NanoVNA V2 / Forum

Note: this page is a mirror of https://groups.io/g/NanoVNAV2.
Click here to join and see most recent posts.

Duplexer Tuning


Sam Reaves 2020/08/03 13:50

Someone had asked about using the NanoVNA for duplexer tuning.
There is an excellent video on that by David Houser KG5RDF here:

https://www.youtube.com/watch?v=y7PEDRN1PpY

Sorry but I could not find the original poster of the question about being able to do this.

73,

Sam
W3OHM

Seth Miller 2020/12/31 14:17

Here's another that I think is a little more straightforward. The author is rightfully skeptical of the NanoVNA's ability to replace professional gear for these kinds pf applications due to limited dynamic range, but the home builder can solve this issue by using the NanoVNA to get everything close and then using a radio receiver (which should be good down to -115 dBm) to make fine peaking adjustments.

https://www.youtube.com/watch?v=GipCVEsiqXc

Seth Miller 2020/12/31 14:17

[Edited Message Follows]

Here's another that I think is maybe a little more straightforward. The author is rightfully skeptical of the NanoVNA's ability to replace professional gear for these kinds of applications due to limited dynamic range, but the home builder can solve this issue by using the NanoVNA to get everything close and then using a radio receiver (which should be good down to -115 dBm) to make fine peaking adjustments.

https://www.youtube.com/watch?v=GipCVEsiqXc

Doug Reed 2021/01/02 10:25

10+ years ago I was using my DG8SAQ VNWA to tune duplexors. I would tune them as well as possible using the VNWA sweep. The final step for tuning the notch was to replace the TX signal from the VNWA with a TX signal from a handheld radio tuned to the notch frequency. That typically gave me an extra 20-30dB of signal to fine tune the notch depth. This worked great until one day I used my VX1 HT as the signal source. That was the day I found out the VX1 would key up and transmit before the VCO was locked. As the VX1 swept across the band, it put full TX power into the VNWA and burned out the input mixer. Other than this little problem, the technique worked great.

73, Doug Reed, N0NAS.

Seth Miller 2021/01/03 13:03

Ah yes, one must be very careful about such things. I had a similar occurrence in tuning a notch duplexer where I found that when using Motorola RSS software to set RF output power the test transmission is on a generic test frequency, not the front panel selected frequency... and there went the front end of my receiver.

But that issue with the VX1 is a nasty one, not sure how you'd know that in advance. I guess you could loosely couple a prospective test radio to the VNC in advance to verify frequency stability, before making a direct connection. All part of the fun.

nanovnav2 2021/01/04 05:35

On Sun, Jan 3, 2021 at 01:15 AM, Doug Reed wrote:
> 10+ years ago I was using my DG8SAQ VNWA to tune duplexors. I would tune them
> as well as possible using the VNWA sweep. The final step for tuning the notch
> was to replace the TX signal from the VNWA with a TX signal from a handheld
> radio tuned to the notch frequency.
>

I always find that tuning duplexers just by looking at a tracking generator (or a VNA if you will) gets you in the ball park but is never quite close enough, no matter how sharp your dips will display. So always touch up the band-rejects by feeding the signal generator from the service monitor that is capable of +13dBm output into the antenna port and then a portable radio tuned to the desired reject frequency on the other end port. That way I can tune the rejects on both sides by ear for maximum attenuation by keeping the signal level from the generator just enough into the noise as I go to be able to discern subtle changes while making adjustments. I've been using this for many many duplexer tunes over the years and with this technique I can usually quickly achieve 70-75dB of rejection. Btw while using the TG or VNA for initial tuning never forget to properly load the open port of the other side of the duplexer with a 50-ohm load while tuning the other end otherwise you might see some trace distortion (phantom dips or peaks I call them) due to small reflections caused by the mismatch on the open port.

Siegfried Jackstien 2021/01/04 15:07

i use my kc901 ... can maesure dips down to 120 db!! (if you use
supergood cables with very low leaking and goooood connectors)

greetz sigi dg9bfc

Am 04.01.2021 um 13:35 schrieb nanovnav2@miwww.com:

Doug Reed 2021/01/04 19:10

After the fact I did confirm the problem using a spectrum analyzer set to peak hold and sweeping a range near the fatal test frequency. A radio that comes up on frequency will have a single peak. But the VX1 had multiple peaks that happened when I keyed the radio multiple times. Each peak was when a spectrum sweep was in progress and the VX1 transmitter signal went by... Peak hold would show multiple peaks, but not catch every key up. You do of course want to use an attenuator or distance to make sure you don't overload the test equipment when doing this.

73, Doug Reed, N0NAS.

nanovnav2 2021/01/05 02:08

On Mon, Jan 4, 2021 at 03:07 PM, Siegfried Jackstien wrote:
> i use my kc901 ... can maesure dips down to 120 db!! (if you use
> supergood cables with very low leaking and goooood connectors)
>
> greetz sigi dg9bfc

Of course, and RG142 double shielded cable works best for me.
Are you referring to achieve a 120db reject in a BpBr duplexer?
Duplexers are generally limited to 70-80dB reject, and this is usually
already pushing the limits of the external interconnects.

nanovnav2 2021/01/05 02:14

On Tue, Jan 5, 2021 at 03:10 AM, Doug Reed wrote:

>
> After the fact I did confirm the problem using a spectrum analyzer set to peak
> hold and sweeping a range near the fatal test frequency. A radio that comes up
> on frequency will have a single peak. But the VX1 had multiple peaks that
> happened when I keyed the radio multiple times. Each peak was when a spectrum
> sweep was in progress and the VX1 transmitter signal went by... Peak hold
> would show multiple peaks, but not catch every key up. You do of course want
> to use an attenuator or distance to make sure you don't overload the test
> equipment when doing this.
>
> 73, Doug Reed, N0NAS.
>

Never use TX to adjust a duplexer, and if you must because of
a lack of a proper RF generator, then at the very least use an
attenuator between the TX and the duplexer to be adjusted to
keep signal levels low enough and within hopefully safe levels
for your measuring equipment. But even so I would not recommend
using TX power.

Stephen Laurence 2021/01/05 02:20

Yes, my kc901c and 901s (the first one, and the second one but it does not have decent rbw) are quite good if you know their limitations.

Steve L

Siegfried Jackstien 2021/01/05 11:02

i have a very very good 70cm duplex filter with bandpasses and notches
in both filter chains ... with 6 chambers in each chain (means 12
chambers in sum for that filter!!) ... each chamber has two tuning
screws (resonance ... and coupling to next chamber) so its 6+5 screws
for each filter chain ... or in sum 22 tuning screws ... not easy to get
the notch that deep but its possible to get them very deep down to 100
db or a bit lower ... but then that notch is very narrow and if
temperature does change much (summer to winter on repeater site) you do
not want to do a retune every few month ...

i set the notches a tiny bit broader to around 90 db and be done with it
(also lowers the through attenuation)

tuning is a bit of a compromize but the end result is a superb filter
with low trough loss and high isolation between ports

as an example you have to look for super good swr on the tx side ... not
so much on the rx side ... but you sure all know that ... double
shielded cable and good connectors is a must if we go down into the -100
db area ... and when using a normal rig to use as a sensitive receiver
... then it should be a good shielded mobile or station rig and not a
handheld trx with plastic case

to say it short baofengwoxunchinacheap is forbidden ... grin

if i need best isolation ... i use two different devices (say my kc901
and a pluto) ... one placed on leftside and one on the right side of the
bench ... no interconnections ... no ground couplings between the units
(besides the connections to the to be tuned filter) ... then i can easy
reach the needed 100db plus isolation ... even up to 120 db (but thats
the absolute maximum cause then you see common mode currents on shields
and other effects)

i also tune receiver frontends ... with a low level out from test gear
... and a 60 db attenuator on antenna socket of the rig but also then
... use good cables and connectors

all that is in the minus 100 db area or lower becomes tricky (and is for
sure impossible to get lower on a cheap 60 bucks unit!)

the v2plus4 is at that border ... 105 db below 1.5 gig (with high
average and very slow scanning) and around 90 db at 13 cm

so if you need that high isolation of over 100 db to tune 70cm duplexers
the plus4 would be my choice (if i would not own the much more expensive
kc901!)

greetz sigi dg9bfc

Am 05.01.2021 um 10:08 schrieb nanovnav2@miwww.com:

Stephen Laurence 2021/01/07 12:06

I should add to my previous post, that the two Deepace 901 devices cost me as much in total as 20 or more Nanovnas. That is quite a few toys to blow up with Tx power straight in or esd accidents.

The 901 devices are passable handheld spectrum analyses 0 -3gkz though, albeit with their own limitations. I sorted lots of problems with them.

Steve L

To reply to this topic, join https://groups.io/g/NanoVNAV2