Coax Cable/Connector Signal Loss

This is the place for any magazine-related discussions that don't fit in any of the column discussion boards below.
Post Reply
User avatar
Sambuchi
Posts: 366
Joined: Tue Jan 18, 2005 1:01 am
Location: Orlando FL
Contact:

Coax Cable/Connector Signal Loss

Post by Sambuchi » Mon Jun 29, 2009 6:28 am

Hello all,

I would like to calculate the signal loss in a coax cable and connector. I don't have a particular connector in mind. I was hoping to find an online calculator that would allow me to plug in some numbers to see what the loss would be if I were to use a certain length RG178 cable with some interconnects to a connector.

Does anyone have experience with this?

Thanks

Tony

Robert Reed
Posts: 2276
Joined: Wed Nov 24, 2004 1:01 am
Location: ASHTABULA,OHIO
Contact:

Re: Coax Cable/Connector Signal Loss

Post by Robert Reed » Mon Jun 29, 2009 9:02 am

What frequency? What style connector? How long is the cabling?
In general a rule of thumb is less than 1/2 db per connector. But if properly installed with proper tools can be much less. To expand on that, the assembly is installed exactly as Manu specifies using fairly expensive calibration tools . The higher the operating frequency, the more critical installation becomes and is quite critical up in the very high GHz range.
For most of us that are woking in sub UHF range the loss should not be of much concern even with bargain basement connectors. I have seen little difference in attenuation from expensive to cheap connectors for <200 MHz use. If your project is not instrumentation grade or is not critical for group delay, you won't need to concern yourself too much with those inherant losses. Your BIG loss is going to be in the RG-174 cable if frequency is high and cable is long relative to operating wavelength. Need more info Tony. Also there is a lot of info on the web about this. Google up RG-174 and seach the sites. If you come up with nothing worthwhile, I can dig thru my archives and post some tables.

gerty
Posts: 309
Joined: Mon Dec 06, 2004 1:01 am
Location: Tennessee
Contact:

Re: Coax Cable/Connector Signal Loss

Post by gerty » Mon Jun 29, 2009 9:16 am

I've always been told that a screw on (bnc) connector is more succeptible to loss than a crimp or solder type. I have no way to prove this, other than the fact that when I do have a problem with CCTV it's almost always a screw on connector. All I usually have to do is tighten it, or in some cases, cut it off and re-terminate it.

User avatar
Sambuchi
Posts: 366
Joined: Tue Jan 18, 2005 1:01 am
Location: Orlando FL
Contact:

Re: Coax Cable/Connector Signal Loss

Post by Sambuchi » Mon Jun 29, 2009 10:17 am

What frequency?
Very low and low power. ~ 200kH

What style connector?
I would be using some standard SMA stuff.

How long is the cabling?
I don't know! I cant make it long if I loose too much of the signal.

But if properly installed with proper tools can be much less.
I will be getting the proper tools for this cable assembly.

I want to calculate the difference of using a setup where I have a box with 5 sensors looped with RG178. To where I have a box loop with 5 detachable sensors in a loop. This setup would help me in case where a sensor/cable gets damaged. To make it detachable I wanted interconnects for the cables.

like this



------

box |~~~cable~~~~~~Connector~~~-Sensor-~~~Connector~~~-Sensor~~~Connector~~~-Sensor

------


Make sense?

Robert Reed
Posts: 2276
Joined: Wed Nov 24, 2004 1:01 am
Location: ASHTABULA,OHIO
Contact:

Re: Coax Cable/Connector Signal Loss

Post by Robert Reed » Mon Jun 29, 2009 8:09 pm

At 200 KHz, Don't even concern yourself with losses as this is practically audio frequency. You can use any connector you desire such as RCA phono plugs. Unless you have an abundance of SMA connectors on hand, they are way overkill for this job, expensive and somewhat more difficult to work with. RG-174 is 50 0hm cable and normally should be matched at each end. This is only critical when the wave length/cable length becomes critical. At 200 KHz the wavelength is 1500 meters. Given a 0.67 velocity factor of RG, one wavelength in the cable will be 1000 meters. Matching really isn't too important for cable lengths <<1/4 wavelength. Long story short - if your cables are less than a football feild long you are very probably in good shape without the need for proper matching.

PS- when I mentioned "proper tools", I was referring to sets that cost upwards of $500 dollars and are not installation tools, but rather calibration tools for pin set, pressure point, etc. and normally geared for SMA and type N connectors. Definately not the stuff you or I want to invest in. Of course you still need a good set of crimps for the cables you are working with.

Robert Reed
Posts: 2276
Joined: Wed Nov 24, 2004 1:01 am
Location: ASHTABULA,OHIO
Contact:

Re: Coax Cable/Connector Signal Loss

Post by Robert Reed » Tue Jun 30, 2009 8:13 am

I got to thinking about your signal content. My thoughts prior were in regards to a 200 KHz sine wave carrier. If its actually a digital stream that requires fast rise times, the the frequency span could be much greater tha 200k and may be better served with proper matching (assuming the system requires long cable runs). A few tens of feet are not important. Most likely the input will see the cable as a lumped component (capacitive) so in any case you will want to drive it with a lo impedance source - any Op-Amp will do. If the capacitive load is high enough it can cause instability in the amp, but is easily cured with a lo value series resistor tied to its output. Any where from 30 to 100 ohms should do the trick.
BTW C-Mos logic also has lo output impedance for slower rise times but will suffer loss at very fast rise times with increasing capacitive load. However,very hi Tr should not be of much importance in your case.

User avatar
Sambuchi
Posts: 366
Joined: Tue Jan 18, 2005 1:01 am
Location: Orlando FL
Contact:

Re: Coax Cable/Connector Signal Loss

Post by Sambuchi » Tue Jun 30, 2009 9:11 am

Thank you Robert for putting my mind at ease.

Yes the signal will be digital

The total length of the cable wont be longer than 40 feet. I think with what you said I should be ok.

I am designing the board that the cables connect to and I included 50 termination and matching.

I want to thank you again for all this great info to take in consideration as well.

:D

Tony

Post Reply

Who is online

Users browsing this forum: Google [Bot] and 36 guests