if only the mail service would hurry up and deliver my new MOSFET drivers...

[bob.rennips]

Some tests I have done show the mossfet signal has ossilations that kill the pulse. The change in voltage seems to produce the spike where as the ossilation and duty do not help. How do you kill a pulse just after the change in voltage has happened?

http://66.222.229.160:800/pic/pic.html
probe 1 is *10 5volt div.

Cam

Do you have a 10K resistor from the mosfet signal input to earth (negative) ?

Do you have a 1nF,10nf and 100nf in parallel across the power pins of the driver chip, as close to the power pins as possible ? (Try this first ifyou don't have them) use non-electrolytic, ceramic, nonolithic. Earl recommends the larger one as a tantalum correctly connected! I didn't have a tantalum around, and used ceramics.


Are you probing at one of the coil pins to see the spike ?

Not sure if it makes a difference, but I have my mosfet switching to earth. i.e. The coil is held at 12V and then one of the coil leads is swithced to earth.

I'm using a Freq. gen and keep the freq. below 500khz so the input signal to the mosfet driver is very well defined. I change the freq. until I get a crisp single 'kick'.

What's your input signal like when it's not connected to your mosfet driver ? This will eliminate your signal generator circuit as the cause of the problem.

You could try putting your input signal through an AND gate - tie one of the AND input pins to Vdd. or use two NAND gates in series and tie one of the input pins of each gate to zero ?  The gate output will normally go high to just below the extent of Vdd but check specs. of whatever chip you're using.

[bob.rennips]

I decided to see what happens if you pulse a coil with two diodes connected in series to both ends of a simple air coil. In other words one diode for current in and the other for current out. So what happens when the pulse goes off ?

Normally the magnetic field collapses and current is induced in the opposite direction but the diodes stop this from happening. So what happens to the magnetic field and all that energy ? Well....

The pulse goes off, which disconnects the coil from the negative. The magnetic field collapses. You get a large voltage oscillation but the oscillation is centered around 300Volts! Once the oscillation finishes the coil holds the voltage level at a steady state of 300Volts. And I really do mean rock steady horizontal. With a 10% duty cycle and pulsing at a frequency of 120khz it oscillates for around 10% of the cycle and spends the next 80% rock steady at 300volts. In other words the magnetic energy gets converted into pure capacitance.

This pure electric field oscillation can be picked up on an oscilloscope probe placed on metal objects several feet away from the coil and shows voltage levels of 3-5 volts. Now obviously this is voltage without amperage. If you short out the diodes and leave everything else the same, and probe the same metal objects several feet way you get no reading on the oscilloscope.

If you have this diode arrangement on one coil of a bifilar wound coil you can extract this energy from the coil at any point in the 80% of the steady state cycle simply by shorting out the other coil of the bifilar coil. In other words we can now capture the BEMF and use it whenever we want. This includes coils at 90 degrees to the bifilar coil. You short out a coil at 90 degrees and the energy is transferred to this coil.

Let's say we set up 3 separate coils in a circle all with this diode arrangement as per my previously diagrammed arrangement. We are creating a rotating electric field WITHOUT THE MAGNETIC FLUX. I think we are all aware of SM's comments concerning the cancelling of the effects of the magnetic flux.

For 80% of the duty cycle electrons will now be free to whizz around the collector - attracted to the rotating position of the electric field!!!

Because we are capturing the BEMF as an electric field and transferring it to the next pulse, the voltage of the electric field will rapidly get bigger. A rotating electric field, will entrain other electric fields, such as the one around the earth!!! If this occurs you will get a lightening strike on your TPU. Think of the lightening that occurs in a tornado due to rotating electric fields.

A large electric field will play havoc with solid state electronics (ICs) but tubes will be fairly immune. (That's another SM tick).

The capacitance is effectively held between the coil windings in the insulation. This is why you see thick insulated wire in the SM TPU's it's to hold as large an electric charge as possible. (That's another SM tick).

Also the effective voltage will be low at one end and maximum at the other end of the coil.

Another way to increase the effective capacitance of a coil is to wind a bifilar coil and then connect these together in series. This ensures the greatest difference in voltage from one winding to the next, which is a requirement for increaseing capacitance of a coil.

Incidently another way of increasing the capacitance is to have a coil of larger circumference. (That's another SM tick).


Enough of the talking...

I'll post results as soon as I have any. Good or bad. If anyone else gets there first - please post.