Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

   Home   Help Search  
Pages: [1]   Go Down
  Print  
Author Topic: Homebrew HF distribution amplifier  (Read 5665 times)
AA4HA
Member

Posts: 1386




Ignore
« on: September 07, 2011, 08:40:58 AM »

I have "a bunch" of high end HF receivers that I am trying to feed from an assortment of antennas (Beverages, fan-dipoles, loops, etc...) and am piecing together a distribution amp/ bandpass filter for my setup. Here is a written block diagram;

Antenna --> Surge Protection --> Low Pass Filter --> Push-Pull J310 amplifier --> Low Pass Filter --> Power Splitter/Combiner --> Receivers

The low pass filters are eBay specials from people who have been parting out Racal 6790/GM receivers. These are the four pole low pass filters that roll off at around 35 MHz and have around 2 dB of in-band insertion loss. I have spent about $25/each on the low-pass filters,originally as spare parts on a couple of 6790/GM radios I already have in use.

The Push-Pull J310 amplifier has about 15 dB of gain with a noise figure of around 2 dB from 0.5 to 50 MHz.

The Power Splitter/Combiner is a Mini-Circuits ZFSC-16-3 (16 port) and is spec'ed out from 1 to 30 MHz. Insertion loss is around 12.5 dB with 40-50 dB of isolation between ports.

The surge protection is a hybrid device of my own design using a coupling capacitor, low voltage gas tube, high speed diodes, etc...

Working up the calculations (losses/ gains/ connector-jumper losses) it is about "a wash" as far as performance. I am not trying to make a high gain pre-amp for a single radio as most receivers out there do not suffer from a lack of gain. My concern is in dealing with the out-of-band high power signals that can drive things into non-linear behavior. I will be aligning the two low pass filters to clip out the AM BCB (<1800 KHz) and high frequencies (>32 MHz). I am using two filters as the first one is to keep the pre-amp from freaking out from high signal levels and the second filter is to keep any sort of out-of-band "junk" from the push-pull J310 amplifier from making it to a receiver.

Oh, all unterminated BNC connections on the splitter/combiner will be terminated with 50 ohm plugs.

My questions are;
1. Do you think that 40-50 dB of port-port isolation is sufficient between the receivers?
2. For those who are familiar with the low-pass filters on the Racal 6790/GM do you think the shape factor will be sufficient to minimize non-linear action? The spectrum display shows the insertion loss >45 MHz is around 65 dB down.
3. Should I make some provisions for a small antenna tuner between the splitter/combiner and each receiver? They will not all be on the same frequency (but sometimes a few may).

This setup is feeding the following receivers;
(2) R-390A's
(1) Hammarlund SP-600JX17
(1) W-J 8718A
(2) Racal 6790/GM
(1) Harris RF-590
(1) Cubic CDR-3250
(1) AOR AR-8600MKII
(1) Ten-Tec RX-320

Any suggestions would be appreciated.

thanks,
Tisha Hayes, AA4HA
« Last Edit: September 07, 2011, 08:44:01 AM by AA4HA » Logged

Ms. Tisha Hayes, AA4HA
Lookout Mountain, Alabama
WB6BYU
Member

Posts: 13125




Ignore
« Reply #1 on: September 07, 2011, 09:06:22 AM »

What is the IP3 of your amplifier?

How strong are the strongest local AM and FM broadcast stations?  (These might need
additional notching if they are too strong.)

Do you have any local SW broadcast stations?  A couple switchable 10/20dB pads in the
input path might not be a bad idea, just in case.

I remember a story in Technical Topics about one of the tube rigs that had a strong
LO output at the antenna port.  (Possibly a HRO?)  You might want check your rigs -
probably not a problem with newer ones using balanced mixers, but a pentagrid converter
doesn't give a lot of isolation.  I think that would be the limiting factor, otherwise 40 to
50dB should be adequate.
Logged
AA4HA
Member

Posts: 1386




Ignore
« Reply #2 on: September 07, 2011, 11:31:34 AM »

What is the IP3 of your amplifier?

How strong are the strongest local AM and FM broadcast stations?  (These might need
additional notching if they are too strong.)

Do you have any local SW broadcast stations?  A couple switchable 10/20dB pads in the
input path might not be a bad idea, just in case.

I remember a story in Technical Topics about one of the tube rigs that had a strong
LO output at the antenna port.  (Possibly a HRO?)  You might want check your rigs -
probably not a problem with newer ones using balanced mixers, but a pentagrid converter
doesn't give a lot of isolation.  I think that would be the limiting factor, otherwise 40 to
50dB should be adequate.

The IP3 is somewhere around 9-11 dB

Using the AOR 8600MKII in spectrum analyzer mode I show three significant signals in the 88-108 MHz band;
89.9 MHz at -40 dBm
91.5 MHz at -6 dBm (really strong)
105.1 MHz at -10 dBm

There are 21 of other stations ranging from -65 to -110 dBm. The noise floor in the FM broadcast band is down around -115 dBm. (measured using a discone at 40' with no preamplifier).

In the AM band I have signals at;
570 KHz at -72 dBm
860 KHz at -78 dBm
1154 KHz at -82 dBm
1730 KHz at -77 dBm

Generally the noise floor in the AM BCB is around -110 dBm

The AOR is not really all that good for HF/BCB as the noise floor is not that great. Across the 1.8 to 30 MHz band the internally generated noise (with a 50 ohm terminator in place of the antenna connection) is right at -110 dBm. With a random length of wire for a HF antenna signal levels are between -60 and -80 dBm (radio in CW mode). There are four or five SWL broadcast stations between 15300 and 15700 KHz that come in around -40 dBm. For a comparison, WWV at 15000 KHz is at around -60 dBm with the same chunk of wire.

I could do the same thing with the HP or Hameg spectrum analyzer but using an SDR radio is just too tempting for a quickie email as I just need to fire up Spectrum Commander on my computer.

Back of the napkin numbers show me that the received signal for the FM broadcast stations should be down around -70 dBm for the NPR affiliate (91.5 MHz and the strongest FM signal in the area). I could make a quarter wave coaxial stub filter to attenuate that down further as it would have a low enough Q that it could also attenuate down the 89.9 MHz signal as well.

Into any of the decent receivers I usually see -120 to -130 dBm noise floors across HF. I do not have neighbors (at least who use 'lectricity <j/k>) who are closer than a few hundred meters. My locally generated RF noise is caused by the porch light that I can turn off with a switch and my computers.

I would think that if I have a receiver that is generating spurious signals that are visible at the antenna it would be the Hammarlund SP-600. Later designs like the R-390A (if the navy mods were done) and the newer surveillance receivers were specifically designed to be stealthy and the RF usually goes right into a mixer stage. Someone must have learned that a torpedo, missile or bomb could follow your receiver IF signal right back to the bull's eye. I knew that was of concern for the navy as there were worries that the German U-Boats were DF'ing convoys through that technique.
Logged

Ms. Tisha Hayes, AA4HA
Lookout Mountain, Alabama
WB6BYU
Member

Posts: 13125




Ignore
« Reply #3 on: September 07, 2011, 12:25:39 PM »

And the Japanese DFed them in the Pacific as well.  The SW BC sets with a single 455kHz IF
and no RF amp were some of the biggest problems, since the input tuned circuits gave
little LO rejection on the higher bands.  Though I don't know if they were nearly as bad
as the regens they replaced - there were many cases of stations making contacts
using a regenerative receiver as a transmitter as well.  Adding an RF amp and using a
higher IF make a big difference, as does a balanced mixer arrangement and better
front-end filters to reduce LO radiation.

Since you have the SDR handy, you should be able to quickly check the LO output level
of the receivers.  An output of -50dBm with 40dB isolation across your switch still puts
it well above the noise floor in the other receivers.  Fortunately that is a source of
interference that you also have control over, though it might not always be obvious.


There was a study done in England (I believe) that a rhombic pointed at the center of Europe
would generate about 1 watt output during the night, due to summing all the signals across
the HF bands at the antenna.  That was probably 30 to 40 years ago.  I don't know if any of
your wideband antennas would pick up quite that much RF, but it might suggest including an
attenuator just in case.

Otherwise, as long as the FM broadcast signals are attenuated so they don't contribute
to overload of the distribution amplifier, it sounds like you have a pretty good setup.
Logged
AA4HA
Member

Posts: 1386




Ignore
« Reply #4 on: September 07, 2011, 02:18:08 PM »

Actually I have a Beverage antenna pointed at Europe that is 3200 feet long. It just happens to be the longest diagonal distance across my property. I did have to custom wind a 9:1 balun as my flimsy little Litz wire baluns were burning up every time the sky flashed with lightning. It had even va-poo-rized a few ferrites and left a fine copper coating from the wire inside of the balun box.

I should put a bridge rectifier on the antenna and look at the voltage, that might be interesting.
Logged

Ms. Tisha Hayes, AA4HA
Lookout Mountain, Alabama
WB6BYU
Member

Posts: 13125




Ignore
« Reply #5 on: September 07, 2011, 04:37:25 PM »

But the beverage signal levels are typically much lower than a rhombic for the same wire length.
That's why they are often used with a preamp.

One of the big log periodic wire arrays would be interesting, though.  I'm thinking of some of the ones
that I saw in the salt flats of San Francisco bay, some of which were vertically polarized, hanging
from a catenary sloping down from a tall tower.

With an IP3 of +10dBm, the preamp will go into compression somewhere below that level.  I'd probably
design for a maximum input of -10dBm if that is an issue.
Logged
G3RZP
Member

Posts: 4445




Ignore
« Reply #6 on: September 08, 2011, 12:38:51 AM »


>There was a study done in England (I believe) that a rhombic pointed at the center of Europe
would generate about 1 watt output during the night, due to summing all the signals across
the HF bands at the antenna.  That was probably 30 to 40 years ago.<

It was. 1974, and if I remember correctly, I did the measurements in the autumn of that year when I was working on a high performance rx development at KW-Decca. I left there in March 1975 for a job at Racal with more than double the pay.

Is that IP3 the input IP3 or the output IP3?
Logged
M0HCN
Member

Posts: 473




Ignore
« Reply #7 on: September 08, 2011, 06:01:10 AM »

You might want to consider going to a bipolar Norton feedback design like the Clifton Labs product, 10db gain and IIP3 somewhere in the +40dBm region.

For the most part preamps before the bandpass filters are IMHO the kiss of death on HF unless the aerial is electrically very short or inherently very narrowband.

regards, Dan.
Logged
G3RZP
Member

Posts: 4445




Ignore
« Reply #8 on: September 08, 2011, 06:37:11 AM »

An amplifier prior to the BPF using 3 off 6EH7 in a distributed amplifier giving about 17 dB gain should be OK on IMD.
Logged
W8JI
Member

Posts: 9304


WWW

Ignore
« Reply #9 on: September 08, 2011, 04:12:25 PM »

The DX Engineering preamp has one of the highest OIP3 around. Typically around 45 dBm. If you run 15 volts or so, they will do about 50 dBm OIP3. I know, because that is what I use here. I can light a 12 volt lamp off the output here on my beverages.

As for the push pull J310's, you should be able to get an OIP3 of about 30 dBm. That's what I get without much fuss.

Look at this system:

http://www.w8ji.com/images/New%20Contest%20Room/Contest%20station%20CQWW2007/receiver-switch-matrix.jpg

on this page:

http://www.w8ji.com/contest_station_w8ji.htm

I run my beverages into six different amps that split 4 ways with magic T splitters. The four way split is about 7 dB loss or slightly less. The net gain is about +10 dB. That's more than enough for a beverage.

20 dB or so isolation is far more than enough in practice. Been there, do that all the time.  :-)

We run four or more receivers on this system, and run multiple 1500 watt transmitters on site while receiving. I can full duplex on 40 meters with about 5 kHz spacing between TX and RX. That's with no BP filters either.

73 Tom
« Last Edit: September 08, 2011, 04:15:38 PM by W8JI » Logged
Pages: [1]   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!