Cable parameters? 2

[Dmitry]

Sirs, colleagues,
the CSI-2400 spectraanalyzer cable, for accelerometers with the built-in amplifiers has broken down, - it is necessary to replace. Are there any pitfalls besides that cable mast be shielded (15 ì is required)?

With the best regards,
Dmitry Semenov

 

[ronfrend]

If your accelerometer is a Piezo-Electric type with built in amplifier and not a charge-coupled accelerometer then you will be OK with just about any type of cable. However, to avoid interference I would strongly suggest that the cable be screened twisted pair. If you have multiple pairs (for multiple accelerometers) then each pair should be individually screened. You can use coaxial cable but I would only use that as a last resort and only for portable data collector applications.

ron.frend@predicon.net

 

[Dmitry]

Greetings ronfrend!
Doing cable repair, I have found out, to my surprise, that its screen is not connected to anything. So it cannot be grounded. Is it normal?
I would like to learn if cable entrance resistance is of importance. How to choose entrance resistance of a cable?

Regards
D. Semenov

 

[ronfrend]

It has become common practice to ground the screen at one end only - this is mainly for fixed accelerometer systems. This is because if you ground at both ends and there is a difference in ground potential you will get a ground loop. With very small voltages you can very large loop currents flowing which is very undesirable. You should find that your cable is grounded at the collector/analyser end only - NOT at the accelerometer.

By cable entrance resisance I assume you mean nominal impedance? If so don't worry too much about it. This is only really critical if you are using charge coupled accelerometers (fairly rare nowadays) or if you are going to very high frequencies over a long cable length.

ron.frend@predicon.net

 

[Dmitry]

Hi ronfrend!

Can you say how high mast be frequencies to be affected by nominal impedance of 10 m cabel? The accelerometer is a Piezo-Electric type with built in amplifier.


Regards
D. Semenov

 

[ronfrend]

Fmax = 10^9/(2*3.1402*C*V/(Ic-1))


where, Fmax = maximum frequency
C = cable capacitance (picoFarads)
V = maximum output from sensor (volts) assumed @ 5V
Ic = constant current from signal conditioner (mA) - 2.4 mA
109 = scaling factor to equate units

This is for cable length in feet and cable capacitance in microfarads per foot.

If you don't want to do the calc I have attached a table to give the max transmittable frequencies for cable length and cable capacitance. Note that this is frequency transmitted along the cable - NOT Fmax. This should equate to your sampling frequency or about double your Fmax. I got this info from IMI in New York.

Cable Capacitance (picoFarads/foot)
Cable Length (feet) 16 26 30 42
10 142084 87436 75778 54127
20 71042 43718 37889 27064
30 47361 29145 25259 18042
40 35521 21859 18945 13532
50 28417 17487 15156 10825
75 18945 11658 10104 7217
100 14208 8744 7578 5413
150 9472 5829 5052 3608
200 7104 4372 3789 2706
300 4736 2915 2526 1804
400 3552 2186 1894 1353
500 2842 1749 1516 1083
1000 1421 874 758 541
2000 710 437 379 271
3000 474 291 253 180
4000 355 219 189 135
5000 284 175 152 108
10000 142 87 76 54

 

[NeedHelp]

Speaking of Vibration cables, we are currently using a High Temp Low Noise cable, which as a layer of Kapton, but not as the primary insulator or outer cover. The Electrical guy's are having fits about this cable being utilized in an aircraft because of the controverse over Kapton as it relates to aircraft fires. Any help with this one?

 

[kevS]

Kapton is being looked at with a lot of scruitiny by aircraft manufacturers. There has been plenty of focus on it by the FAA as well. I'd suggest a PTFE insulation such as MIL W-22759/8 as an alternative.

 

[Driven]

Hello Dmitry, can one determine the cable parameter from the a dc step test? That is by applying a step dc input voltage and monitoring the output. Thank you
Lotten, South Africa

 

[Dmitry]

Hi, Lotten!
I think you are right. Applying sine signal and wiring its frequency also can do it. But to do test??? – It never came to my mind because some one should know the answer with appropriate accuracy.

Regards,
D.Semenov

 

[bemein]

Speaking of noise, does anyone know of a testing laboratory that can help us develop a test specification for cable with low triboelectric noise? We are using ISA RP 37.2 which applies to accelerometers but has a paragraph devoted to cable. Our sensors have high impedances (up to a couple of gigohms) in common with accelerometers, but the ISA standard does not adequately screen out noisy cable. Our current spec is 5mV peak to peak.