Topics

BPF too good? #filter

Bob Benedict, KD8CGH
 

I built a 20 m BPF. Since I have many to make, I made a holder with SMA connectors and decided to use my recently acquired  NanoVNA for the first time with a filter. Filter analysis shows a center at 14.09 MHz and a 3 db BW of 1.39 MHz, which seems fine. Neither variable cap is at its extreme.
My concern is that the loss looks too low. I'm seeing 0.4 dB S21 loss. The VSWR is also 2. After calibration check with straight through wires scan reads VSWR 1.03, return loss -35 dB and S21 gain -.077 dB.




--
  73
    KD8CGH

Dave
 

Looks to me as if the filter is perfect.  Well done and may all your filters work exactly as designed!

Dave


On Jul 7, 2020, at 09:02, Bob Benedict, KD8CGH <rkayakr@...> wrote:

I built a 20 m BPF. Since I have many to make, I made a holder with SMA connectors and decided to use my recently acquired  NanoVNA for the first time with a filter. Filter analysis shows a center at 14.09 MHz and a 3 db BW of 1.39 MHz, which seems fine. Neither variable cap is at its extreme.
My concern is that the loss looks too low. I'm seeing 0.4 dB S21 loss. The VSWR is also 2. After calibration check with straight through wires scan reads VSWR 1.03, return loss -35 dB and S21 gain -.077 dB.

<20mBP.jpg>
<holder.jpg>



--
  73
    KD8CGH

Bob Benedict, KD8CGH
 

Thanks for the vote of confidence. What gives me pause is that the docs list the insertion loss as 1.75 dB. While it is not impossible that I built one of Hans's filters with a lower loss than he did, I think the probability is about like that of me outplaying LeBron James in basketball (or like outplaying Beckham in football for those of you in the rest of the world).
--
  73
    KD8CGH

Dave
 

I would have to guess who LeBron James is and know well who David Beckham is. And Victoria too. :-)

Dave


On Jul 7, 2020, at 11:51, Bob Benedict, KD8CGH <rkayakr@...> wrote:

Thanks for the vote of confidence. What gives me pause is that the docs list the insertion loss as 1.75 dB. While it is not impossible that I built one of Hans's filters with a lower loss than he did, I think the probability is about like that of me outplaying LeBron James in basketball (or like outplaying Beckham in football for those of you in the rest of the world).
--
  73
    KD8CGH

ajparent1/KB1GMX
 

I'd be suspect of the cal.   If the cables and jig called at 20M the loss should
be zero and the SWR  also perfect (cal absorbs errors).

The S11 should be a good match inside the pass band, its a symmetrical
filter.

For S21 its too good.  I can imagine and model a better filter loss but
only with perfect caps and coils.  See ELSIE.

1.25Db is best I could ever do for double tuned and that was with
3/8th" ceramic slug tuned inductor ( with enough turns to keep
core barely inserted) with silver mica caps critically coupled.

That was verified with both VNA (agilent 8573) NanoVNA, and
S11/S21 setup on Rigol DSA815T.  I tend to do that want I get
numbers close to the book.  I have found the NanoVNA I have
to be a bit more optimistic at HF by variable amounts but under
.1db (good enough).

Allison
-------------------------------
Please reply on list so we can share.
No private email, it goes to a bit bucket due address harvesting

Bob Benedict, KD8CGH
 

I suspected the cal also. When I repeated a sweep with the wires I used for "through" I got what I listed above - almost perfect cal for 10 MHz to 30 MHz:
VSWR 1.03, return loss -35 dB and S21 gain -.077 dB.
As stated in OP - this is my first try at measuring a filter with the NanoVNA, I would not be surprised to find a measurement error obvious to others well versed.
--
  73
    KD8CGH

Jim Painter
 

It looks like your test jig has two large metal brackets which are very close to the coils. Could they be screwing with the inductance? 


Jim...kq3s



On Tue, Jul 7, 2020 at 9:02 AM, Bob Benedict, KD8CGH
<rkayakr@...> wrote:
I built a 20 m BPF. Since I have many to make, I made a holder with SMA connectors and decided to use my recently acquired  NanoVNA for the first time with a filter. Filter analysis shows a center at 14.09 MHz and a 3 db BW of 1.39 MHz, which seems fine. Neither variable cap is at its extreme.
My concern is that the loss looks too low. I'm seeing 0.4 dB S21 loss. The VSWR is also 2. After calibration check with straight through wires scan reads VSWR 1.03, return loss -35 dB and S21 gain -.077 dB.




--
  73
    KD8CGH

Dave
 

Toroids are self shielding, as almost all the magnetic flux is enclosed in the toroid.  There may be some capacitive linkage to the nearby metal, but it would be very minimal.

Dave


On Tue, Jul 7, 2020 at 4:48 PM Jim Painter via groups.io <Jfpainter=yahoo.com@groups.io> wrote:
It looks like your test jig has two large metal brackets which are very close to the coils. Could they be screwing with the inductance? 


Jim...kq3s



On Tue, Jul 7, 2020 at 9:02 AM, Bob Benedict, KD8CGH
<rkayakr@...> wrote:
I built a 20 m BPF. Since I have many to make, I made a holder with SMA connectors and decided to use my recently acquired  NanoVNA for the first time with a filter. Filter analysis shows a center at 14.09 MHz and a 3 db BW of 1.39 MHz, which seems fine. Neither variable cap is at its extreme.
My concern is that the loss looks too low. I'm seeing 0.4 dB S21 loss. The VSWR is also 2. After calibration check with straight through wires scan reads VSWR 1.03, return loss -35 dB and S21 gain -.077 dB.




--
  73
    KD8CGH

KEN G4APB
 

Your results will depend on what other losses you have taken out in the CAL. The 4 pole non-50 ohm connections to the filter should NOT be calibrated out if you want a true 'in circuit' reading as they appear in the actual U3S. I found if you measure filters in-situ, especially if you start using the switched relay board with multiple filters (BPFs), the filter responses soon become misshapen and losses increase rapidly. I measure some 5-6dB in circuit losses in my mutiband U3S Rx.
Having said that, your tune and test up method looks good.

73 Ken G4APB

Bob Benedict, KD8CGH
 

Well reversion to the mean finally hit me. I tested a 30 m LP filter that also looked great. Then I tested my 20 m LP filter and it has a first shoulder at about 10 MHz and then a second shoulder at 15 MHz so the loss at 14 MHz is about 4.4 dB. I rechecked the parts placement, torroid winding and solder connections and all seem fine.  I made a model of the filter in SimSmith and played around with the values. The only way I got something like what I saw was to reduce L1 and L3 down to about 500nH.
I then pulled L1 and tested it in the impedance tester daughter board of an Analog Discovery 2. Doing a sweep after calibration it says the inductance at 14 MHz is 876 nH, too high! If I remove one turn the measured inductance goes down to 797nH. Unfortunately, according to the SimSmith model T1 being too high doesn't lead to the two shoulders, it only moves the shoulder a bit down in frequency.

Any suggestions?


--
  73
    KD8CGH

Ben Bangerter, K0IKR
 

Bob,

Coincidentally I constructed a QRP-Labs 20M BPF just last night.  When I saw your post today, I went to work characterizing my filter.  I don't have a VNA, but I do have a Siglent SA/TG here, and a home-made resistive return loss bridge.  I tuned the filter for minimum loss, and it peaked at 14.227 MHz with a loss of 2.54 dB.  The 3 dB bandwidth was 1.38 MHz, which seems reasonable (Hans aims for a BW of ~10% i think).  At 10 dB down the width was 2.72 MHz.  The loss is higher than I like, and may be due to the resonators being undercoupled.  I will mess with a gimmick capacitor to see if I can reduce the loss while keeping the filter shape nice.

Then I did return loss measurements.  My bridge is not very good, and shows only a 35 dB RL with an accurate 50 Ohm load.  I observed a 16.2 dB RL at 14.227 MHz, but the bottom of the RL curve was 14.32 MHz, with an RL of 17.2 dB.  When I touched up the tuning to put the minimum RL at the 14.227 MHz center frequency, the RL was 18.6 dB, corresponding to a VSWR of 1.22.  But the filter was degraded slightly.  The center frequency was unchanged, but the loss was 2.87 dB and the BW was 1.44 MHz.  The attached plots are from after the filter was tuned for maximum RL at 14.227 MHz.

73 Ben K0IKR


ajparent1/KB1GMX
 

The LPPF for 20M should be 773nh for the ends and 904 for the middle.

Start playing with that and you get bumps and lumps.

Allison
-------------------------------
Please reply on list so we can share.
No private email, it goes to a bit bucket due address harvesting

ajparent1/KB1GMX
 



I really hate the on line editor as it has enough lag to get old versions.
(groups IO and net issue).

Seems there is a lot of mixing of LPF and BPF, they are not the same thing
and expected result is different.

For LPF loss should not exceed  .5Db in the pass band is expected and correct.
Its curve for S21 (through loss) should be mostly flat from some low frequency
though the band to the corner frequency and then a slope down to -50db
or greater.  Lumps or steps indicate wrong values or grounding issue as
well as lumps and steps in the region that is sloping should not be there.

For band pass the band width and loss will vary from band to band but
under 2db (s21) is nominal for the filters.  One must use a wide enough
sweep to insure all resonators are in the band pass and there are no
external peaks that would result in excess loss.   Note he position of
the input and output links should be the same relative to each other
and nearest to the cold (grounded) end of the inductor for both best
S11 and S21.  Reminer if measuring S11 the far end of the filter does
have to be terminated in the design resistance (50 ohms for these).

Allison
-------------------------------
Please reply on list so we can share.
No private email, it goes to a bit bucket due address harvesting

 

 
-------------------------------
Please reply on list so we can share.
No private email, it goes to a bit bucket due address harvesting

Adrian Scripcă
 

Hi all,

I also get a good IL of ~1dB when properly tuning the 40meter BPF version. I used a HP70000 series SA+TG combo for characterising; please find the plot attached.

All the best,
Adrian, YO6SSW


On Wed, Jul 8, 2020 at 11:28 PM ajparent1/KB1GMX <kb1gmx@...> wrote:


I really hate the on line editor as it has enough lag to get old versions.
(groups IO and net issue).

Seems there is a lot of mixing of LPF and BPF, they are not the same thing
and expected result is different.

For LPF loss should not exceed  .5Db in the pass band is expected and correct.
Its curve for S21 (through loss) should be mostly flat from some low frequency
though the band to the corner frequency and then a slope down to -50db
or greater.  Lumps or steps indicate wrong values or grounding issue as
well as lumps and steps in the region that is sloping should not be there.

For band pass the band width and loss will vary from band to band but
under 2db (s21) is nominal for the filters.  One must use a wide enough
sweep to insure all resonators are in the band pass and there are no
external peaks that would result in excess loss.   Note he position of
the input and output links should be the same relative to each other
and nearest to the cold (grounded) end of the inductor for both best
S11 and S21.  Reminer if measuring S11 the far end of the filter does
have to be terminated in the design resistance (50 ohms for these).

Allison
-------------------------------
Please reply on list so we can share.
No private email, it goes to a bit bucket due address harvesting

 

 
-------------------------------
Please reply on list so we can share.
No private email, it goes to a bit bucket due address harvesting