question arround a mystic microphonic problem in capacitor

[B&KMan]

Hello with all,

Since nothing is definitively finished, I said myself: which is the real sensitivity of the condensers??? After all it is normal to be skeptic on phenomena which are that too seldom exposed.

Also by personal curiosity and to make benefit many members from this
splendid forum. Here an experiment which, I hope for it, with a certain rigour the phenomenon of distortion of the condensers by the vibration will expose.

Initially the set up test is simpler and more complex: (that starts well not)

On a side we have a generator of vibration of reference which
contains a sensor of reference which measures the level and the
frequency of the head of the discharger. this sensor is connected to a system compression of the exiting signal in order to make linear the vibratory excitation of the head.

On this head, I fix a condenser with traditional a tarap. Then one produces a sinusoidal sweeping of frequency to the head of
excitation. the condenser is thus found subjected to a sinusoidal vibration of constant power at all the frequencies.

At the same time one sends a sinusoidal signal in the condenser which is not attenuated.
One thus measures in time 1 the signal years vibration and in time 2 the signal with vibration.

The results are in fact interesting:

See by yourselves...

[B&KMan]

tribo-electric noise in condenser or capacitor
------------------------------------------------------------------------
the thin line is the condenser without vibration and the whiter line
is the electric response of the condenser with the vibration.

Several remarks can made beings of this simple measurement.

1 --- the curve is not a curve but the representation of the whole of
the electric vibratory reaction of a condenser exposed to a vibratory
situation.

2 --- the effect of pollution is in connection with the power of the
electric signal and the vibration induced by the mechanical system.

Thus more the electric signal is strong and that the vibration is weak
minus the effect of pollution is present.
In example aove, the condenser received an electric signal of 50 Mv and one vibration equivalent to an acoustic pressure of 90 dB.
The condenser is a polyester of origin JBL in the networks of origin of the 4343.
If the LF receives 80 % of the electric power and that the condensers
high and ultra high receive only fractions of dB, one includes/understands easily why this approach which insulates the condensers has a better dynamic response.
Indeed if the LF receives 2.83 V to produce 93 dB with 1 meter, the 2420 receives from it to him only one fraction for the same acoustic pressure.
93 - > 2,83
96 - > 2,83 X 2
99 - > 2,83 X 4
101 - > 2,83 X 8
104 - > 2,83 X 16
107 - > 2,83 X 32

thus ex: for a power of 104 dB of sensitivity, theoretical HF to play 93 dB would not receive that 1/32 of 2.83V is 88 mV do not forget that the harmonics are small things which is propagated with low power. Variations of 30-40- 50 dB make all the difference in the transparency of a system of reproduction.

Here it is noted that a pollution of more than 16 dB on the fundamental signal... In fact the harmonics produced by the hum of 60 cycles of the current is occulted by the noise tribo electric.


This is expose any upgrade of condenseur placement for reduce exposure of vibration is a real good upgrade.

in fact if you put a caps on light structure or non isolated environnement of sound or vibration present, you introduce a big noise (moe than 15 Db in your delicate harmonic structure of high and uhf response on your system.

you destroy the quality of caps or other investement of high end product.

so esoteric caps in poor environnement is poor response than good strait caps in good environment.

I hope this is help you to discerne why certain modification is better than other.

My recommendation is if you not possibility to put caps out of environnement polluted by sound and vibration, create a mass inertia factor for reduce vibration at miximum.

In this, the silent is more deep in your audio electrical system and the dynamic is proportionnaly increase...

This is expose any upgrade of condenseur placement for reduce exposure of vibration is a real good upgrade.

The next question is if all differents nature of caps is react on vibration...

well actually the film and foil Solens tested here is less sensitivity of vibration of polyester or polypropylene metalised Solens. less 6 db of the polyester. It is not surprise, the side effect of this is the caps is better electric silence in same environment.

add of the better dielectric transfert and less loose energy and you have a real better caps with no esoteric fees.



I dont have tested the electrolytic because I dont have one in hand for compare.

:cheers:

jean.

[B&KMan]

Ok now for few crazy is interesting by this point and is want more information

go to

http://www.audioheritage.org/vbulletin/showthread.php?t=1388&page=20&pp=15&highlight=vibration post 300 and more

and start your reading with look the picture of hammer and caps test

I perform test vibration of caps fixed in thin massonite vs thick massonite. or mdf or composite...

:cheers:

[Earl K]

Hi Jean,

(i) So I think I see in that top trace , some induced microphonics whose amlpitudes are frequency dependant . Frequency areas higher than 300 hz seem to have tighter packed disturbances as well as microphonics of greater amplitude ( ignoring the AC line pollution that is manifested by those large spikes in both traces )

- Is this the correct interpretation ?

:)

[B&KMan]

Hi Jean,

(i) So I think I see in that top trace , some induced microphonics whose amlpitudes are frequency dependant . Frequency areas higher than 300 hz seem to have tighter packed disturbances as well as microphonics of greater amplitude ( ignoring the AC line pollution that is manifested by those large spikes in both traces )

- Is this the correct interpretation ?

:)

ok I send new pict of same mesure with localisation of element:

[B&KMan]

ok I send new pict of same mesure with localisation of element:

A: end of noise introducing on electrical signal by sweep sinewave vibration.

b: normal noise floor

c: tone signal electric 50mV or 70.04 Db output power...

d: bad reaction impedance in generator output and caps ( if I mesure only generator no trace of this....

e net difference amplitude in dB on level of electrical noise produce by the vibration.

N,B, the floor of noise introduce by vibration is a sweep sinewave so one tone in introduce at time. the result is the summation of influence of each fr in signal.

better understand ???

:cheers:

[johnaec]

better understand ???

:cheers:Even with perfect English translation, your understanding of the significance of these "effects" is far beyond my comprehension, (and I suspect many others here).

John

[B&KMan]

Even with perfect English translation, your understanding of the significance of these "effects" is far beyond my comprehension, (and I suspect many others here).

John

Ok I will try to make that simplest possible...

when you listen to music your system produced an electric signal
converted by your HP into sound.

a musical note produces a tone and a series of harmonics according to
the nature and the nature of the musical instrument.

The problem is that this acoustic production generates pure vibrations
according to the produced music...

however if the note fundamental, by their power, are affected very little by the noise tribo-electric. the harmonics if have a power much smaller...

This small power is largely affected by a pollution of the signal. Also all the manufacturers try to reduce to the maximum the effect of
distortion and to drop the noise to the minimum possible. It is of one of the great success of the digital. Alas very numerical must finish by being converted into analogical and finds its sensitivity... a harmonic of 40 dB exposed to a background noise from 25 to 30 dB will be clearly to pollute and denatured. so music natural and realistic is not good: in particulary case of pressure driver. the sensitivity and fastest expose more this effect ...

Thus it is absolutely necessary to attenuate as much as possible this
distortion by a larger mass of inertia. less vibrations = less noise in the signal...

this is clearer.... say to me what you do not understand.

so more the response curve grah is down better than !!
look the y scale for more evaluation of effect.

:cheers: