Use for the Tesla Switch

[lamare]

OCR version of Mueller report

Hi all,

I have made a new, digital version of the Mueller report, upgraded with some pictures I found at John's site.

@John: I have sent you a PM via the forum to ask permission for sharing this file a couple of days ago, but so far I have not received any response. Just now I found out you have publsihed a version of this file on your own site too:

http://www.esmhome.org/library/john-...ery_switch.pdf

Considering this, I assume you don't mind this file to be posted here. If you do, please let me know, and I will delete my attachment immediately.

@all: please don't distribute this file outside of the members of this forum, unless John B. gives us permission to do so.
(attachments can only be read by the members of the forum...)

[ldissing]

Balancing

Leroy,
Yes I understand that, what do you have for measurements on the batteries, are they all equal in charge? And yes the batteries will get to a point where they balance and I'm doing the same thing with mine. I'm still working on that at high frequency.
John B

Well, I've seen some interesting things.

When running the small load, I've seen the battery voltage rise .7 volts in 15 seconds, but I was a little frightened at that, probably that was the right place to be. And with the small load, they all continually charge at certain frequencies. I probably didn't use that small load long enough.

The battery voltages seem a little low to me, all are now at approximately 12.6 within .02 volts of each other when running, however at resting 8 hrs, one of the batteries is about 0.12V lower this morning, the others at 12.6V. You know how to do some battery measurements for true charge that I only dream about. I shut it down last night, because I was using a higher V load and I didn't want anything bad to happen over night. I'm not afraid to run the small load all the time.

These are not the newest batteries on the block. But most have been through the SSG a few times, but I'm not sure all of them have though. These were just batteries I had on hand that were all the same...when a normal person have 4 supposedly identical 12V batteries. I'm guessing I need to go buy some more brand new batteries, but if this does what I think it does, it should not matter.

Oh, I think I know the third purpose, kind of relates to some problems you've spoken about in the past. Could be wrong, been there before. Love to see your underwater experiments.

Leroy

[hherby]

This has become the side show to the transistor version, but I thought I'd post here anyway.

YouTube - TS commutator

Love and light

Hello Inquorate and all. I have been a member of this forum for quite some time reading all of the interesting information here and also a member of all of JB's yahoo groups. I see many familiar faces from the yahoo groups here. I mainly post on the yahoo groups and have found this site to be a wealth of information and decided to post here as well. I saw JB said hello to all the lurkers in the thread such as myself and decided I should introduce myself.

Back to the topic. In the summer I started building a mechanical switch for the TS. On my first basic test my frequency was much lower than I expected. Then slapping myself squarely on the forehead, I realized to obtain 100hz, I would have to spin the commutator at 6000 rpm. Relating back to the electrodyne experiments, they were running between 100hz to 800 hz. For 800hz my commutator would have to spin at 48,000 rpm.

I could not even get near 2000 rpm without the commutator vibrating and brushes bouncing even though I had it trued and polished and balanced as well as possible in a lathe. It seemed the shaft would start flexing at higher speeds. I ended up abandoning the mechanical switch after several weeks of work building it.

It could be done with a commercially made slip ring commutator in concert with a split ring commutator with many more segments on a larger diameter with odd segments not connected to prevent shorting switch phases. This would allow several switching cycles per revolution reducing the required rpm for the desired switching frequency. I emailed a couple of commutator manufacturers in Canada and the US for pricing but did not get a response. I can only assume they didn't want to deal with a low volume experimentor.

I then decided to try it with electronic switching. I did some experiments driving the base-emitter junction with a 120v-12v transformer wired backwards like JB had in his original TS using audio transformers as shown in the Mueller report that I purchased from Rick. Depending on polarity, the BJT would fire either on the rising pulse or falling pulse on the 8ohm secondary. Primary was hooked up to base and emmiter. I found at low frequency I could not control the pulse width to the BJT. A pulse on the secondary would only fire the BJT on the rising or falling edge of the pulse (again depending on polarity on the secondary) regardless of pulsewidth. Note I was testing this trying to get a bulb to light up that was wired between 2 NPN's. Dummy that I am when it comes to transistors, I was trying to figure out how to bias the transistors when using NPN's on both the High and low sides of the load.

I wanted to duplicate this using opto's but had a hard time figureing out how to keep the grounds isolated. Then I did an experiment where the output of the primary winding of the transformer charged up a cap. The cap negative was tied to the emmiter of the NPN BJT and the positive of the cap went to pin 5 of an H11D1 opto and the output of pin 4 was fed to the base of the NPN through a resistor. As long as the cap was large enough and kept charged by independently pulsing the secondary, the pulse width on the BJT would match the pulse width on the opto. I had my setup of the cap/opto driver half built when the weather started getting too cold to work in the shed. I finally setup a workspace in my basement and moved everything indoors. Getting back to reading all the new posts on the TS, I was delighted to see JB posting in the thread. John B has posted examples of how to forward Bias the NPN using a 100ohm resistor from collector to base and also an opto with the collector of the opto connected to the collector of NPN and the emitter of the opto connected to the base of the NPN through a diode with a paralleled cap.

John B. I was wondering if you could help me understand the function of the combination of a diode in parallel with a cap feeding the base of the NPN?

Thanks

Alex L (h2ou123 on the yahoo groups)

[hherby]

Hello team, here is an updated schematic for the PICAXE-18X controlled TS. I have included its own on board power that will need to be "battelfield tested" in the words of Vtech. As always, use at your own risk and please let me know if there are any changes I need to make.

Thanks

Bit's

Hello Bit's,

Nice layout. Can the PIC handle greater than 5v on its ADC inputs without having to use a voltage divide resistor arrangement?

Thanks,

Alex

[stonewater]

Tom,
The experiment is to understand the direction of the potential the switch is in. But you are rotating the batteries why is battery 2 down is something wrong with it? What is going on with battery 3. they should all balance a little higher in voltage.
John B

I think part of my problem is that none of the batteries were fully charged when I started, I had my switch wired wrong, leading to just some current being moved around, and I did not make sure that the higher voltage batteries were in spots 1 and 2 with the lowest on 3 powering the load when I first started. I also am not sure of the condition of these batts, they came out of computer back up power supplies.

thanks for you time john, I need to be more careful, I was just really excited about the whole process.

I am charging up all theree, doing a load test on all 3 to see their condition then charging two and doing the ts charge on batt 3

I will keep this process carefully documented.:

Tom C

[Bit's-n-Bytes]

Hello Bit's,

Nice layout. Can the PIC handle greater than 5v on its ADC inputs without having to use a voltage divide resistor arrangement?

Thanks,

Alex

Hi Alex Thanks, Here is what I pulled from the manual;
The ADC range is the power supply voltage range. The maximum recommended
input impedance is 20k. Unconnected ADC will ‘float’ giving varying false
readings.

Thanks
Bit's

[LutherG]

Hi John,

Hello from a "watcher"

I'm gathering parts - I've got the TI SG3524N and saw your fix for this. Also have a few H11D3's and 4's... Will begin the 3-battery rotation this weekend. Thanks for posting the link for quick etching circuit boards. Would really like to touch base with you regarding a circuit idea we once discussed over the phone... best regards - Luther

[Jetijs]

Hi all,
here is my report from the battlefield

Today I tried various loads on the scalar charger circuit and also made many scope shots at various places in the circuit at different switching frequencies. I found that the load makes a huge difference on the scope shots. A small 6v dc motor creates spikes of 36V peak to peak, but this happens at higher frequencies, but if I replace the motor with a small light bulb, there is the average square waveform much like the one using the motor, but without any spikes. Here is the complete circuit I am using now:



I put numbers on each transistor and cap so that you can see what the waveform looks like at each component. I used transistor emitter and collector to capture the waveforms. So here we go:


Waveforms at 0.5Hz using 6v motor as a load

Transistor 1:

Transistor 2:

Transistor 3:

Cap 1:
http://www.emuprim.lv/bildez/thumbs/...0.5hz_cap1.JPG
Cap 2:
http://www.emuprim.lv/bildez/thumbs/...0.5hz_cap2.JPG


Waveforms at 7Hz using 6v motor as a load

Transistor 1:

Transistor 2:

Transistor 3:

Cap 1:
http://www.emuprim.lv/bildez/thumbs/...2-7hz_cap1.JPG
Cap 2:
http://www.emuprim.lv/bildez/thumbs/...9-7hz_cap2.JPG


Waveforms at 160Hz using 6v motor as a load

Transistor 1:

Transistor 2:

Transistor 3:

Cap 1:
http://www.emuprim.lv/bildez/thumbs/...160hz_cap1.JPG
Cap 2:
http://www.emuprim.lv/bildez/thumbs/...160hz_cap2.JPG



Waveforms at 1.56KHz using 6v motor as a load

Transistor 1:

Transistor 2:

Transistor 3:

Cap 1:
http://www.emuprim.lv/bildez/thumbs/...56khz_cap1.JPG
Cap 2:
http://www.emuprim.lv/bildez/thumbs/...56khz_cap2.JPG

At higher frequencies spikes start to appear if inductive load is being used. If I replace the motor with a light bulb, the waveform is essentially the same but without spikes. At 160Hz the motor runs faster than at 1.56KHz, but the spikes are in the same peak to peak range. Also I noticed the on transistors 1 and 3 the spikes are more positive, but on transistor 2 the spikes are more negative. Without the load I ran the circuit all night at 0.5Hz and the battery voltage dropped. Now I will let the circuit run using the motor as a load and running at 160Hz, as this is the frequency where the motor ran the fastest. I also noticed that at this frequency if a motor is used as a load, I can hear the noise it makes on a radio
Will see how it performs.
Sorry for the large post.
Thanks,
Jetijs

[Bit's-n-Bytes]

Hi all,
here is my report from the battlefield

Sorry for the large post.
Thanks,
Jetijs

Awesome work Jetijs This is exactly the type of data mining that is needed to determine the "Range" of oscillation frequency for what type of load is being serviced. This such a great forum with such brillant minds.

Keep up the great work, Team!!!!

Bit's

[Bit's-n-Bytes]

@Jetijs

I am working on the trace placements on the circuit board design as we speak. Can you tell me what the current looks like in the base of the tranny's?

Thanks

Bit's

[hherby]

Hi all,
here is my report from the battlefield

Jetijs

Hello Jetijs,

In your schematic, shouldn't the emitter of transistor 1 be going to the plus side of cap 2?

Do you have the load hooked up between the positive of the battery and the positive of cap 1?

Are you running the motor through a bridge?

Thanks

Alex

[blackchisel97]

Hello Jetijs,

In your schematic, shouldn't the emitter of transistor 1 be going to the plus side of cap 2?

Do you have the load hooked up between the positive of the battery and the positive of cap 1?

Are you running the motor through a bridge?

Thanks

Alex

I'm thinking the same..

Vtech

[Jetijs]

Alex, you are right, there was a mistake on my picture, just fixed it
I am using the load between the battery positive terminal and the positive terminal of the cap1

Bits, will do that tomorrow, I am out of my shop now
Thanks,
Jetijs

[hherby]

I am working on the trace placements on the circuit board design as we speak

Bit's,

I noticed from a pinout diagram I found that this PIC has only 3 ADC inputs. I attached the pinout I found. Is this the correct pinout diagram for this chip?
In6 and In7 do not appear to be ADC inputs. (Please bear in mind I have no knowledge of the Pic beyond the pinouts. I have worked with AVR micros and I am thinking of trying something similar using an ATMEGA88 in 28 dip pkg)

Are you planning on measuring voltage on the 4 sense points in your circuit as feedback?
It looks like they will all be referenced to the negative side of battery1. It looks to me that Sense D1 will always be zero because it is also common ground for the pic.

Would the Zener/resistor/cap power regulation circuit be fast enough to filter the spikes from hitting the pic on both positive and negative rails?

What about hooking it up as part of the output load like Leroy did with an auxilliary connection to precharge the output cap on DC side of output bridge where the AC input of the bridge would be connected to J5?

Just some observations. Hope this helps.

Alex

[blackchisel97]

rotating batteries

Hi guys,

I keep testing 3 batt. manual switch with rotation. I noticed slow but steady gain. They aren't great and haven't been fully charged (should they be?) before.
Their starting voltage was different however, I put the highest one in a large bank. Net gain per cycle is about 0.05V with 5W 12V bulb. I kept them "ON" for 15min and let it rest for 30min. When I keep them "ON" longer - 30 min, charging process slows down and "equalization" takes place. Also, while disconnected and swapped in places, they gain 0.01 - 0.02V immediately in their new "spot" without switch even turned on. This is something I don't get it.
I'll drag my big LABs and try circuit with them. Don't like SLABs too much.


Vtech