Tonight I tried it with decent voltage regulators and was able to get very good results out of the ECL85. I set things up so I could vary the triode grid bias, the triode plate bias and the triode cathode bias. By controling all these I was able to get very linear results with the different triode grid points by changing the triode plate bias thus changing where the AC sat on the curve and then using the grid voltage to adjust the curve. By changing the plate bias I could go from needing positive to negative grid bias to get linear output from the pentode. I think I would rather have -1.2V grid rather than 16V positive with 15mA going into the grid!
The one thing I found was that it was very touchy! I was using single turn puts, just the slightest change in pot position would go from highly distorted one way to highly distorted the other. When I managed to get it just right, itwas wonderful. I'm thinking I need a multi turn pot with fixed resistors to get the voltage into the ball park.
It was very tricky to adjust things using this circuit since every adjustment caused the others to change.
With it setup the way I had it I was getting a current gain of about 15 which was actually working quite well for what I was doing. If I can get things a bit more stable I might actually take this off the bench and hook it up to the Lowthers and see how it sounds. That very linear output is looking very good.
Its interesting looking at the BDT. As you adjust the triode bias point the waveform on the plate is radically changed (it should of course since the transconductance needs to change the voltage to match the nonlinear resistance) , but the other plate not hooked up to a triode has the inverse wave form, I guess I should expect this since I'm using a PP plate choke.
I've been encouraged by these experiments, this looks like a very viable techique. Next I want to try a 6L6 output tube.
Thanks again for bringing this up. The combination of the BDT as the transconductance amp, the linearized pentode driving Lowthers is going to be a killer combination.
John S.
The one thing I found was that it was very touchy! I was using single turn puts, just the slightest change in pot position would go from highly distorted one way to highly distorted the other. When I managed to get it just right, itwas wonderful. I'm thinking I need a multi turn pot with fixed resistors to get the voltage into the ball park.
It was very tricky to adjust things using this circuit since every adjustment caused the others to change.
With it setup the way I had it I was getting a current gain of about 15 which was actually working quite well for what I was doing. If I can get things a bit more stable I might actually take this off the bench and hook it up to the Lowthers and see how it sounds. That very linear output is looking very good.
Its interesting looking at the BDT. As you adjust the triode bias point the waveform on the plate is radically changed (it should of course since the transconductance needs to change the voltage to match the nonlinear resistance) , but the other plate not hooked up to a triode has the inverse wave form, I guess I should expect this since I'm using a PP plate choke.
I've been encouraged by these experiments, this looks like a very viable techique. Next I want to try a 6L6 output tube.
Thanks again for bringing this up. The combination of the BDT as the transconductance amp, the linearized pentode driving Lowthers is going to be a killer combination.
John S.
all those pots to adjust!!
Yes, I know what you mean about all those pots to adjust and interacting too!
Using series diodes simplifies things a bit once you get the right number of diodes figured out. Only have to adjust the pentode bias then. Varying the input (or triode plate bias current) current only makes it operate at different operating points in the linear range then.
I guess saving tubes means gaining pots.
But once the design is settled, it should be better.
I am hoping to get around to using an X-Y plotter to graph a set of linearity curves (current in versus current out) for the VT current mirror for both the diode case and the triode case versus some of those voltage settings. Then I can scan them and post them. Everyone will finally know what all this esoterica is we are talking about then!
Might be interesting to listen to the amplifier while slightly adusting the linearity correction from the concave thru the linear to the convex curve. Maybe even a user controlled knob on the amplifier chassis to play with.
(sufficiently limited in range so as not to be able to blow the speaker out though)
Don
Yes, I know what you mean about all those pots to adjust and interacting too!
Using series diodes simplifies things a bit once you get the right number of diodes figured out. Only have to adjust the pentode bias then. Varying the input (or triode plate bias current) current only makes it operate at different operating points in the linear range then.
I guess saving tubes means gaining pots.
But once the design is settled, it should be better.
I am hoping to get around to using an X-Y plotter to graph a set of linearity curves (current in versus current out) for the VT current mirror for both the diode case and the triode case versus some of those voltage settings. Then I can scan them and post them. Everyone will finally know what all this esoterica is we are talking about then!
Might be interesting to listen to the amplifier while slightly adusting the linearity correction from the concave thru the linear to the convex curve. Maybe even a user controlled knob on the amplifier chassis to play with.
(sufficiently limited in range so as not to be able to blow the speaker out though)
Don
Haa! I love the "distortion knob" idea. Stright up, low distortion, you want a little second harmonic in there, go to the left, you want some third, go to the right.
My big problem now is I have run out of tube sockets, I need to go buy a case so I can continue experimenting. I could canabalize already existing stuff, but I like to compare what I did before with the new one, thats tough to do when you reuse the parts. I'm also going to buy a hundred 15 turn trimpots!!
The ECL85 SE works great but is not very high powered, I need a little more power even with the Lowthers, so on with the 6L6s.
John S.
My big problem now is I have run out of tube sockets, I need to go buy a case so I can continue experimenting. I could canabalize already existing stuff, but I like to compare what I did before with the new one, thats tough to do when you reuse the parts. I'm also going to buy a hundred 15 turn trimpots!!
The ECL85 SE works great but is not very high powered, I need a little more power even with the Lowthers, so on with the 6L6s.
John S.
I just use some cheap variable power supplies for setting the assorted bias voltages in my experiments:
http://www.mpja.com/productview.asp?product=14600+PS
http://www.mpja.com/productview.asp?product=14602+PS
Nice for working on SS amplifiers too, both have enough current capability to handle an odd filament voltage as well. The 50V model has a fine voltage adjust (5V delta) beside the coarse adjust, that works good for the sensitive corrector bias settings.
I also use a couple old Heathkit IP-17 tube voltage power supplies for plate and screen voltage.
Plus a large phenolic board with a bunch of assorted tube sockets with short tinned wire leads, and clip leads to connect them. Filament pins pre-wired. Of course, nothing permanent this way. But easy enough to set up something again if documented on paper.
I have some pre wired assembies of series connected 6JU8 and 9006 diodes with a string of terminals for tap points. Makes for easy adjustment of the number of series connected diodes using a clip lead. Maybe I will set up a pre-wired triode with adjustment pot once I find a triode I like for variable diode use.
Don B.
http://www.mpja.com/productview.asp?product=14600+PS
http://www.mpja.com/productview.asp?product=14602+PS
Nice for working on SS amplifiers too, both have enough current capability to handle an odd filament voltage as well. The 50V model has a fine voltage adjust (5V delta) beside the coarse adjust, that works good for the sensitive corrector bias settings.
I also use a couple old Heathkit IP-17 tube voltage power supplies for plate and screen voltage.
Plus a large phenolic board with a bunch of assorted tube sockets with short tinned wire leads, and clip leads to connect them. Filament pins pre-wired. Of course, nothing permanent this way. But easy enough to set up something again if documented on paper.
I have some pre wired assembies of series connected 6JU8 and 9006 diodes with a string of terminals for tap points. Makes for easy adjustment of the number of series connected diodes using a clip lead. Maybe I will set up a pre-wired triode with adjustment pot once I find a triode I like for variable diode use.
Don B.
Variable geometry diodes at last!!
After working with fixed biased triodes as diode correctors, I am still unhappy with their performance. Always some kind of kink in the transfer at lower currents. I am beginning to think that changing the grid bias just causes a changeover from a cathode to plate geometry, to one of grid as cathode to plate geometry. With the changeover happening at different current thresholds depending on grid bias.
Some further experimenting has lead to a true variable diode geometry technique with excellent linear results. What I have found is that one can use a ferrite donut magnet from an old speaker to modify a diode's characteristics. Placing the donut magnet over a 6JU8 diode tube causes the current gain of the VT current mirror stage to increase up to about 3 times depending on position of the magnet. Also, the equivalent "# of diodes in series" factor increases as well, but not by as much. Rough guess would be by the square root of the current gain factor.
My guess is that the magnetic field causes the low potential difference cathode electrons to spiral around the cathode on their way to the plate, effectively increasing the distance from cathode to plate, which would give the same results as observed. Linearity remains excellent. (of course one still has to get the # of diodes in series factor and the diode biasing voltage correct for a straight line linear transfer curve)
One can also observe the opposite (but not so desirable) effect by placing the donut magnet over the pentode tube in the current mirror. The current gain decreases but remains linear (with a small touch up of bias voltages). The magnetic field is effectively decreasing the transconductance, or Gm, of the pentode.
The magnet can also be used on the "triode as diode" corrector element with the upper current portion of the transfer curve increasing in gain. But the "triode as diode" curve still is somewhat messy at low currents.
To make practical use of this magnet technique, one will probably want to place a 3 inch section of large diameter steel pipe over the outside of the magnet to prevent its field from affecting other nearby tubes. An electromagnet could of course be used as well, but would be inefficient, requiring an additional power supply, but might be made more compact. What one really needs is a sleeve like ring, permanent magnet, that just fits over the tube. Maybe could tape a bunch of small ferrite rods or neodymium bar magnets around the tube. Still needs shielding.
Don B.
After working with fixed biased triodes as diode correctors, I am still unhappy with their performance. Always some kind of kink in the transfer at lower currents. I am beginning to think that changing the grid bias just causes a changeover from a cathode to plate geometry, to one of grid as cathode to plate geometry. With the changeover happening at different current thresholds depending on grid bias.
Some further experimenting has lead to a true variable diode geometry technique with excellent linear results. What I have found is that one can use a ferrite donut magnet from an old speaker to modify a diode's characteristics. Placing the donut magnet over a 6JU8 diode tube causes the current gain of the VT current mirror stage to increase up to about 3 times depending on position of the magnet. Also, the equivalent "# of diodes in series" factor increases as well, but not by as much. Rough guess would be by the square root of the current gain factor.
My guess is that the magnetic field causes the low potential difference cathode electrons to spiral around the cathode on their way to the plate, effectively increasing the distance from cathode to plate, which would give the same results as observed. Linearity remains excellent. (of course one still has to get the # of diodes in series factor and the diode biasing voltage correct for a straight line linear transfer curve)
One can also observe the opposite (but not so desirable) effect by placing the donut magnet over the pentode tube in the current mirror. The current gain decreases but remains linear (with a small touch up of bias voltages). The magnetic field is effectively decreasing the transconductance, or Gm, of the pentode.
The magnet can also be used on the "triode as diode" corrector element with the upper current portion of the transfer curve increasing in gain. But the "triode as diode" curve still is somewhat messy at low currents.
To make practical use of this magnet technique, one will probably want to place a 3 inch section of large diameter steel pipe over the outside of the magnet to prevent its field from affecting other nearby tubes. An electromagnet could of course be used as well, but would be inefficient, requiring an additional power supply, but might be made more compact. What one really needs is a sleeve like ring, permanent magnet, that just fits over the tube. Maybe could tape a bunch of small ferrite rods or neodymium bar magnets around the tube. Still needs shielding.
Don B.
Hello Nixie,
The idea is working well with thermionic diodes as input devices, but not so well using a triode to make a variable geometry parameter diode. The magnetic axial field on a cylindrical diode works fine for making a variable geometry diode, but is not so convenient in practice due to the extra power supply for an electromagnet coil.
The requirement for several diodes in series, in practice (and generally requiring custom selecting of how many for each new gain tube to be linearized), for ordinary fixed geometry tubes available (6JU8, ...) makes the idea somewhat impractical except for the die hard experimenter.
Imperfections in tube construction also cause practical tube current mirrors to only correct lower order harmonics well, only reaching perfect linearity in theory.
A practical fixup for this problem does exist however: Hawksford style error correction loop, an unusual form of negative feedback that does not require high gain (in fact, only unity gain). Error Correction has been applied to solid state current mirrors to fix their imperfections successfully (published results are on the internet).
For the tube current mirror case, this would require two more tube gain elements (couple of pentodes in diff. amp. config.). Now we are up to a total of three tube gain elements and a series string of thermionic diodes to replace the original single conventional gain stage. Purists, who may have been attracted to the original current mirror's ideal linearity without feedback, are likely to be put off by the Error correction loop at this increased difficulty level. (you also need some form of distortion analyzer when working at the low distortion residuals remaining using the error correction add-on.) Doing the error correction loop with a couple of transistors instead would likely give excellent results (especially with matched transistors like the MAT02 series or similar), but then we have the silicon in the tube circuit issue.
Nevertheless, a fun idea to play with on the bench!
Don
The idea is working well with thermionic diodes as input devices, but not so well using a triode to make a variable geometry parameter diode. The magnetic axial field on a cylindrical diode works fine for making a variable geometry diode, but is not so convenient in practice due to the extra power supply for an electromagnet coil.
The requirement for several diodes in series, in practice (and generally requiring custom selecting of how many for each new gain tube to be linearized), for ordinary fixed geometry tubes available (6JU8, ...) makes the idea somewhat impractical except for the die hard experimenter.
Imperfections in tube construction also cause practical tube current mirrors to only correct lower order harmonics well, only reaching perfect linearity in theory.
A practical fixup for this problem does exist however: Hawksford style error correction loop, an unusual form of negative feedback that does not require high gain (in fact, only unity gain). Error Correction has been applied to solid state current mirrors to fix their imperfections successfully (published results are on the internet).
For the tube current mirror case, this would require two more tube gain elements (couple of pentodes in diff. amp. config.). Now we are up to a total of three tube gain elements and a series string of thermionic diodes to replace the original single conventional gain stage. Purists, who may have been attracted to the original current mirror's ideal linearity without feedback, are likely to be put off by the Error correction loop at this increased difficulty level. (you also need some form of distortion analyzer when working at the low distortion residuals remaining using the error correction add-on.) Doing the error correction loop with a couple of transistors instead would likely give excellent results (especially with matched transistors like the MAT02 series or similar), but then we have the silicon in the tube circuit issue.
Nevertheless, a fun idea to play with on the bench!
Don
Hi Nixie,
I agree with your assessments, but keep in mind that error correction sits right on the edge of oscillation with crossover from stability at the unity gain setting. So it requires a critical gain pot. setting which might later drift into instability. So is not likely to catch on commercially any time soon. But seeing as we Diyers are allowed to break the manufacturablity laws:
see here for the Hawksford error corrected current mirror writeup (page 4 , figure 6 of the .pdf file or the .jpg picture below )
http://www.diyaudio.com/forums/showthread.php?postid=562818#post562818
For our tube mirror version, we would flip all the p and n types to n and p respectively and supply polarities too, with T1 being our thermionic diode string, T2 our gain stage (pentode), T3 and T4 and Rx are the error correction circuitry. T3 and T4 should be a matched pair and best if contained in the same case for temperature tracking.
Our mirror design has current gain, unlike the one shown, so the R,R resistors in the emitter circuit of T2 (to Vs) need to be scaled down by the expected current mirror gain factor to get the same voltages to appear at T3 and T4 bases (as well as T1 and T2 emitters).
The DC idle currents from T3 and T4 collectors balance in the original no gain mirror design, but present a problem in our design with the different R,R resistors now. A high value resistor from the collector of T3 to B+ (or better a regulated supply) could be added to act as a DC current source to compensate (needs to supply DC idle current of T3 (or T4) times the expected current gain - 1. (ie, same DC idle voltages to appear on the mid points of each pair R,R from side to side)
The Rx (a gain degeneration resistor for the diff. amp. stage T3,T4 to drop down to unity loop gain for the error correction) can be split I think, and just a resistor to ground or some bias Voltage (from the split point, like a long tail pair config.) substituteded for the two Io current sources.
We also need some provision for negative grid bias for T2 in our tube version, and this should be do-able with the usual cathode bias RC shunt pair in the T2 "emitter" lead. (between T2 cathode and T4 base) or could try just a high value grid bias resistor from the T1 emitter (diode string cathode end) and cap couple signal from T1 collector (diode string anode end) to T2 (grid).
I haven't actually built this circuit, so the usual caveats are in order here. Good luck! I would be working on this myself now except some of my benchtesting equipment got put into storage mistakenly due to a recent move. I will back on the air I hope by late Spring here.
Don
I agree with your assessments, but keep in mind that error correction sits right on the edge of oscillation with crossover from stability at the unity gain setting. So it requires a critical gain pot. setting which might later drift into instability. So is not likely to catch on commercially any time soon. But seeing as we Diyers are allowed to break the manufacturablity laws:
see here for the Hawksford error corrected current mirror writeup (page 4 , figure 6 of the .pdf file or the .jpg picture below )
http://www.diyaudio.com/forums/showthread.php?postid=562818#post562818
For our tube mirror version, we would flip all the p and n types to n and p respectively and supply polarities too, with T1 being our thermionic diode string, T2 our gain stage (pentode), T3 and T4 and Rx are the error correction circuitry. T3 and T4 should be a matched pair and best if contained in the same case for temperature tracking.
Our mirror design has current gain, unlike the one shown, so the R,R resistors in the emitter circuit of T2 (to Vs) need to be scaled down by the expected current mirror gain factor to get the same voltages to appear at T3 and T4 bases (as well as T1 and T2 emitters).
The DC idle currents from T3 and T4 collectors balance in the original no gain mirror design, but present a problem in our design with the different R,R resistors now. A high value resistor from the collector of T3 to B+ (or better a regulated supply) could be added to act as a DC current source to compensate (needs to supply DC idle current of T3 (or T4) times the expected current gain - 1. (ie, same DC idle voltages to appear on the mid points of each pair R,R from side to side)
The Rx (a gain degeneration resistor for the diff. amp. stage T3,T4 to drop down to unity loop gain for the error correction) can be split I think, and just a resistor to ground or some bias Voltage (from the split point, like a long tail pair config.) substituteded for the two Io current sources.
We also need some provision for negative grid bias for T2 in our tube version, and this should be do-able with the usual cathode bias RC shunt pair in the T2 "emitter" lead. (between T2 cathode and T4 base) or could try just a high value grid bias resistor from the T1 emitter (diode string cathode end) and cap couple signal from T1 collector (diode string anode end) to T2 (grid).
I haven't actually built this circuit, so the usual caveats are in order here. Good luck! I would be working on this myself now except some of my benchtesting equipment got put into storage mistakenly due to a recent move. I will back on the air I hope by late Spring here.
Don
Hi Nixie,
Arghh! Missed your reference before it got censored. Which paper?
I guess you are not referring to the stage gain being greater than one , but maybe to splitting the correction into part error feedforward, so the error feedback can be a little less than unity. Looks like the error corrected mirror design actually does that by using both collector outputs from the diffl. amp. stage. That would nicely fix the stability problem. Would be interesting to apply that idea to a SS ECd output stage too maybe.
Don
Arghh! Missed your reference before it got censored. Which paper?
I guess you are not referring to the stage gain being greater than one , but maybe to splitting the correction into part error feedforward, so the error feedback can be a little less than unity. Looks like the error corrected mirror design actually does that by using both collector outputs from the diffl. amp. stage. That would nicely fix the stability problem. Would be interesting to apply that idea to a SS ECd output stage too maybe.
Don
Regarding your signature, even the primary researcher of 2T physics, Itzhak Bars, states in this preprint that "The physical phenomena in 1T or 2T physics are not different, but the spacetime formalism used to describe them is," and, "2T-physics could be viewed as a device for gaining a better understanding of 1T-physics." So even he views 2T physics, as someone commented on a different forum, "as a tool rather than theories of independent interest."
Well, I wouldn't cancel the time travel preamp just yet!
GUT physics and string theory both posit extra spatial dimensions, which may at this point be considered as "theoretical tools" also, but may prove to be real as well (waiting for LHC results on this). Supersymmetry allows for 2 time dimensions in its maximal extension, becoming "F" theory, or related, in string theory.
Even in the currently popular "M" theory, the "single" time dimension is a "complex" dimension. This may allow one to circumvent the speed of light limitation by "going around it" with v=ki (i=sqrt(-1)) This will require some mastery of spin physics and the weak force.
Once around the v=c pole, travel at v>c is mathematically ordinary and would allow travel into the past by the usual accelerated traveler scheme.
However, I must point out that the usual sci fi image of time travel to the past is incorrect. Time travel will modify the traveler to be consistent with the time, just as we are now being modified to be consistent with the future, as we currently time travel at the "speed of light" or "speed of time" into it. (We simply all have the same linear time momentum.)
This all works much more practically in a 2T space however, where time linear momentum and time angular momentum conservation rules allow much more interesting but still causality preserving pathes. (the big bang then becoming the result of our all having the same angular time momentum as well, producing a time horizon from curvature of about 14 billion years.) Due to the time momentum conservation rules, one will have to interact with some time-reversed or time orthogonal dark matter in order to change ones time "course". I would suggest the hard to find "will-o-the-wisp" lights or "spook" lights as suitable linkages to this time altered dark matter. (Please leave your tube supplies to me for storage while exiting the currently observable universe! --- no one has been seen again after stepping thru one of these!)
Don
Was just thinking a little about the error correction idea. It works well for fixing gain stages that are nearly linear. How about using it to fix up the non-linearity imposed on a triode stage by a real load impedance.
For that matter, maybe there is even a theoretically perfect fix, using a negative impedance in the circuit somewhere. Maybe have to consider some impedance in the grid circuit when driven in the grid current range. (yes, its cheating to just put the the neg. impedance across the load to cancel it.)
smoking-amp said:
Well, I wouldn't cancel the time travel preamp just yet!
Cool! Let the my soon-to-be patented Temporal Topology™ (based around causal-followers, naturally) roll on!
I don't wish to divert this thread any longer, but I'm surprised to hear that you can say this with some certainty. It implies that you could never take a time-travel machine back past before it existed too, no?smoking-amp said:
Perhaps you could instead, point me in the direction of some literature that specifically adresses this?
Well, in 2T space-time(s) there is no preferred direction in time that can be called a future or past. (same as relativity) Our local initial conditions determine the direction we call the future. But no matter what direction one is traveling in 2T time, one can only electromagnetically observe similarly traveling objects, hence the illusion of fixed future and past directions. They are only OURS. Someone elses (dark matter aliens say) future might be in the direction of our past for example. So we ARE time traveling right now in the fully general sense. Let me know if you find someone who is NOT changing in time!
The second issue of closed time loop causality is a little more difficult to show. Looking at quantum physics from the path integral perspective, the outcome of an interaction is the most probable sum over ALL possible paths. When no preferred direction in time is allowed, then interaction outcomes must be compatible with all directions in time, hence the " local future" is always compatible with the "local past" as well as ALL possible pasts and vice versa. (ie, you cannot put a preferred arrow on causality either, if it is true in one time direction it must be true in all time directions.) (the present 3D+T QM theory still manages to work when ignoring the other time directions because of the typical lack of interactions in those directions in "low" (below GUT level) energy experiments)
I don't believe there is any constraint on time machines (by which I mean, a device capable of altering time momentum direction of some conventional matter via an interaction with differing time momentum matter), they either have always (our past, future or any other time direction) been here if they exist OR they are ruled out everywhere (TBD). To rule them out would be kinda like ruling out steering on cars though, so probably are around. You need GUT physics capability to build one though, since simple electromagnetic interactions won't do. (try jumping thru a "spook" light!, might be a naturally occuring one, but who knows where or when you might end up, or even if in one piece!)
Don
The second issue of closed time loop causality is a little more difficult to show. Looking at quantum physics from the path integral perspective, the outcome of an interaction is the most probable sum over ALL possible paths. When no preferred direction in time is allowed, then interaction outcomes must be compatible with all directions in time, hence the " local future" is always compatible with the "local past" as well as ALL possible pasts and vice versa. (ie, you cannot put a preferred arrow on causality either, if it is true in one time direction it must be true in all time directions.) (the present 3D+T QM theory still manages to work when ignoring the other time directions because of the typical lack of interactions in those directions in "low" (below GUT level) energy experiments)
I don't believe there is any constraint on time machines (by which I mean, a device capable of altering time momentum direction of some conventional matter via an interaction with differing time momentum matter), they either have always (our past, future or any other time direction) been here if they exist OR they are ruled out everywhere (TBD). To rule them out would be kinda like ruling out steering on cars though, so probably are around. You need GUT physics capability to build one though, since simple electromagnetic interactions won't do. (try jumping thru a "spook" light!, might be a naturally occuring one, but who knows where or when you might end up, or even if in one piece!)
Don
Hmmm... Unfortunately my book and journal library got put in storage 700 miles away when I moved recently. Most of these ideas I have seen separately in speculative articles in Scientific American over the last few years, although I have not seen the whole logically put together. I am sure most any of the popular science journals (New Scientist, Physics World, Physics Today, Discover,....) would have covered these areas under the topics of Cosmology, String Theory, Extra dimensions, etc.
I know I have seen a few speculations on the big bang being just a curved surface instead of a singularity. Also several mentions of classes of dark matter being time altered ordinary matter. I think it was Freund's book on supersymmetry that I first saw the 2T case in. There are a number of old journal articles (50s) covering the inclusion of the time reversed solutions of electromagnetic interactions and how they could be used to explain causality and spooky action at a distance in quantum interactions. Putting this together with 2T space and no absolute time direction seems obvious, QM even starts to make sense.
On causality time loops I have only seen the usual throwing up of the hands in despair and head banging against the wall comments. But I think this is the result of missing the significance of interactions along the time path having no preferred direction other than initial conditions. There was an article I heard mentioned on arXiv.org I think that performed some particle collision experiment (in theory) with a time reversed particle attempting to disrupt the initial conditions and they showed, using string theory I believe, that it couldn't do that, ie self consistency or causality was preserved. Just becomes mysterious for everyone when extrapolated to macroscopic intelligent action.
1T physics time travel would be just like playing a movie backwards, and you don't see characters murdering their parents before they were born in backward movies. 2T physics might allow one to go back and be someone else, but the past is already written and no strangers murdered ones parents before we were born, so you just have to get used to multi-directional causality interaction along the path constraining free will or memory somehow.
There might even be some rule such as identical quantum states cannot occupy the same time and place and consciousness could be a type of quantum state. Then you would never be able to get back to exactly the same place you had been, freak interactions would keep bumping you off course. And the closer you got to the destination, the stronger the directional corrections needed to get back on course, until you would be reaching Planckian energy levels (something like the total energy of the universe required) within a Planck unit of your destination.
Don't forget that changing ones time momentum requires an equal but opposite effect on some dark matter and you would still be constrained to be consistent with its altered course. At Planck distances to your objective you would be fighting the gravitational forces of the entire universe. OK, so enough wild speculation.
Better get back to vacuum tubes before we get censored by someone with time travel capabilities on the thread server.
Don
I know I have seen a few speculations on the big bang being just a curved surface instead of a singularity. Also several mentions of classes of dark matter being time altered ordinary matter. I think it was Freund's book on supersymmetry that I first saw the 2T case in. There are a number of old journal articles (50s) covering the inclusion of the time reversed solutions of electromagnetic interactions and how they could be used to explain causality and spooky action at a distance in quantum interactions. Putting this together with 2T space and no absolute time direction seems obvious, QM even starts to make sense.
On causality time loops I have only seen the usual throwing up of the hands in despair and head banging against the wall comments. But I think this is the result of missing the significance of interactions along the time path having no preferred direction other than initial conditions. There was an article I heard mentioned on arXiv.org I think that performed some particle collision experiment (in theory) with a time reversed particle attempting to disrupt the initial conditions and they showed, using string theory I believe, that it couldn't do that, ie self consistency or causality was preserved. Just becomes mysterious for everyone when extrapolated to macroscopic intelligent action.
1T physics time travel would be just like playing a movie backwards, and you don't see characters murdering their parents before they were born in backward movies. 2T physics might allow one to go back and be someone else, but the past is already written and no strangers murdered ones parents before we were born, so you just have to get used to multi-directional causality interaction along the path constraining free will or memory somehow.
There might even be some rule such as identical quantum states cannot occupy the same time and place and consciousness could be a type of quantum state. Then you would never be able to get back to exactly the same place you had been, freak interactions would keep bumping you off course. And the closer you got to the destination, the stronger the directional corrections needed to get back on course, until you would be reaching Planckian energy levels (something like the total energy of the universe required) within a Planck unit of your destination.
Don't forget that changing ones time momentum requires an equal but opposite effect on some dark matter and you would still be constrained to be consistent with its altered course. At Planck distances to your objective you would be fighting the gravitational forces of the entire universe. OK, so enough wild speculation.
Better get back to vacuum tubes before we get censored by someone with time travel capabilities on the thread server.
Don
Oh, just remembered. The time reversed interaction theory papers were referred to as the Wheeler-Feynman theory papers.
Might be able to find something on that on the WWW. Seem to recall that causality traced back to the initial singularity of the big bang. Also explains entropy in thermodynamics due to expansion and cooling of "universe".
Might be able to find something on that on the WWW. Seem to recall that causality traced back to the initial singularity of the big bang. Also explains entropy in thermodynamics due to expansion and cooling of "universe".
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.