开云体育

I’ve have two NanoVNAs, and both were supplied with a female-female SMA
adapter like on the right. It quite a bit longer than most SMA
female-female adapters. The one on the left is much shorter. There’s
nothing special about the one on the left - it just the first I found when
I opened a drawer.

The photograph of a VNA screen is an HP 8720D VNA. I didn’t perform a fresh
calibration of the VNA, but recalled one previously saved. That was good
enough. The supplied thru, left open has a return loss of 0.368 dB, but I
would expect it to be *much* smaller - certainly under 0.05 dB at 1.5 GHz.
The female-female adapters supplied with my NanoVNAs are particularly
poor.

You may notice a huge dip at 6.42 GHz. Thats some internal resonance. It’s
well outside the range of the NanoVNA, but is indicative of a poor adapter.

There are actually two traces shown on the VNA. One is stored in memory. It
is so close to the zero line you can barely see it, but if you look at the
far right, near 7 GHz, you can just see it dipping below zero due to its
loss

*Experience tells me that the thru supplied with the NanoVNA is very poor.*
If you have a thru that’s physically shorter, it will almost certainly be
better.

Although I don’t show it here, I did put a short on the end of the adapter.
The response looks awful. Essentially the thrus I have been supplied are
the worst I have ever seen. Just about any thru will be better.

Perhaps someone else with a calibrated VNA, who has received a thru like
that on the right, will confirm it is a particularly poor one.



--
Dr. David Kirkby,
Kirkby Microwave Ltd,
drkirkby@...

Telephone 01621-680100./ +44 1621 680100

Registered in England & Wales, company number 08914892.
Registered office:
Stokes Hall Lodge, Burnham Rd, Althorne, Chelmsford, Essex, CM3 6DT, United
Kingdom

On Tue, 8 Oct 2019 at 22:33, Dr. David Kirkby <
drkirkby@...> wrote:

There are actually two traces shown on the VNA. One is stored in memory.
It is so close to the zero line you can barely see it, but if you look at
the far right, near 7 GHz, you can just see it dipping below zero due to
its loss.

I forgot to say, the trace that’s barely visible is the thru on the left of
the photograph. It’s return loss is very close to 0 dB even at 7 GHz.

Dave
--
Dr. David Kirkby,
Kirkby Microwave Ltd,
drkirkby@...

Telephone 01621-680100./ +44 1621 680100

Registered in England & Wales, company number 08914892.
Registered office:
Stokes Hall Lodge, Burnham Rd, Althorne, Chelmsford, Essex, CM3 6DT, United
Kingdom

Hello David,
I had heard rumours of there being bad Female-Female adaptors included with
some NanoVNAs. I have a bunch of cheap chinese adaptors of that kind, the
one that came with the NanoVNA, and what I consider a good one from
Amphenol, kindly donated to me.

I calibrated my NanoVNA with the Amphenol one and checked the included
"long" F-F adaptor, and a random one from my drawer.

The included one I received checks out to be very close to the Amphenol
version, at least to 900 MHz. Beyond that I get too much noise to get
really good readings. The random connector from the drawer showed a drop
almost from 50kHz, of about -0.03 dB. As it's very little, and the
connectors were done up by hand, I would tend to conclude that I was just a
measurement uncertainty.

Interesting to see that yours is so much worse.

--
Rune / 5Q5R

On Tue, 8 Oct 2019 at 23:33, Dr. David Kirkby from Kirkby Microwave Ltd <
drkirkby@...> wrote:

I’ve have two NanoVNAs, and both were supplied with a female-female SMA
adapter like on the right. It quite a bit longer than most SMA
female-female adapters. The one on the left is much shorter. There’s
nothing special about the one on the left - it just the first I found when
I opened a drawer.

The photograph of a VNA screen is an HP 8720D VNA. I didn’t perform a fresh
calibration of the VNA, but recalled one previously saved. That was good
enough. The supplied thru, left open has a return loss of 0.368 dB, but I
would expect it to be *much* smaller - certainly under 0.05 dB at 1.5 GHz.
The female-female adapters supplied with my NanoVNAs are particularly
poor.

You may notice a huge dip at 6.42 GHz. Thats some internal resonance. It’s
well outside the range of the NanoVNA, but is indicative of a poor adapter.

There are actually two traces shown on the VNA. One is stored in memory. It
is so close to the zero line you can barely see it, but if you look at the
far right, near 7 GHz, you can just see it dipping below zero due to its
loss

*Experience tells me that the thru supplied with the NanoVNA is very poor.*
If you have a thru that’s physically shorter, it will almost certainly be
better.

Although I don’t show it here, I did put a short on the end of the adapter.
The response looks awful. Essentially the thrus I have been supplied are
the worst I have ever seen. Just about any thru will be better.

Perhaps someone else with a calibrated VNA, who has received a thru like
that on the right, will confirm it is a particularly poor one.



--
Dr. David Kirkby,
Kirkby Microwave Ltd,
drkirkby@...

Telephone 01621-680100./ +44 1621 680100

Registered in England & Wales, company number 08914892.
Registered office:
Stokes Hall Lodge, Burnham Rd, Althorne, Chelmsford, Essex, CM3 6DT, United
Kingdom

Dave,
I have two of the longer SMA F-F adapters, from two similar NanaVNA kits. I checked them on my 8753D. Both look reasonable to the 6GHz limit of my 8753D. Each shows a 900MHz through loss of about 0.07dB, and a reflection loss of about 0.3dB. A random adapter from my lab looks slightly better, but not much.
--John Gord

Hi David

Exactly, the supplied open is pretty poor and has high loss. See attached measurement with the NanoVNA aand NanoVNA saver calibrated with my HP85033C . That is why I have not yet published my characterization as I am searching for confirmation that mine is not a special poor one, and you just confirmed. The NanoVNA calibration setup does not have loss (yet) so using the supplied female female adaptor together with the supplied short, load and open (which I tend to omit but has data for) as a female kit is not providing the best calibration due to this excessive loss. I attach the Sweep of my APC7 20cm airline so you can see the NanoVNA is not that bad ?
The Dip you call internal resonance is simply the adaptor being a 1/4 wavelength resonance and the loss transformed to a low Z.

Kind regards

Kurt



-----Oprindelig meddelelse-----
Fra: [email protected] <[email protected]> P? vegne af Dr. David Kirkby from Kirkby Microwave Ltd
Sendt: 8. oktober 2019 23:33
Til: nanovna-users <[email protected]>
Emne: [nanovna-users] Very poor thru (barrel) supplied with NanoVNAs



I’ve have two NanoVNAs, and both were supplied with a female-female SMA adapter like on the right. It quite a bit longer than most SMA female-female adapters. The one on the left is much shorter. There’s nothing special about the one on the left - it just the first I found when I opened a drawer.



The photograph of a VNA screen is an HP 8720D VNA. I didn’t perform a fresh calibration of the VNA, but recalled one previously saved. That was good enough. The supplied thru, left open has a return loss of 0.368 dB, but I would expect it to be *much* smaller - certainly under 0.05 dB at 1.5 GHz.

The female-female adapters supplied with my NanoVNAs are particularly poor.



You may notice a huge dip at 6.42 GHz. Thats some internal resonance. It’s well outside the range of the NanoVNA, but is indicative of a poor adapter.



There are actually two traces shown on the VNA. One is stored in memory. It is so close to the zero line you can barely see it, but if you look at the far right, near 7 GHz, you can just see it dipping below zero due to its loss



*Experience tells me that the thru supplied with the NanoVNA is very poor.* If you have a thru that’s physically shorter, it will almost certainly be better.



Although I don’t show it here, I did put a short on the end of the adapter.

The response looks awful. Essentially the thrus I have been supplied are the worst I have ever seen. Just about any thru will be better.



Perhaps someone else with a calibrated VNA, who has received a thru like that on the right, will confirm it is a particularly poor one.







--

Dr. David Kirkby,

Kirkby Microwave Ltd,

<mailto:drkirkby@...> drkirkby@...

<>

Telephone 01621-680100./ +44 1621 680100



Registered in England & Wales, company number 08914892.

Registered office:

Stokes Hall Lodge, Burnham Rd, Althorne, Chelmsford, Essex, CM3 6DT, United Kingdom

Hi John
The one 0.3dB is not much less the what David and I measure so the one from your lab is not optimal. A Rosenberger female female was used as S21 thru during my test but the open was from my HP 85033C Kit and the Rosenberger has some 0.05dB at 900MHz and a Amphenol RF has slightly less about 0.08dB as hard to determine due to noise contribution
Kind regards
Kurt

-----Oprindelig meddelelse-----
Fra: [email protected] <[email protected]> P? vegne af johncharlesgord via Groups.Io
Sendt: 9. oktober 2019 02:07
Til: [email protected]
Emne: Re: [nanovna-users] Very poor thru (barrel) supplied with NanoVNAs

Dave,
I have two of the longer SMA F-F adapters, from two similar NanaVNA kits. I checked them on my 8753D. Both look reasonable to the 6GHz limit of my 8753D. Each shows a 900MHz through loss of about 0.07dB, and a reflection loss of about 0.3dB. A random adapter from my lab looks slightly better, but not much.
--John Gord

On Wed, 9 Oct 2019 at 02:04, Kurt Poulsen <kurt@...> wrote:

Hi David

Hi Kurt

Exactly, the supplied open is pretty poor and has high loss. See attached

measurement with the NanoVNA aand NanoVNA saver calibrated with my HP85033C
.

Yours looks worse than mine, and I thought mine was bad enough!

That is why I have not yet published my characterization as I am searching

for confirmation that mine is not a special poor one, and you just
confirmed.

Yeah, it seems there are some dodgy ones around.

I would not even bother to attempt to characterise one with the intention
to share the results, as yours, Rune’s and mine all look physically like
same, but are electrically very different.

It would be interesting to know why they are so bad. I wonder if the
dielectric is paper that’s painted white to look like PTFE!

Kind regards

Kurt

Dave
--
Dr. David Kirkby,
Kirkby Microwave Ltd,
drkirkby@...

Telephone 01621-680100./ +44 1621 680100

Registered in England & Wales, company number 08914892.
Registered office:
Stokes Hall Lodge, Burnham Rd, Althorne, Chelmsford, Essex, CM3 6DT, United
Kingdom

On Wed, 9 Oct 2019 at 01:06, johncharlesgord via Groups.Io <johngord=
[email protected]> wrote:

Dave,
I have two of the longer SMA F-F adapters, from two similar NanaVNA kits.
I checked them on my 8753D. Both look reasonable to the 6GHz limit of my
8753D. Each shows a 900MHz through loss of about 0.07dB, and a reflection
loss of about 0.3dB. A random adapter from my lab looks slightly better,
but not much.
--John Gord

There is obviously quite a bit of variation among them. You and Rune find
them okay, I find it poor and Kurt even worse.

I expect George Orwell would have said, all adapters are equal, but some
are more equal than others.???

Dave.
--
Dr. David Kirkby,
Kirkby Microwave Ltd,
drkirkby@...

Telephone 01621-680100./ +44 1621 680100

Registered in England & Wales, company number 08914892.
Registered office:
Stokes Hall Lodge, Burnham Rd, Althorne, Chelmsford, Essex, CM3 6DT, United
Kingdom

On Wed, 9 Oct 2019 at 02:17, Kurt Poulsen <kurt@...> wrote:

Hi John
The one 0.3dB is not much less the what David and I measure so the one
from your lab is not optimal.

Yes, I didn’t actually look ar John’s figures - just accepted his comment
they were okay.

A Rosenberger female female was used as S21 thru during my test but the

open was from my HP 85033C Kit and the Rosenberger has some 0.05dB at
900MHz and a Amphenol RF has slightly less about 0.08dB as hard to
determine due to noise contribution
Kind regards
Kurt

Dave.

--

Dr. David Kirkby,
Kirkby Microwave Ltd,
drkirkby@...

Telephone 01621-680100./ +44 1621 680100

Registered in England & Wales, company number 08914892.
Registered office:
Stokes Hall Lodge, Burnham Rd, Althorne, Chelmsford, Essex, CM3 6DT, United
Kingdom

On Wed, 9 Oct 2019 at 01:06, johncharlesgord via Groups.Io <johngord=
[email protected]> wrote:

Dave,
I have two of the longer SMA F-F adapters, from two similar NanaVNA kits.
I checked them on my 8753D. Both look reasonable to the 6GHz limit of my
8753D. Each shows a 900MHz through loss of about 0.07dB, and a reflection
loss of about 0.3dB. A random adapter from my lab looks slightly better,
but not much.
--John Gord

I didn’t take too much notice of the reflection figure you gave - just
your quafigures you gave initially, kurt was more observant. A return loss
of 0.3 dB at 900 MHz is poor.

Dave.
--
Dr. David Kirkby,
Kirkby Microwave Ltd,
drkirkby@...

Telephone 01621-680100./ +44 1621 680100

Registered in England & Wales, company number 08914892.
Registered office:
Stokes Hall Lodge, Burnham Rd, Althorne, Chelmsford, Essex, CM3 6DT, United
Kingdom

Kurt,
You are right. I rechecked my random lab F-F and get more like 0.03dB reflection loss at 900MHz. (Operator malfunction.)
--John Gord

I think this is not adapter is bad. The cable that come with NanoVNA is very bad :)

Just tested with NanoVNA, my adaptor works ok up to 792 MHz with S21=0.00 dB. At 900 MHz it has S21=0.21 dB. I'm not sure if this is some measurement glitch, but with 10 cm RG405 it has S21=0.00 dB at 900 MHz.

Also, I have another Chinese sma-sma adapter, it has almost the same behavior, but it has S21=0.12 dB at 900 MHz.

What have you calibrated it against? You need a known good reference to
see if there's any loss.

I know that both Kurt and David have good references to compare against,
and they are seeing clear problems with their particular samples.

I have a decent reference connector, and I am not seeing significant
problems with my particular sample.

--
Rune

There is some asymmetry in some barrel connectors causing problems if the male pins are also a bit different. If I reverse the barrel connector there is no connection with the LOAD, the other way around it works....

On Wed, Oct 9, 2019 at 03:01 PM, Rune Broberg wrote:

What have you calibrated it against?

I calibrated NanoVNA through 10 cm RG405 cable with soldered SMA connectors. I measured S21 of sma-female to sma-female adaptor with another adaptor sma-male to sma-male. 100-200 ps delay doesn't matter here, because I measure S21.

Hi qrp.ddc
Wrong, the cables are part of the NanoVNA hardware and cancelled out during calibration.
The may cause phase changes when bended during practical measurements
Kind regards
Kurt

-----Oprindelig meddelelse-----
Fra: [email protected] <[email protected]> P? vegne af qrp.ddc@...
Sendt: 9. oktober 2019 13:40
Til: [email protected]
Emne: Re: [nanovna-users] Very poor thru (barrel) supplied with NanoVNAs

I think this is not adapter is bad. The cable that come with NanoVNA is very bad :)

Dave,
I cut a high loss barrel apart. The construction does not look terrible for the frequency range involved. (It has a bulge in the center of the inner contact to help lock it in place in the insulator, but it is under 1 mm long.) The dielectric does not feel like PTFE or PE. It burns fairly easily with a possibly chlorine like smell, so I suspect it may be PVC, which would explain the high loss.
--John Gord

On Thu, Oct 10, 2019 at 04:46 AM, johncharlesgord wrote:

I cut a high loss barrel apart. The construction does not look terrible for
the frequency range involved.

Could you please share the photo what is inside?

On Thu, 10 Oct 2019 at 01:17, Kurt Poulsen <kurt@...> wrote:

Hi qrp.ddc
Wrong, the cables are part of the NanoVNA hardware and cancelled out
during calibration.
The may cause phase changes when bended during practical measurements
Kind regards
Kurt

I have not extensively tested the cables supplied with my NanoVNAs, but
giving *one* cable 20 seconds to relax after bending, shows the reflected
signal not changing any more than -35 dB up to 1.5 GHz. That is reasonable.
Maybe if one tugged on the cable a bit, which could easily happy in
practical measurements, things would get poorer. This really was based on a
test taking one minute, so hardly conclusive.

In contrast, a test taking one minute showed me the adapter is very poor.


I can’t imagine myself wanting to use SMA connectors much with the NanoVNA,
as I want it for antennas measurements outside - I have better instruments
available for lab use. But if I did use the NanoVNA in a lab
environment, *based
on the smallest possible sample of tests*, I would not be overly concerned
about the cables.

Anyone can easily test the cables using the NanoVNA. You don’t need
expensive lab equipment to test them.

Dave
--
Dr. David Kirkby,
Kirkby Microwave Ltd,
drkirkby@...

Telephone 01621-680100./ +44 1621 680100

Registered in England & Wales, company number 08914892.
Registered office:
Stokes Hall Lodge, Burnham Rd, Althorne, Chelmsford, Essex, CM3 6DT, United
Kingdom

On Wed, 9 Oct 2019 at 13:04, <erik@...> wrote:

There is some asymmetry in some barrel connectors causing problems if the
male pins are also a bit different. If I reverse the barrel connector there
is no connection with the LOAD, the other way around it works....

Which basically results in the same conclusion that the adapter is only fit
for bin. ??????



--
Dr. David Kirkby,
Kirkby Microwave Ltd,
drkirkby@...

Telephone 01621-680100./ +44 1621 680100

Registered in England & Wales, company number 08914892.
Registered office:
Stokes Hall Lodge, Burnham Rd, Althorne, Chelmsford, Essex, CM3 6DT, United
Kingdom