# Experiment to demonstrate mutual observed time dilation

Discussion in 'Physics & Math' started by superluminal, Apr 11, 2005.

1. ### superluminalI am MalcomRValued Senior Member

Messages:
10,876
Ok. I have this experiment in mind to show mutually observed time dilation. It is rough but provides a place to start (assuming anyone is interested). Can this be simplified? Is there a much simpler way to do it that involves less sophisticated electronics? (My first approach is to go for the electronics since that's my field).

Check my calculations. Overall, is this feasible? Comments, criticisms, enhancements requested.

[hit your browsers STOP button if the motion makes you queasy]

----------------------------------------------------------------------------------------------------------------------------
BRIEF DESCRIPTION:

C1 spins about C2 at high speed. They each measure the others tick rates and compare to their own. The comparison results are digitally transmitted to R and after a day we see what we will see.

Expected results per SRT:

C1 will see the frequency from C2 as slower and transmit that fact, as digital data, to R.
C2 will see the frequency from C1 as slower and transmit that fact, as digital data, to R.

----------------------------------------------------------------------------------------------------------------------------
ELECTRONIC HARDWARE DESCRIPTION:

C1 and C2 are precision clock/counter/transceivers. R is a two channel receiver (9600 baud FSK modulated over 70Mhz/72MHz).

C1 contains a 1.3 GHz oscillator/counter, a 1.3GHz receiver/counter, a 1GHz transmitter, and a 70MHz data transmitter.

C2 contains a 1.0 GHz oscillator/counter, a 1.0GHz receiver/counter,a 1.3GHz transmitter and a 72MHz data transmitter.

C1 receives and counts C2’s 1.3GHz transmission (f1) and generates a comparison to it's own 1.3GHz counter. This comparison is digitally transmitted to R. (upper left diagram)

C2 receives and counts C1’s 1.0GHz transmission (f2) and generates a comparison to it's own 1.0GHz counter. This comparison is also digitally transmitted to R.

The diameter of the C1 orbit is 200 meters and C2 is considered a point source.

Each timing assembly (TA) is temperature controlled to +/-1C

The dual frequencies for comparison and data transmission are simply for channel seperation.

----------------------------------------------------------------------------------------------------------------------------
ERROR SOURCES:

1) Crystal oscillators change frequency under acceleration. Determine by calibration.
2) Crystal oscillators drift with time. A 1GHz osc with 1ppm/year aging rate may drift 2.7ns per day. Determine by calibration.

----------------------------------------------------------------------------------------------------------------------------
BASIC ASSUMPTIONS:

1) Since there is no radial motion between C1 and C2 (it is all transverse), the only time difference effects will be due to relativistic dilation and error sources.

2) We are not trying to quantify to any accuracy the magnitude of the dilation effect which we don't have the timing resolution for. We are looking for significant, unambiguous mutually observed time dilation.

----------------------------------------------------------------------------------------------------------------------------
CALCULATIONS:

- C1 rotation rate:

RPM = 15
RPS = RPM/60 = 0.25

Transverse velocity for C1 = (pi x 200m) x 0.25 = 157.079m/s

- Centripetal acceleration:

A = v^2/r = (157.079m/s)^2/100m = 246.73m/s = 25g’s

So the electronics must function at this g level. (easily doable for a correctly built SMT assembly)

- Dilation results:

Gamma = 1/sqrt(1-(157.079^2/300e6^2)) = 1.000000000000137

Over 24 hours a time difference of 86400 x Gamma – 86400 = 11.843e-9s could be expected due to velocity alone.

----------------------------------------------------------------------------------------------------------------------------
MOTION HARDWARE DESCRIPTION:

How to get a mass of a pound or so moving in a circle at 157m/s?
- Linear accelerator?
- High speed motors?

----------------------------------------------------------------------------------------------------------------------------
EXPERIMENTAL PROCEDURE:

1) Calibrate:
a) Quantify stationary oscillator drift over 1 week for C1 and C2.
b) Run C1 at speed and quantify the magnitude of acceleration frequency shift.
c) Repeat cal (a,b) three times to develop consistency in results.​

2) Run:
a) Reset all counters at rest.
b) Run system at speed for 86400 seconds while recording difference measurements at R.
c) Analyze data by incorporating error sources to isolate relativistic dilation effect.​

Last edited: Apr 13, 2005

3. ### kevinalmRegistered Senior Member

Messages:
993
Error source 3.

Quartz is temp sensitive. You're probably going to have to use crystal ovens.

>>edit Temp control for the whole assembly would be best.

Last edited: Apr 12, 2005

Messages:
10,876
Thanks kev,

7. ### superluminalI am MalcomRValued Senior Member

Messages:
10,876
I agree, but a 200m diameter track might have to be outside so I only thought to control the timing assy's themselves.

8. ### Neddy BateValued Senior Member

Messages:
1,589
Superluiminal,

I think this experiment is a great idea, but I am afraid that you will be disappointed if you are expecting to observe mutual time dilation. Since C1 will be the clock with the slower rate, C2 will have to be the clock with the faster rate. The reason is because all observations are viewed in the laboratory's frame of reference.

I don't think the laboratory can ever really obtain data "as viewed from" C1. Rearranging the clocks so that they both revolve would only create the illusion of mutual time dilation since both clocks would dilate evenly relative to a clock at R.

I believe that we will never have empirical evidence for reciprocal time dilation, because we can only exist in one reality at a time. Reciprocal time dilation seems to require that there are multiple realities.

Sunday night, April 17th, there will be a one hour program about Einstein on the Science channel. I think it is at 9:00pm EST. I am hoping that it (or someone on this board) will set me straight because the idea that the universe is bifurcating into multiple realities is causing me great difficulty.

Last edited: Apr 12, 2005
9. ### kevinalmRegistered Senior Member

Messages:
993
Yeah, I just worry about temp drift, bias levels changing junction capacitance and pulling the osc freq. That sort of thing. Just an old e-tech thinking about how to fix the stuff you EE types dream up.

Seriously, it's an interesting problem in practicle electronics. I was thinking though that instead of an outdoor track NASA or the USAF has to have a centrifuge for pilot testing they don't use much. Considerably smaller r though so the g forces will be much higher.

10. ### superluminalI am MalcomRValued Senior Member

Messages:
10,876
Well, Ned,

C1 is doing a comparison in it's own frame and transmitting the results to R in the lab via a digital data stream. The data from C1 will reflect what C1 measures as the comparison. Yes?

11. ### superluminalI am MalcomRValued Senior Member

Messages:
10,876
We depend on you guys for that!

The existing centrifuge in a good idea, but you are right, the g forces get to be a problem for small diameters that still give a measurable dilation as compared to crystal aging and such..

12. ### superluminalI am MalcomRValued Senior Member

Messages:
10,876
I think everyone is putting too much weight into the 'different realities' regarding the frames of SRT. Yes, time, length, and apparent mass changes for objects moving very fast wrt an observer, but so what? Reality has always been relative.

What you should really worry about is the 'many worlds' interpretation of QM. How many realities can you handle?

:m:

13. ### Neddy BateValued Senior Member

Messages:
1,589
I apologize; in my previous post I had C1 and C2 reversed since I could not see your animation while I was typing. I edited that post to correct that.

I would like to agree with you that the data from C1 is somehow 'current' for lack of a better term. But loss-of-simultanaety precludes us from assuming that it is. I think the results of the experiment will show that there is no way to circumvent the fact that the transmision data is always in the lab frame.

From the perspective of C1, it is the lab and C2 which are revolving around it. It would be from that reference frame only that C2 ticks slower than C1. But please note that for all moments of time in which C1 observes C2 as ticking slower, the time in C1's reference frame is ahead of anything that can ever be seen from the lab frame. You are hypothesising that C1 will send a transmission from what can only be considered as 'the future' in the lab frame.

I believe you will put more weight on the concept of 'different realities' if your experiment does not provide you with evidence of reciprocal time dilation. I hope you will remember to credit me for providing you with an explanation, still consistent with SRT, as to why your device did not detect reciprocity.

It is becoming apparent to me that there must be an infinite number of realities. I would prefer to be wrong about that, but so far it seems to be the only explanation.

14. ### Neddy BateValued Senior Member

Messages:
1,589
I propose one small change to your experiment so that it will not suffer from the simultanaety problems I described in my two posts above. Unfortunately it makes the experiment more difficult from an engineering point of view. You might have to let the centrifuge run for a longer time.

Mount a video camera behind each clock such that it can record its own clock and the distant clock simultaneously. Playing each tape back later will provide a direct comparison of what appeared to be occurring in each frame of reference. Deduct the tiny delay caused by the light having to traverse the distance between the clocks, and you have a real comparison. The lab frame is no longer influencing the the state of the data.

15. ### MacMRegistered Senior Member

Messages:
10,104
Superluminal,

Good to see you trying to get your feet wet on this. I notice others too that are starting to think for a change. That is worthwhile.

I do have to question this however:

If this data is correct I would have to ask why we have gone to Atomic clocks to be able to simply measure a few us/day. You have these crystals performing 1,000 times better?

I understand that to get data that is not at the error level the atomic clocks must be better than the output data but are you sure of your crystal accuracies and stability figures?

One other suggestion. Move C1 out of the center into a smaller orbit and have each transmit their accumulated data to each other as well as R.

As configured C1 has no velocity. By giving each a velocity you will have a direct comparison between "Relative Velocity" and "R" which would be the local common rest frame for each.

You can now compare computations of both views. Also you can eliminate the "Simultaneity" issue by merely reviewing tick rates and not accumulated time which requires synchronized start/stop times.

That is all time counting can be done by "R" using a counter to process the broadcast tick rate of each clock.

Last edited: Apr 12, 2005
16. ### Quantum QuackLife's a tease...Valued Senior Member

Messages:
18,224

SL correct me if I am wrong here but I can not see this experiment working for the simple reason that to do so the center of rotation has to change when you switch frames.

For some reason we seem to have missed this vital point.

Remembering that the frame you use to discover mutual dilations has to be at rest.

C1 rotates around C2 .....C1 has all the velocity.[thus dilated]

C2 rotates around C1........C2 has all the velocity.[thus dilated]

To do the test as is you will not get reciprication of dilation simply because one frame only is in the center of roatation therefore at rest.

17. ### superluminalI am MalcomRValued Senior Member

Messages:
10,876
This is a fine idea, however, I think the idea is to measure the mutual dilation while the observers are in motion, not after C1 stops. The comparison data generated in each frame will not be affected by being transmitted to R since it is in a digital format. This is the essence of "Bob says Mary's clock is slow, while Mary says Bob's clock is slow". Since all motion in this experiment is transverse, there is no radial doppler to worry about (time dilation due to transverse motion is called 'transverse doppler').

This is the oscillator I'm referencing:

http://www.vectron.com/products/xo/co286w.htm

I can't speak about the accuracy of atomic clocks vs crystal oscillators. At aging of 1ppm (which is the same as 1000ppb) after one year, a 1e9 hz (1GHz) oscillator will have drifted 1000ns. 1000/365 = 2.74ns/day. Check me. Do you agree with this?

If I expect 11.8ns per day dilation effect, then 2.7ns osc drift is acceptable. A note I neglected to put in the assumptions section (which I will add) is that I am not trying to quantify the dilation effect. I am only looking for each clock to show as unambiguously dilated wrt the other. Probably atomic clocks are much more stable as necessary to actually quantifying the effects.

As I've argued before, the relative transverse velocity between C1 and C2 is 157m/s. If this is an issue let's discuss as it does complicate the experiment.

Unfortunately the instantaneous tick-rate difference at a gamma of
1.000000000000137 is 1.37e-13 sec. I have no way to measure that, hence the one day accumulated difference. If we sync the clocks at rest and then spin C1 up, we don't have a problem right?

Good input! Keep going!

Messages:
2,471

Geistkiesel

19. ### superluminalI am MalcomRValued Senior Member

Messages:
10,876
Geist,

Don't know if that's simpler. I think simple circular motion will be easier to pull off. Plus I don't like the fact that you are shifting frames at each 'turn around'. Complicates things. As far as measuring both clocks running slower than the other, that's the whole point of the experiment. SRT says it must happen.

I'm actually seriously considering doing this. I can build the electronics (I have lots of contacts locally, in the industry, and at the university). The harder part seems to be the physical setup. A large diameter track with an object moving at approx. 352 MPH is a challenge...

20. ### MacMRegistered Senior Member

Messages:
10,104
Yes, but I question your assumption that "drift" equates to stability and accuracy. I too have an electronics background and I can say that the stability and accuracy you are seeking is not going to be easy to achieve.

Ovens will certainly be required and you won't be using a simple relaxation oscillator. Note the 6db noise degradation per octave above 500MHz!

I don't see drift as a problem. Your testing periods will be short term and a comparison at mutual rest would suffice even without synchronization of the frequency.

One issue which I think you may have forgotten however, using crystals is the impact of acceleration forces on the stress hence frequency of the crystal.

One can apply a stress and shift frequency with no motion such that motion could infact show a change in frequency unrelated to velocity perse.

I do believe it is an issue. If I were only interested in winning a debate I would encourage you to build your unit as described since from C2's perspective C1 has no velocity, it is the rotational rest axis for C2.

The modification I proposed is to give both C1 and C2 comperable but different velocities such that the C1 velocity relative to C2, and vice versa, affect can be compared to the individual composite velocity affects relative to the R axis monitor.

If you assume the acceleration doesn't alter the crystal frequency; which I believe it will. That old "ficticous" centrifugal force will put stress n the crystal changing its frequency.

One method around these issues would be independant velocity controls of C1 and C2 such that you could run each at equal centrifugal force and then at equal velocity, plus run a series of veloicty data runs for both C1 and C2.

It is somewhat complex but comparison of data should indicate the affect you seek if it is there. That is any change in frequency may not correspond to an SRT calculation but you are not interested in absolute data but "relative data".

So this arrangement would work for your purpose. Keep going. One suggestion would be the leasing of some centrifuge time with C2 in the capsule and C1 mounted midway on the radial arm. To run the equal velocity and equal centrifugal force test you could run each at the same time in the capsule and each on the arm mount to establish comparative data under exactly identical stress and velocity conditions.

You would now have baseline data to run one in the capsule and the other on the arm for the velocity affect.

if you make the mount articulated (i.e. a ball screw positioner) as you move the clocks out along the arm together you should see no differance between the two but see an affect of both centrifugal force and velocity.

Understanding this it amazes me that you would expect to see different results of C1 at any given location along the arm simply because C2 was located in the capsule. But that is the issue.

My prediction is that what you will get (assuming acceptable stability) is that C1 will replicate its tick rate for positions along the arm regardless if C2 is in the capsule or not.

21. ### PeteIt's not rocket surgeryRegistered Senior Member

Messages:
10,166
What?
No one is suggesting that the presence or absence of C2 has any affect on C1's tick rate, are they?

Do you think that is the prediction of SRT?

22. ### geistkieselValued Senior Member

Messages:
2,471
OK, but have you designed any data sheets and proposed calculations?
I mean, how are you going to show that C1 clock is moving slower than C2 and at the same time show C2 is moving slower than C1?

Are you saying that C1 will measure C2 radiated tick rate and get 90 say, while C1 measures his own at 100?

And similarly C2 will measure C1 clock trate at 95 while his own is 110?

Forget the specific numbers and gamma, but is this the general approach?

Geistkiesel ​

23. ### geistkieselValued Senior Member

Messages:
2,471

Superluminal:

OK, but have you designed any data sheets and proposed calculations?

I mean, how are you going to show that C1 clock is moving slower than C2 and at the same time show C2 is moving slower than C1?

Are you saying that C1 will measure C2 radiated tick rate and get 90 say, while C1 measures his own at 100?

And similarly C2 will measure C1 clock trate at 95 while his own is 110?

Forget the specific numbers and gamma, but is this the general approach?

PS -a slightly different question, but it comes from your statement above that acceleration affected the timing circuits. Isn't this a question that came up in GPS where the arguements were all centered around 'velocity'? I was screaming "acceleration. acceleration", well something to think about.

Geistkiesel ​