Use for the Tesla Switch

[Rl2003]

Ok,
I have tested the board and I have the two batteries (1 and 2) and
(3 and 4) that give me the 24volts, when testing each side,
But only when the MJL..94's are not connected to the curcuit.
When I connect the 94's the volts jump to 24+
and then slowly drop to 20 or so volts before the next pulse.
I checked and rechecked and check and rechecked the circuit. I also checked
the pulses to the MJL's and every thing looks right, and the 94's are good..
Has anyone come across this?

Mark

[Rl2003]

Up N Running

It turn out, nothing is wrong.. at least not that I've found.
Here is a short video of the switch running with the small 20W load.
It really helps to have a way to re-program as it's all hooked up. I have
already been trying different codes.

I used one of Bit's boards he designed and connected to outputs
0,1 without making any other changes to the board. Want to keep it
99% un-modified.
So far it works as Matt had said..I would like it make some other transformers
to test with this, now that I know it's working right.

VIDEO0039 - YouTube

Voltages after 1 hr 15 min. as follows/ taken while switch was under load.
Start Finish
Batt 1 12.56 12.51v
Batt2 12.41 12.11v
Batt3 12.55 12.23v
Batt4 12.45 11.95v
It's a start..It can only get better from here.

Mark

[jpolakow]

Finished build, now troubleshooting

Hi Caroll,

I finished my build of the tesla switch as per matt's specs in the pdf. I performed all the preliminary tests on the safety check list of the pdf and passed fine. The switches are working perfectly with the stamp and I have continuity at all points I am supposed to. I put 5 amp fuses on the negative side of all 4 batteries, that's the max current my ssr's are supposed to have. The problem is as soon as I hook up the transformer, the fuses for the bottom two batteries blow pretty quickly. I've already checked for a short in the windings and I don't have one. The windings each have about a half ohm of resistance so for 12 volts running through there I see why the fuses are blowing. What would you suggest?

Thanks for your help!

[citfta]

Hi John,

If the fuses are blowing and you are not pulsing the SSRs then you have a short somewhere. There should be no current flowing until you start pulsing the SSRs. If the fuses are blowing after you start pulsing the SSRs then you need to shorten your pulse times. I would start with pulse times of around a one half a millisecond and slowly work your way up to see what you get. As you go for longer pulse times your output will go up. When your output starts to drop off again you have found your best output frequency.

Another thing to check is to make sure you have SSRs that are designed for DC. There are SSRs that only work on AC. If you turn one of them on and have DC flowing through them they will not turn off again until you interrupt the DC. These kind will not work on this TS. I just checked Matt's PDF and he recommends 10 - 20 amp SSRs so yours may be too small to get this to work unless you have a very small load.

If I think of anything else I will post it later.

Carroll

[jpolakow]

Thanks Caroll,

Yea the fuses are only popping when the circuit is getting pulses - no shorts. I'm using the ssr's matt listed in the schematic in his pdf:
http://www.clare.com/home/pdfs.nsf/www/CPC1718.pdf/$file/CPC1718.pdf

They say 6.75 amps without a heat sink and 17.5 amps with a heat sink... I guess I'll try and get a heat sink! I have an old computer I can prolly butcher one from.

Otherwise is there a different ssr you would reccomend? I'm not too good at finding stuff on digikey. Oh and the pulse I was using started at 1250 as matt indicated. Which is I believe 2.5 milliseconds (units are 2 microseconds). Thanks for all the help! Have a good day!

[citfta]

OK John,

I would shorten those pulses and try again. You can probably get by with a slightly higher fuse too. If you are using those SSrs then you could use a 10 amp or so and just watch the heat as you lengthen your pulses. Each transformer will be a little different as to what pulse length works best. Start with maybe 100 us instead of 1 ms. If you start getting heat on the SSRs then find a heat sink to protect them. Most electronics can run pretty warm without damage as long as they don't get blistering hot. I like to keep mine cool with a big heat sink because I think over all they will last longer. But for short tests a little heating won't hurt them.

Carroll

[Sobakin]

the fuses for the bottom two batteries blow pretty quickly.

Maybe they blows at too low frequency?
If frequency is too low, transformer's core saturates and primary windings works closer to short circuit (reactive resistance get lowered).
How many turns has transformer's primary (turns, not feet)?
At which frequency fuses blows?

[jpolakow]

Yay!

Good news!
I got my switch up and running. I put some heat sinks on my chips and am running 10 amp fuses. It wasn't really neccesary though because the chips aren't even getting warm. The only time the 5 amp fuses were blowing was during the safety testing and I had 2 second long pulsewidths.

Anyways so nowI get the best voltage out of the transformer at a pulseout of 5100- which is 10,200 microseconds or about 10 milliseconds. Is this in the ballpark of what everyone else is getting?

Secondly, The AC voltage I'm getting out of the battery is about 17 volts with no load connected. At first I thought it would be higher because you get 12 volt peaks in each direction so I figured AC would be around 24 volts or more considering the output winding is 20% longer. But then I remembered meters don't measure peak to peak they measure RMS so 17 sounds about right. Is that what you guys are getting too?

I have yet to actually start load testing times, but will start that soon. Attached is a pic of my setup

[jpolakow]

Also the VDC out at the capacitors is 14.2 volts. Again does this sound ballpark to you guys?

[citfta]

Congratulations!

Glad you got it going. That is a nice looking build. Yes all of your values seem to be right in the ballpark for a rewound transformer. So now you get to have fun testing. Just take your time and document everything so you can see what works best.

Carroll

[Sobakin]

I still can't get positive effect.
Just built H-bridge driver for another baseline test with same transformator for clear compare TS and parallel 4 batteries work time, now it's running, paralleled both primaries for lower resistance. It maybe half-bridge because of already 2 identical primaries in transformer, but I would need full bridge for another purposes, so built it.
I also bought two MJL21194 for another tests, because mosfets works great - no heat and no voltage drop, but no positive effect too, maybe BJT will work better? I'll try them.
I will drive it by BD139, not by another MJL, because for drive MJL's base current 1 amp may be more than enough, I found this MJL's are expensive -- about $6 per one, compared to same power mosfet (IRF640) of about 50 cents.

[jpolakow]

I still can't get positive effect.
Just built H-bridge driver for another baseline test with same transformator for clear compare TS and parallel 4 batteries work time, now it's running, paralleled both primaries for lower resistance. It maybe half-bridge because of already 2 identical primaries in transformer, but I would need full bridge for another purposes, so built it.
I also bought two MJL21194 for another tests, because mosfets works great - no heat and no voltage drop, but no positive effect too, maybe BJT will work better? I'll try them.
I will drive it by BD139, not by another MJL, because for drive MJL's base current 1 amp may be more than enough, I found this MJL's are expensive -- about $6 per one, compared to same power mosfet (IRF640) of about 50 cents.

Hi Sobakin,
May I humbly suggest something for you? I'm new to this forum and somewhat new to electronics (I study EE) so take my humble opinion with a grain of salt. You seem pretty smart and have some good ideas and no doubt know tons more about electronics than i do. When I did this build I decided to follow Matthew Jones' advice to the LETTER. I had some different ideas for making the switch more efficient, but I wanted to stick to a plan I KNEW worked. There were a lot of things I could have changed about the circuit, but I knew Matt had been down this road already so I figured he knew what works and doesn't work. If you are having trouble with your build, maybe try something new and faithfully make an exact replica of Matt's circuit. You may get better results with a circuit that has already been proven to work for a number of people. After you get that going, then try some of your ideas and see if they give an improvement in performance. I don't know for sure yet if my tesla switch circuit is giving better run times than just paralleling the batteries, but so far it seems so. Just an idea for you if you are getting stuck with your circuit. Also, Carroll has been extremely helpful to me in getting mine to work. I think he may be willing to help you too if you stick to Matt's build. Have a good day Sobakin!

[jpolakow]

Preliminary Results

Well after preliminary testing the project definitely seems a success! I did a 4hour load testing of the tesla switch vs parallel batteries. The results were better than expected. I'm not 100% sure all variables were eliminated during the testing so I'm not going to exhaustively post the data, but I will publish the results.

First a description of test:
The load was an 8 watt incandescent auto bulb. As stated I ran both tests for 4 hours. I took voltage readings at the batteries every 15 minutes, and then averaged the readings to get a value to compare with. I took all readings with the load connected. For the next test I will take readings with and without the load connected. After reviewing the results I'm not sure 4 hours is long enough for conclusive results to be drawn, so more testing will done. Also I would like to use closer to a 20 watt load for a larger voltage draw on the batteries. Here's the results:

4 batteries in parallel:
This was pretty straightforward. For hours 1 and 2 the
batteries lost voltage in a linear fashion. For hours 3 and 4 the voltage started dropping off dramatically. The batteries would lose in 15 mins what would normally take 30-45 minutes during hour 1. The longer the test ran the more dramatic the voltage drop was. This is why I will do a longer test next time.

Tesla switch:
This offered some baffling results. The batteries lost the most charge during the first hour, and then incrementally smaller during the succesive hours. Obviously the reverse of the parallel batteries. The results got better the longer the test ran. The average voltage seemed to level out and drop very little. What's more, all 4 batteries actually CHARGED during the 4th hour compared to the third. This was inexplicable to me. The charging brought the batteries almost up to the same state as when the test began. Very confusing.

Comparison of data:
Looking at the results the tesla switch in effect lost zero charge(.002) after 4 hours, while batteries in parallel lost .1 volts. That is a huge magnitude of difference. I wasn't comfortable with this so I looked at the results for the first 3 hours. In this time period the tesla switch lost .04 volts on average vs .09 for batts in parallel. So this is closer to what i was expecting, a factor of 2.

Possible anomalies in data:
Here's some possible factors for the data being inconclusive:
1. The tests began with the batteries at different starting voltages. The tesla switch began with an average batt voltage of 12.33 volts vs 12.46 for parallel batts. Not sure what effect this would have but it is a variable.
2. Accuracy of the meter. The DMM I used displays voltage to 2 digits after the decimal place. I'm not sure how accurate the meter. I will have to check the manual. However the meter does give repeatable results. For instance if I check a batteries voltage, and then recheck it a couple more times the meter will give the same reading. So for our purposes it does show fairly accurate differential voltages.
3. Taking readings- I'm not sure what the best protocol is for taking the readings- ie with or without the load connected. I took all my readings with the load connected. Without the load connected the batteries might regain their charge during a resting period.
4. Small difference in starting vs. finishing voltage. A .1 voltage difference is not quite a big enough differential in my eyes to draw any conclusive results from. That is why I'm stepping up to a bigger load and longer run times.

So that's it for now, I assure you more exhaustive tests will follow!

[lamare]

@jpolakow :

You may want to check out the stuff I wrote earlier about what I called "the electret effect":

Article:Free Electric Energy in Theory and Practice - PESWiki


What is going on with Bedini's batteries is that the layers that are formed on the battery plates are very different from what happens when charging them normally. Bedini shows this in this video, and Rogers also wrote some on this phenomenon:

BatteryForming_2008_04_25_16_16_47.wmv
Directory contents of /pdf/Reference_Material/Rogers/

Also see my earlier post here:
http://www.energeticforum.com/renewa...tml#post107866

The thing that has really intrigued me is the report by John Bedini that he has batteries "cold boiling" for up to more than half an hour *after* he has shut down the charger. How on earth can that be happening?

Then there's an also intriguing experiment by MIT where they dissect a leyden jar, a capacitor, and clearly show that the "charge" is not on the capacitor plates, but somehow stored in the dielectric: http://www.energeticforum.com/renewa...html#post76020

And we have this also intriguing effect that electrolytic capacitors are able to re-charge themselves after having been shortcut, upto a certain extent:
http://www.energeticforum.com/renewa...rch-group.html
http://www.energeticforum.com/renewa...html#post76208

The interesting thing with these (standard) electrolytic capacitors is that the extent to which they are able to spontaneously recharge themselves can be greatly extended by "conditioning" them with "radiant energy". That is: feed them with high-voltage short duration spikes.

Putting these things together, I first of all concluded that in these "conditioned" capacitors the dielectric layer, where the energy is actually stored given the MIT leyden jar experiments, is somehow being super polarized. That means such a very thin dielectric film on the aluminum anode can apparantly be pushed into high gear, which means this super-polarized film you will generate a very strong electric field in its vicinity. Not because there is a high voltage across this dielectric film, but because the film is very thin. Normally, when a electrolytic capacitor is charged, this field is being nutralized, because charges are being drawn towards / pushed away from the surface of this thin film at both sides, until eventually the fields generated by these charges balances the field generated by the dielectric. It is because of these fields that oppoze one another that the capacity of a capacitor becomes bigger due to the presence of a dielectric.

Apparantly, it is possible to disrupt this balance by shortcutting the capacitor, after which the capacitor will re-establish the balance by spontaneously recharging itself. The interesting thing, however, is that just after this shortcutting, the only field that is present in the vicinity of the electrolyte, the water, is the field generated by the dielectric. And because this dielectric layer is very thin (a few micrometers) you will for some time have a very strong electric field in the vicinity of the dielectric, especially if you have polarized the dielectric using high voltage spikes/pulses, because the polarization of a dielectric depends on the applied field strength, not on the applied current or something.

Now you may say that the hydrogen produced in these old day electrolytic capacitors is just plain old electrolysis, but that would not explain why a glow can be observed with these things: Borax or Baking Soda Rectifier and the glow. :

As mentioned earlier, there is a faint glow associated with these borax (or baking soda) rectifiers that can be observed in a dark room. It seems that moderately high voltages are necessary in order to produce the glow. The glow is produced on the aluminum plate when it is at the positive (reverse bias) part of the cycle and minimum current is flowing.

Why would these things glow? Could it be that actually the water dielectric breaks down causing a corona discharge by which hydrogen is being produced, albeit very little, because first of all the voltages used are not that high and secondly, this particular setup uses plates at a considerable distance. In my experiments I could not produce a significant amount of gas when anode/cathode were placed at a distance, but I could produce hydrogen easily using concentric tubes. I know this does not prove anything, but it once again does suggest there may be more to it than it seems.

Given that Bedini's "cold boiling", i.e. H2 and O2 generating, batteries are being fed with the same kind of energy, HV spikes, and we also have similar "spontaneous" recharging, albeit with different metals and chemicals, I concluded that we are really looking at the same thing: a thin dielectric film is apparantly also formed inside batteries and that is apparantly also being super-polarized.

Bringing this together, I concluded that this must be the key. The construction of dielectric capacitors and batteries is very similar, especially in the old days and we get similar effects, even though hydrogen production does not always occur.

Given that the water contained in batteries is full of ions and the "cold boiling" effect observed by Bedini and others when the batteries are being charged with pretty powerfull Bedini chargers, I concluded that in this case the fields generated by the thin dielectric films on the plates is that strong that somehow electrolysis is being achieved, even after power is being fed to the battery. Based on that, I concluded that the presence of a strong static electric field inside the electrolyte is the sole requirement for this kind of electrolysis to occur and that that has nothing to do with any resonance effects in the water, as Meyer says. I mean, it just can't be if we can trust Bedini's observations and I do trust these are real accurate observations. Now this doesn't mean that it is impossible that you could also have interesting effects using resonance, but if that is the case, it has nothing this particular mechanism.

So, in my view, it may be much more difficult to create strong polarized dielectric films when the anode/cathodes are submerged in water containing lots of ions, but eventually all that really means it that you need more muscles to polarize your dielectric, because you have to drag more ions around. So, it may take more energy to accomplish the effects we want, but the energy it takes to free electrons from the negative ions in the half reaction, and feed them to the other half reaction to get elecrolysis is exactly the same.

So, the bad news is that you will have to spend considerably more effort to see these effects with non-pure water. The good news is that it is achievable anyway and when you succeed in creating self-healing dielectric films on aluminum tubes as is being done in electrolytic capacitors all the time, I am convinced you can really get to a point that such a super-polarized dielectric releases great amounts of hydrogen gas for which you only have to pay the energy needed to maintain the polarization field. However, even that energy does not have to be lost, since you can re-use this energy when "discharging" the capacitor, which will not only return most of the energy, but will also enhance the disbalance of the fields. That means you will actually get more hydrogen production if you reuse this energy!

And given the long times Bedini has reported his batteries to "cold boil" after shutting of the power, I think you really don't have to pay much in terms of energy to keep the dielectric polarized, especially if you re-use and re-apply the same energy over and over again.

So, it may very well be that (under certain conditions) you get the same kinds of changes in your batteries with the Tesla switch, which might explain why you see them (self?)charging after a few hours of operation.

So, if you want to compare normal operation with operation in a TS, it may very well be that you cannot operate a TS "conditioned" battery normally once this special dielectric layer as reported by Bedini is present at the plates of your battery.

And if this is what really happens, then you should see changes in the performance of your batteries, until the point where a stable "Bedini" dielectric layer has been formed. And at that point, the performanc of your batteries with normal chargers should also be much less. Bedini says that "conditioned" batteries can hardly be charged with a normal charger.

[Sobakin]

Hi Sobakin,
May I humbly suggest something for you?

Hi jpolakow,
thanks for your advice.
As You see, after all I came to same idea - replicate closer to Matt's schematic.
But in his pdf he write that there is no need to 1:1 copy their schematics, but the principle, so I built same circuit but with mosfets instead of bipolar MJL's and had no positive effect, tried different transformers (I wired 4 of them - 3 iron, small, big and bigger, and one small ferrite for 10+kHz).
Now I want to try MJL's - as delusive hope - maybe I dont know something and MJLs has some "magic" properties - knowing their price I can belive they are magic