ProSoundWeb Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: [1] 2  All   Go Down

Author Topic: Reverse Engineering Fade Margin Calculations  (Read 6237 times)

Michael Fisher

  • Newbie
  • *
  • Offline Offline
  • Posts: 7
Reverse Engineering Fade Margin Calculations
« on: July 26, 2016, 07:31:49 PM »

Hi,

(I'm a looong time lurker, first time poster. I've read the rules, and I think I'm in compliance.)

I've been following the wireless posts with some interest recently, as we're planning a long overdue upgrade to our wireless infrastructure at my venue.

We have four Sennheiser EM 300 G3 receivers mounted in a metal rack in an amplifier closet beside the stage. I'd like to remote two omni antennas and I'm working through the math based on Henry Cohen's informative posts here on PSW. We're in Canada and own equipment in the Sennheiser A-band exclusively (~516-558), but in my calculations I've used the E-band frequencies (~823-865) as a worst case. We likely won't bandpass the antennas right out of the gate.

Here's my math; am I barking up the right tree? Unless I messed up a sign somewhere, my calculations show that this should work well without amplifiers or other complicating bits. I'm ready to go out and buy the equipment, test it out for a few weeks in a temporary setup to confirm that it does work, and then install it permanently.  Before I do that, I'm very open to corrections and clarifications, for example with the "coupling loss" variable. I've otherwise tried to keep my +/- signs and dB and dBm units straight.  :)

Constants and Variables
FSPL(air)=52.72 dB
Quote
Free space path loss. How much RF energy is lost to the attenuation in air due to the distance between Rx and Tx antennas:
FSPL(dB)=20 log10⁡(d)+ 20log10⁡(f)-27.55
d = maximum distance in meters between transmitting and receiving antennas (20 meters)
f = lowest frequency used in mHZ (Sennheiser A-band, 516 mHz)
FSPL(coupling)=10.22 dB  ???
Quote
I have to confess I haven’t been able to find out what this is. I saw another thread where Henry quoted coupling loss at 0.5’ = 12 db for a Sennheiser A-band application. When I use the FSPL calculations for ½ a foot (0.1524 meters) I get around 10 dB of loss which seems close to what Henry calculated. But I can’t say I understand what coupling loss even is, let alone the purpose of the ½ foot from the other thread.
CableLoss = 3.2 dB
Quote
Calculated based on the longest run (75’), at the highest frequency (Sennheiser E-band at 865 mHz), using LMR-400 cable and includes connector loss. The other run will be considerably shorter than this, closer to 25'.
BeltpackBodyAttenuation = 20 dB (Rule-of-thumb)
SplitterGain = 0 dB (Lectrosonics UMC16)
RxAntennaGain = 0 dB (Sennheiser A1013 Omni). Omni to cover excursions beyond the stage.
TransmitterERP = 10 dBm (ERP for a 30mW beltpack will typically be 10-15mW (10-11dBm) due to a nominally tuned ¼ wave antenna with insufficient ground plane.)
RxSensitivity = -100 dBm (Sennheiser EM 300, <2uV at 50 ohms = <-100 dBm, converted)



Net Transmission Loss
NetTxLoss = FSPL(air)+ FSPL(coupling)+ CableLoss + BeltpackBodyAttenuation + SplitterGain + RxAntennaGain
NetTxLoss (dB) = 86.14 dB

Level at Receiver Input
RxInputLevel= TransmitterERP – NetTxLoss
RxInputLevel = 10 – 86.14
RxInputLevel = -76.14 dB


Fade Margin
FadeMargin= RxInputLevel – RxSensitivity
FadeMargin= -76.14 – (-100)
FadeMargin= -76.14 + 100
FadeMargin= 23.86 dB
Quote
Which should be (and is) greater than 20dB for safety.
Logged

Brad Harris

  • Full Member
  • ***
  • Offline Offline
  • Posts: 194
Re: Reverse Engineering Fade Margin Calculations
« Reply #1 on: July 27, 2016, 02:05:29 PM »

So what cable and how long are you planning on using?


BRad
Logged

Michael Fisher

  • Newbie
  • *
  • Offline Offline
  • Posts: 7
Re: Reverse Engineering Fade Margin Calculations
« Reply #2 on: July 27, 2016, 02:35:34 PM »

So what cable and how long are you planning on using?


BRad

Haven't purchased it yet, but something like LMR-400 or similar (I hear that Wireman 106 is good and shares similar attenuation characteristics). If we use something different, I'll have to re-calculate the cable loss - but I'm very happy to observe the wisdom that low-loss cable is preferred.

Probably 25' on one end and 75' on the other.  The amp closet is SR, so the SR run will be shorter than the SL run.

It's a thrust stage at the "top" of a diamond-shaped room and we'll be mounting the antennas on the back wall at USL and USR.
« Last Edit: July 27, 2016, 02:39:12 PM by Michael Fisher »
Logged

Ike Zimbel

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1054
  • I'm not a newbie, I just play one on the internet!
    • Zimbel Audio Productions
Re: Reverse Engineering Fade Margin Calculations
« Reply #3 on: July 27, 2016, 04:36:34 PM »

Haven't purchased it yet, but something like LMR-400 or similar (I hear that Wireman 106 is good and shares similar attenuation characteristics). If we use something different, I'll have to re-calculate the cable loss - but I'm very happy to observe the wisdom that low-loss cable is preferred.

Probably 25' on one end and 75' on the other.  The amp closet is SR, so the SR run will be shorter than the SL run.

It's a thrust stage at the "top" of a diamond-shaped room and we'll be mounting the antennas on the back wall at USL and USR.
One thing to keep in mind is that LMR-400 and similar low-loss cables are really, really stiff. This can make it a challenge to route into a rack, or on to the front of your antenna distro if you are front-mounting the BNC's. Likewise at the antenna end. I have seen these cables place significant strain on the BNC's they are connected to, and do things like point your antennas in an unintended direction because that's the way the cable wants to bend. At the risk of alienating the entire interweb, I'm going to say that for the 25' run, a good quality RG-58 would be just fine and an RG-8U would be more than adequate for the longer run. It's been my experience that if you are worried about signal loss over shorter cable runs, you've got other problems.
Best,
Ike
Logged
~Ike Zimbel~
Wireless frequency coordination specialist and educator.
Manufacturer's Representative (Canada)
Radio Active Designs
Pro Audio equipment repair and upgrades.
~416-720-0887~
ca.linkedin.com/pub/ike-zimbel/48/aa1/266

Mac Kerr

  • Old enough to know better
  • Administrator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 7546
  • Audio Plumber
Re: Reverse Engineering Fade Margin Calculations
« Reply #4 on: July 27, 2016, 04:54:57 PM »

One thing to keep in mind is that LMR-400 and similar low-loss cables are really, really stiff. This can make it a challenge to route into a rack, or on to the front of your antenna distro if you are front-mounting the BNC's. Likewise at the antenna end. I have seen these cables place significant strain on the BNC's they are connected to, and do things like point your antennas in an unintended direction because that's the way the cable wants to bend. At the risk of alienating the entire interweb, I'm going to say that for the 25' run, a good quality RG-58 would be just fine and an RG-8U would be more than adequate for the longer run. It's been my experience that if you are worried about signal loss over shorter cable runs, you've got other problems.
Best,
Ike

On the other hand, it's a permanent install. Running stiff cable shouldn't be that big a problem, and as long as you're spending the money and effort to run the cable you may as well run the good stuff. LMR400 is overkill for a 25' run, but in a permanent install cable runs tend to be longer than you think, and the 75' might easily be 100' or more where LMR400 may be the right choice.

Mac
Logged

Michael Fisher

  • Newbie
  • *
  • Offline Offline
  • Posts: 7
Re: Reverse Engineering Fade Margin Calculations
« Reply #5 on: July 27, 2016, 06:52:39 PM »


...for the 25' run, a good quality RG-58 would be just fine and an RG-8U would be more than adequate for the longer run. It's been my experience that if you are worried about signal loss over shorter cable runs, you've got other problems.
Best,
Ike

...LMR400 is overkill for a 25' run, but in a permanent install cable runs tend to be longer than you think, and the 75' might easily be 100' or more where LMR400 may be the right choice.

Thanks gentlemen!

Looks like Belden 8214 and 8240 can be found locally, but not much cheaper than CQ106 or LMR400UF

For the minimal price difference here in the great white north, I'd be inclined to stick with low-loss variants. Would these more flexible variants of be more suitable?

Another thought was to use pre-made cables to take the termination quality out of the equation.
Logged

Mac Kerr

  • Old enough to know better
  • Administrator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 7546
  • Audio Plumber
Re: Reverse Engineering Fade Margin Calculations
« Reply #6 on: July 27, 2016, 08:38:54 PM »

For the minimal price difference here in the great white north, I'd be inclined to stick with low-loss variants. Would these more flexible variants of be more suitable?

Another thought was to use pre-made cables to take the termination quality out of the equation.

In a permanent install you should be able to deal with the stiffer cable, it's not like you're going to be unwinding it and coiling it up every night. You just need to be aware of the slightly larger bend radius so you don't damage the cable while installing it.

Mac
Logged

Michael Fisher

  • Newbie
  • *
  • Offline Offline
  • Posts: 7
Re: Reverse Engineering Fade Margin Calculations
« Reply #7 on: July 27, 2016, 11:26:32 PM »

In a permanent install you should be able to deal with the stiffer cable, it's not like you're going to be unwinding it and coiling it up every night. You just need to be aware of the slightly larger bend radius so you don't damage the cable while installing it.

Great, thanks!

Just to confirm, does my math above make sense? I'm still not sure what to use for "coupling loss", and want to make sure my plan has the best chance of surviving first contact with reality. :)
Logged

Jason Glass

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 918
    • CleanWirelessAudio.com
Re: Reverse Engineering Fade Margin Calculations
« Reply #8 on: July 28, 2016, 12:55:36 AM »

Hi Michael,

Kudos to you for doing the job right!

Coupling loss is tough to predict and is closely related to polarization mismatch loss. This is why circularly polarized IEM TX antennas like the CP Beam and PWS HA-8089 are so popular; they guarantee that this loss is fixed at -3.1dB, no matter what the angular orientation is between the TX and RX antennas.

If you're willing to save a lot of dough by terminating your own connectors, check out Tessco's cost on Belden 9913F7 cable. It's low loss and on par with LMR-400 and very flexible.

Sent from my mobile phone. Please excuse the inevitable spelling and grammatical errors.

Michael Fisher

  • Newbie
  • *
  • Offline Offline
  • Posts: 7
Re: Reverse Engineering Fade Margin Calculations
« Reply #9 on: July 29, 2016, 03:46:28 PM »

If you're willing to save a lot of dough by terminating your own connectors, check out Tessco's cost on Belden 9913F7 cable. It's low loss and on par with LMR-400 and very flexible.

Thanks for the vote of confidence Jason! One of the reasons I've lurked so long here without posting is that I've never yet had a question that couldn't be answered with the search function. Maybe my post can now help others with a similar question. Hopefully that "coupling loss" variable is not too far off from typical for sharkfin/omni antennas.

I had noticed 9913F7 in the Ham radio cable loss reference table from Nasa, and it seems to be a popular product at RF-oriented sites. It looks like it is popular for a reason!

However, when I checked Tessco's prices I found them to be higher than I can find elsewhere though - perhaps they offer better pricing for established accounts? (As a Canadian, I also have to fight with USD-CAD conversion and cross-border shipping costs, unless I can find things locally, so the list price isn't always the whole story for me.) So far, I've found the best Canadian or USD-converted pricing (including shipping) for bulk cable or cable assemblies at Amateur Radio Supplies, with flat-rate shipping to Canada. (I don't work for them or have any other relationship, this isn't a shill post.) Anyone have experience with them recently? It seems their reputation in the past was not so great.

I'm finding a typical cost adder around +$10-14 USD per terminated cable assembly (plus connectors, which I'd have to purchase either way, but avoiding the cost of the necessary crimpers and the skill to use them, neither of which  have).  I think I'm okay having them terminated professionally - then I don't have to worry about a hidden source of suboptimal performance. This is definitely a cry-once, buy-once item for us. Fortunately, BNC connectors are low-profile, and it's not like we'll have to pull a VGA connector through conduit.

« Last Edit: July 29, 2016, 03:54:29 PM by Michael Fisher »
Logged

ProSoundWeb Community

Re: Reverse Engineering Fade Margin Calculations
« Reply #9 on: July 29, 2016, 03:46:28 PM »


Pages: [1] 2  All   Go Up
 



Site Hosted By Ashdown Technologies, Inc.

Page created in 0.025 seconds with 25 queries.