Pre duplexer vs post duplexer power

[emtprt]

What is a reasonable amount of loss through a duplexer?

Currently I am losing 10 watts 45/35 from a gr500 repeater. I added a 100 watt amp set at 85 watts out and I am losing approx 18 watts. Is this normal?

I am also considering purchasing a new duplexer (compact- 50watt) for a second repeater. Any recommendations.

Thanks

Mark

[HumHead]

85 watts in for 67 watts out comes out to just a hair over 1dB of insertion loss, which is a perfectly respectable number.

By "compact" duplexer, I assume (dangerous word there) that you are talking about the small notch-notch type mobile duplexers. You should be aware that the notch type duplexers will not provide a lot of selectivity if you are at a crowded site with a lot of RF "noise". If you are at a remote and quiet site, you may be able to get away with it.

[emtprt]

The repeater will be the only radio at the site. It is a R1225 UHF repeater. with a 8db omni directional .

[kcbooboo]

"Reasonable Loss" is a loaded question. It does depend on several factors. However, a good rule of thumb is about 0.5dB per section.

Just about everything is done using dB for relative measurements, rather than actual power in watts. To convert power to dB, you need a calculator or a table of logarithms. There are several reference levels, but for most two-way radio stuff, the level known as 0dBm is one milliwatt across 50 ohms. A 10dB change is equivalent to a 10X power change. Using this same reference, a 1 watt signal is +30dBm. A 10 watt signal is +40dBm. Your 85 watt input power is equal to +49.3dBm, and the 67 watt output power is equal to +48.3dBm, so the loss is 1.0dBm. As mentioned above, this is respectable and normal loss for most two-section duplexers.

Bob M.

[HumHead]

Bob:

Good explaination. I'm afraid that I didn't "show my work" like the math teacher used to expect.

When figuring dB ratios, you can also figure them directly for two values, without having to first go to the conversion tables, assuming that both values are expressed using the same unit of measurement.

A dB power ratio (voltage is a slightly different story) is defined as 10 times the common log of the first value divided by the second value or 10[log(V1/V2)].

In this case, I just took 85watts (input power) and divided it by 67watts (output power). That yields a ratio of 1.2687. Hit the log button and you get .1034. Bump the decimal, and you have 1.03dB.

[kcbooboo]

Ah So, grasshopper, but your equation is flawed !

10 log (P1/P2) is good for POWER. If you use voltage ratios, the correct equation is 20 log (V1/V2). It has something to do with voltage being squared (hence twice the log value) compared to power. I didn't want to muddy the waters by throwing in voltage ratios.

But you are absolutely correct, that you can use just the power ratios to figure the dB gain or loss. I just chose the 0dBm reference level to make it easier for me.

Bob M.

[Wowbagger]

Ah So, grasshopper, but your equation is flawed !

10 log (P1/P2) is good for POWER. If you use voltage ratios, the correct equation is 20 log (V1/V2). It has something to do with voltage being squared (hence twice the log value) compared to power. I didn't want to muddy the waters by throwing in voltage ratios.

GO TO THE BACK OF THE CLASS, YOUNG MAN.

Decibels are ALWAYS 10log10(N).

Now, when you are dealing with dB(power) - e.g. dBmW or dBW, a doubling of voltage will cause a 6 dB increase in power for a constant impedance due to P=V^2*R, but that is because you are confabulating units, not because of some strange definition of dB.

The term "decibel" comes from "deci-" (one tenth of) and "Bel" - the base unit. Researchers at Bell labs found that for many things, you needed a logrithmic unit, so they defined the Bel - one Bel increase is a ten times increase in the magnitude of the thing being measured. The then found that a Bel was too big for most of the uses they needed, so they went with one tenth of a Bel - by normal metric prefixes that became the deci-Bel (which is why the abbreviation for the unit is dB, not db).

It is perfectly valid to have dBV, dBmV and the like, wherein a doubling of the voltage gives you a 3dB increase in dBV.

You can also get into weirdos like dBHz and dBkHz when looking at the dB scale of an audio analzyer looking at the demodulated output of an FM receiver.

(Then you get into dBmW-EMF where you account for the power lost across the source impedance, and dBmW-P where you only consider the power disspated in the load, and.... Metrology is a :o.)

[HumHead]

If you read that again, what I said was:

A dB power ratio (voltage is a slightly different story) is defined as 10 times the common log of the first value divided by the second value or 10[log(V1/V2)]

The voltage equation is different because the voltage to power formula is P=E^2/R, however, I also skipped over that here, for the sake of clarity.

Harumph!

[kcbooboo]

You're right, you did say that, however you also said 10log(V1/V2), and the last time I looked, the letter "V" was used mainly for voltage, and for a voltage ratio, I think it's still 20log(V1/V2).

I was careful to use the letter "P" for power in my equation.

I think we're both right, one way or the other, eventually !

Bob M.

[HumHead]

D'oh! Sloppy me.

I used "V" for Value, where most of the world would have read it as Volts (where I was taught to use "E").

That's probably why I do much better in the real world than I ever did in math class.

[Will]

HumHead writes:
"That's probably why I do much better in the real world than I ever did in math class. "

Me too!

We call them Dog Biscuits.

[Karl NVW]

Good, the dispute about 10 P vs 20 V is over. Now back to the original inquiry:

The duplexer insertion loss does indeed look acceptable from a dB standpoint. However, the power being lost in the cans is twice what it was before, so you should expect some thermal drifting. The individual filter sections detune as they get hot or cold, which will reduce the isolation from transmit to receive. It could also eventually cause permanent damage to the duplexer if the transmitter noise into the receiver keeps the repeater keyted continuously.

[ASTROMODAT]

I would recommend that you refer to your duplexer manufacturer's specs for loss on the TX and RX sides of your duplexer, and pay carfeful attention as to how they define the loss . Generally, the insertion loss due to the connectors is about 0.1 dB per connector, and the interconnecting cables themselves will easily add another 0.1 dB or so. The filters (cans) themselves are where most of the loss occurs. It's true that a notch reject system, such as TX/RX Sytems duplexers, will run about 1 dB total, or a bit more/less of loss, per side. However, a true bandpass duplexer, like a Motorola T1507, will run about 2.5 dB per side (which involves two cans), all told, with all effects taken into account (connectors, cans, inter cabling, etc.), so duplexer loss varies widely.

Again, I'd check your manufacturer's specs, and when you determine the actual loss, make sure you do an actual apples to apples comparison when you calculate total dB of loss. If its way off from the vendor's nominal specs, than you either have a poorly tuned duplexer, broken cable fittings and the like, or some other problem.

Good luck.

[Big Towers]

More IMPORTANT than the math calculations is this, a small notch only duplexer usually found in GR300s and 500s is rated for 50 WATTS, not 85! The celwave made moto duplexers are typically notch only as well.

[Dan562]

I don't know if any of you backed up and looked at this side of the equation from the original post.

Fact: /\/\ GR500 Repeater 45 Watts Output with 50 Watt Duplexer

Fact: +8 dB Antenna

Fact: 100 Watt FM RF Class "C" Amplifier Set at 85 Watts Output not quite +3 dB Gain

Fact: Nothing provided for type of Repeater Application, Coaxial Feedline, Height of Antenna Above Ground Level, Height Above Average Terrain, Buildings, Trees, Hills or Subcriber (Portables) Path Losses.

Fact: Doubling the RF Wattage Output equaling +3 dB Gain will only increase the overall RF Transmit coverage 25%. It does not increase the Talk-In Coverage Range or Receiver Sensitivity!

With providing this information, I'll pick a range out of the air, 15 miles and multiply it by 25% equals 3.75 miles farther or 18.75 miles Total RF Transmitted Radius from the Repeater. I'll bet you can not even tell the Transmitted Range difference out in the real world.

Fact: You've got to increase RF Power Output by +6 dB, +9 dB or +10 dB to notice an increase in Transmitted Range Coverage. There are many ways to increase the Receiver's Talk-in Coverage but do not attempt to install a Pre-amplifier with the /\/\ GR500s repeater's receiver as you are asking for more problems.

[ASTROMODAT]

Very well stated, Dan.

[emtprt]

More info

The duplexer in use is not the stock Motorola duplexer. It is a rack mount 100watt Decibal.

The cable run is approx. 30' with RG213 cable. The omnidirectional antenna is located on the roof of a 3 story building 142' above sea level.

The amplifier was installed to improve pager coverage.

Thanks for all the info so far.

[Dan562]

emtprt wrote: The amplifier was installed to improve pager coverage.

My exact point, you must increase the RF signal level by +6 dB minimum, in your case from 45 Watts to 180~200 Watts output to detect any difference for RF signal coverage on the Radio Pager / Receiver.

When considering UHF Radio Paging systems, the Pager has an automatic -16 dB Path Loss (worn at Belt Level) in itself. Then you must consider the Buildings, Trees (Foliage), Hills, Feedline and any RF filtering (Band Pass Cavities or Duplexers) and add all of the Minus numbers together totaling "X" amount of Negative Losses. Then you take the Plus Gains being the Antenna, Height above Ground Level and the RF Wattage Output. Take the two sub-total number subtracting the Plus or Gain amount from the Negative or Losses, ending up with the Total System Losses and calculate the RF coverage area from that Negative number.

I know there are software programs available for doing these RF coverage calculations. You might find a few individuals here on Batlabs that can provide this service.

Adding +3 dB Gain to ANY RF System will ONLY provide 25% more transmitted coverage, no matter if it's a 2-way or Paging system. You can not defy the Law of Physics.

Here's another item you need to consider, depending the UHF coverage that your customer wants or needs and this +8 dB Gain antenna that you are using, will determine how well the RF system coverage actually is. Using a high gain antenna will provide further out lying signal coverage but at a price. The exception is the closer in signal coverage penetration, I'll say in a hospital's building complex will fail more often because you're sending the majority of the signal away from the building and creating what is known as the "Umbrella Effect." Ninety percent of the signal is radiated outward while only 10% is penetrating down into the building.

There's an alternative, you have a building 3 stories tall, roughly 45 Feet high. Purchase 50 Feet of Rohn 25 or 45 Tower installing and guying it on the Flat Roof of the building. Install a +6 dB Gain antenna at the top of the Tower with a minimum of 1/2" or better yet 7/8" Andrews Heliax Feedline cable between the antenna and Duplexer common port. By raising the antenna Height, Doubling the Height (the Antenna combined with the Height are key issues here) will automatically provide +6 dB of System Gain. Even with lowering the Antenna Gain by 2~3 dB will provide the outlying System Coverage and better RF Signal Penetration into the building below. Now if you want to add the Class "C" RF Amplifier to the installation, you may detect the signal difference.

Dan