Donald Smith Devices too good to be true

[Dave45]

Since its essentially a diode and being connected like this it would work more like a diode, would the output be DC?

[Dave45]

This is interesting http://www.faraday.ru/magnetron.pdf

[Gedfire]

Pulsed DC and Capacitors

Hello All,

Has anyone experimented with pulsed DC to a capacitor?


I think and tried it and the cap got hot and exploded.It was an electrolytic @400 v with 119 volts 50 hz applied.

I attached a diode bridge to the capacitor.What happened? Too much current or overvoltage?

What are the ways in which this explosion could have been prevented?


Best regards,
Ged

[janost]

Hello All,

Has anyone experimented with pulsed DC to a capacitor?


I think and tried it and the cap got hot and exploded.It was an electrolytic @400 v with 119 volts 50 hz applied.

I attached a diode bridge to the capacitor.What happened? Too much current or overvoltage?

What are the ways in which this explosion could have been prevented?


Best regards,
Ged

Most likely overvolt.

C1/D1 is a voltage doubler circuit so the output from it is a peak 333v.

And the 400v capacitor cannot handle it with high current when it is initially empty and charged.

[Solarlab]

I am not sure what your point is, as I fully understand that RMS values can be determined by the amount of heat being produced, and at least some better quality true rms meters use a method similar to that. Anyway, for your example, if the source is say 120VAC and the current through the 10 ohm resistor is say 0.25A, we will be dissipating 0.625W in the 10 ohm source resistor R1. If the resistor is a 10W power resistor, it will likely not be too hot. If the black box contains a step down transformer, and steps the voltage down to say 24VAC, and let's say the load equals about 10 ohms as well. Then we would have 12V across the 10 ohm resistor R2, giving a current of 1.2A. The power dissipated by the 10 ohm resistor R2 would be 14.4 watts. If it is also a 10 watt resistor, it wouldn't take long for the resistor to over heat and maybe burn up and fail. So, you don't need another source in the black box. Just a transformer could produce exactly this sort of results. Again, not sure what your point was with that.

In the example, you can not arbitrarily isolate the input from the output; even when a transformer is assumed to be in the Black Box; consider the overall circuit in-toto [for your example: a quick "ballpark" using I=V/R & P=V*I yields Pri = 1440W, Sec = 57.6W (assume a loss-less transformer) ???, yea I know it's AC & ... but we're just "ball parking" it for quick logic].

These links may be instructive, at least to some extent:

Transformer Circuits

AC Power Measurement

power-circle.jpg

tracir5.jpg

tracir4.jpg

[level]

In the example, you can not arbitrarily isolate the input from the output; even when a transformer is assumed to be in the Black Box; consider the overall circuit in-toto [for your example: a quick "ballpark" using I=V/R & P=V*I yields Pri = 1440W, Sec = 57.6W (assume a loss-less transformer) ???, yea I know it's AC & ... but we're just "ball parking" it for quick logic].

Hi Solarlab. Your calculated numbers don't make sense to me. Although the measured numbers likely wouldn't come out exactly the same with a real transformer circuit setup due to real world losses, I was just throwing out some example numbers to illustrate that this shouldn't be anything unusual. In my example, assuming a power factor of 1, the input power to the transformer would be 0.25A x (120V - (0.25A x 10 ohms)) = 29.375 Watts. The power delivered to the load would be 1.2A^2 x (10 + 10) = 28.8 Watts. That's an efficiency of about 98%. Nice efficient transformer.

No idea how you came up with those numbers of 1440W in and 57.6W out, but you obviously are making an error somewhere. With a supply of 120V and a current of 0.25A in, it should be obvious that your calculated power numbers don't make sense. Assuming a power factor of 1, it is basic ohm's law and basic power calculations. Anyway, I don't want to distract too much from the topic at hand which is Don Smith devices. If you still have doubts, you can always setup a test circuit with a 120VAC to 24VAC transformer and a 10 ohm series resistor in the primary and two 10 ohm resistors in series in the secondary, and measure for yourself.

[level]

Hello All,
Has anyone experimented with pulsed DC to a capacitor?
I think and tried it and the cap got hot and exploded.It was an electrolytic @400 v with 119 volts 50 hz applied.
I attached a diode bridge to the capacitor.What happened? Too much current or overvoltage?
What are the ways in which this explosion could have been prevented?
Best regards,
Ged

Hi Ged. Your full wave diode bridge rectifier is not shown connected up properly in your attached circuit diagram. If you connected it up as is drawn in the circuit diagram, you would be getting a near dead short across your AC source on negative swings, depending on what the actual capacitance value is for C1. If C1 is a small value, it should be OK though. Is that how you actually had everything connected? Anyway, if you did not connect your fullwave bridge rectifier up correctly to an electrolytic capacitor such that it wasn't always getting positive DC to its positive lead and negative DC to its negative lead, this can make electrolytic caps explode. You need to be careful with that as an exploding cap can be dangerous. Also, what type of capacitor were you using for C1, and what is its voltage rating? Was it C1 or C2 that exploded?
This is how a full wave bridge rectifier should be connected in:

[janost]

If you connected it up as is drawn in the circuit diagram, you would be getting a near dead short across your AC source on negative swings, depending on what the actual capacitance value is for C1.

No it's not a dead short. It's a Villard circuit.
On the negative swing it charges the C1 cap and on the positive swing it puts the charged cap in series with the AC source effectively adding to it's voltage.

The peak to peak is still the same but the positive peak is Vac x 2 and the negative peak is 0volts.

[level]

No it's not a dead short. It's a Villard circuit.
On the negative swing it charges the C1 cap and on the positive swing it puts the charged cap in series with the AC source effectively adding to it's voltage.
The peak to peak is still the same but the positive peak is Vac x 2 and the negative peak is 0volts.

Yes, it wouldn't be a short if you are using a small value for the capacitance, as I mentioned. If a very large value of capacitance is used it would act more like a short, but then you would pop a breaker or blow a fuse if the current draw was too high. Gedfire didn't mention what value he is using for his C1. Also, Gedfire mentioned a 'diode bridge', so I was assuming he was intending a full wave bridge rectifier setup. The way he has it connected in now would not be a proper full wave bridge setup, but maybe that is how he intended it. I wouldn't think that 340V peak would cause a 400V electrolytic to get damaged, but I haven't played around with circuits like that, so not certain. Maybe there is something else going on there. Accidentally hooking up an electrolytic capacitor backwards could definitely cause it to explode though, especially at those kind of voltages.

[level]

Interesting effects with a tesla coil setup

I have been doing some more experimenting with a tesla coil arrangement, but using a commercially wound coil on a form of about 2 3/4" diameter and about 110 turns, as the secondary. I have the primary of about 10 or 12 windings wound at the lower end of the secondary. I have the bottom lead of the secondary connected to earth ground.

Tests with GDT's in place of a sparkgap:
I tried replacing my sparkgap with one or more 1KV GDTs in series. Got poor performance with just one or two GDTs, but the performance started getting better when I put 3 1KV GDT's in series. The output to the secondary coil of the 'tesla coil' arrangement is different than with a sparkgap though. I test the voltage on the secondary coil using a neon bulb connected with one lead to the lower end of the secondary (ground), and I use the other neon bulb lead to touch various points on the secondary coil to test the voltage by how brightly the neon bulb glows, and also by how much arcing and what quality of arcing I get to the neon bulb lead. With my 12V NST transformer which has short circuit protection, I can't quite get as much drive as I really need, but I got the best performance with 4 1 KV GDTs in series, and I could get 5 1KV GDTs to fire, but it was starting to get somewhat intermittent. Even with 5 GDTs in series, the output on the secondary didn't seem quite as energetic overall as with a single sparkgap, but the qualtiy of the output was different as well.

With the GDTs in place, when I bring the neon bulb 'hot' lead near different points of the secondary, the discharges are mainly white, but also very much more 'static' like, or DC like discharges, compared to when I use a sparkgap. The discharges are sort of erratic and make a bit of a crackling noise, and when I touch the neon bulb to a point on the secondary and leave it there, the neon bulb glows very erratically and makes a kind of erratic clicking noise like some of the those gieger counters make. Also, it is mainly the top end of the secondary that produces the most energetic static like discharges. Also, I should mention, it only takes 30 seconds or so for all the GDTs to get hot, even when using 4 or 5 in series, whereas I can leave my sparkgap running for a long time and it only gets a bit warm.

With a sparkgap:
Output on the secondary is energetic in the whole top 3/4 of the secondary. I can't test the very lower portion of the secondary, because I have the primary wound over it, but with the bottom lead of the secondary grounded, the voltage slowly rises as you go from the lower part of the secondary towards the top of the secondary.

One interesting thing I have noticed is that although the arc lengths get longer as I move from the lower portion of the secondary towards the top of the secondary, the quality of the arcs changes as well. The arcs are very white and 'dense' looking towards the lower end of the tesla coil, and start to become more violet and a bit thinner looking as I move more towards the top. As near as I can guess, the brighter white arcs are because there is more current flowing, and as the arcs turn more and more violet colored there is less and less current flowing. Also as the arcs turn more violet, I can draw longer and longer arcs.
That would seem to indicate that although the voltage is lower at the lower end of my grounded secondary, there is more current available in this area, and nearer the top, the voltage is a lot higher but there is less current available. That would seem to make sense though anyway.

So, the secondary of a tesla coil must be at least something like a transmission line or antenna wire with current and voltage nodes, and I believe that is consistent with what I have read elsewhere about tesla coils. So, one possible simpler way to draw power at lower voltages from Don Smith's tesla coil type setups, is to tap the secondary coil at a lower voltage point, nearer to a current node, which in my setup would be nearer to the bottom of the secondary. This might be a simpler way to draw power off the secondary without having to resort to more complicated ways of stepping down the high secondary output voltage. Just an idea. There may well be no improvement in performance however, or performance might be worse, and there may be other problems that arise.

Also, I noticed something else that is interesting when I was testing my secondary coil with a neon bulb. If I touch the 'hot' end of my neon bulb to the lower portion of the secondary coil where I get the more intense white and thicker arcs, it feels like the lead is being pulled towards the secondary coil and being held there. The neon bulb lead is actually stuck to the secondary coil a bit and I have to pull it a bit for it to come loose. At the top of the coil where the voltage is higher and the arcing is thinner and more violet looking, I don't feel the lead being pulled towards the tesla secondary, and it feels like it is being pushed away a little bit. When I checked up and down the secondary coil really closely, there is one spot about 3/4 of the way up the secondary where the I can feel the neon bulb lead being pushed away from the tesla coil quite noticeably. It is not super strong like a magnet, but there is a very noticeable attraction and repulsion feeling at different points along the tesla coil. No idea what causes this attraction and repulsion though. First I have ever heard of arcing causing attraction or repulsion, but maybe that is some well known effect in plasma physics. Anyone know what might cause a copper wire to be pulled towards or pushed away from another copper wire when there is arcing between the wires? Anyway, it is an interesting effect. It might be interesting to try an experiment to see if there is any difference in the amount of power that can be tapped from the points where there is stronger attraction and points where there is stronger repulsion.

[level]

...
I think anyone who has a solid state Tesla coil could do this instantly just by removing components from their build. But it should be pretty simple
it would be cool to see someone's 500 or 1000 watt Tesla coil drive a load using the ground as in the schematic

It looks like this guy has it all down to an art. He builds 900V solid state IGBT 'sparkgap' modules and connects them in series for the desired breakover voltage. He states that the effective resistance is quite a bit lower than a typical sparkgap as well (if the circuit is built well), so that would mean less power loss in the sparkgap.
SIDAC / IGBT SPARK GAP - by Terry Fritz - PDF File

Also, another cool feature he describes is that if you use a potentiometer for R4 in his circuit, you can adjust the 'spark' firing on time.
His circuit might just be the cat's meow.

[janost]

Yes, it wouldn't be a short if you are using a small value for the capacitance, as I mentioned. If a very large value of capacitance is used it would act more like a short, but then you would pop a breaker or blow a fuse if the current draw was too high. Gedfire didn't mention what value he is using for his C1.

It doesn't matter what value you have on the C1 cap.
It's a singel stage voltage multiplier.

Voltage doubler - Wikipedia, the free encyclopedia

And it acts like one of the quadrants in a fullwave rectifier bridge with a smoothing cap after it.

C1 could pop if you draw above the C1 ripple current rating but that is a differnet story.

[level]

It doesn't matter what value you have on the C1 cap.
It's a singel stage voltage multiplier.

Yeah, I understood that. No worries mate.
By the way, did the voltage divider help at all with your FET joule thief circuit?

[janost]

Yeah, I understood that. No worries mate.
By the way, did the voltage divider help at all with your FET joule thief circuit?

I have been working all week.
I'll try it tomorrow when I'm off from work.

But I understood the voltage divider and it should work

[janost]

By the way, did the voltage divider help at all with your FET joule thief circuit?

Something like this would move the switching point for the mosfet?