Am 30.04.20 um 03:01 schrieb Bob kb8tq:

... lots of ebay offers snipped

The problem is, they will do their Job at RF, but the GPS receiver

usually has to feed the antenna pre-amplifier with DC through the cable.

The 4 RX candidates should not have to fight for the right / duty to do
this,

with their possibly conflicting offers. Given that DC makes it through the

transformers (or whatever) of the splitter of choice at all.

So, probably a bias tee is needed also, and some replacement dc power

source, which is the easy part.

Also, the 4 receivers should not see each other DC-wise, although there

is usually a protection against catastrophic abuse.

Cheers, Gerhard

There is a Greek ham who sells a nice, very small, 4 way splitter with amp on eBay.  Sorry, no URL handy.  It works well, but is L1 only, if that matters.  It costs a lot more than a Mini-circuits splitter, but a lot less than a "real" GPS one.

John

On Apr 30, 2020, 3:16 AM, at 3:16 AM, Bob kb8tq kb8tq@n1k.org wrote:

Hi

Pretty much all the Mini Circuits splitters will pass DC. The only ones that
don’t are the ones that show “goes to DC” on the data sheet. Those are
resistive splitters. I’ve run thousands of GPS modules through a lot of different
Mini Circuit splitters. None of the “right”. ones have ever had a problem passing
DC (or failed in any way).

DC blocking ports is generally a good idea. You can get “cable TV” dc blocks
for <. $1 each. They have F connectors on them so total cost to get to SMA
is more than $1. For a large splitter setup ( like say 64 ports … ummm … errr …)
that can be a lot of DC blocks.

I tend to block all DC to the splitter and feed the antenna from a bias tee. They
also are pretty cheap. It eliminates a “single point of failure” when the device that
does pass DC gets unplugged. ( Actually it just moves to the bias supply wall wart.
Practically the failure goes away).  The bias tee approach also provides a very tiny
improvement in lightning immunity. (you loose a 5V wall wart and not a GPSDO).

=======

Devices with gain generally also have filters in them. We now are in a world that
has cheap multi GNSS / multi band modules in it. Both are already in production
GPSDO’s. None of the filtered amps will pass multi band GPS / GNSS. Many of them
are sharp enough to take out Glonass based mulit-GNSS.

Does this matter if your splitter is a $30 (delivered) sort of unit? Maybe not. It’s
all worth considering before you move up into the “hundreds of dollars” range.

Bob

On 4/30/2020 7:33 AM, Bob kb8tq wrote:

Hi Bob and all,

That's absolutely correct.  Something you might consider is what I did -
I got a pair of Mini-Circuits ZC6PD-1900W's cheap on the e-place.
Frequency-wise and loss-wise, not perfect, but good enough for me.
Problem was, like you note, all the ports are DC-pass, and I wanted only
output port 1 to be supplying DC to the antenna.

So I opened mine up, carefully cut the trace going to outputs 2 and up,
and bridged them with, IIRC, a 0.01uF ceramic capacitor to block DC.

Before and after plots on the VNA at work showed very little increase in
loss after the mod compared to before the mod.

Just some food for thought.  :)

thanks much and 73,
ben, kd5byb

Hi

At 1.5 GHz something like a 50 to 100 pf NPO is probably the best
pick for a blocking cap.

Why? It does a fine job at the intended frequency. It acts as a high pass
down lower where lightning energy is a bit higher ….

Bob

So, being a heartless butcher I opened up one of the HP 8-port GPS
splitters (forget the number) and removed the sawtooth filters so the
unit would pass both L1 and L2.

I found that to bridge the filter pads even the smallest cap I had (I
think 7pf) caused problems.  I ended up making "gimmick" capacitors out
of a couple of inches of #26 magnet wire twisted together.  That has
worked well; there's a little difference in gain between ports, but not
enough to worry about.

John

On 4/30/20 11:39 AM, Bob kb8tq wrote:

I use splitters made for satellite TV with good results, I think. They
have response 5 MHz to >2 GHz and pass DC from any of the outputs to the
input, so any or all GPS receivers can feed the antenna.  Some cabling
can be done with F connectors, or I add adapters as necessary (F to BNC
adapters are common).  I even have some amplified splitters, though I
modify them so they pass the normally 5V from GPS RXs to antenna and
have external 12V power for the amplifier.

73,

David

On 4/29/20 8:18 PM, Frank O'Donnell wrote:

This InStock Wireless 4-Way splitter is a passive device, meaning no
amplification is included. The headline spec of "Insertion Loss 0.4 dB
typical" may be true but does not tell the whole story. Below that spec,
is the column heading of "Insertion Loss Above 6.02 dB" explaining that
each output is down 6 dB due to the 4-way power split without
amplification. Thus their Marketing Department is emphasizing the low
insertion loss but downplaying the loss due to power splitting without
amplification.

This may be adequate if you have enough gain in your GPS antenna and low
enough loss in the coax down feed but it may not be what you expect when
casually reading the InStock Wireless spec. Power splitting is splitting
power; if you divide power 4 ways, each output has to be 6 dB down.

On the good side, this splitter does work on other GNSS systems.

And, they offer (SMA in this case) 50 Ohm terminators, which should be
used on any port that is not connected to a receiver.

Very few GPS splitters need a Bias-T because they typically have one
port that passes through the antenna voltage provided by a receiver,
usually the one connected to Port 1. The other ports are dc blocked.

The GPS Source S14WI 4-Way splitter (as configured) I described here
previously is amplified and has a more sophisticated method of deriving
the source for the antenna voltage, detecting which port, starting at
Port 1, has source voltage available and using that port while dc
blocking the other output ports. When I bought my NOS units several
years ago on eBay, I paid $65 each plus some modest international
postage; I considered that a great price!

Output ports connected to a receiver but not the one powering the
antenna should present a dc load to each receiver so that the antenna
failure detector in the receiver does not falsely report an antenna
failure. Most GPS splitters include a dc load, often 200 Ohms, to output
ports not powering the antenna.

I do have Type N 50 Ohm terminators but currently each output port on my
splitter is connected to a GPS device.

There is a GPS Source 4-Way multi-GNSS splitter offered on eBay now as
Item 401835679864. However, I don't have a data sheet explaining the
model number letters shown to know if this is an amplified splitter or
not; both passive and amplified versions of the S14 splitter are
available. The eBay listing title reads, "GPS Source 1x4 Stamdard
Amplified Splitter - S14 (S14S-E-SF)" but I would not take that as a
guarantee this particular model is amplified since both versions are
offered with that same basic model number. Am I a bit cynical? Well,
maybe...

Larry

On 4/30/2020 2:38 AM, Chris Burford wrote:

--
Best wishes,

Larry McDavid W6FUB
Anaheim, California  (SE of Los Angeles, near Disneyland)

On 4/30/2020 10:39 AM, Bob kb8tq wrote:

Hi Bob and all,

You are 100% correct!  I decided to look up my notes on my modification
and I see that I did use a 100 pF NPO cap versus the 0.01 uF value I
originally mentioned.  I should know better than to try to remember
anything off of the top of my head like that.  :(

I'll bet that I used 100 pF based on discussions here, as it's almost
impossible I would have made the right choice on my own!  I'm a
mechanical engineer, after all...  ;)

I also found my VNA test results:  my modified splitter was -9.1 dB loss
at 1.525 GHz.  :)  Theoretical should be 7.8 dB, so it's not great, but
it's not horrible either.

Works just fine - feeding two Trueposition units, an NTP server, a
lightning detector, and a uBlox M8N.  When my HP Z3801 was working, it
was on the sixth port...but it went nuts about six months ago and since
I had the TruePosition units, I never bothered to do more than basic
troubleshooting on it.

thanks much and 73,
ben, kd5byb