Lords of the Ring

[Thaelin]

@ Mannix:
     
        You wont say it but I will,    TUBES!.

         

[bolt]

In SM's chat doc he covers a whole chapter on his revelations of making audio amps of great purity. This is another clue I'm sure. Anyway i still think the 5.76KHz looks ok to me. But how do we divide this down? It kinda defeats the object but we might have to start square and sine it after but use a clock IC divider as its the only way i can think to make sure the phase is correct. Im sure there are other ways.  If we drift at all even by a nats  it wont work. Some control windings will become your hi fi speakers:)  The mos fets do make good audio amps i seen one many years ago that used the speaker hanging direct of the source to + supply. The advantage is it becomes a very wide bandwidth amp DC to VHF amp driving the control coil as our load. As the sine ramps up it looks like a magnet is passing by as in a conventional gen or SSG and i think this is what we need to achieve and needs to replicate this as faithfully as possible without spikes or harmonics. In effect its a DC power op amp and should do at least a 1 watt per channel. I think SM controlled all this on one switch then switch 2 was the heterodyne offset osc to produce the 7.83 Hz. Where does this go i dunno via a mixer to the 720Hz control coil maybe but this osc MUST be detuned and brought into focus very slowly. SM described the use of IC's in his control circuit and again i speculate it required some smart stuff to ensure it doesn't actually drift perfectly in tune to prevent a melt down. There will have to be feedback from the collector also. As we send 720 to one segment that will produce a small kick around the loop which will be seen by the next control coil. We take this small signal down from the collector coil and use this signal and send it another frequency and wait for the response on the next control coil and so on. After some seconds the collector wheel is "spinning".

This is why he uses a magnet! We are replicating a mechnical magnet spinning around at 120,000 RPM However, when the circuit is first turned on the circuit is waiting for a control feedback pulse but there is none until we move a magnet manually ourselves to the control coil. This produces the first tiny kick of the day which is picked out by the next collector control coil which then fires out the first frequncy then the sequence has perpetual feedback.

........more thoughts laters

[otto]

Hello,

why not multiplie it with 7??

Otto

[bolt]

bit more

"I have taken a high voltage power supply as follows:
500 v-0-500 v 300 mV plate transformer run it through a full wave silicon circuit then run it through a 5U4 electron tube rectifier.
Now you know that the 5U4 requires 5 volts AC at 3 amps for its heater to gather the electrons and complete the circuit.
Well, I measured the output from the tube and the result is 500 volts DC at 250 mV.
The loss is due to the high impedance of the tube and its limited ability to dissipate more than 250 mA..
The point I wish to make here is that also along with the 500 volt DC is, yes, you guessed it, the 5 volts three amp AC current!"

He is explaining that multiple signal can co exist on the loop. It just occurred to me that the 7.83 Hz heterodyne may coexist on all 3 frequencies and not just one. In practice this means using 2 master local oscillators with the second offset by 7.83 Hz.. Mmm maybe not as this difference error will be reduced by the divider. Well plenty of smart guys on here i let you figure this bit out but the division clock remains the same and all 3 frequencies are reproduced faithfully but mixed into the firing control coil. The result is the collector will always have this 7.83 Hz component to it. So its here we produce the scalar wave. So far the theory ties up with everything i seen.
Well the firing order and duration etc will have to be worked out on the bench.

[bolt]

Otto the ratio of the frequencies must be correct although we can change them. In addition its always better to start with a high master clock and divide down then errors are reduced on the division. The constructive wave components fired into the collector should add back up to the master clock at 5.76KHz. I feel this is correct because the units churn this out when measured by the inspectors. Of course there plenty of room for discussion but so many things make sense running these lower sine waves rather then whacking it with pulses even if these suggested frequencies are completely wrong.