Tubes?

[Jdo300]

Hi Earl,

Thanks for the Tesla excerpts. I suppose one could get quite nice pulses with 12 tubes all driving in parallel  .

I just happen to get a hold of one 6SN7 tube yesterday so I tried it in the simple resistor circuit and also in your circuit. The single triode alone still did not pull the inductor to ground, though it switched on twice as fast as the 12J5 that I was using earlier. According to the datasheet, the tube has a plate resistance of about 8 KOhms which gave me a very big drop when I used my 20k resistor in series with it for the resistor sheet.

But after paralleling the two triodes together, I got way better results. After resetting up your circuit, I put the 20k resistor on in place of the coil to measure the fall time of the tube. I posted scope shots of my observations below. The top trace is the tube and the bottom trace is the MOSFET. With the resistor, I got the tubes to switch on in roughly 200 ns while the FET switched on in 8-15 ns.

I included multiple shots so you can see wave at different zoom levels. The last scope shot is with the inductor in series. By the way, I actually measured it and found out that the inductance was actually 324 uH and not 30 mH like I was thinking. Oh one more thing, the tube was power from my power supply at 468V.

I'm not sure what to do at this point. I'm wondering if I should keep looking for a better tube, or find a different way to wire this one up to get better results? The circuit you gave me definitely works as prescribed though.

God Bless,
Jason O

[Earl]

Jason,

I assume that you have checked the heater voltage right at the tube socket for correct value.

I would try unwinding the torroid choke turns until only one layer around the complete circumference remains.

If that doesn't give good results, try removing some more turns and spacing the winding so there is just a little bit of space between each turn.  The idea here is to reduce inter-turn capacity.

A grounded-grid or -gate amplifier has low input impedance; the FET sees a low impedance looking into the cathode and switches quickly.  The tube plate unfortunately for digital switching has a large resistance.

If the above circuit tries still have too slow switching, the only other possibility that I see is to use a cathode-follower circuit.  This means a battery-powered circuit and because the tube grid has high-impedance and low capacity, you probably can get away with NO FET driver, just the IC alone.  This assumes that your IC has enough voltage swing to turn the tube on and off.  If the 6V swing from a 74HC is insufficient, then you could use the older 4000 series or the faster Philips HEF series, both of which work to 15 Volts.  The IC would normally be at a zero logic level (GRID = - V) only switching to a logical one for short pulses.  The tube cathode "output" has a low output impedance, orders of magnitude below plate resistance.

Another possibility is to use a tube circuit like the Seike G-strain.

Earl

the circuit below is a start, but not correct, it has to be changed a bit.
Will have to think about it so more, but maybe the IC circuit part needs to have Vdd connected to cathode (think flipped upside down).
Yes, the first schematic is not correct, the 2nd one should be OK.  The grid should always be at minus V with respect to the cathode, only going to zero Volts for a short pulse to turn on the tube.

REMEMBER the following 2nd schematic will only function correctly, IF the tube totally turns off a a grid voltage of minus 6V (74HC logic).

I have decided to edit this post and remove the first schematic to avoid confustion.

HIGH VOLTAGE CAN KILL - BE VERY VERY CAREFUL !

[sparks]

   @ Earl

       Tesla was a master of the standing wave.  If the receiver in the car was capturing a hf signal and storing this power in a standing wave then he has on board a battery and charging system.  Directing hf energy into the standing wave battery and controlling rf flash over current so that it is directed through the motor circuit.  The only hf energy scource I can think of that would be around in those days is infrared. Still is.  He could have been running the car on energy from his breakfast.   Maybe that is all we need to do.  Develop resonant circuits in the infrared spectrum and convert this energy into low frequency power to drive our grid transformers.  The Earth converting solar energy into a massive infrared transmitter.  Better than just letting it get reflected by greenhouse gases and melting ice.  Maybe we can collect enough of it and send it on it's way and cool things down around here. Change it into a fantastic light show for the viewing enjoyment of all.

[turbo]

Hey Jason nice going on the bench 
Did you try to smack the signals so they combine at or around 7.X Hertz?
You can also try to create a setup like a radio in reverse.
Normally the HF signal is mixed with the MF signal because this is more easy to amplify.
In reverse this would mean the LF is mixed with the MF to give a lower power HF output....
This is especially intresting when the LF wave isn't capable of inducing current in a too small inductor where HF is quite good at this.
Nice to see you are doing great work.
Keep it up 

Marco.

[Jdo300]

Hi Earl,

I just tried your circuit setup and, though I didn't have tubes that would switch completely off at or under 15V, I substituted the IC for a power supply and was able to throttle the tube up and down. However, when I used my MOSFET driver IC to try and drive it (partially since I could only input 15, and not the 25-30V that It needed), after a few seconds, the tube must of overloaded because it made a lout "POOF" sound and the driver IC preceded to explode  .

Then I got the idea to use a push-pull setup of FETs to drive the grid at a higher voltage with the power supply (that was a mistake I made later), but for this experiment, the FETs I used at the time weren't as fast as the IRF840's, plus, since my driver was extinct, I was forced to use the function generator to drive them directly, which obviously made them turn on much slower. So while I was attempting to do this, I was also draining over an amp through the setup so I couldn't leave it on long enough to see if the tube responded before the fets got hot. The tube again, must of got mad at me at that point because it made another "POW" from inside. I immediately shut it down and rewired the circuit with the power supply so I could manually adjust the grid bias to "asses" the damage.

After proving to myself that the tube was working fine, I used a single MOSFET to make and break the connection between the supply and the grid....... and thats when the MOSFET exploded and the power supply fried  . But miraculously enough, the tube just kept on humming away like nothing had happened!.... I love tubes, they're impossible to kill!

~Sigh~ anyway, I'm starting to wonder if I should just forget the SS circuits completely and go all tubes.... I'm thinking maybe something along the lines of a blocking oscillator or a tube ring oscillator would be a good idea, especially since I don't have any tubes with a low gate bias that I can directly drive with and IC. Earl, I know you mentioned the Sieke oscillator, but I'm not sure how to simplify the setup to use triodes. To be honest, I'm having a hard time seeing how it is functioning.

God Bless,
Jason O