Just what is a "Polyphase harmonic rejection mixer "


drew231955
 

The main break through with the HF+ is the Polyphase harmonic rejection mixer. As best I can work out, when converting to base-band, it is an effective filter for the desired signal and rejects even strong signals close by with virtually no filtering ahead of the mixer.

It uses multiple phases of the local oscillator to use phasing to reject its harmonics, but at the same time, and because it is to a 200 kHz base-band, it rejects everything else too.

The big advantage is not needing a large number of band pass filters like a direct sampling SDR; the IC-7300 has 15!

The best explanation I have found is a slide show; http://icd.ewi.utwente.nl/temp_files/158b39412cff88a4181bfec0f4449c24.pdf. It is also subject to patent; https://www.google.ch/patents/US20110298521?hl=de. One of the authors wrote the slide show.

Presumably the mixer is a CMOS device, but I have not found one. And the RF cover on the HF+ is too hard to remove!

Any thoughts on this very novel approach?

Regards Drew VK4ZXI


drew231955
 

The PDF is icd.ewi.utwente.nl/publications/get_file.php?pub_id=563

Drew


prog
 

On Sun, Jan 14, 2018 at 10:41 pm, drew231955 wrote:
The main break through with the HF+ is the Polyphase harmonic rejection mixer. As best I can work out, when converting to base-band, it is an effective filter for the desired signal and rejects even strong signals close by with virtually no filtering ahead of the mixer.

It uses multiple phases of the local oscillator to use phasing to reject its harmonics, but at the same time, and because it is to a 200 kHz base-band, it rejects everything else too.

The big advantage is not needing a large number of band pass filters like a direct sampling SDR; the IC-7300 has 15!

The best explanation I have found is a slide show; http://icd.ewi.utwente.nl/temp_files/158b39412cff88a4181bfec0f4449c24.pdf. It is also subject to patent; https://www.google.ch/patents/US20110298521?hl=de. One of the authors wrote the slide show.

Presumably the mixer is a CMOS device, but I have not found one. And the RF cover on the HF+ is too hard to remove!

Any thoughts on this very novel approach?

Regards Drew VK4ZXI
You can see it as a "super Tayloe mixer". The problem with the original Tayloe Mixer is the harmonic responses at multiples of the LO frequency. The fix is to mathematically suppress these responses by adding more phases. The LO will no longer look like a square wave, but rather like a quantized sine wave. Basically, the more phases you add, the more harmonics you cancel.
This method is combined with narrow band filtering at the mixer itself. There is a switched-capacitor N-Path filter built into the mixer that is tuned using the same LO phases, which provides additional selectivity.
When you see it, all the ingredients required to implement this architecture can be implemented using CMOS silicon, and have a very good "horizontal" and "vertical" scalability: Horizontal with more phases (hence, less harmonics); Vertical with better fab processes (better linearity and NF).

The icing on the cake: This same technology can also work for TX.

The future looks bright!



doug
 


On 01/16/2018 07:29 AM, prog wrote:
On Sun, Jan 14, 2018 at 10:41 pm, drew231955 wrote:
The main break through with the HF+ is the Polyphase harmonic rejection mixer. As best I can work out, when converting to base-band, it is an effective filter for the desired signal and rejects even strong signals close by with virtually no filtering ahead of the mixer.

It uses multiple phases of the local oscillator to use phasing to reject its harmonics, but at the same time, and because it is to a 200 kHz base-band, it rejects everything else too.

The big advantage is not needing a large number of band pass filters like a direct sampling SDR; the IC-7300 has 15!

The best explanation I have found is a slide show; http://icd.ewi.utwente.nl/temp_files/158b39412cff88a4181bfec0f4449c24.pdf. It is also subject to patent; https://www.google.ch/patents/US20110298521?hl=de. One of the authors wrote the slide show.
There is something wrong with this url. I think it's the second /  What should the character be?

--doug

Presumably the mixer is a CMOS device, but I have not found one. And the RF cover on the HF+ is too hard to remove!

Any thoughts on this very novel approach?

Regards Drew VK4ZXI
You can see it as a "super Tayloe mixer". The problem with the original Tayloe Mixer is the harmonic responses at multiples of the LO frequency. The fix is to mathematically suppress these responses by adding more phases. The LO will no longer look like a square wave, but rather like a quantized sine wave. Basically, the more phases you add, the more harmonics you cancel.
This method is combined with narrow band filtering at the mixer itself. There is a switched-capacitor N-Path filter built into the mixer that is tuned using the same LO phases, which provides additional selectivity.
When you see it, all the ingredients required to implement this architecture can be implemented using CMOS silicon, and have a very good "horizontal" and "vertical" scalability: Horizontal with more phases (hence, less harmonics); Vertical with better fab processes (better linearity and NF).

The icing on the cake: This same technology can also work for TX.

The future looks bright!




Alberto I2PHD
 

On 1/16/2018 6:00 PM, doug wrote:

The best explanation I have found is a slide show; http://icd.ewi.utwente.nl/temp_files/158b39412cff88a4181bfec0f4449c24.pdf. I
There is something wrong with this url. I think it's the second /  What should the character be?

The PDF of the slide show can be found here :

http://icd.ewi.utwente.nl/publications/get_file.php?pub_id=563

--
73 Alberto I2PHD
Credo Ut Intelligam



David Eckhardt
 

Simply put, the old phasing method of SSB generation is somewhat similar.  It was capable of nulling the carrier (the LO, so to speak) and nulling one of the sidebands (the image).  It required a quadrature LO source.  But the mixer technology referred to takes that to a whole different level.

If I'm in error, someone please correct.

Dave - WØLEV

On Tue, Jan 16, 2018 at 12:29 PM, prog <info@...> wrote:
On Sun, Jan 14, 2018 at 10:41 pm, drew231955 wrote:
The main break through with the HF+ is the Polyphase harmonic rejection mixer. As best I can work out, when converting to base-band, it is an effective filter for the desired signal and rejects even strong signals close by with virtually no filtering ahead of the mixer.

It uses multiple phases of the local oscillator to use phasing to reject its harmonics, but at the same time, and because it is to a 200 kHz base-band, it rejects everything else too.

The big advantage is not needing a large number of band pass filters like a direct sampling SDR; the IC-7300 has 15!

The best explanation I have found is a slide show; http://icd.ewi.utwente.nl/temp_files/158b39412cff88a4181bfec0f4449c24.pdf. It is also subject to patent; https://www.google.ch/patents/US20110298521?hl=de. One of the authors wrote the slide show.

Presumably the mixer is a CMOS device, but I have not found one. And the RF cover on the HF+ is too hard to remove!

Any thoughts on this very novel approach?

Regards Drew VK4ZXI
You can see it as a "super Tayloe mixer". The problem with the original Tayloe Mixer is the harmonic responses at multiples of the LO frequency. The fix is to mathematically suppress these responses by adding more phases. The LO will no longer look like a square wave, but rather like a quantized sine wave. Basically, the more phases you add, the more harmonics you cancel.
This method is combined with narrow band filtering at the mixer itself. There is a switched-capacitor N-Path filter built into the mixer that is tuned using the same LO phases, which provides additional selectivity.
When you see it, all the ingredients required to implement this architecture can be implemented using CMOS silicon, and have a very good "horizontal" and "vertical" scalability: Horizontal with more phases (hence, less harmonics); Vertical with better fab processes (better linearity and NF).

The icing on the cake: This same technology can also work for TX.

The future looks bright!





--
Dave - WØLEV
Just Let Darwin Work


jdow
 

Use this URL http://icd.ewi.utwente.nl/publications/get_file.php?pub_id=563

The site creates a temporary copy of the file for viewing. It goes away when you disconnect. So you have to use the real URL to get there.

{^_^}

On 2018-01-16 09:00, doug wrote:
On 01/16/2018 07:29 AM, prog wrote:
On Sun, Jan 14, 2018 at 10:41 pm, drew231955 wrote:

The main break through with the HF+ is the Polyphase harmonic rejection
mixer. As best I can work out, when converting to base-band, it is an
effective filter for the desired signal and rejects even strong signals
close by with virtually no filtering ahead of the mixer.

It uses multiple phases of the local oscillator to use phasing to reject
its harmonics, but at the same time, and because it is to a 200 kHz
base-band, it rejects everything else too.

The big advantage is not needing a large number of band pass filters like
a direct sampling SDR; the IC-7300 has 15!

The best explanation I have found is a slide show;
http://icd.ewi.utwente.nl/temp_files/158b39412cff88a4181bfec0f4449c24.pdf.
It is also subject to patent;
https://www.google.ch/patents/US20110298521?hl=de. One of the authors
wrote the slide show.
There is something wrong with this url. I think it's the second / What should the character be?
--doug


Presumably the mixer is a CMOS device, but I have not found one. And the
RF cover on the HF+ is too hard to remove!

Any thoughts on this very novel approach?

Regards Drew VK4ZXI

You can see it as a "super Tayloe mixer". The problem with the original Tayloe Mixer is the harmonic responses at multiples of the LO frequency. The fix is to mathematically suppress these responses by adding more phases. The LO will no longer look like a square wave, but rather like a quantized sine wave. Basically, the more phases you add, the more harmonics you cancel.
This method is combined with narrow band filtering at the mixer itself. There is a switched-capacitor N-Path filter built into the mixer that is tuned using the same LO phases, which provides additional selectivity.
When you see it, all the ingredients required to implement this architecture can be implemented using CMOS silicon, and have a very good "horizontal" and "vertical" scalability: Horizontal with more phases (hence, less harmonics); Vertical with better fab processes (better linearity and NF).

The icing on the cake: This same technology can also work for TX.

The future looks bright!


drew231955
 

Hi Dave

Yes, I thought the same thing about phasing SSB modulators too.

The puzzle for me to understand is how the mixer operates as a selective RF filter, not just for the harmonics. "Prog"'s reply makes some sense, its just very hard to visualize.

The polyphase harmonic rejection mixer can be used for TX to generate a very clean signal, as has been noted.

The Airspy HF+ mentions the cooperation of ST Microelectronics, but I couldn't find anything on their site as a device.

Regards Drew


David Eckhardt
 

I found the following diagram online which may help.  Just run through the operations and you will ultimately visualize how the critter works.  Far more than quadrature is required of the LO signal.  
Inline image 2

On Wed, Jan 17, 2018 at 8:32 PM, drew231955 <drew_wollin@...> wrote:
Hi Dave

Yes, I thought the same thing about phasing SSB modulators too.

The puzzle for me to understand is how the mixer operates as a selective RF filter, not just for the harmonics. "Prog"'s reply makes some sense, its just very hard to visualize.

The polyphase harmonic rejection mixer can be used for TX to generate a very clean signal, as has been noted.

The Airspy HF+ mentions the cooperation of ST Microelectronics, but I couldn't find anything on their site as a device.

Regards Drew




--
Dave - WØLEV
Just Let Darwin Work


Brian Machesney <nekvtster@...>
 

The ST part is the STA709. If you can find the datasheet, please let us know. The only doc available on the web site is an uber-brief description of electricals and physicals, no detailed functional description.

-- Brian K1LI

On Wed, Jan 17, 2018 at 3:32 PM, drew231955 <drew_wollin@...> wrote:
Hi Dave

Yes, I thought the same thing about phasing SSB modulators too.

The puzzle for me to understand is how the mixer operates as a selective RF filter, not just for the harmonics. "Prog"'s reply makes some sense, its just very hard to visualize.

The polyphase harmonic rejection mixer can be used for TX to generate a very clean signal, as has been noted.

The Airspy HF+ mentions the cooperation of ST Microelectronics, but I couldn't find anything on their site as a device.

Regards Drew



 

I think the full STA709 datasheet is currently only available under NDA.

 

-Ray WB6TPU

 

 

From: main@airspy.groups.io [mailto:main@airspy.groups.io] On Behalf Of Brian Machesney
Sent: Wednesday, January 17, 2018 3:31 PM
To: main@airspy.groups.io
Subject: Re: [airspy] Just what is a "Polyphase harmonic rejection mixer "

 

The ST part is the STA709. If you can find the datasheet, please let us know. The only doc available on the web site is an uber-brief description of electricals and physicals, no detailed functional description.

-- Brian K1LI

 

On Wed, Jan 17, 2018 at 3:32 PM, drew231955 <drew_wollin@...> wrote:

Hi Dave

Yes, I thought the same thing about phasing SSB modulators too.

The puzzle for me to understand is how the mixer operates as a selective RF filter, not just for the harmonics. "Prog"'s reply makes some sense, its just very hard to visualize.

The polyphase harmonic rejection mixer can be used for TX to generate a very clean signal, as has been noted.

The Airspy HF+ mentions the cooperation of ST Microelectronics, but I couldn't find anything on their site as a device.

Regards Drew

 


drew231955
 

A brief data sheet is available, but I don't think it is the right device. It has 4 mixers but not the polyphasing of the local oscillator.
Drew VK4ZXI 


 

Drew-

 

The diagram shown in the databrief is a very simplified, conceptual block diagram. Many details are not shown (on purpose). The STA709 is the device used in the HF+

 

-Ray WB6TPU

 

 

From: main@airspy.groups.io [mailto:main@airspy.groups.io] On Behalf Of drew231955
Sent: Thursday, January 18, 2018 11:37 AM
To: main@airspy.groups.io
Subject: Re: [airspy] Just what is a "Polyphase harmonic rejection mixer "

 

A brief data sheet is available, but I don't think it is the right device. It has 4 mixers but not the polyphasing of the local oscillator.
Drew VK4ZXI 


jdow
 

The block diagrams I've seen from Youssef et al show a 16 sampler block diagram. At least a couple are show labeled 1 and 2 and maybe 3. Then the last one is labeled 16 and has an elipsis above it indicating items left out that are the same as the ones above it.

{^_^}

On 2018-01-18 11:47, Ray WB6TPU wrote:
Drew-
The diagram shown in the databrief is a very simplified, conceptual block diagram. Many details are not shown (on purpose). The STA709 is the device used in the HF+
-Ray WB6TPU
*From:*main@airspy.groups.io [mailto:main@airspy.groups.io] *On Behalf Of *drew231955
*Sent:* Thursday, January 18, 2018 11:37 AM
*To:* main@airspy.groups.io
*Subject:* Re: [airspy] Just what is a "Polyphase harmonic rejection mixer "
A brief data sheet is available, but I don't think it is the right device. It has 4 mixers but not the polyphasing of the local oscillator.
Drew VK4ZXI