Wardenclyffe - Tesla's true intention

[gyula]

There is probably someone here with more MOSFET experience than I have. Perhaps you can help me out.

Of course I am working on a proof of concept of the above system. But as every item has to be hand made this takes a while.
Meanwhile, I thought to experiment a bit with the 'plasma bridge' and receiver coil.
The biggest coil I have (84 cm diam, 72 cm height) has a SRF of about 160 KHz.
So, to quickly have a rectified 160KHz high voltage source I was thinking about a TV/monitor fly-back transformer. I have a TC4421 MOSFET driver and a IPP60R380 MOSFET. I connect the driver directly to the MOSFET and I feed the driver directly with a TTL level signal.
Using 5 V for the FBT everything holds pretty well, though the driver does get a bit warm.
Using 12 V the driver gets a bit hot but the MOSFET (factory calls it 'cool mos' ) gets so hot in seconds, that I burned my finger on its heatsink.
All voltages, currents and frequencies are well within specs. The only thing I can think of is that the driver is a bit too rough on the FET, pushing the dV/dt too high.

Anyone? Suggestions?

Hi Ernst,

My first suggestion is to use plentiful decoupling capacitors very near to the supply pins of the TC4421 (1uF, 470nF, 47nF, all in directly parallel with the supply pins if you have not done so. I say this I suspect the nearfield of the
coil(s) may induce unwanted voltages in the driver circuit. Also try place a 1 kOhm or so 'terminating' resistor directly across the input pin and the neg. supply pin of the TC4421 to prevent it becoming high impedance for any moment, to be able to pick-up any nearfield induction, this will not affect the TTL drive level at all.

If these are of no help, then it may be about duty cycle, you wrote all the voltages and currents are within spec, the MOSFET dissipation simply exceeds the power level which could be allowed for free air cooling without using any amount of heat sink. At 12V supply voltage and at the same ON time for the MOSFET the self-dissipation is many times higher than at 5V, just consider the FBT's drain coil's DC + AC impedances: how much peak current can flow when the FET is ON, that current heats the FET.

What you wrote on the driver's high speed switching and due to this the flyback pulse may approach or exceed in amplitude the limit of the MOSFET drain-source max voltage rating, well this could surely increase the FET's dissipation too. An oscilloscope test on the drain peak pulse voltage is in order of course.

So maybe using a normal heat sink for the MOSFET is the solution, if the peak drain voltage is also ok, that is all, once you are satisfied with the currents and voltages. ?

rgds, Gyula

[Ernst]

Thank you, Gyula, for your suggestions!
I will surely implement all of them in the next attempt.
Unfortunately the MOSFET that I was using is no more...
But I think I still have 2 spares of this same type.

Extra information:
The duty cycle is pretty high, about 90% or so. This gives the best performance with an FBT.
I did use a small heat sink, but that was obviously not enough.
I use an old PC power supply (400W or there about).
I tried with a 7 Ohm 60W resistor between the driver and the FET in order to limit this current a bit. That is when the FET went to FET-heaven.
There is an extra diode (1KV/60A) across the drain-source to take care of the pulse that the FBT returns.

Anyway, thanks again for your help!

Ernst.

[boguslaw]

Ah, this is the question ... how to set parameters to allow mosfets pass capacitive discharge without damage :-(
I hate burning mosfets.

[citfta]

FET overheating

Hi Ernst,

All of gyula's suggestions are good ones. So be sure and follow them. Also you may not know what killed you FET when you added the 7 Ohm resistor. I will try to explain if you don't mind. When a transistor is turned fully on it is dissipating the least amount of power compared to being only partially turned on. When the transistor or FET is only partially turned on it is acting like a resistor so it will heat up a lot more than when it is turned fully on. As you reduced the gate voltage by putting the 7 Ohm resistor in the circuit you restricted the amount the FET could turn on and thus created even more heat in the FET. Mosfet driver chips help to overcome this problem by making sure the FET is either all the way on or all the way off so as to dissipate the least amount of power in the FET. We want our power to be dissipated in our load not our FET. Also you should be aware it is normal for power FETs to get pretty warm when running. Even with a heat sink they will feel pretty warm. They should not get hot enough to burn you though. Hope this helps.

I tried with a 7 Ohm 60W resistor between the driver and the FET in order to limit this current a bit. That is when the FET went to FET-heaven.

Respectfully,
Carroll

[Ernst]

Thanks for explaining, Carroll!

Yes, when you put it this way, it makes perfect sense.
I have always been pretty succesfull, designing and building digital/computer circuits, but outside of the digital world, I fail where everyone else seems to succeed

Imagine my surprise when my first 3-coil Tesla transformer worked (almost) straight away.

The power supply for my new Tesla coil will be ready soon. Perhaps I should just stay with that.

Thanks again!

Ernst.

[jake]

I always thought a ZVS driver would make a good hfhv power supply. Anyone try one of these?

[Ernst]

Can a ZVS be tuned to 160 KHz?

It's not just HF-HV.
I need to be able to tune it's frequency to the SRF of the receiving coil.

[gyula]

Ah, this is the question ... how to set parameters to allow mosfets pass capacitive discharge without damage :-(
I hate burning mosfets.

Hi Boguslaw,

If I got your question correctly you mean you use MOSFETs as switches to discharge capacitors, right?
If yes, then you may wish to consider ALL the resistances and impedances that are included in the discharge circuit path. First thing is the capacitor itself you wish to discharge, see this link and in the Table you can see typical series resistance values for different electrolytic capacitor values and at different max working voltage ratings: Capacitance ESR typical values chart

Suppose you have a capacitor of 220uF. 250V DC rated and you have charged it up to ,say, 200V DC voltage. From the Table in the link
you can read 0.5 Ohm series resistance for a 250V rated 220uF capacitor.

Now you discharge this with a decent srewdriver, that surely has some milliOhm maximum resistance which is negligible with respect to the inner resistance of the 0.5 Ohm. What current is involved when you short this capacitor? Simply apply Ohm's law and you get 200V/0.5 Ohm = 400 Amper peak current.
And in case you happen to use a MOSFET switch that has an ON resistance under 1 Ohm (of course there are much lower values available) say you use a power FET with a 0.1 Ohm Drain-Source ON resistance, adding this to the 0.5 Ohm and using Ohm law again, you get 200V/0.6= 333.3 Amper.

Of course I do not know your actual circuit total inner resistances and impedances to figure out the actual peak current which however must have surely exceeded your MOSFET max allowed drain-source current ratings. Even if you use 20-30 Amper drain current rated MOSFETs, you may easily exceed that limit and the device fails.
So what you can do is study all the inner resistances and impedances of the components in advance which are in the path of the discharge current. Capacitors are very, very good voltage sources, due to their normally very low inner impedances, meaning huge current can flow out from them till the full discharge.
Another issue can be the drain source voltage rating of the MOSFET because in case there is an inductance in the discharge path, the switch-off spike may exceed its rating.

rgds, Gyula

[boguslaw]

Hi Boguslaw,

If I got your question correctly you mean you use MOSFETs as switches to discharge capacitors, right?
If yes, then you may wish to consider ALL the resistances and impedances that are included in the discharge circuit path. First thing is the capacitor itself you wish to discharge, see this link and in the Table you can see typical series resistance values for different electrolytic capacitor values and at different max working voltage ratings: Capacitance ESR typical values chart

Suppose you have a capacitor of 220uF. 250V DC rated and you have charged it up to ,say, 200V DC voltage. From the Table in the link
you can read 0.5 Ohm series resistance for a 250V rated 220uF capacitor.

Now you discharge this with a decent srewdriver, that surely has some milliOhm maximum resistance which is negligible with respect to the inner resistance of the 0.5 Ohm. What current is involved when you short this capacitor? Simply apply Ohm's law and you get 200V/0.5 Ohm = 400 Amper peak current.
And in case you happen to use a MOSFET switch that has an ON resistance under 1 Ohm (of course there are much lower values available) say you use a power FET with a 0.1 Ohm Drain-Source ON resistance, adding this to the 0.5 Ohm and using Ohm law again, you get 200V/0.6= 333.3 Amper.

Of course I do not know your actual circuit total inner resistances and impedances to figure out the actual peak current which however must have surely exceeded your MOSFET max allowed drain-source current ratings. Even if you use 20-30 Amper drain current rated MOSFETs, you may easily exceed that limit and the device fails.
So what you can do is study all the inner resistances and impedances of the components in advance which are in the path of the discharge current. Capacitors are very, very good voltage sources, due to their normally very low inner impedances, meaning huge current can flow out from them till the full discharge.
Another issue can be the drain source voltage rating of the MOSFET because in case there is an inductance in the discharge path, the switch-off spike may exceed its rating.

rgds, Gyula

I'm trying to discharge 8uF AC motor run capacitor to the low inductance coil like 6uH with low resistance (I have to measure it) with capacitor rated at 500V but charged to 430V max (from 290V to 430V depending on how it may charge)

[gyula]

I'm trying to discharge 8uF AC motor run capacitor to the low inductance coil like 6uH with low resistance (I have to measure it) with capacitor rated at 500V but charged to 430V max (from 290V to 430V depending on how it may charge)

Normally AC motor run capacitors are of much better quality than electrolytic caps so a 8uF 500V rated run cap can have a series resistance (ESR) even under 1 Ohm. Your 6uH coil's DC resistance is surely well under 1 Ohm, its inductive reactance may also be low, this depends on of course what frequency you use for discharging the 8uF cap (rate of discharge) but if it is in the some ten Hertz as the maximum (time is also needed for charging it up), then the 6uH represent just a fraction of an Ohm reactance. This may not be true when you have another coil mutually coupled to this 6uH of course (I mean transformed impedance from another coil the 6uH is coupled to) but you can figure out this for yourself knowing your own setup.
Let me assume 1 Ohm for the cap inner ESR and let me assume a worst case 1 Ohm reactance for the 6uH and let me assume a 1 Ohm drain-source ON resistance for your MOSFET switch and let me assume 290V in the capacitor, the discharge current in the moment of the switch-on is 290/(1+1+1) = 96.6 Amper, this immediately starts reducing as the stored energy in the cap is reducing of course. Question is what is the drain-source current rating for your MOSFET, is it able to cope with the near 100A peak current and if yes, then for how long time (consult data sheet, it normally includes SOA, the maximum Safe Operational Area figure for pulsed conditions too). IF your 8uF is charged up for 430V, then the discharge current in the very first moment is 430/3=143.3A!

Notice 1: my numbers above are assumptions of course but the 8uF capacitor's ESR is surely around 1 Ohm or lower, this could be known from the cap's data sheet if the type is a decent one, identifyable at all or could be measured by an ESR meter if available.

Notice 2: unfortunately, for best power MOSFETs in the 500V or higher drain-source voltage rated ranges they have at least 45-50 milliOhm ON resistance at 30 Amper drain current. See this data sheet for instance: http://www.st.com/internet/com/TECHN...CD00002391.pdf or see these here:
http://www.fairchildsemi.com/ds/FD/FDL100N50F.pdf and
http://ixapps.ixys.com/DataSheet/DS1...00N50Q3%29.pdf
(you can have them at Digikey).

Notice 3: Consider the 8uF cap as a voltage source, this has the assumed 1 Ohm inner ESR resistance, then comes in series with the ESR the assumed 1 Ohm coil impedance, then in series comes the FET 1 Ohm ON resistance, these form a voltage divider whereby the capacitor voltage is divided to 3 parts between these resistances or impedances and you may wish to use a FET with a few milliOhm ON resistance only (not the 50 milliOhm type) so that the huge current should not cause big unwanted power loss (the power loss I*I*Rds across the switch surely cannot reach the 6uH coil at all, it is a waste).

Hope this helps in some way.

rgds, Gyula

[Ernst]

Anyone interested in down scaling Wardenclyffe to something that could be used in our homes?

Let's first look at the full scale system.

What would have happened in Wardenclyffe?
- a 3 coil Tesla transformer creates a HF HV. The primairy and the shunt capacitor determine the main frequency.
- the spark gap fires at a different frequency which causes an amplitude modulation on this main frequency.
- the top load of this transformer (Tesla calls it 'the free system') consists of a vacuum bulb which has a small capacity resulting in very high voltages.
- this gives longitudinal electrical waves that can easily be picked up by another vacuum bulb
- this type of connection has a rectifying effect. Probably only a small effect but due to the very high voltages there will be a significant net result.
- because of this rectifying effect the modulated (spark gap) frequency becomes visible and can be used to bring a secondary system into resonance
- everything that happens in this secondary system has no consequence for the free system (because of the difference in frequency) so we can use this energy for free
- this energy is used to set an electrical charge in motion which is coming from the earth, going to the top of the tower and back again.
- this causes an electrical disturbance in the earths charge which can be picked up and used anywhere on earth

Extra note: There are unconfirmed (?) rumours that Tesla used a large amount of UV lights in this tower. I was thinking that if those were in some way linked to the spark gap and if they would have an effect on the amount of electricity radiated from the vacuum bulb that might strongly increase the rectifying effect.

Before we start down scaling, any comments on the full scale system as I have described it here?

Ernst.

[GSM]

Extra note: There are unconfirmed (?) rumours that Tesla used a large amount of UV lights in this tower. I was thinking that if those were in some way linked to the spark gap and if they would have an effect on the amount of electricity radiated from the vacuum bulb that might strongly increase the rectifying effect.
Ernst.

Would the UV be from mercury arc rectifiers, or especial spark gaps.

UVC + metal (especially tin) = free electrons.

Arcadia the leaders in pet-care lighting technology - Ultra Clear UVC fluorescent lamps
Tin Styling Sheets | Homecrafts.co.uk

Cheers ........ Graham.

[Ernst]

Would the UV be from mercury arc rectifiers, or especial spark gaps.

I have been thinking about that too, but I have not found any confirmation in Tesla's own writings that he was using UV or an extra component not described in the 'rare notes'.
I would like to believe that the schematics in those notes are complete, but of course there is always a chance that while building this tower he thought up some further improvements.

Eric Dollard showed in his video that vacuum tubes (lights) radiate longitudinal electricity when subjected to RF HV. I was wondering would this effect be stronger when irradiated with UV(C)? And what if those vacuum tubes contained tin?
I would like to experiment with this a bit, but at this time I do not have the necessary equipment. Does anyone know about experiments in this area?

For our garage-Wardenclyffe however this is not a necessity because today we can use HV diodes to accomplish the same.
Yesterday I tested my magnetically quenched, pre-heated HF spark gap. Not as beautiful as some work that I have seen here, but it worked like a (bit violent ) dream.

Ernst.

[dR-Green]

May I suggest that you begin by taking on the Crystal Radio Initiative as set forth by Eric Dollard

[Ernst]

May I suggest that you begin by taking on the Crystal Radio Initiative as set forth by Eric Dollard

You may

Thanks, dR-Green!
Do you have a link?

Ernst.