-
#3860 by WarpTech on 14 Jul, 2016 20:09
-
...
So why there is not more thrust than for a photon rocket using a simple (50 Ohm)load and a HF source? (Where normally total net-thrust is zero because of radiation and attenuation in such a closed(perfect) system is equal).
Regarding that last part there on the EM Drive. What is going on only mimics gravity over a narrow bandwidth. It's not interacting with the vacuum. It's interacting with itself. It is a resonant cavity. If it were symmetrical, I would not expect any acceleration to result, but given that dissipative systems do not need to conserve momentum because heat loss is a random momentum transfer, AND it is asymmetrical....
Todd
Edit: added last para.
If I'm understanding you correctly. My answer is that where the damping factor is greater, a greater amount of momentum can be transferred than where damping factor is lower. If I use the PV Model terminology, where K is the refractive index.
p(K) = p0*sqrt(K)
If the damping is higher, K is higher. If there is asymmetry in K, then momentum transfer will also be asymmetrical. This could provide more thrust than a photon rocket, because it depends on the rate of change in K, i.e., the asymmetry. In the EM drive, damping should be highest at the small end. Right? Where attenuation is highest and the induction is highest.
Todd -
#3861 by dustinthewind on 14 Jul, 2016 20:31
-
A little over a year ago, I conjectured that the EM Drive was a type of warp drive because gravity is related to dissipation and attenuation. I wrote and published a paper called the The Electromagnetic Quantum Vacuum Warp Drive that showed exactly how gravity is related to the available driving power from the ZPF, but it didn't really talk much about dissipation or attenuation. Just that the available power was a variable.
Well, I'm now finishing up a new paper that shows without a doubt, that gravity is the result of increased local energy density CAUSING increased radiative damping within the harmonic oscillations that make up all matter. The ZPF is the driving power that inflates matter, and radiative damping is the loss of power that causes gravitational length contraction and time dilation. Matter in an inertial frame is in equilibrium between the source and sink of this power. Essentially then, gravitational acceleration is caused by asymmetrical variations in radiative damping. Not only did the math fall out beautifully such that it could easily be understood how this could be misinterpreted as space-time curvature. It also shows the Planck force acting on a sphere with a Planck length radius, is EQUAL to the force on the Hubble sphere, with the Hubble radius. I find this very interesting, because why should these two extreme length scales be related in any way unless one is derived from the other? I show that it is!
In any case, you all can gather what this implies. That if this EM Drive thing moves without ejecting any momentum, this is why! It is asymmetrical damping and attenuation creating a gravitational effect within a limited bandwidth of the EM spectrum. DeAquino's ideas of using ferro metal at the small end may indeed be the ticket to higher thrust, because it will cause a greater asymmetry in the dissipation taking place in the shell of the frustum.
Todd
edit: small was big
How would you answer to the following argument against any useful interaction with the Quantum Vacuum:
Virtual particles in the Quantum Vacuum (for example electron-positron pairs) that fluctuate into and out of existence violate energy conservation and so must be governed by the Heisenberg energy-time Uncertainty relation.
That limits their fluctuational lifetimes to 3 X 10-22 sec.
As of May 2010, the smallest time interval uncertainty in direct measurements is on the order of 12 attoseconds (1.2 × 10−17 seconds). (See: http://phys.org/news/2010-05-attoseconds-world-shortest.html ). In the visible domain, a single oscillation of the electric field takes about 1200-2500 attoseconds, which is 154 times longer than this. And that is in the visible domain (not in the GHz domain of the EM Drive, which takes several orders of magnitude longer).
So the smallest time interval uncertainty in direct measurements is about 40 000 times longer than the time of the QV fluctuational lifetimes.
Another comparison: the atomic unit of time is 24 attoseconds (https://en.wikipedia.org/wiki/Atomic_units) (the fastest possible time scale of processes in the outer shells of an atom) which is 80 000 times longer than the maximum possible lifetime of virtual pairs from the Quantum Vacuum.
This life-time (of virtual pairs in the Quantum Vacuum) is so short ( 3 X 10-22 sec) that it seems impossible to transfer anything to/from a given virtual pair before it vanishes. To do anything useful with such a Quantum Vacuum, it would need to have a gigantic density.
If you calculate how large that density would need to be, it is so large that it would have been detected long, long ago. [We discussed this previously, for example with Dr. Notsosureofit] So something like a microwave cavity operating at GHz of small dimensions with energy input of less than 1kW seems quite unlikely to be able to detect such a huge density that has not been previously detected, by much more sophisticated means of detection.
This paper discuses the quantum vacuum and an associated interaction between matter and said vacuum. Maybe it relates to QV particle creation annihilation also? I find it fascinating that they discuss a connection to space propulsion and they have a different take on it. They propose an asymmetrical capacitor and it looks like a frustum. My initial thought is at microwave frequencies doesn't the frustum act as both an inductor and capacitor? That is it being a resonant circuit.
Quantum vacuum energy, gravity manipulation and the force generated by the interaction between high potential electric fields and ZPF by LM Caligiuri, T Musha
https://scholar.google.com/scholar?cluster=8287855046649846200&hl=en&as_sdt=0,48 -
#3862 by dustinthewind on 14 Jul, 2016 21:13
-
That sounds really similar to saying where the damping factor is greater, K being greater, the mass of a photon increases making a photon more able to transfer energy upon impact....
So why there is not more thrust than for a photon rocket using a simple (50 Ohm)load and a HF source? (Where normally total net-thrust is zero because of radiation and attenuation in such a closed(perfect) system is equal).
Regarding that last part there on the EM Drive. What is going on only mimics gravity over a narrow bandwidth. It's not interacting with the vacuum. It's interacting with itself. It is a resonant cavity. If it were symmetrical, I would not expect any acceleration to result, but given that dissipative systems do not need to conserve momentum because heat loss is a random momentum transfer, AND it is asymmetrical....
Todd
Edit: added last para.
If I'm understanding you correctly. My answer is that where the damping factor is greater, a greater amount of momentum can be transferred than where damping factor is lower. If I use the PV Model terminology, where K is the refractive index.
p(K) = p0*sqrt(K)
If the damping is higher, K is higher. If there is asymmetry in K, then momentum transfer will also be asymmetrical. This could provide more thrust than a photon rocket, because it depends on the rate of change in K, i.e., the asymmetry. In the EM drive, damping should be highest at the small end. Right? Where attenuation is highest and the induction is highest.
Todd
I remember discussing this with you I think, it was that while p(K) and E(K) are not conserved p(K)*E(K) is conserved. The reason for p(K) increasing was that while velocity decreases mass increases at a faster rate. Trying to remember it exactly I think it was that v(K)=v0/K and m(K)=m0*K^(3/2)? So then v0*m0=p0*sqrt(K).
Polarizable-Vacuum (PV) representation of general relativity by H. E. Puthoff
https://scholar.google.com/scholar?cluster=11891316723593604093&hl=en&as_sdt=0,48 -
#3863 by Monomorphic on 14 Jul, 2016 21:14
-
Pretty sure this is a confirmation of resonance in Dave's frustum. Image taken from latest video and threshold filter applied. Compared to FEKO prediction. Looks like the endplates may be slightly out of alignment, but it's defintely the same pattern forming.
-
#3864 by Rodal on 14 Jul, 2016 21:15
-
...
Thanks but my concerns about this paper were raised quite high when I saw that this paper quotes the claims of the Russian ceramics engineer Plodkletnov as serious "experimental evidence"
This paper discuses the quantum vacuum and an associated interaction between matter and said vacuum. Maybe it relates to QV particle creation annihilation also? I find it fascinating that they discuss a connection to space propulsion and they have a different take on it. They propose an asymmetrical capacitor and it looks like a frustum. My initial thought is at microwave frequencies doesn't the frustum act as both an inductor and capacitor? That is it being a resonant circuit.
Quantum vacuum energy, gravity manipulation and the force generated by the interaction between high potential electric fields and ZPF by LM Caligiuri, T Musha
https://scholar.google.com/scholar?cluster=8287855046649846200&hl=en&as_sdt=0,48QuoteThe first experimental evidence about this concrete possibility came from the experiment conducted by
Podkletnov [13] in which a composite bulk YBa2Cu3O7-x superconductor, at a temperature below 70°K interacting with a suitable e.m. field, manifested “weak gravitation shielding” in the space below and above the superconductor apparatus
the concern-meter went even higher when reading this, quoting work by Ning Li:QuoteThe same experiment was subsequently reproduced by Li et al. [14] and analyzed by Modanese [15].
According to Li’s conclusions “rotating superconductors in an alternate magnetic field would generate gravity”, so suggesting a primary role of electromagnetic energy in modifying the quantum vacuum ZPE and then, for the above considerations, its energy density QV r . -
#3865 by Rodal on 14 Jul, 2016 21:18
-
Pretty sure this is a confirmation of resonance in Dave's frustum. Image taken from latest video and threshold filter applied. Compared to FEKO prediction. Looks like the endplates may be slightly out of alignment, but it's defintely the same pattern.
Evidently the problem with rfmwguy's visualization of the mode shape is not due to the conductivity of copper (since diffusivity is a slower process than induction heating) but to his choosing an unconventional mode shape, for his test, not reported as tested before by Shawyer, Yang or NASA, a mode shape that upon heating of the ends is difficult to show (in great contrast to the mode shape TM212 chosen by NASA that displays an unmistakable quadrupole that is much easier to measure with an infrared camera) -
#3866 by dustinthewind on 14 Jul, 2016 21:23
-
...
Thanks but my concerns about this paper were raised when I saw that this paper quotes the claims of the russian ceramics engineer Plodkletnov as "experimental evidence"
This paper discuses the quantum vacuum and an associated interaction between matter and said vacuum. Maybe it relates to QV particle creation annihilation also? I find it fascinating that they discuss a connection to space propulsion and they have a different take on it. They propose an asymmetrical capacitor and it looks like a frustum. My initial thought is at microwave frequencies doesn't the frustum act as both an inductor and capacitor? That is it being a resonant circuit.
Quantum vacuum energy, gravity manipulation and the force generated by the interaction between high potential electric fields and ZPF by LM Caligiuri, T Musha
https://scholar.google.com/scholar?cluster=8287855046649846200&hl=en&as_sdt=0,48QuoteThe first experimental evidence about this concrete possibility came from the experiment conducted by
Podkletnov [13] in which a composite bulk YBa2Cu3O7-x superconductor, at a temperature below 70°K interacting with a suitable e.m. field, manifested “weak gravitation shielding” in the space below and above the superconductor apparatus
I am assuming then that, that particular experiment by said character, was not reproducible or that he is known for falsifying information? Well I am curious the exact reason for his questionably. Maybe his claims are too far out there? I'll have to look back over their paper to see how much it depends on his work.
some quick information on him here: https://en.wikipedia.org/wiki/Eugene_Podkletnov
oh, ok thanks Ning Li (physicist), so it looks like her proposal didn't pan out. -
#3867 by X_RaY on 14 Jul, 2016 21:25
-
β is equivalent to k...
So why there is not more thrust than for a photon rocket using a simple (50 Ohm)load and a HF source? (Where normally total net-thrust is zero because of radiation and attenuation in such a closed(perfect) system is equal).
Regarding that last part there on the EM Drive. What is going on only mimics gravity over a narrow bandwidth. It's not interacting with the vacuum. It's interacting with itself. It is a resonant cavity. If it were symmetrical, I would not expect any acceleration to result, but given that dissipative systems do not need to conserve momentum because heat loss is a random momentum transfer, AND it is asymmetrical....
Todd
Edit: added last para.
If I'm understanding you correctly. My answer is that where the damping factor is greater, a greater amount of momentum can be transferred than where damping factor is lower. If I use the PV Model terminology, where K is the refractive index.
p(K) = p0*sqrt(K)
If the damping is higher, K is higher. If there is asymmetry in K, then momentum transfer will also be asymmetrical. This could provide more thrust than a photon rocket, because it depends on the rate of change in K, i.e., the asymmetry. In the EM drive, damping should be highest at the small end. Right? Where attenuation is highest and the induction is highest.
Todd
jβ=2πj/λ same thing as k=2π/λ
γ=α+jβ |-α --> γ-α=jβ
(γ=propagation constant, β=phase constant, α= attenuation factor)
At the small end β increases since λ increases
α must decrease, for getting γ constant.
Means that the momentum would be smaller at the small end? Greater wavelength(at same frequency) --> less momentum?
Just thinking out loud here and its late night in germany, therefore PLESE correct me if I am wrong.
-
#3868 by Rodal on 14 Jul, 2016 21:27
-
...
Thanks but my concerns about this paper were raised when I saw that this paper quotes the claims of the russian ceramics engineer Plodkletnov as "experimental evidence"
This paper discuses the quantum vacuum and an associated interaction between matter and said vacuum. Maybe it relates to QV particle creation annihilation also? I find it fascinating that they discuss a connection to space propulsion and they have a different take on it. They propose an asymmetrical capacitor and it looks like a frustum. My initial thought is at microwave frequencies doesn't the frustum act as both an inductor and capacitor? That is it being a resonant circuit.
Quantum vacuum energy, gravity manipulation and the force generated by the interaction between high potential electric fields and ZPF by LM Caligiuri, T Musha
https://scholar.google.com/scholar?cluster=8287855046649846200&hl=en&as_sdt=0,48QuoteThe first experimental evidence about this concrete possibility came from the experiment conducted by
Podkletnov [13] in which a composite bulk YBa2Cu3O7-x superconductor, at a temperature below 70°K interacting with a suitable e.m. field, manifested “weak gravitation shielding” in the space below and above the superconductor apparatus
I am assuming then that, that particular experiment by said character, was not reproducible...
[https://en.wikipedia.org/wiki/Ning_Li_(physicist)] -
#3869 by WarpTech on 14 Jul, 2016 21:38
-
...
Thanks but my concerns about this paper were raised quite high when I saw that this paper quotes the claims of the Russian ceramics engineer Plodkletnov as serious "experimental evidence"
This paper discuses the quantum vacuum and an associated interaction between matter and said vacuum. Maybe it relates to QV particle creation annihilation also? I find it fascinating that they discuss a connection to space propulsion and they have a different take on it. They propose an asymmetrical capacitor and it looks like a frustum. My initial thought is at microwave frequencies doesn't the frustum act as both an inductor and capacitor? That is it being a resonant circuit.
Quantum vacuum energy, gravity manipulation and the force generated by the interaction between high potential electric fields and ZPF by LM Caligiuri, T Musha
https://scholar.google.com/scholar?cluster=8287855046649846200&hl=en&as_sdt=0,48QuoteThe first experimental evidence about this concrete possibility came from the experiment conducted by
Podkletnov [13] in which a composite bulk YBa2Cu3O7-x superconductor, at a temperature below 70°K interacting with a suitable e.m. field, manifested “weak gravitation shielding” in the space below and above the superconductor apparatus
the concern-meter went even higher when reading this, quoting work by Ning Li:QuoteThe same experiment was subsequently reproduced by Li et al. [14] and analyzed by Modanese [15].
According to Li’s conclusions “rotating superconductors in an alternate magnetic field would generate gravity”, so suggesting a primary role of electromagnetic energy in modifying the quantum vacuum ZPE and then, for the above considerations, its energy density QV r .
Yeah, mine too. The ideas are about where things were at 10 years ago, by my time line.
-
#3870 by WarpTech on 14 Jul, 2016 21:49
-
...
β is equivalent to k
jβ=2πj/λ same thing as k=2π/λ
γ=α+jβ |-α --> γ-α=jβ
(γ=propagation constant, β=phase constant, α= attenuation factor)
At the small end β increases since λ increases
α must decrease, for getting γ constant. Momentum would be smaller at the small end?
Just thinking out loud here and its late night in germany, therefore PLESE correct me if I am wrong.
Per Zeng & Fan, α/k increases as kr decreases.
From the dispersion relationship. As attenuation increases, frequency decreases for a constant phase angle, and this loss of energy is transferred to the frustum.
-
#3871 by Rodal on 14 Jul, 2016 22:02
-
...
β is equivalent to k
jβ=2πj/λ same thing as k=2π/λ
γ=α+jβ |-α --> γ-α=jβ
(γ=propagation constant, β=phase constant, α= attenuation factor)
At the small end β increases since λ increases
α must decrease, for getting γ constant. Momentum would be smaller at the small end?
Just thinking out loud here and its late night in germany, therefore PLESE correct me if I am wrong.
Per Zeng & Fan, α/k increases as kr decreases.
From the dispersion relationship. As attenuation increases, frequency decreases for a constant phase angle, and this loss of energy is transferred to the frustum.
Todd, please remember that big error we found in the paper of Zeng and Fan -
#3872 by rfmwguy on 14 Jul, 2016 22:23
-
Pretty sure this is a confirmation of resonance in Dave's frustum. Image taken from latest video and threshold filter applied. Compared to FEKO prediction. Looks like the endplates may be slightly out of alignment, but it's defintely the same pattern forming.
Thanks Jamie, nice to see the new mag resonating and creating the same general pattern. Yes, the sub-mm imperfections are likely the culprit here. Good thing though it is close enough to get the forming pattern at the edges (man they disappear quick in copper).
So, TM013 is on the table for a possible mode for the EmDrive effect...have to wait to make that claim once testing resumes on the PCM modified magnetron core. To be continued... -
#3873 by FailTrain on 14 Jul, 2016 22:50
-
Found an interesting paper on Superconducting RF Cavities.
I Haven't been able to follow the forums as much lately because of life but I figured you guys might enjoy reading it if someone else hasn't posted.
Please forgive me if it was already posted!
-
#3874 by frobnicat on 14 Jul, 2016 22:54
-
The ADC output in Monomorphic's test P11 is granular as might be expected. The difference between consecutive measurements is always an exact multiple of 10/8192 volts, or 10/(2^13).
The values tend to 'stick' every 16 such units (once in a while every 20), or every 10/512 volts.
.../...
Yes the values of the "sticking" flat areas are in fact 20/1024 = 20/(2^10) so the flat areas are at the discrete levels of the ADC :
ADC characteristic are
±10 V Differential Analog Inputs
10-bit Resolution
Now on the mystery why there is some kind of intermediate levels and the flat areas are smoothed...
Zen-in thinks that "There has to be at least 8 bits in the output at max. amplitude because digitization is barely visible". I think this is wrong (with all due respect to someone working in Digilab when I was still learning to read Alice in wonderland) : the hardware digital conversion is moving only 3 least significant bits given the amplitude of input. Then I guess WinDaq (or other acquisition software used) does post processing of the raw digital values on 10 bits with a moving average on a window of 16 successive samples. Giving the illusion of added 4 significant bits in the values of the spreadsheet : 10/(2^13) is actually =20/(2^14)=20/(2^10)/(2^4)=20/(2^10)/16, 20 volts full scale, 10 bits of real hardware precision in the ADC, running average on 16 successive samples.
The hardware ADC is not in cause, but the amplitude of analog input is way too low : 160 mV of range (max-min in discussed plot n°11), the quantum being 20V/(2^10)≈20mV, 8 discrete levels to play with, 3 real hardware bits of precision, later "extended" to 7 bits by the averaging that also "erodes" the steps that should appear more clearly.
This explains why "The values tend to 'stick' every 16 such units" but I have no explanation as to why "once in a while every 20" (some cyclic rounding in the moving average procedure ? Other software numerical treatment artefact ?)
As illustration I attach some illustration (the parameters used are not meant to be the ones of the actual data acquisition system) of that :
A : a smooth signal (illustrative)
B : some analog noise on top of it (thermal noise of measurements...)
C : hardware digital conversion on a few discrete levels (range of analog input not many times the quantum as it should)
D : software moving average (with a window of 6 samples, again illustrative)
Notice as the number of values between two successive "sticky values" (integers on vertical scale) is the same as the width of the averaging (6 in my illustration, 16 in monomorphic's plot n°11). The averaging erodes the steps, some noise remains, more or less flattish areas appear at regular vertical intervals, and when an inversion of slope occurs in the smooth analog signal there is a high chance to see flattish area(s) of great length compared to other flat areas when signal crosses hardware discrete levels more rapidly.
I agree with warptech that monomorphic needs to amplify analog input to the ADC by better fitting to its full scale, that will increase noise but noise is much more preferable than quantization errors at 3 real hardware bits !
"10 bits is fine if the input signal fills the range, but if the signal is only using 2 or 3 bits, then it requires a front end OpAmp to boost the signal."
-
#3875 by WarpTech on 15 Jul, 2016 03:05
-
....
I agree with warptech that monomorphic needs to amplify analog input to the ADC by better fitting to its full scale, that will increase noise but noise is much more preferable than quantization errors at 3 real hardware bits !
"10 bits is fine if the input signal fills the range, but if the signal is only using 2 or 3 bits, then it requires a front end OpAmp to boost the signal."
The signal is small but it is also slow varying. Monomorphic can put a ~10nF capacitor across the resistor to filter out the high frequency noise. A real signal should hardly be affected by it. Even a little larger if necessary, but more capacitance will make quick transitions slower, so don't add more than you need to cure the jitters. -
#3876 by Bob Woods on 15 Jul, 2016 03:45
-
....
I agree with warptech that monomorphic needs to amplify analog input to the ADC by better fitting to its full scale, that will increase noise but noise is much more preferable than quantization errors at 3 real hardware bits !
"10 bits is fine if the input signal fills the range, but if the signal is only using 2 or 3 bits, then it requires a front end OpAmp to boost the signal."
The signal is small but it is also slow varying. Monomorphic can put a ~10nF capacitor across the resistor to filter out the high frequency noise. A real signal should hardly be affected by it. Even a little larger if necessary, but more capacitance will make quick transitions slower, so don't add more than you need to cure the jitters.
Todd,you're a very sharp mind. I can't follow your math on your theories, but I'm impressed nevertheless.
Keep on Truck'in. -
#3877 by zen-in on 15 Jul, 2016 06:52
-
The ADC output in Monomorphic's test P11 is granular as might be expected. The difference between consecutive measurements is always an exact multiple of 10/8192 volts, or 10/(2^13).
The values tend to 'stick' every 16 such units (once in a while every 20), or every 10/512 volts.
.../...
Yes the values of the "sticking" flat areas are in fact 20/1024 = 20/(2^10) so the flat areas are at the discrete levels of the ADC :
ADC characteristic are
±10 V Differential Analog Inputs
10-bit Resolution
Now on the mystery why there is some kind of intermediate levels and the flat areas are smoothed...
Zen-in thinks that "There has to be at least 8 bits in the output at max. amplitude because digitization is barely visible". I think this is wrong (with all due respect to someone working in Digilab when I was still learning to read Alice in wonderland) : the hardware digital conversion is moving only 3 least significant bits given the amplitude of input. Then I guess WinDaq (or other acquisition software used) does post processing of the raw digital values on 10 bits with a moving average on a window of 16 successive samples. Giving the illusion of added 4 significant bits in the values of the spreadsheet : 10/(2^13) is actually =20/(2^14)=20/(2^10)/(2^4)=20/(2^10)/16, 20 volts full scale, 10 bits of real hardware precision in the ADC, running average on 16 successive samples.
The hardware ADC is not in cause, but the amplitude of analog input is way too low : 160 mV of range (max-min in discussed plot n°11), the quantum being 20V/(2^10)≈20mV, 8 discrete levels to play with, 3 real hardware bits of precision, later "extended" to 7 bits by the averaging that also "erodes" the steps that should appear more clearly.
This explains why "The values tend to 'stick' every 16 such units" but I have no explanation as to why "once in a while every 20" (some cyclic rounding in the moving average procedure ? Other software numerical treatment artefact ?)
As illustration I attach some illustration (the parameters used are not meant to be the ones of the actual data acquisition system) of that :
A : a smooth signal (illustrative)
B : some analog noise on top of it (thermal noise of measurements...)
C : hardware digital conversion on a few discrete levels (range of analog input not many times the quantum as it should)
D : software moving average (with a window of 6 samples, again illustrative)
Notice as the number of values between two successive "sticky values" (integers on vertical scale) is the same as the width of the averaging (6 in my illustration, 16 in monomorphic's plot n°11). The averaging erodes the steps, some noise remains, more or less flattish areas appear at regular vertical intervals, and when an inversion of slope occurs in the smooth analog signal there is a high chance to see flattish area(s) of great length compared to other flat areas when signal crosses hardware discrete levels more rapidly.
I agree with warptech that monomorphic needs to amplify analog input to the ADC by better fitting to its full scale, that will increase noise but noise is much more preferable than quantization errors at 3 real hardware bits !
"10 bits is fine if the input signal fills the range, but if the signal is only using 2 or 3 bits, then it requires a front end OpAmp to boost the signal."
I don't know enough about how your experiment uses the LDS or A/D. I also don't understand why anyone is saying "the signal is only using 2 or 3 bits". If that was the case you would only see 4 or 8 discrete values in the waveform. It's possible someone has gotten confused with how a delta sigma A/D operates. Those A/Ds can use just 2 or 3 bits in the conversion cycle. However over time the accumulated sample can be 10, 16 or even 21 bits. This is all done inside the IC. They are not fast A/Ds but they are well suited for low noise measurements. This is just speculation on my part because I have not seen a schematic of the apparatus.
Below I have annoted the plot that shows flat areas. Each flat area, starting with the lowest value, has a number next to it. I see 9 flat areas that are each separated from the ones above and below them by approximately 2 horizontal hash marks, or approximately .01 You would have to look back into the raw data to determine what is causing these flat areas. Also, if were possible to feed a ramp or triangle wave into the A/D maybe these flat areas would show up again. They do look like some artifact that is created after the A/D. But it could also be something about how the A/D is programmed. I don't have enough information to know.
-
#3878 by frobnicat on 15 Jul, 2016 08:09
-
Zen_in : my explanation (moving average) why there are intermediate levels between flat areas is speculation because I don't know what does the software after digital conversion, but even if we had not seen any plot or take any measurements we could know beforehand that :
The digitized signal is spanning 160 mV of range, the hardware ADC converter is documented as being 10 bits on a range of 20V, that makes discrete steps every 20V/(2^10)≈20mV,( that's your twice " approximately .01" separating each flattish levels), 160mV/20mV = 8 discrete levels to play with, 3 real hardware bits of precision. This is a rather direct and straightforward prediction of known parameters, even if we noticed the problem after looking at recorded value. -
#3879 by RERT on 15 Jul, 2016 11:26
-
Yes the values of the "sticking" flat areas are in fact 20/1024 = 20/(2^10) so the flat areas are at the discrete levels of the ADC......
Very convincing, your chart D is what we are seeing! Further, if the ADC smoothing is trailing MA, it will introduce a delay. If it is a central MA it will appear 'psychic' and anticipate changes. Finding a way to turn it off might be helpful.
On another topic, any idea why noise might be suppressed during power-on, as looks to be the case?
Means that the momentum would be smaller at the small end? Greater wavelength(at same frequency) --> less momentum?