Author Topic: High isolation LNA for the HF band direct conversion receiver  (Read 6670 times)

0 Members and 1 Guest are viewing this topic.

Offline szszjdbTopic starter

  • Regular Contributor
  • *
  • Posts: 208
  • Country: cn
High isolation LNA for the HF band direct conversion receiver
« on: September 23, 2021, 05:53:00 am »
Dear All,

I am going to add a high isolation LNA before the common quadrature detector to protect the local signal leaking to the ANT port. But it seems few of the existing LNA design sample meet the requirement. The most famous elecraft KX3 use the dedicated isolation operational amplifier for that purpose which look like more expensive.  And the MCHF project just use the common emitter LNA ,which have poor reverse isolation. 
So I search around and find the cascode amplifier ,mostly in the modern 3-5Ghz LNA design might have much better reverse isolation performance than the above.  But few of the existing circuit or experience about the cascode amplifier in HF band application. Why? Any risk?

Does anyone could give more guide on the cascode amplifier in HF band application or the other information about the isolation LNA for the direct  conversion receiver?

Best Regards & 73!
James
« Last Edit: September 23, 2021, 06:25:44 am by szszjdb »
 

Offline UR5FFR

  • Contributor
  • Posts: 23
  • Country: ua
    • UR5FFR ham radio
Re: High isolation LNA for the HF band direct conversion receiver
« Reply #1 on: September 23, 2021, 10:31:16 am »
Hi

Try simple common gate amplifier with J310. It have good reverse isolation.
If you need high gain you can use any amplifier before it. By example ERA MMIC.
73 Andrey
 
The following users thanked this post: szszjdb

Offline A.Z.

  • Frequent Contributor
  • **
  • Posts: 838
  • Country: it
Re: High isolation LNA for the HF band direct conversion receiver
« Reply #2 on: September 23, 2021, 01:41:03 pm »
Hi there, I agree with UR5FFR about using a common gate JFET stage, it's simple and offers good input-output isolation, another idea may be starting with the tuned preamplifier circuit presented in the February 2018 issue of QEX and found here

http://www.arrl.org/files/file/QEX_Next_Issue/Jan-Feb2018/Steber.pdf



the preamp is pretty simple/straightforward and offers good performances, willing to improve it, an idea may be adding a common gate JFET stage between the antenna and the tuned circuit input, such a stage should offer pretty LOW gain, it will only be used as a buffer to both isolate the antenna and offer a low input impedance suitable for feeding the stage using a standard coax; being the stage wideband, it will be better keeping the gain low to avoid overload from strong "off band" signals, the following varactor tuned circuit may be left as-is, or if aiming for multiband operations, modified to include a band switch to allow the circuit to operate over the whole desired frequencies range, the preamplifier may be modified too, if desired, an idea in such a case may be replacing the whole preamp (starting from R3 in the schematic) with an AD605 IC

https://www.analog.com/media/en/technical-documentation/data-sheets/AD605.pdf

such an IC offers a 40MHz bandwidth, which is more than enough to cover the whole HF range, plus it offers gain control, as is the gain control ranges from -14dB to +34dB, but adding a -10dB attenuator, it will be possible to bring the preamp to a -24dB to +24dB range which should be more than enough to deal with strong or weak signals, plus it would then be easy to add an AGC control to the AD605, in such a case the operator may still have a "gain" control knob, but it would just act on the max gain, while the effective gain (up to the max set by the operator) would then be controlled by the AGC

That being said, I'd avoid using a wideband RF preamplifier, in my opinion if you're seeking gain, better adding it where it's needed, that is after a tuned circuit, so that the preamp won't be overloaded by offband (off the DESIRED band) signals and offer much better performances

My 2 cents

[edit]

Forgot, if going to build the Steber tuned preamp (either using the original preamp or another one), it will be advisable to put a voltage regulator and some bypass capacitors on the line going to R2/VR1 to ensure the voltage driving the varactor will be stable and that it won't be modulated by RF leakage

« Last Edit: September 23, 2021, 03:55:13 pm by A.Z. »
 
The following users thanked this post: szszjdb

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16510
  • Country: us
  • DavidH
Re: High isolation LNA for the HF band direct conversion receiver
« Reply #3 on: September 23, 2021, 03:07:03 pm »
But few of the existing circuit or experience about the cascode amplifier in HF band application. Why? Any risk?

HF cascode amplifiers were more common in the past when transistor performance was worse and the most performance had to be extracted from each stage.  Cascode amplifiers are still found where ultra-broadband performance is required like cable systems and baseband applications like oscilloscopes.  Integrated circuit amplifiers of all kinds use them including many operational amplifiers.

Quote
Does anyone could give more guide on the cascode amplifier in HF band application or the other information about the isolation LNA for the direct  conversion receiver?

A common base/gate stage provides the same benefit of high isolation.  Common base/gate HF RF amplifiers are common.
 
The following users thanked this post: szszjdb

Offline szszjdbTopic starter

  • Regular Contributor
  • *
  • Posts: 208
  • Country: cn
Re: High isolation LNA for the HF band direct conversion receiver
« Reply #4 on: September 23, 2021, 05:37:56 pm »
Dear All,
ManY Thanks for your kindly help!

I have tried the common gate amplifier with J310 and found good isolation but gain dropping for about 6db when working on 6m band and above. The J310 were loaded with 200 Ohm and gain 6db on and below the 10m band. As for my current circuit,there have a bandpass filter in the front end ,following with the switchable 20db common emitter ampliifer, then the J310 and the quadrature detector. The sensitivity were poor above the 10m band due to the J310. So I am seeking for the more wider bandwidth circuit.

I will try the core of the Steber tuned preamp. The ADI/TI chip is so expensive now and I also heard it is a little noisy than the other type. I will try it later.

I found the BGA416 is the cascode type LNA and have 2 BJT inside it. So I built the similar one like the circuit below, but found poor gain when load with 50 Ohm in the output side.  The transformer in the circuit were replaced with just a inductor 10uH and the bias were 0.6mA. It seems the output impedence of the common base is far large than the 50 Ohm load. Still in tuning.

Any further comment are welcome ! Many Thanks!

Best Regards,
James
« Last Edit: September 23, 2021, 05:45:07 pm by szszjdb »
 

Offline A.Z.

  • Frequent Contributor
  • **
  • Posts: 838
  • Country: it
Re: High isolation LNA for the HF band direct conversion receiver
« Reply #5 on: September 23, 2021, 05:45:09 pm »
Hi James, the common gate (or common base) amplifier isn't generally used for gain, but to offer good I/O isolation, I will repeat it, if you are seeking for gain AND I/O isolation, a wideband, untuned common gate stage, followed by a tuned stage, which in turn feeds a good gain preamp is the way to go imVVHo, and I believe that the simple, yet effective preamp designed by Steber may be a good start
 
The following users thanked this post: szszjdb

Offline szszjdbTopic starter

  • Regular Contributor
  • *
  • Posts: 208
  • Country: cn
Re: High isolation LNA for the HF band direct conversion receiver
« Reply #6 on: September 23, 2021, 06:13:00 pm »
Hi,A.Z.,
Many Thanks!

Agree with you that the J310 is just for isolation. But it will introduce 6db or more gain drop on the 6m band and above, which make it hard to compensate for it.
Does you mean to change the cascade order to improve the gain loss in the higher band ? Move the J310 to right behind the frontend BPF instead of before the quadrature detector stage?

Best Regards,
James

 
 

Offline A.Z.

  • Frequent Contributor
  • **
  • Posts: 838
  • Country: it
Re: High isolation LNA for the HF band direct conversion receiver
« Reply #7 on: September 23, 2021, 06:21:32 pm »
Hi,A.Z.,
Many Thanks!

Agree with you that the J310 is just for isolation. But it will introduce 6db or more gain drop on the 6m band and above, which make it hard to compensate for it.
Does you mean to change the cascade order to improve the gain loss in the higher band ? Move the J310 to right behind the frontend BPF instead of before the quadrature detector stage?

Best Regards,
James

Hi there again, James, are you sure that the gain decrease (loss if you want) isn't due to the common gate/base design you are using ? A schematic would greatly help.
 
The following users thanked this post: szszjdb

Offline szszjdbTopic starter

  • Regular Contributor
  • *
  • Posts: 208
  • Country: cn
Re: High isolation LNA for the HF band direct conversion receiver
« Reply #8 on: September 23, 2021, 07:14:24 pm »
Hi,A.Z.,
Many Thanks!
The circuit is as attached. And I could find the voltage output of the J310 decreased to the half of the voltage on 10m band ,when switch to 6m band and above. The MDS test also prove that gain drop.  The transformer were proven to be 0.3db loss below the 150Mhz range. It seems that the gain loss might introduce by the mismatch of the impedence between J310 and the following quadrature detector FST3253 , but have no idea to compensate for it.

Best Regards,
James
 

Offline UR5FFR

  • Contributor
  • Posts: 23
  • Country: ua
    • UR5FFR ham radio
Re: High isolation LNA for the HF band direct conversion receiver
« Reply #9 on: September 23, 2021, 07:47:07 pm »
Hi

Common gate cascade have gain depended from |Z| load. Are you sure that your quadrature detector on FST3253 have same input impedance in range 2..50MHz? I think not have.
Try to insert simple emitter follower between J310 and detector. And place resistor with 100-200ohm parallel to inductor L51 in j310 drain.
73 Andrey
 
The following users thanked this post: szszjdb

Online KE5FX

  • Super Contributor
  • ***
  • Posts: 1872
  • Country: us
    • KE5FX.COM
Re: High isolation LNA for the HF band direct conversion receiver
« Reply #10 on: September 23, 2021, 10:41:55 pm »
A J310 should still be capable of providing good performance at 6 meters.   Try supplying the drain voltage through the transformer primary itself, rather than the 47 uH choke.  It  may have enough parasitic capacitance to hose you at 50 MHz.
 
The following users thanked this post: szszjdb

Offline szszjdbTopic starter

  • Regular Contributor
  • *
  • Posts: 208
  • Country: cn
Re: High isolation LNA for the HF band direct conversion receiver
« Reply #11 on: September 24, 2021, 07:34:52 am »
Hi,UR5FFR & KE5FX,
Many Thanks!
It's great! I will test and feedback to you.

Update:
1. Remove the choke and connect the J310 dirrectly to the transformer, a little improvement in the j310 drain voltage.
2. Add 470ohm parallel to inductor L51 in j310 drain, disconnect the transformer ,no adding the follower, still found 6db gain drop on the drain voltage comparing to 20m band and 3db to 10m band.
It seems common gate J310 have no voltage gain on 6m ,3db on 10m and 6db on 20m band.  That is really a headache.

Any further advice are welcome.

Best Regards & 73!
James
« Last Edit: September 24, 2021, 06:16:08 pm by szszjdb »
 

Offline UR5FFR

  • Contributor
  • Posts: 23
  • Country: ua
    • UR5FFR ham radio
Re: High isolation LNA for the HF band direct conversion receiver
« Reply #12 on: September 24, 2021, 08:20:01 pm »
Hi

J310 very well working on high frequency. If you open datasheet you can find that NF specified for 450MHz.
My suggestion still unchanged - insert emitter follower between J310 and detector
73 Andrey
 
The following users thanked this post: szszjdb

Offline A.Z.

  • Frequent Contributor
  • **
  • Posts: 838
  • Country: it
Re: High isolation LNA for the HF band direct conversion receiver
« Reply #13 on: September 26, 2021, 09:39:07 am »
my guess is that the input and output circuitry is cutting gain when going up in frequency, I'd try modifying the values of c207/c218, l52, r83 and eventually c202, notice that the input impedance of the j310 in common gate config is around 70 Ohm (if memory serves me correctly)
 
The following users thanked this post: szszjdb

Offline szszjdbTopic starter

  • Regular Contributor
  • *
  • Posts: 208
  • Country: cn
Re: High isolation LNA for the HF band direct conversion receiver
« Reply #14 on: September 26, 2021, 06:06:36 pm »
Hi,UR5FFR & A.Z. ,
Many Thanks!

I found out the gain drop is due to the ceramic inductor L51/L52. Replace them with the SMD Wire Wound Chip one, the gain recovered.
The isolation is around -30db, reasonable for such simple circuit.

Best Regards & 73!
James
 

Offline A.Z.

  • Frequent Contributor
  • **
  • Posts: 838
  • Country: it
Re: High isolation LNA for the HF band direct conversion receiver
« Reply #15 on: September 26, 2021, 06:46:51 pm »
Happy to know you found a solution, James; now I'm curious, are you building a receiver ?
 
The following users thanked this post: szszjdb

Offline A.Z.

  • Frequent Contributor
  • **
  • Posts: 838
  • Country: it
Re: High isolation LNA for the HF band direct conversion receiver
« Reply #16 on: September 26, 2021, 07:27:39 pm »
forgot, if you're still seeking for ideas, have a look at this

https://www.qsl.net/g3oou/highdynamicranferfamplifiers.html

;)
 
The following users thanked this post: szszjdb

Offline szszjdbTopic starter

  • Regular Contributor
  • *
  • Posts: 208
  • Country: cn
Re: High isolation LNA for the HF band direct conversion receiver
« Reply #17 on: September 27, 2021, 01:07:11 pm »
Hi, A.Z. ,
Many Thanks for your kindly help!

Yes I am designing in my own transciever in the ZIF archetechture. In the RX side, it is include the BPF , LNA , isolation , QSD and the audio ADC. I am focusing on the frontend these days. Now is the LNA. I am trying the common emmiter amplifier as attached, but not satisfied with the high power comsumption. Actually , the following quadrature detector and the ADC will saturate arround -5 dbm. I am wondering if I need so high IP3 amplifier or just like the KX3 to decrease the Vcc to 3-5V to minimise the current drain. Or just using some up to date integrated LNA , like BGA416 , only 5 mA. 
Would like to have your advice!

Best Regards & 73!
James
 

Offline A.Z.

  • Frequent Contributor
  • **
  • Posts: 838
  • Country: it
Re: High isolation LNA for the HF band direct conversion receiver
« Reply #18 on: September 27, 2021, 01:41:59 pm »
Hi James, thanks for satisfying my curiosity !

As for the preamp, that W7IUV is a "classic" (check this) but it has been designed for optimal performances in the 160 to 40 meters bands, given that you're aiming to go up to the 6m band, I wonder how well it may work "up there", also consider that if the plan is putting a wideband, untuned preamp at the receiver antenna input, then I believe you'd better revise your approach; in my very humble opinion, it would be a better idea placing a low gain stage there, and a stage capable of both preamplifying the signal a bit (say +6dB or the like) or attenuating it (say -10dB or more), following such a stage with the preselector (bandpass) for the given band and after the preselector adding another preamp with the desired max gain, such a second preamplifier may also (optionally) include a tuned circuit (like the "Steber" one - see my previous post) with the ability to bypass the tuned stage; that stage, if inserted, may allow to further narrow the input preselector and improve the input signal; that way you'll avoid issues from IMD/Overload caused by strong offband signals since the gain will only come AFTER the preselector and the input stage will anyways allow to both recover the loss from the bandpass filter and attenuate strong signals (if needed) to avoid overloading the input stage (and the preamp sitting after the preselector); anyhow, for other preamp ideas, have a look here, another (very simple, don't expect stellar performance) input circuit may be the one in the attached image

[edit]

further preamp ideas may also be found here

My 2 cents.
« Last Edit: September 27, 2021, 02:28:29 pm by A.Z. »
 
The following users thanked this post: szszjdb

Offline UR5FFR

  • Contributor
  • Posts: 23
  • Country: ua
    • UR5FFR ham radio
Re: High isolation LNA for the HF band direct conversion receiver
« Reply #19 on: September 27, 2021, 07:59:14 pm »
Hi James

Amplifier with collector-base negative feedback have poor isolation. It input impedance depended from load (see EMRFD and other). Any MMIC have very similar topology and need 50R load. Good choice for isolation amplifier is common gate or common emitter with emitter-base negative feedback.
You can find more different amplifier design on my site here
73 Andrey
 
The following users thanked this post: szszjdb

Online KE5FX

  • Super Contributor
  • ***
  • Posts: 1872
  • Country: us
    • KE5FX.COM
Re: High isolation LNA for the HF band direct conversion receiver
« Reply #20 on: September 27, 2021, 10:03:50 pm »
Hi James

Amplifier with collector-base negative feedback have poor isolation. It input impedance depended from load (see EMRFD and other). Any MMIC have very similar topology and need 50R load. Good choice for isolation amplifier is common gate or common emitter with emitter-base negative feedback.
You can find more different amplifier design on my site here

Very nice site, Andrey -- I like the way you use the message board to organize your personal notes. 

(Edit: I added a link to your page, via Google Translate, at the bottom of my "Sort-of-but-not-really-a-Norton amplifier" page here.)
« Last Edit: September 27, 2021, 10:06:37 pm by KE5FX »
 
The following users thanked this post: szszjdb, UR5FFR

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16510
  • Country: us
  • DavidH
Re: High isolation LNA for the HF band direct conversion receiver
« Reply #21 on: September 27, 2021, 11:50:19 pm »
Usually I think of two transistor cascode amplifier for high isolation, and they are used for exactly this in precision timing applications, but while I was searching, I found no examples of *reverse* cascode RF amplifiers.  Is this because they rely on complementary transistors?

Update: I should have written folded cascode instead of reverse cascode.
« Last Edit: September 28, 2021, 10:10:55 pm by David Hess »
 
The following users thanked this post: szszjdb

Offline Gerhard_dk4xp

  • Frequent Contributor
  • **
  • Posts: 321
  • Country: de
Re: High isolation LNA for the HF band direct conversion receiver
« Reply #22 on: September 28, 2021, 08:08:02 am »
I have made an isolation amplifier as a replacement for a NIST circuit that used
RF PNPs which are a dieing-out species. OK, I still have a reel of BFG31 :-)

I'm still not happy with it; backward isolation is only 80 dB @ 100 MHz and
S22 is a mess. The input tantalums are to short the bias voltage noise through
the low impedance source. The tantalums in the bias chain make a difference
in isolation.

Reverse isolation is 130 dB @ 5 MHz.

On the UR5FFR web site I have seen the Rohde amplifier with emitter feedback.
For me, that was a power oscillator at least with BFQ19S and MACOM transformers.

Gerhard DK4XP
« Last Edit: September 28, 2021, 08:14:34 am by Gerhard_dk4xp »
 

Online KE5FX

  • Super Contributor
  • ***
  • Posts: 1872
  • Country: us
    • KE5FX.COM
Re: High isolation LNA for the HF band direct conversion receiver
« Reply #23 on: September 28, 2021, 08:38:00 am »
What does the reverse isolation do when you disconnect the power?  It should get worse.

What about when you remove the metal shield?  Obviously enough, it should get worse then, too.

If doing something that should make the isolation worse makes it better instead, that's always a good clue.  Sometimes trying to make the problem worse is the best troubleshooting strategy of all. :)

There's a good ground plane on the other side of that board, right?

What bad things happen if you install a cap bigger than 10000 pF at C10?
 

Offline Gerhard_dk4xp

  • Frequent Contributor
  • **
  • Posts: 321
  • Country: de
Re: High isolation LNA for the HF band direct conversion receiver
« Reply #24 on: September 28, 2021, 09:14:17 am »
Backward isolation is not really a problem at 130 dB?
I had to use 1 Hz resolution bandwidth on the VNA to see it.  :-)

Top & bottom of the board should be quite rf-tight. My home-etched
boards have usually an unbroken bottom ground. That makes
aligning the masks so much easier.

The weak point is probably the bias string. The 100u tantalums
are a great step forward. 680uF has no added value.

Increasing the C10 capacitor too much seems to introduce
bias instability.

That looks ugly. I should wash the flux off.
« Last Edit: September 28, 2021, 09:26:19 am by Gerhard_dk4xp »
 
The following users thanked this post: szszjdb


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf