LMX2541


Robin Szemeti - G1YFG
 

Generally speaking, how good is the phase noise on these things for low noise, SSB stylee applications?

I was thinking of using a couple of these to get one of the Ionica units down to 145 MHz.  I'd just like a "yep, they are great providing the loop filter is done right" ranging to "avoid like plague, only suitable for consumer electronics"  sort of ballpark idea, so I know if I am wasting my time pasting them onto a PCB with some mixers.

--
Best regards,

Robin Szemeti

Redpoint Consulting Limited

E: robin@...
T: +44 (0) 1299 405028
M: +44 (0) 7971 883371

CONFIDENTIALITY NOTICE
The information contained in this e-mail is intended only for the
confidential use of the above named recipient. If you are not the
intended recipient or person responsible for delivering it to the
intended recipient, you have received this communication in error
and must not distribute or copy it.
Please accept the sender's apologies, notify the sender immediately
by return e-mail and delete this communication.

Thank you.


Andy G4JNT
 

Perfectly OK 
There are one or two UHF beacons using them, generating JTxx modulation by reprogrammong (Reg 2 !!!) in real time


G8BKE (or is it G8ACE) is even using one at 47GHz




On Tue, 7 Jul 2020 at 12:03, Robin Szemeti - G1YFG <robin@...> wrote:
Generally speaking, how good is the phase noise on these things for low noise, SSB stylee applications?

I was thinking of using a couple of these to get one of the Ionica units down to 145 MHz.  I'd just like a "yep, they are great providing the loop filter is done right" ranging to "avoid like plague, only suitable for consumer electronics"  sort of ballpark idea, so I know if I am wasting my time pasting them onto a PCB with some mixers.

--
Best regards,

Robin Szemeti

Redpoint Consulting Limited

E: robin@...
T: +44 (0) 1299 405028
M: +44 (0) 7971 883371

CONFIDENTIALITY NOTICE
The information contained in this e-mail is intended only for the
confidential use of the above named recipient. If you are not the
intended recipient or person responsible for delivering it to the
intended recipient, you have received this communication in error
and must not distribute or copy it.
Please accept the sender's apologies, notify the sender immediately
by return e-mail and delete this communication.

Thank you.


Robin Szemeti - G1YFG
 

Perfect, that's all I need to know :)

On Tue, 7 Jul 2020 at 12:23, Andy G4JNT <andy.g4jnt@...> wrote:
Perfectly OK 
There are one or two UHF beacons using them, generating JTxx modulation by reprogrammong (Reg 2 !!!) in real time


G8BKE (or is it G8ACE) is even using one at 47GHz




On Tue, 7 Jul 2020 at 12:03, Robin Szemeti - G1YFG <robin@...> wrote:
Generally speaking, how good is the phase noise on these things for low noise, SSB stylee applications?

I was thinking of using a couple of these to get one of the Ionica units down to 145 MHz.  I'd just like a "yep, they are great providing the loop filter is done right" ranging to "avoid like plague, only suitable for consumer electronics"  sort of ballpark idea, so I know if I am wasting my time pasting them onto a PCB with some mixers.

--
Best regards,

Robin Szemeti

Redpoint Consulting Limited

E: robin@...
T: +44 (0) 1299 405028
M: +44 (0) 7971 883371

CONFIDENTIALITY NOTICE
The information contained in this e-mail is intended only for the
confidential use of the above named recipient. If you are not the
intended recipient or person responsible for delivering it to the
intended recipient, you have received this communication in error
and must not distribute or copy it.
Please accept the sender's apologies, notify the sender immediately
by return e-mail and delete this communication.

Thank you.


Andy G4JNT
 

Need to add ...   providing the loop filter is done right  :-)
With a loop bandwidth of a few tens of kHz then the output will take on the (multiplied) phase noise of the reference.  Use a decent  reference, and you've got yourself a decent uWave  signal; at least out to +/- the loop filter bandwidth.



On Tue, 7 Jul 2020 at 13:05, Robin Szemeti - G1YFG <robin@...> wrote:
Perfect, that's all I need to know :)

On Tue, 7 Jul 2020 at 12:23, Andy G4JNT <andy.g4jnt@...> wrote:
Perfectly OK 
There are one or two UHF beacons using them, generating JTxx modulation by reprogrammong (Reg 2 !!!) in real time


G8BKE (or is it G8ACE) is even using one at 47GHz




On Tue, 7 Jul 2020 at 12:03, Robin Szemeti - G1YFG <robin@...> wrote:
Generally speaking, how good is the phase noise on these things for low noise, SSB stylee applications?

I was thinking of using a couple of these to get one of the Ionica units down to 145 MHz.  I'd just like a "yep, they are great providing the loop filter is done right" ranging to "avoid like plague, only suitable for consumer electronics"  sort of ballpark idea, so I know if I am wasting my time pasting them onto a PCB with some mixers.

--
Best regards,

Robin Szemeti

Redpoint Consulting Limited

E: robin@...
T: +44 (0) 1299 405028
M: +44 (0) 7971 883371

CONFIDENTIALITY NOTICE
The information contained in this e-mail is intended only for the
confidential use of the above named recipient. If you are not the
intended recipient or person responsible for delivering it to the
intended recipient, you have received this communication in error
and must not distribute or copy it.
Please accept the sender's apologies, notify the sender immediately
by return e-mail and delete this communication.

Thank you.


Robin Szemeti - G1YFG
 

I was planning on using one of those 0.28ppm crystals you noted the other day at 10MHz, or possibly the 10MHz reference out of the 70cm linear.


On Tue, 7 Jul 2020 at 21:56, Andy G4JNT <andy.g4jnt@...> wrote:
Need to add ...   providing the loop filter is done right  :-)
With a loop bandwidth of a few tens of kHz then the output will take on the (multiplied) phase noise of the reference.  Use a decent  reference, and you've got yourself a decent uWave  signal; at least out to +/- the loop filter bandwidth.



On Tue, 7 Jul 2020 at 13:05, Robin Szemeti - G1YFG <robin@...> wrote:
Perfect, that's all I need to know :)

On Tue, 7 Jul 2020 at 12:23, Andy G4JNT <andy.g4jnt@...> wrote:
Perfectly OK 
There are one or two UHF beacons using them, generating JTxx modulation by reprogrammong (Reg 2 !!!) in real time


G8BKE (or is it G8ACE) is even using one at 47GHz




On Tue, 7 Jul 2020 at 12:03, Robin Szemeti - G1YFG <robin@...> wrote:
Generally speaking, how good is the phase noise on these things for low noise, SSB stylee applications?

I was thinking of using a couple of these to get one of the Ionica units down to 145 MHz.  I'd just like a "yep, they are great providing the loop filter is done right" ranging to "avoid like plague, only suitable for consumer electronics"  sort of ballpark idea, so I know if I am wasting my time pasting them onto a PCB with some mixers.

--
Best regards,

Robin Szemeti

Redpoint Consulting Limited

E: robin@...
T: +44 (0) 1299 405028
M: +44 (0) 7971 883371

CONFIDENTIALITY NOTICE
The information contained in this e-mail is intended only for the
confidential use of the above named recipient. If you are not the
intended recipient or person responsible for delivering it to the
intended recipient, you have received this communication in error
and must not distribute or copy it.
Please accept the sender's apologies, notify the sender immediately
by return e-mail and delete this communication.

Thank you.


Robin Szemeti - G1YFG
 

Just reading your design notes, looks like the hard work is done.  Nice work. Even the PIC code ... lovely, and the loop filter .xls ... nice, thank you!

For a (roughly) 800MHz output,  from 10MHz,  is the 32KHz loop bandwidth about right?  Is there a good rule of thumb for choosing loop bandwidth?



On Wed, 8 Jul 2020 at 00:06, Robin Szemeti - G1YFG via groups.io <robin=redpoint.org.uk@groups.io> wrote:
I was planning on using one of those 0.28ppm crystals you noted the other day at 10MHz, or possibly the 10MHz reference out of the 70cm linear.


On Tue, 7 Jul 2020 at 21:56, Andy G4JNT <andy.g4jnt@...> wrote:
Need to add ...   providing the loop filter is done right  :-)
With a loop bandwidth of a few tens of kHz then the output will take on the (multiplied) phase noise of the reference.  Use a decent  reference, and you've got yourself a decent uWave  signal; at least out to +/- the loop filter bandwidth.



On Tue, 7 Jul 2020 at 13:05, Robin Szemeti - G1YFG <robin@...> wrote:
Perfect, that's all I need to know :)

On Tue, 7 Jul 2020 at 12:23, Andy G4JNT <andy.g4jnt@...> wrote:
Perfectly OK 
There are one or two UHF beacons using them, generating JTxx modulation by reprogrammong (Reg 2 !!!) in real time


G8BKE (or is it G8ACE) is even using one at 47GHz




On Tue, 7 Jul 2020 at 12:03, Robin Szemeti - G1YFG <robin@...> wrote:
Generally speaking, how good is the phase noise on these things for low noise, SSB stylee applications?

I was thinking of using a couple of these to get one of the Ionica units down to 145 MHz.  I'd just like a "yep, they are great providing the loop filter is done right" ranging to "avoid like plague, only suitable for consumer electronics"  sort of ballpark idea, so I know if I am wasting my time pasting them onto a PCB with some mixers.

--
Best regards,

Robin Szemeti

Redpoint Consulting Limited

E: robin@...
T: +44 (0) 1299 405028
M: +44 (0) 7971 883371

CONFIDENTIALITY NOTICE
The information contained in this e-mail is intended only for the
confidential use of the above named recipient. If you are not the
intended recipient or person responsible for delivering it to the
intended recipient, you have received this communication in error
and must not distribute or copy it.
Please accept the sender's apologies, notify the sender immediately
by return e-mail and delete this communication.

Thank you.


Andy G4JNT
 

HI Robin
Choosing loop bandwidth is complicated - it involves trying to make close in and far out phase noise overlap properly. I think teh optimum loop BW is at the crossover  I just choose a value of 20 - 40kHz or so when Fcomp is the highest (10MHz usually)  and just go with it.  Changing charge pump current to give what looks the best result.    Not very scientific, but can generally get a good enough looking spectrum.

Chris 'DGU knows all about optimising PLLs to the last degree and has posted here before giving more details.  But it was a long while back



On Wed, 8 Jul 2020 at 01:42, Robin Szemeti - G1YFG <robin@...> wrote:
Just reading your design notes, looks like the hard work is done.  Nice work. Even the PIC code ... lovely, and the loop filter .xls ... nice, thank you!

For a (roughly) 800MHz output,  from 10MHz,  is the 32KHz loop bandwidth about right?  Is there a good rule of thumb for choosing loop bandwidth?



On Wed, 8 Jul 2020 at 00:06, Robin Szemeti - G1YFG via groups.io <robin=redpoint.org.uk@groups.io> wrote:
I was planning on using one of those 0.28ppm crystals you noted the other day at 10MHz, or possibly the 10MHz reference out of the 70cm linear.


On Tue, 7 Jul 2020 at 21:56, Andy G4JNT <andy.g4jnt@...> wrote:
Need to add ...   providing the loop filter is done right  :-)
With a loop bandwidth of a few tens of kHz then the output will take on the (multiplied) phase noise of the reference.  Use a decent  reference, and you've got yourself a decent uWave  signal; at least out to +/- the loop filter bandwidth.



On Tue, 7 Jul 2020 at 13:05, Robin Szemeti - G1YFG <robin@...> wrote:
Perfect, that's all I need to know :)

On Tue, 7 Jul 2020 at 12:23, Andy G4JNT <andy.g4jnt@...> wrote:
Perfectly OK 
There are one or two UHF beacons using them, generating JTxx modulation by reprogrammong (Reg 2 !!!) in real time


G8BKE (or is it G8ACE) is even using one at 47GHz




On Tue, 7 Jul 2020 at 12:03, Robin Szemeti - G1YFG <robin@...> wrote:
Generally speaking, how good is the phase noise on these things for low noise, SSB stylee applications?

I was thinking of using a couple of these to get one of the Ionica units down to 145 MHz.  I'd just like a "yep, they are great providing the loop filter is done right" ranging to "avoid like plague, only suitable for consumer electronics"  sort of ballpark idea, so I know if I am wasting my time pasting them onto a PCB with some mixers.

--
Best regards,

Robin Szemeti

Redpoint Consulting Limited

E: robin@...
T: +44 (0) 1299 405028
M: +44 (0) 7971 883371

CONFIDENTIALITY NOTICE
The information contained in this e-mail is intended only for the
confidential use of the above named recipient. If you are not the
intended recipient or person responsible for delivering it to the
intended recipient, you have received this communication in error
and must not distribute or copy it.
Please accept the sender's apologies, notify the sender immediately
by return e-mail and delete this communication.

Thank you.


Paul Randall G3NJV
 

Robin,  Hi, not sure of your intentions but if you get  a pcb made for the lmx2541/ pic etc or purchase chips,  i would be very interested in several boards and chips and happy to share costs. This is such a useful item I imagine a lot of people "could"  be interested in a group buy, so perhaps worth asking?
I really enjoy following your posts,  thanks,  Paul, G3NJV



Sent from my Samsung Galaxy smartphone.



-------- Original message --------
From: Robin Szemeti - G1YFG <robin@...>
Date: 08/07/2020 00:06 (GMT+00:00)
To: UKMicrowaves@groups.io
Subject: Re: [UKMicrowaves] LMX2541

I was planning on using one of those 0.28ppm crystals you noted the other day at 10MHz, or possibly the 10MHz reference out of the 70cm linear.


On Tue, 7 Jul 2020 at 21:56, Andy G4JNT <andy.g4jnt@...> wrote:
Need to add ...   providing the loop filter is done right  :-)
With a loop bandwidth of a few tens of kHz then the output will take on the (multiplied) phase noise of the reference.  Use a decent  reference, and you've got yourself a decent uWave  signal; at least out to +/- the loop filter bandwidth.



On Tue, 7 Jul 2020 at 13:05, Robin Szemeti - G1YFG <robin@...> wrote:
Perfect, that's all I need to know :)

On Tue, 7 Jul 2020 at 12:23, Andy G4JNT <andy.g4jnt@...> wrote:
Perfectly OK 
There are one or two UHF beacons using them, generating JTxx modulation by reprogrammong (Reg 2 !!!) in real time


G8BKE (or is it G8ACE) is even using one at 47GHz




On Tue, 7 Jul 2020 at 12:03, Robin Szemeti - G1YFG <robin@...> wrote:
Generally speaking, how good is the phase noise on these things for low noise, SSB stylee applications?

I was thinking of using a couple of these to get one of the Ionica units down to 145 MHz.  I'd just like a "yep, they are great providing the loop filter is done right" ranging to "avoid like plague, only suitable for consumer electronics"  sort of ballpark idea, so I know if I am wasting my time pasting them onto a PCB with some mixers.

--
Best regards,

Robin Szemeti

Redpoint Consulting Limited

E: robin@...
T: +44 (0) 1299 405028
M: +44 (0) 7971 883371

CONFIDENTIALITY NOTICE
The information contained in this e-mail is intended only for the
confidential use of the above named recipient. If you are not the
intended recipient or person responsible for delivering it to the
intended recipient, you have received this communication in error
and must not distribute or copy it.
Please accept the sender's apologies, notify the sender immediately
by return e-mail and delete this communication.

Thank you.


Robin Szemeti - G1YFG
 

I'm doing a board for the Ionica units .. so  144 in/out  .. and then a 947->144 receive convertor, and a LO feed to drive the Ionica receive block to do the 3456 to 947 conversion.    On the transmit side it will do the same trick in reverse and produce 10mw max of 3456 to drive the Ionica PA. I've sourced a SAW filter for the 947 segment, I just need to figure out the best solution for a 3456 filter ...   the 144/947 will be done with active mixers, the MAX2680 / MAX2671, I'm thinking something from the Minicircuits lineup for the 947->3456 upconvertor.

Slapping one of the LMX2451 units onto a PCB would be simple enough, I think JNT has done this already though.  If there is interest, I could do that, but probaby best to let me iron the mistakes out on the 9cm board and then pick up the lump of LMX off that board and slap it onto a PCB. I have an SMT machine in the shed, assembling them would be simple enough.


On Wed, 8 Jul 2020 at 23:12, Paul Randall G3NJV <paulfrandall@...> wrote:
Robin,  Hi, not sure of your intentions but if you get  a pcb made for the lmx2541/ pic etc or purchase chips,  i would be very interested in several boards and chips and happy to share costs. This is such a useful item I imagine a lot of people "could"  be interested in a group buy, so perhaps worth asking?
I really enjoy following your posts,  thanks,  Paul, G3NJV



Sent from my Samsung Galaxy smartphone.



-------- Original message --------
From: Robin Szemeti - G1YFG <robin@...>
Date: 08/07/2020 00:06 (GMT+00:00)
Subject: Re: [UKMicrowaves] LMX2541

I was planning on using one of those 0.28ppm crystals you noted the other day at 10MHz, or possibly the 10MHz reference out of the 70cm linear.


On Tue, 7 Jul 2020 at 21:56, Andy G4JNT <andy.g4jnt@...> wrote:
Need to add ...   providing the loop filter is done right  :-)
With a loop bandwidth of a few tens of kHz then the output will take on the (multiplied) phase noise of the reference.  Use a decent  reference, and you've got yourself a decent uWave  signal; at least out to +/- the loop filter bandwidth.



On Tue, 7 Jul 2020 at 13:05, Robin Szemeti - G1YFG <robin@...> wrote:
Perfect, that's all I need to know :)

On Tue, 7 Jul 2020 at 12:23, Andy G4JNT <andy.g4jnt@...> wrote:
Perfectly OK 
There are one or two UHF beacons using them, generating JTxx modulation by reprogrammong (Reg 2 !!!) in real time


G8BKE (or is it G8ACE) is even using one at 47GHz




On Tue, 7 Jul 2020 at 12:03, Robin Szemeti - G1YFG <robin@...> wrote:
Generally speaking, how good is the phase noise on these things for low noise, SSB stylee applications?

I was thinking of using a couple of these to get one of the Ionica units down to 145 MHz.  I'd just like a "yep, they are great providing the loop filter is done right" ranging to "avoid like plague, only suitable for consumer electronics"  sort of ballpark idea, so I know if I am wasting my time pasting them onto a PCB with some mixers.

--
Best regards,

Robin Szemeti

Redpoint Consulting Limited

E: robin@...
T: +44 (0) 1299 405028
M: +44 (0) 7971 883371

CONFIDENTIALITY NOTICE
The information contained in this e-mail is intended only for the
confidential use of the above named recipient. If you are not the
intended recipient or person responsible for delivering it to the
intended recipient, you have received this communication in error
and must not distribute or copy it.
Please accept the sender's apologies, notify the sender immediately
by return e-mail and delete this communication.

Thank you.


Paul G8AQA
 

Hi Robin,

I made a pipecap filter for 3.6214 Ghz It was quite narrowband and had an approx 3dB loss.  You can play with loss and bandwidth by adjusting the probes.  I have a bag of 28mm pipecaps here if you want one. A bit old school but quick and easy.

73
Paul G8AQA

On 08/07/2020 23:43, Robin Szemeti - G1YFG wrote:
I'm doing a board for the Ionica units .. so  144 in/out  .. and then a 947->144 receive convertor, and a LO feed to drive the Ionica receive block to do the 3456 to 947 conversion.    On the transmit side it will do the same trick in reverse and produce 10mw max of 3456 to drive the Ionica PA. I've sourced a SAW filter for the 947 segment, I just need to figure out the best solution for a 3456 filter ...   the 144/947 will be done with active mixers, the MAX2680 / MAX2671, I'm thinking something from the Minicircuits lineup for the 947->3456 upconvertor.

Slapping one of the LMX2451 units onto a PCB would be simple enough, I think JNT has done this already though.  If there is interest, I could do that, but probaby best to let me iron the mistakes out on the 9cm board and then pick up the lump of LMX off that board and slap it onto a PCB. I have an SMT machine in the shed, assembling them would be simple enough.

On Wed, 8 Jul 2020 at 23:12, Paul Randall G3NJV <paulfrandall@...> wrote:
Robin,  Hi, not sure of your intentions but if you get  a pcb made for the lmx2541/ pic etc or purchase chips,  i would be very interested in several boards and chips and happy to share costs. This is such a useful item I imagine a lot of people "could"  be interested in a group buy, so perhaps worth asking?
I really enjoy following your posts,  thanks,  Paul, G3NJV



Sent from my Samsung Galaxy smartphone.



-------- Original message --------
From: Robin Szemeti - G1YFG <robin@...>
Date: 08/07/2020 00:06 (GMT+00:00)
Subject: Re: [UKMicrowaves] LMX2541

I was planning on using one of those 0.28ppm crystals you noted the other day at 10MHz, or possibly the 10MHz reference out of the 70cm linear.


On Tue, 7 Jul 2020 at 21:56, Andy G4JNT <andy.g4jnt@...> wrote:
Need to add ...   providing the loop filter is done right  :-)
With a loop bandwidth of a few tens of kHz then the output will take on the (multiplied) phase noise of the reference.  Use a decent  reference, and you've got yourself a decent uWave  signal; at least out to +/- the loop filter bandwidth.



On Tue, 7 Jul 2020 at 13:05, Robin Szemeti - G1YFG <robin@...> wrote:
Perfect, that's all I need to know :)

On Tue, 7 Jul 2020 at 12:23, Andy G4JNT <andy.g4jnt@...> wrote:
Perfectly OK 
There are one or two UHF beacons using them, generating JTxx modulation by reprogrammong (Reg 2 !!!) in real time


G8BKE (or is it G8ACE) is even using one at 47GHz




On Tue, 7 Jul 2020 at 12:03, Robin Szemeti - G1YFG <robin@...> wrote:
Generally speaking, how good is the phase noise on these things for low noise, SSB stylee applications?

I was thinking of using a couple of these to get one of the Ionica units down to 145 MHz.  I'd just like a "yep, they are great providing the loop filter is done right" ranging to "avoid like plague, only suitable for consumer electronics"  sort of ballpark idea, so I know if I am wasting my time pasting them onto a PCB with some mixers.

--
Best regards,

Robin Szemeti

Redpoint Consulting Limited 

E: robin@... 
T: +44 (0) 1299 405028 
M: +44 (0) 7971 883371 

CONFIDENTIALITY NOTICE 
The information contained in this e-mail is intended only for the
confidential use of the above named recipient. If you are not the
intended recipient or person responsible for delivering it to the
intended recipient, you have received this communication in error 
and must not distribute or copy it. 
Please accept the sender's apologies, notify the sender immediately 
by return e-mail and delete this communication. 

Thank you. 


Paul Randall G3NJV
 

Robin, my immediate need is for a receive converter for Eshail, taking the ~700 MHz lnb o/p down to 144, a common enough task to warrant a board I'd think. I've already sourced a Toko helical filter @ 700 and bought the MAX26xx mixers so was thinking pretty much on the same lines as you. A 432 to 2400 upconvertor would also be very useful.
I'm interested to follow your progress and if you do decide to chop out parts of it or just the LO to some boards .  I must admit I did a double take when viewing the LMX2541 pdf, quite challenging to mount by hand I'd say. Am I right in thinking it can take a 100MHz reference? I was already contemplating a 100MHz Crystal/PLL  as a route to get 25MHz from a 10 MHz gpsdo hopefully keeping pn low. 
Paul



Sent from my Samsung Galaxy smartphone.



-------- Original message --------
From: Robin Szemeti - G1YFG <robin@...>
Date: 08/07/2020 23:43 (GMT+00:00)
To: UKMicrowaves@groups.io
Subject: Re: [UKMicrowaves] LMX2541

I'm doing a board for the Ionica units .. so  144 in/out  .. and then a 947->144 receive convertor, and a LO feed to drive the Ionica receive block to do the 3456 to 947 conversion.    On the transmit side it will do the same trick in reverse and produce 10mw max of 3456 to drive the Ionica PA. I've sourced a SAW filter for the 947 segment, I just need to figure out the best solution for a 3456 filter ...   the 144/947 will be done with active mixers, the MAX2680 / MAX2671, I'm thinking something from the Minicircuits lineup for the 947->3456 upconvertor.

Slapping one of the LMX2451 units onto a PCB would be simple enough, I think JNT has done this already though.  If there is interest, I could do that, but probaby best to let me iron the mistakes out on the 9cm board and then pick up the lump of LMX off that board and slap it onto a PCB. I have an SMT machine in the shed, assembling them would be simple enough.

On Wed, 8 Jul 2020 at 23:12, Paul Randall G3NJV <paulfrandall@...> wrote:
Robin,  Hi, not sure of your intentions but if you get  a pcb made for the lmx2541/ pic etc or purchase chips,  i would be very interested in several boards and chips and happy to share costs. This is such a useful item I imagine a lot of people "could"  be interested in a group buy, so perhaps worth asking?
I really enjoy following your posts,  thanks,  Paul, G3NJV



Sent from my Samsung Galaxy smartphone.



-------- Original message --------
From: Robin Szemeti - G1YFG <robin@...>
Date: 08/07/2020 00:06 (GMT+00:00)
Subject: Re: [UKMicrowaves] LMX2541

I was planning on using one of those 0.28ppm crystals you noted the other day at 10MHz, or possibly the 10MHz reference out of the 70cm linear.


On Tue, 7 Jul 2020 at 21:56, Andy G4JNT <andy.g4jnt@...> wrote:
Need to add ...   providing the loop filter is done right  :-)
With a loop bandwidth of a few tens of kHz then the output will take on the (multiplied) phase noise of the reference.  Use a decent  reference, and you've got yourself a decent uWave  signal; at least out to +/- the loop filter bandwidth.



On Tue, 7 Jul 2020 at 13:05, Robin Szemeti - G1YFG <robin@...> wrote:
Perfect, that's all I need to know :)

On Tue, 7 Jul 2020 at 12:23, Andy G4JNT <andy.g4jnt@...> wrote:
Perfectly OK 
There are one or two UHF beacons using them, generating JTxx modulation by reprogrammong (Reg 2 !!!) in real time


G8BKE (or is it G8ACE) is even using one at 47GHz




On Tue, 7 Jul 2020 at 12:03, Robin Szemeti - G1YFG <robin@...> wrote:
Generally speaking, how good is the phase noise on these things for low noise, SSB stylee applications?

I was thinking of using a couple of these to get one of the Ionica units down to 145 MHz.  I'd just like a "yep, they are great providing the loop filter is done right" ranging to "avoid like plague, only suitable for consumer electronics"  sort of ballpark idea, so I know if I am wasting my time pasting them onto a PCB with some mixers.

--
Best regards,

Robin Szemeti

Redpoint Consulting Limited

E: robin@...
T: +44 (0) 1299 405028
M: +44 (0) 7971 883371

CONFIDENTIALITY NOTICE
The information contained in this e-mail is intended only for the
confidential use of the above named recipient. If you are not the
intended recipient or person responsible for delivering it to the
intended recipient, you have received this communication in error
and must not distribute or copy it.
Please accept the sender's apologies, notify the sender immediately
by return e-mail and delete this communication.

Thank you.


Andy G4JNT
 

Good magnifier and steady hand is all you need
The reference input can be up to 700MHz, according to the data sheet.   The comparison frequency up to 104MHz

25MHz from 10MHz.   I'd be tempted to to do X5, filter, divide by 2




On Thu, 9 Jul 2020 at 08:25, Paul Randall G3NJV <paulfrandall@...> wrote:
... I must admit I did a double take when viewing the LMX2541 pdf, quite challenging to mount by hand I'd say. Am I right in thinking it can take a 100MHz reference? I was already contemplating a 100MHz Crystal/PLL  as a route to get 25MHz from a 10 MHz gpsdo hopefully keeping pn low. 
Paul



Sent from my Samsung Galaxy smartphone.



-------- Original message --------
From: Robin Szemeti - G1YFG <robin@...>
Date: 08/07/2020 23:43 (GMT+00:00)
Subject: Re: [UKMicrowaves] LMX2541

I'm doing a board for the Ionica units .. so  144 in/out  .. and then a 947->144 receive convertor, and a LO feed to drive the Ionica receive block to do the 3456 to 947 conversion.    On the transmit side it will do the same trick in reverse and produce 10mw max of 3456 to drive the Ionica PA. I've sourced a SAW filter for the 947 segment, I just need to figure out the best solution for a 3456 filter ...   the 144/947 will be done with active mixers, the MAX2680 / MAX2671, I'm thinking something from the Minicircuits lineup for the 947->3456 upconvertor.

Slapping one of the LMX2451 units onto a PCB would be simple enough, I think JNT has done this already though.  If there is interest, I could do that, but probaby best to let me iron the mistakes out on the 9cm board and then pick up the lump of LMX off that board and slap it onto a PCB. I have an SMT machine in the shed, assembling them would be simple enough.

On Wed, 8 Jul 2020 at 23:12, Paul Randall G3NJV <paulfrandall@...> wrote:
Robin,  Hi, not sure of your intentions but if you get  a pcb made for the lmx2541/ pic etc or purchase chips,  i would be very interested in several boards and chips and happy to share costs. This is such a useful item I imagine a lot of people "could"  be interested in a group buy, so perhaps worth asking?
I really enjoy following your posts,  thanks,  Paul, G3NJV



Sent from my Samsung Galaxy smartphone.



-------- Original message --------
From: Robin Szemeti - G1YFG <robin@...>
Date: 08/07/2020 00:06 (GMT+00:00)
Subject: Re: [UKMicrowaves] LMX2541

I was planning on using one of those 0.28ppm crystals you noted the other day at 10MHz, or possibly the 10MHz reference out of the 70cm linear.


On Tue, 7 Jul 2020 at 21:56, Andy G4JNT <andy.g4jnt@...> wrote:
Need to add ...   providing the loop filter is done right  :-)
With a loop bandwidth of a few tens of kHz then the output will take on the (multiplied) phase noise of the reference.  Use a decent  reference, and you've got yourself a decent uWave  signal; at least out to +/- the loop filter bandwidth.



On Tue, 7 Jul 2020 at 13:05, Robin Szemeti - G1YFG <robin@...> wrote:
Perfect, that's all I need to know :)

On Tue, 7 Jul 2020 at 12:23, Andy G4JNT <andy.g4jnt@...> wrote:
Perfectly OK 
There are one or two UHF beacons using them, generating JTxx modulation by reprogrammong (Reg 2 !!!) in real time


G8BKE (or is it G8ACE) is even using one at 47GHz




On Tue, 7 Jul 2020 at 12:03, Robin Szemeti - G1YFG <robin@...> wrote:
Generally speaking, how good is the phase noise on these things for low noise, SSB stylee applications?

I was thinking of using a couple of these to get one of the Ionica units down to 145 MHz.  I'd just like a "yep, they are great providing the loop filter is done right" ranging to "avoid like plague, only suitable for consumer electronics"  sort of ballpark idea, so I know if I am wasting my time pasting them onto a PCB with some mixers.

--
Best regards,

Robin Szemeti

Redpoint Consulting Limited

E: robin@...
T: +44 (0) 1299 405028
M: +44 (0) 7971 883371

CONFIDENTIALITY NOTICE
The information contained in this e-mail is intended only for the
confidential use of the above named recipient. If you are not the
intended recipient or person responsible for delivering it to the
intended recipient, you have received this communication in error
and must not distribute or copy it.
Please accept the sender's apologies, notify the sender immediately
by return e-mail and delete this communication.

Thank you.


Andy G4JNT
 

For my 739MHz downconverter for QO100 I use a U2794 quadrature downconverter chip to 28MHz.  That removes the need for helical filters as it gives something like 40dB opposite sideband rejection.   The U2794 is highly obsolete now, but ther eis a plethora of quad downconverter chips out there.

The only downside to all of them is they have balanced differential output, so the I/Q outputs need to go via balun transformers to the 90 deg hybrid combining at IF.   I used a mInicircuit hybrid coz I had one, but 90 degree hybrids are a simple network to build.
As image rejection is not a function of filtering you can have a lower If if you want - baseband perhaps.  A direct conversion 739MHz receiver using the LMX2541 as the tuneable LO.




On Thu, 9 Jul 2020 at 08:25, Paul Randall G3NJV <paulfrandall@...> wrote:
Robin, my immediate need is for a receive converter for Eshail, taking the ~700 MHz lnb o/p down to 144, a common enough task to warrant a board I'd think. I've already sourced a Toko helical filter @ 700 and bought the MAX26xx mixers so was thinking pretty much on the same lines as you. A 432 to 2400 upconvertor would also be very useful.
I'm interested to follow your progress and if you do decide to chop out parts of it or just the LO to some boards .  I must admit I did a double take when viewing the LMX2541 pdf, quite challenging to mount by hand I'd say. Am I right in thinking it can take a 100MHz reference? I was already contemplating a 100MHz Crystal/PLL  as a route to get 25MHz from a 10 MHz gpsdo hopefully keeping pn low. 
Paul



Sent from my Samsung Galaxy smartphone.



-------- Original message --------
From: Robin Szemeti - G1YFG <robin@...>
Date: 08/07/2020 23:43 (GMT+00:00)
Subject: Re: [UKMicrowaves] LMX2541

I'm doing a board for the Ionica units .. so  144 in/out  .. and then a 947->144 receive convertor, and a LO feed to drive the Ionica receive block to do the 3456 to 947 conversion.    On the transmit side it will do the same trick in reverse and produce 10mw max of 3456 to drive the Ionica PA. I've sourced a SAW filter for the 947 segment, I just need to figure out the best solution for a 3456 filter ...   the 144/947 will be done with active mixers, the MAX2680 / MAX2671, I'm thinking something from the Minicircuits lineup for the 947->3456 upconvertor.

Slapping one of the LMX2451 units onto a PCB would be simple enough, I think JNT has done this already though.  If there is interest, I could do that, but probaby best to let me iron the mistakes out on the 9cm board and then pick up the lump of LMX off that board and slap it onto a PCB. I have an SMT machine in the shed, assembling them would be simple enough.

On Wed, 8 Jul 2020 at 23:12, Paul Randall G3NJV <paulfrandall@...> wrote:
Robin,  Hi, not sure of your intentions but if you get  a pcb made for the lmx2541/ pic etc or purchase chips,  i would be very interested in several boards and chips and happy to share costs. This is such a useful item I imagine a lot of people "could"  be interested in a group buy, so perhaps worth asking?
I really enjoy following your posts,  thanks,  Paul, G3NJV



Sent from my Samsung Galaxy smartphone.



-------- Original message --------
From: Robin Szemeti - G1YFG <robin@...>
Date: 08/07/2020 00:06 (GMT+00:00)
Subject: Re: [UKMicrowaves] LMX2541

I was planning on using one of those 0.28ppm crystals you noted the other day at 10MHz, or possibly the 10MHz reference out of the 70cm linear.


On Tue, 7 Jul 2020 at 21:56, Andy G4JNT <andy.g4jnt@...> wrote:
Need to add ...   providing the loop filter is done right  :-)
With a loop bandwidth of a few tens of kHz then the output will take on the (multiplied) phase noise of the reference.  Use a decent  reference, and you've got yourself a decent uWave  signal; at least out to +/- the loop filter bandwidth.



On Tue, 7 Jul 2020 at 13:05, Robin Szemeti - G1YFG <robin@...> wrote:
Perfect, that's all I need to know :)

On Tue, 7 Jul 2020 at 12:23, Andy G4JNT <andy.g4jnt@...> wrote:
Perfectly OK 
There are one or two UHF beacons using them, generating JTxx modulation by reprogrammong (Reg 2 !!!) in real time


G8BKE (or is it G8ACE) is even using one at 47GHz




On Tue, 7 Jul 2020 at 12:03, Robin Szemeti - G1YFG <robin@...> wrote:
Generally speaking, how good is the phase noise on these things for low noise, SSB stylee applications?

I was thinking of using a couple of these to get one of the Ionica units down to 145 MHz.  I'd just like a "yep, they are great providing the loop filter is done right" ranging to "avoid like plague, only suitable for consumer electronics"  sort of ballpark idea, so I know if I am wasting my time pasting them onto a PCB with some mixers.

--
Best regards,

Robin Szemeti

Redpoint Consulting Limited

E: robin@...
T: +44 (0) 1299 405028
M: +44 (0) 7971 883371

CONFIDENTIALITY NOTICE
The information contained in this e-mail is intended only for the
confidential use of the above named recipient. If you are not the
intended recipient or person responsible for delivering it to the
intended recipient, you have received this communication in error
and must not distribute or copy it.
Please accept the sender's apologies, notify the sender immediately
by return e-mail and delete this communication.

Thank you.


Paul Randall G3NJV
 

Thanks Andy. 
Actually my thinking is totally wrong isn't it? I was going along the lines that locking a 100MHz crystal by dividing it and comparing to the gpsdo at 10MHz would be quieter than directly multiplying the gpsdo to 100,  but actually you'd get the same result? Probably worse if the pll wasn't right due to lack of good test gear. 



Sent from my Samsung Galaxy smartphone.



-------- Original message --------
From: Andy G4JNT <andy.g4jnt@...>
Date: 09/07/2020 08:30 (GMT+00:00)
To: "UK Microwaves groups.io" <UKMicrowaves@groups.io>
Subject: Re: [UKMicrowaves] LMX2541

Good magnifier and steady hand is all you need
The reference input can be up to 700MHz, according to the data sheet.   The comparison frequency up to 104MHz

25MHz from 10MHz.   I'd be tempted to to do X5, filter, divide by 2




On Thu, 9 Jul 2020 at 08:25, Paul Randall G3NJV <paulfrandall@...> wrote:
... I must admit I did a double take when viewing the LMX2541 pdf, quite challenging to mount by hand I'd say. Am I right in thinking it can take a 100MHz reference? I was already contemplating a 100MHz Crystal/PLL  as a route to get 25MHz from a 10 MHz gpsdo hopefully keeping pn low. 
Paul



Sent from my Samsung Galaxy smartphone.



-------- Original message --------
From: Robin Szemeti - G1YFG <robin@...>
Date: 08/07/2020 23:43 (GMT+00:00)
Subject: Re: [UKMicrowaves] LMX2541

I'm doing a board for the Ionica units .. so  144 in/out  .. and then a 947->144 receive convertor, and a LO feed to drive the Ionica receive block to do the 3456 to 947 conversion.    On the transmit side it will do the same trick in reverse and produce 10mw max of 3456 to drive the Ionica PA. I've sourced a SAW filter for the 947 segment, I just need to figure out the best solution for a 3456 filter ...   the 144/947 will be done with active mixers, the MAX2680 / MAX2671, I'm thinking something from the Minicircuits lineup for the 947->3456 upconvertor.

Slapping one of the LMX2451 units onto a PCB would be simple enough, I think JNT has done this already though.  If there is interest, I could do that, but probaby best to let me iron the mistakes out on the 9cm board and then pick up the lump of LMX off that board and slap it onto a PCB. I have an SMT machine in the shed, assembling them would be simple enough.

On Wed, 8 Jul 2020 at 23:12, Paul Randall G3NJV <paulfrandall@...> wrote:
Robin,  Hi, not sure of your intentions but if you get  a pcb made for the lmx2541/ pic etc or purchase chips,  i would be very interested in several boards and chips and happy to share costs. This is such a useful item I imagine a lot of people "could"  be interested in a group buy, so perhaps worth asking?
I really enjoy following your posts,  thanks,  Paul, G3NJV



Sent from my Samsung Galaxy smartphone.



-------- Original message --------
From: Robin Szemeti - G1YFG <robin@...>
Date: 08/07/2020 00:06 (GMT+00:00)
Subject: Re: [UKMicrowaves] LMX2541

I was planning on using one of those 0.28ppm crystals you noted the other day at 10MHz, or possibly the 10MHz reference out of the 70cm linear.


On Tue, 7 Jul 2020 at 21:56, Andy G4JNT <andy.g4jnt@...> wrote:
Need to add ...   providing the loop filter is done right  :-)
With a loop bandwidth of a few tens of kHz then the output will take on the (multiplied) phase noise of the reference.  Use a decent  reference, and you've got yourself a decent uWave  signal; at least out to +/- the loop filter bandwidth.



On Tue, 7 Jul 2020 at 13:05, Robin Szemeti - G1YFG <robin@...> wrote:
Perfect, that's all I need to know :)

On Tue, 7 Jul 2020 at 12:23, Andy G4JNT <andy.g4jnt@...> wrote:
Perfectly OK 
There are one or two UHF beacons using them, generating JTxx modulation by reprogrammong (Reg 2 !!!) in real time


G8BKE (or is it G8ACE) is even using one at 47GHz




On Tue, 7 Jul 2020 at 12:03, Robin Szemeti - G1YFG <robin@...> wrote:
Generally speaking, how good is the phase noise on these things for low noise, SSB stylee applications?

I was thinking of using a couple of these to get one of the Ionica units down to 145 MHz.  I'd just like a "yep, they are great providing the loop filter is done right" ranging to "avoid like plague, only suitable for consumer electronics"  sort of ballpark idea, so I know if I am wasting my time pasting them onto a PCB with some mixers.

--
Best regards,

Robin Szemeti

Redpoint Consulting Limited

E: robin@...
T: +44 (0) 1299 405028
M: +44 (0) 7971 883371

CONFIDENTIALITY NOTICE
The information contained in this e-mail is intended only for the
confidential use of the above named recipient. If you are not the
intended recipient or person responsible for delivering it to the
intended recipient, you have received this communication in error
and must not distribute or copy it.
Please accept the sender's apologies, notify the sender immediately
by return e-mail and delete this communication.

Thank you.


Andy G4JNT
 

Yes.
Keep the LMX2541 for those obscure frequencies a fine tuning-step solutions.    For single e frequencies, keep it simple.

That downconverter was actually designed for a CAT controlled converter for 144MHz to 28MHz when I used an SDR-IQ and wanted 1Hz steps for automatic WSJT EME Doppler correction.  It emulated the IC746 (which could only do 10Hz steps) , and was made redundant when I got the Elad which could do everything directly.
The downconverter chip covers up to 1000MHz and fortunately I'd used a variant of the LMX' that could generate 739MHz.  But beware;  that chip is annoying as different variants are needed for frequency coverage above about 450MHz. 
   
If you don't need ultra-fine steps, the ADF4351 is a better bet now, with 34 - 4400MHz continuous coverage.  And, if you're willing to aid-and-abet them in destroying the rest of the World's technology capability, available from China for peanuts.



On Thu, 9 Jul 2020 at 08:51, Paul Randall G3NJV <paulfrandall@...> wrote:
Thanks Andy. 
Actually my thinking is totally wrong isn't it? I was going along the lines that locking a 100MHz crystal by dividing it and comparing to the gpsdo at 10MHz would be quieter than directly multiplying the gpsdo to 100,  but actually you'd get the same result? Probably worse if the pll wasn't right due to lack of good test gear. 



Sent from my Samsung Galaxy smartphone.



-------- Original message --------
From: Andy G4JNT <andy.g4jnt@...>
Date: 09/07/2020 08:30 (GMT+00:00)
To: "UK Microwaves groups.io" <UKMicrowaves@groups.io>
Subject: Re: [UKMicrowaves] LMX2541

Good magnifier and steady hand is all you need
The reference input can be up to 700MHz, according to the data sheet.   The comparison frequency up to 104MHz

25MHz from 10MHz.   I'd be tempted to to do X5, filter, divide by 2




On Thu, 9 Jul 2020 at 08:25, Paul Randall G3NJV <paulfrandall@...> wrote:
... I must admit I did a double take when viewing the LMX2541 pdf, quite challenging to mount by hand I'd say. Am I right in thinking it can take a 100MHz reference? I was already contemplating a 100MHz Crystal/PLL  as a route to get 25MHz from a 10 MHz gpsdo hopefully keeping pn low. 
Paul



Sent from my Samsung Galaxy smartphone.



-------- Original message --------
From: Robin Szemeti - G1YFG <robin@...>
Date: 08/07/2020 23:43 (GMT+00:00)
Subject: Re: [UKMicrowaves] LMX2541

I'm doing a board for the Ionica units .. so  144 in/out  .. and then a 947->144 receive convertor, and a LO feed to drive the Ionica receive block to do the 3456 to 947 conversion.    On the transmit side it will do the same trick in reverse and produce 10mw max of 3456 to drive the Ionica PA. I've sourced a SAW filter for the 947 segment, I just need to figure out the best solution for a 3456 filter ...   the 144/947 will be done with active mixers, the MAX2680 / MAX2671, I'm thinking something from the Minicircuits lineup for the 947->3456 upconvertor.

Slapping one of the LMX2451 units onto a PCB would be simple enough, I think JNT has done this already though.  If there is interest, I could do that, but probaby best to let me iron the mistakes out on the 9cm board and then pick up the lump of LMX off that board and slap it onto a PCB. I have an SMT machine in the shed, assembling them would be simple enough.

On Wed, 8 Jul 2020 at 23:12, Paul Randall G3NJV <paulfrandall@...> wrote:
Robin,  Hi, not sure of your intentions but if you get  a pcb made for the lmx2541/ pic etc or purchase chips,  i would be very interested in several boards and chips and happy to share costs. This is such a useful item I imagine a lot of people "could"  be interested in a group buy, so perhaps worth asking?
I really enjoy following your posts,  thanks,  Paul, G3NJV



Sent from my Samsung Galaxy smartphone.



-------- Original message --------
From: Robin Szemeti - G1YFG <robin@...>
Date: 08/07/2020 00:06 (GMT+00:00)
Subject: Re: [UKMicrowaves] LMX2541

I was planning on using one of those 0.28ppm crystals you noted the other day at 10MHz, or possibly the 10MHz reference out of the 70cm linear.


On Tue, 7 Jul 2020 at 21:56, Andy G4JNT <andy.g4jnt@...> wrote:
Need to add ...   providing the loop filter is done right  :-)
With a loop bandwidth of a few tens of kHz then the output will take on the (multiplied) phase noise of the reference.  Use a decent  reference, and you've got yourself a decent uWave  signal; at least out to +/- the loop filter bandwidth.



On Tue, 7 Jul 2020 at 13:05, Robin Szemeti - G1YFG <robin@...> wrote:
Perfect, that's all I need to know :)

On Tue, 7 Jul 2020 at 12:23, Andy G4JNT <andy.g4jnt@...> wrote:
Perfectly OK 
There are one or two UHF beacons using them, generating JTxx modulation by reprogrammong (Reg 2 !!!) in real time


G8BKE (or is it G8ACE) is even using one at 47GHz




On Tue, 7 Jul 2020 at 12:03, Robin Szemeti - G1YFG <robin@...> wrote:
Generally speaking, how good is the phase noise on these things for low noise, SSB stylee applications?

I was thinking of using a couple of these to get one of the Ionica units down to 145 MHz.  I'd just like a "yep, they are great providing the loop filter is done right" ranging to "avoid like plague, only suitable for consumer electronics"  sort of ballpark idea, so I know if I am wasting my time pasting them onto a PCB with some mixers.

--
Best regards,

Robin Szemeti

Redpoint Consulting Limited

E: robin@...
T: +44 (0) 1299 405028
M: +44 (0) 7971 883371

CONFIDENTIALITY NOTICE
The information contained in this e-mail is intended only for the
confidential use of the above named recipient. If you are not the
intended recipient or person responsible for delivering it to the
intended recipient, you have received this communication in error
and must not distribute or copy it.
Please accept the sender's apologies, notify the sender immediately
by return e-mail and delete this communication.

Thank you.


Robin Szemeti - G1YFG
 

As Andy says, you can get the ADF4351 boards pre-assembled for less than £12 on eBay ... which is less than I can buy the bare chips for.




On Thu, 9 Jul 2020 at 09:29, Andy G4JNT <andy.g4jnt@...> wrote:
Yes.
Keep the LMX2541 for those obscure frequencies a fine tuning-step solutions.    For single e frequencies, keep it simple.

That downconverter was actually designed for a CAT controlled converter for 144MHz to 28MHz when I used an SDR-IQ and wanted 1Hz steps for automatic WSJT EME Doppler correction.  It emulated the IC746 (which could only do 10Hz steps) , and was made redundant when I got the Elad which could do everything directly.
The downconverter chip covers up to 1000MHz and fortunately I'd used a variant of the LMX' that could generate 739MHz.  But beware;  that chip is annoying as different variants are needed for frequency coverage above about 450MHz. 
   
If you don't need ultra-fine steps, the ADF4351 is a better bet now, with 34 - 4400MHz continuous coverage.  And, if you're willing to aid-and-abet them in destroying the rest of the World's technology capability, available from China for peanuts.



On Thu, 9 Jul 2020 at 08:51, Paul Randall G3NJV <paulfrandall@...> wrote:
Thanks Andy. 
Actually my thinking is totally wrong isn't it? I was going along the lines that locking a 100MHz crystal by dividing it and comparing to the gpsdo at 10MHz would be quieter than directly multiplying the gpsdo to 100,  but actually you'd get the same result? Probably worse if the pll wasn't right due to lack of good test gear. 



Sent from my Samsung Galaxy smartphone.



-------- Original message --------
From: Andy G4JNT <andy.g4jnt@...>
Date: 09/07/2020 08:30 (GMT+00:00)
To: "UK Microwaves groups.io" <UKMicrowaves@groups.io>
Subject: Re: [UKMicrowaves] LMX2541

Good magnifier and steady hand is all you need
The reference input can be up to 700MHz, according to the data sheet.   The comparison frequency up to 104MHz

25MHz from 10MHz.   I'd be tempted to to do X5, filter, divide by 2




On Thu, 9 Jul 2020 at 08:25, Paul Randall G3NJV <paulfrandall@...> wrote:
... I must admit I did a double take when viewing the LMX2541 pdf, quite challenging to mount by hand I'd say. Am I right in thinking it can take a 100MHz reference? I was already contemplating a 100MHz Crystal/PLL  as a route to get 25MHz from a 10 MHz gpsdo hopefully keeping pn low. 
Paul



Sent from my Samsung Galaxy smartphone.



-------- Original message --------
From: Robin Szemeti - G1YFG <robin@...>
Date: 08/07/2020 23:43 (GMT+00:00)
Subject: Re: [UKMicrowaves] LMX2541

I'm doing a board for the Ionica units .. so  144 in/out  .. and then a 947->144 receive convertor, and a LO feed to drive the Ionica receive block to do the 3456 to 947 conversion.    On the transmit side it will do the same trick in reverse and produce 10mw max of 3456 to drive the Ionica PA. I've sourced a SAW filter for the 947 segment, I just need to figure out the best solution for a 3456 filter ...   the 144/947 will be done with active mixers, the MAX2680 / MAX2671, I'm thinking something from the Minicircuits lineup for the 947->3456 upconvertor.

Slapping one of the LMX2451 units onto a PCB would be simple enough, I think JNT has done this already though.  If there is interest, I could do that, but probaby best to let me iron the mistakes out on the 9cm board and then pick up the lump of LMX off that board and slap it onto a PCB. I have an SMT machine in the shed, assembling them would be simple enough.

On Wed, 8 Jul 2020 at 23:12, Paul Randall G3NJV <paulfrandall@...> wrote:
Robin,  Hi, not sure of your intentions but if you get  a pcb made for the lmx2541/ pic etc or purchase chips,  i would be very interested in several boards and chips and happy to share costs. This is such a useful item I imagine a lot of people "could"  be interested in a group buy, so perhaps worth asking?
I really enjoy following your posts,  thanks,  Paul, G3NJV



Sent from my Samsung Galaxy smartphone.



-------- Original message --------
From: Robin Szemeti - G1YFG <robin@...>
Date: 08/07/2020 00:06 (GMT+00:00)
Subject: Re: [UKMicrowaves] LMX2541

I was planning on using one of those 0.28ppm crystals you noted the other day at 10MHz, or possibly the 10MHz reference out of the 70cm linear.


On Tue, 7 Jul 2020 at 21:56, Andy G4JNT <andy.g4jnt@...> wrote:
Need to add ...   providing the loop filter is done right  :-)
With a loop bandwidth of a few tens of kHz then the output will take on the (multiplied) phase noise of the reference.  Use a decent  reference, and you've got yourself a decent uWave  signal; at least out to +/- the loop filter bandwidth.



On Tue, 7 Jul 2020 at 13:05, Robin Szemeti - G1YFG <robin@...> wrote:
Perfect, that's all I need to know :)

On Tue, 7 Jul 2020 at 12:23, Andy G4JNT <andy.g4jnt@...> wrote:
Perfectly OK 
There are one or two UHF beacons using them, generating JTxx modulation by reprogrammong (Reg 2 !!!) in real time


G8BKE (or is it G8ACE) is even using one at 47GHz




On Tue, 7 Jul 2020 at 12:03, Robin Szemeti - G1YFG <robin@...> wrote:
Generally speaking, how good is the phase noise on these things for low noise, SSB stylee applications?

I was thinking of using a couple of these to get one of the Ionica units down to 145 MHz.  I'd just like a "yep, they are great providing the loop filter is done right" ranging to "avoid like plague, only suitable for consumer electronics"  sort of ballpark idea, so I know if I am wasting my time pasting them onto a PCB with some mixers.

--
Best regards,

Robin Szemeti

Redpoint Consulting Limited

E: robin@...
T: +44 (0) 1299 405028
M: +44 (0) 7971 883371

CONFIDENTIALITY NOTICE
The information contained in this e-mail is intended only for the
confidential use of the above named recipient. If you are not the
intended recipient or person responsible for delivering it to the
intended recipient, you have received this communication in error
and must not distribute or copy it.
Please accept the sender's apologies, notify the sender immediately
by return e-mail and delete this communication.

Thank you.


Andy G4JNT
 

I'd go for the SV1AFN boards.   More expensive, but you're not contributing to China's attempt at World domination



On Thu, 9 Jul 2020 at 10:07, Robin Szemeti - G1YFG <robin@...> wrote:
As Andy says, you can get the ADF4351 boards pre-assembled for less than £12 on eBay ... which is less than I can buy the bare chips for.




On Thu, 9 Jul 2020 at 09:29, Andy G4JNT <andy.g4jnt@...> wrote:
Yes.
Keep the LMX2541 for those obscure frequencies a fine tuning-step solutions.    For single e frequencies, keep it simple.

That downconverter was actually designed for a CAT controlled converter for 144MHz to 28MHz when I used an SDR-IQ and wanted 1Hz steps for automatic WSJT EME Doppler correction.  It emulated the IC746 (which could only do 10Hz steps) , and was made redundant when I got the Elad which could do everything directly.
The downconverter chip covers up to 1000MHz and fortunately I'd used a variant of the LMX' that could generate 739MHz.  But beware;  that chip is annoying as different variants are needed for frequency coverage above about 450MHz. 
   
If you don't need ultra-fine steps, the ADF4351 is a better bet now, with 34 - 4400MHz continuous coverage.  And, if you're willing to aid-and-abet them in destroying the rest of the World's technology capability, available from China for peanuts.



On Thu, 9 Jul 2020 at 08:51, Paul Randall G3NJV <paulfrandall@...> wrote:
Thanks Andy. 
Actually my thinking is totally wrong isn't it? I was going along the lines that locking a 100MHz crystal by dividing it and comparing to the gpsdo at 10MHz would be quieter than directly multiplying the gpsdo to 100,  but actually you'd get the same result? Probably worse if the pll wasn't right due to lack of good test gear. 



Sent from my Samsung Galaxy smartphone.



-------- Original message --------
From: Andy G4JNT <andy.g4jnt@...>
Date: 09/07/2020 08:30 (GMT+00:00)
To: "UK Microwaves groups.io" <UKMicrowaves@groups.io>
Subject: Re: [UKMicrowaves] LMX2541

Good magnifier and steady hand is all you need
The reference input can be up to 700MHz, according to the data sheet.   The comparison frequency up to 104MHz

25MHz from 10MHz.   I'd be tempted to to do X5, filter, divide by 2




On Thu, 9 Jul 2020 at 08:25, Paul Randall G3NJV <paulfrandall@...> wrote:
... I must admit I did a double take when viewing the LMX2541 pdf, quite challenging to mount by hand I'd say. Am I right in thinking it can take a 100MHz reference? I was already contemplating a 100MHz Crystal/PLL  as a route to get 25MHz from a 10 MHz gpsdo hopefully keeping pn low. 
Paul



Sent from my Samsung Galaxy smartphone.



-------- Original message --------
From: Robin Szemeti - G1YFG <robin@...>
Date: 08/07/2020 23:43 (GMT+00:00)
Subject: Re: [UKMicrowaves] LMX2541

I'm doing a board for the Ionica units .. so  144 in/out  .. and then a 947->144 receive convertor, and a LO feed to drive the Ionica receive block to do the 3456 to 947 conversion.    On the transmit side it will do the same trick in reverse and produce 10mw max of 3456 to drive the Ionica PA. I've sourced a SAW filter for the 947 segment, I just need to figure out the best solution for a 3456 filter ...   the 144/947 will be done with active mixers, the MAX2680 / MAX2671, I'm thinking something from the Minicircuits lineup for the 947->3456 upconvertor.

Slapping one of the LMX2451 units onto a PCB would be simple enough, I think JNT has done this already though.  If there is interest, I could do that, but probaby best to let me iron the mistakes out on the 9cm board and then pick up the lump of LMX off that board and slap it onto a PCB. I have an SMT machine in the shed, assembling them would be simple enough.

On Wed, 8 Jul 2020 at 23:12, Paul Randall G3NJV <paulfrandall@...> wrote:
Robin,  Hi, not sure of your intentions but if you get  a pcb made for the lmx2541/ pic etc or purchase chips,  i would be very interested in several boards and chips and happy to share costs. This is such a useful item I imagine a lot of people "could"  be interested in a group buy, so perhaps worth asking?
I really enjoy following your posts,  thanks,  Paul, G3NJV



Sent from my Samsung Galaxy smartphone.



-------- Original message --------
From: Robin Szemeti - G1YFG <robin@...>
Date: 08/07/2020 00:06 (GMT+00:00)
Subject: Re: [UKMicrowaves] LMX2541

I was planning on using one of those 0.28ppm crystals you noted the other day at 10MHz, or possibly the 10MHz reference out of the 70cm linear.


On Tue, 7 Jul 2020 at 21:56, Andy G4JNT <andy.g4jnt@...> wrote:
Need to add ...   providing the loop filter is done right  :-)
With a loop bandwidth of a few tens of kHz then the output will take on the (multiplied) phase noise of the reference.  Use a decent  reference, and you've got yourself a decent uWave  signal; at least out to +/- the loop filter bandwidth.



On Tue, 7 Jul 2020 at 13:05, Robin Szemeti - G1YFG <robin@...> wrote:
Perfect, that's all I need to know :)

On Tue, 7 Jul 2020 at 12:23, Andy G4JNT <andy.g4jnt@...> wrote:
Perfectly OK 
There are one or two UHF beacons using them, generating JTxx modulation by reprogrammong (Reg 2 !!!) in real time


G8BKE (or is it G8ACE) is even using one at 47GHz




On Tue, 7 Jul 2020 at 12:03, Robin Szemeti - G1YFG <robin@...> wrote:
Generally speaking, how good is the phase noise on these things for low noise, SSB stylee applications?

I was thinking of using a couple of these to get one of the Ionica units down to 145 MHz.  I'd just like a "yep, they are great providing the loop filter is done right" ranging to "avoid like plague, only suitable for consumer electronics"  sort of ballpark idea, so I know if I am wasting my time pasting them onto a PCB with some mixers.

--
Best regards,

Robin Szemeti

Redpoint Consulting Limited

E: robin@...
T: +44 (0) 1299 405028
M: +44 (0) 7971 883371

CONFIDENTIALITY NOTICE
The information contained in this e-mail is intended only for the
confidential use of the above named recipient. If you are not the
intended recipient or person responsible for delivering it to the
intended recipient, you have received this communication in error
and must not distribute or copy it.
Please accept the sender's apologies, notify the sender immediately
by return e-mail and delete this communication.

Thank you.


Martin Phillips G4CIO
 

Good, and friendly service, but a word of warning: with the current difficulties, the one I ordered from Greece  in May took six weeks to arrive from the date of despatch.

Martin/

On 9/7/20 10:13 AM, Andy G4JNT wrote:
I'd go for the SV1AFN boards.   More expensive, but you're not contributing to China's attempt at World domination



On Thu, 9 Jul 2020 at 10:07, Robin Szemeti - G1YFG <robin@...> wrote:
As Andy says, you can get the ADF4351 boards pre-assembled for less than £12 on eBay ... which is less than I can buy the bare chips for.




On Thu, 9 Jul 2020 at 09:29, Andy G4JNT <andy.g4jnt@...> wrote:
Yes.
Keep the LMX2541 for those obscure frequencies a fine tuning-step solutions.    For single e frequencies, keep it simple.

That downconverter was actually designed for a CAT controlled converter for 144MHz to 28MHz when I used an SDR-IQ and wanted 1Hz steps for automatic WSJT EME Doppler correction.  It emulated the IC746 (which could only do 10Hz steps) , and was made redundant when I got the Elad which could do everything directly.
The downconverter chip covers up to 1000MHz and fortunately I'd used a variant of the LMX' that could generate 739MHz.  But beware;  that chip is annoying as different variants are needed for frequency coverage above about 450MHz. 
   
If you don't need ultra-fine steps, the ADF4351 is a better bet now, with 34 - 4400MHz continuous coverage.  And, if you're willing to aid-and-abet them in destroying the rest of the World's technology capability, available from China for peanuts.



On Thu, 9 Jul 2020 at 08:51, Paul Randall G3NJV <paulfrandall@...> wrote:
Thanks Andy. 
Actually my thinking is totally wrong isn't it? I was going along the lines that locking a 100MHz crystal by dividing it and comparing to the gpsdo at 10MHz would be quieter than directly multiplying the gpsdo to 100,  but actually you'd get the same result? Probably worse if the pll wasn't right due to lack of good test gear. 



Sent from my Samsung Galaxy smartphone.



-------- Original message --------
From: Andy G4JNT <andy.g4jnt@...>
Date: 09/07/2020 08:30 (GMT+00:00)
To: "UK Microwaves groups.io" <UKMicrowaves@groups.io>
Subject: Re: [UKMicrowaves] LMX2541

Good magnifier and steady hand is all you need
The reference input can be up to 700MHz, according to the data sheet.   The comparison frequency up to 104MHz

25MHz from 10MHz.   I'd be tempted to to do X5, filter, divide by 2




On Thu, 9 Jul 2020 at 08:25, Paul Randall G3NJV <paulfrandall@...> wrote:
... I must admit I did a double take when viewing the LMX2541 pdf, quite challenging to mount by hand I'd say. Am I right in thinking it can take a 100MHz reference? I was already contemplating a 100MHz Crystal/PLL  as a route to get 25MHz from a 10 MHz gpsdo hopefully keeping pn low. 
Paul



Sent from my Samsung Galaxy smartphone.



-------- Original message --------
From: Robin Szemeti - G1YFG <robin@...>
Date: 08/07/2020 23:43 (GMT+00:00)
Subject: Re: [UKMicrowaves] LMX2541

I'm doing a board for the Ionica units .. so  144 in/out  .. and then a 947->144 receive convertor, and a LO feed to drive the Ionica receive block to do the 3456 to 947 conversion.    On the transmit side it will do the same trick in reverse and produce 10mw max of 3456 to drive the Ionica PA. I've sourced a SAW filter for the 947 segment, I just need to figure out the best solution for a 3456 filter ...   the 144/947 will be done with active mixers, the MAX2680 / MAX2671, I'm thinking something from the Minicircuits lineup for the 947->3456 upconvertor.

Slapping one of the LMX2451 units onto a PCB would be simple enough, I think JNT has done this already though.  If there is interest, I could do that, but probaby best to let me iron the mistakes out on the 9cm board and then pick up the lump of LMX off that board and slap it onto a PCB. I have an SMT machine in the shed, assembling them would be simple enough.

On Wed, 8 Jul 2020 at 23:12, Paul Randall G3NJV <paulfrandall@...> wrote:
Robin,  Hi, not sure of your intentions but if you get  a pcb made for the lmx2541/ pic etc or purchase chips,  i would be very interested in several boards and chips and happy to share costs. This is such a useful item I imagine a lot of people "could"  be interested in a group buy, so perhaps worth asking?
I really enjoy following your posts,  thanks,  Paul, G3NJV



Sent from my Samsung Galaxy smartphone.



-------- Original message --------
From: Robin Szemeti - G1YFG <robin@...>
Date: 08/07/2020 00:06 (GMT+00:00)
Subject: Re: [UKMicrowaves] LMX2541

I was planning on using one of those 0.28ppm crystals you noted the other day at 10MHz, or possibly the 10MHz reference out of the 70cm linear.


On Tue, 7 Jul 2020 at 21:56, Andy G4JNT <andy.g4jnt@...> wrote:
Need to add ...   providing the loop filter is done right  :-)
With a loop bandwidth of a few tens of kHz then the output will take on the (multiplied) phase noise of the reference.  Use a decent  reference, and you've got yourself a decent uWave  signal; at least out to +/- the loop filter bandwidth.



On Tue, 7 Jul 2020 at 13:05, Robin Szemeti - G1YFG <robin@...> wrote:
Perfect, that's all I need to know :)

On Tue, 7 Jul 2020 at 12:23, Andy G4JNT <andy.g4jnt@...> wrote:
Perfectly OK 
There are one or two UHF beacons using them, generating JTxx modulation by reprogrammong (Reg 2 !!!) in real time


G8BKE (or is it G8ACE) is even using one at 47GHz




On Tue, 7 Jul 2020 at 12:03, Robin Szemeti - G1YFG <robin@...> wrote:
Generally speaking, how good is the phase noise on these things for low noise, SSB stylee applications?

I was thinking of using a couple of these to get one of the Ionica units down to 145 MHz.  I'd just like a "yep, they are great providing the loop filter is done right" ranging to "avoid like plague, only suitable for consumer electronics"  sort of ballpark idea, so I know if I am wasting my time pasting them onto a PCB with some mixers.

--
Best regards,

Robin Szemeti

Redpoint Consulting Limited 

E: robin@... 
T: +44 (0) 1299 405028 
M: +44 (0) 7971 883371 

CONFIDENTIALITY NOTICE 
The information contained in this e-mail is intended only for the
confidential use of the above named recipient. If you are not the
intended recipient or person responsible for delivering it to the
intended recipient, you have received this communication in error 
and must not distribute or copy it. 
Please accept the sender's apologies, notify the sender immediately 
by return e-mail and delete this communication. 

Thank you. 



Chris Bartram G4DGU <chris@...>
 

Multiplication of a good 10MHz reference to a higher frequency is very much a known art in synthesiser design. The key words are 'clock multiplier', and the technique has been used for decades.

In my 10GHz transverter modules, the synthesiser clock was a 40MHz TCXO chosen for its good phase noise performance.* That was locked to 10MHz using a very simple PLL based on a couple of 74ACT parts. With good design that approach can be at least as good as using non-linear frequency multiplication, and it gives opportunities for spectrum shaping. I spent quite a lot of time working on that when designing the modules.

Although the technology around synthesiser design and manufacture has been known and appreciated professionally for many decades, it always rather saddened and surprised me that there has been rather unwarranted resistance from within the amateur radio community. I blame a now SK contributor to Radcom for generating scare stories, based largely on some of the more inept transceiver designs produced in Japan in the 70's, '80s and '90s!

That resistance was real. I saw it in my then workaday world with muTek. When around 1985 we  introduced a HF transverter, a lot of the development time was spent on optimising the synthesiser. Using what is now an antediluvian chip - the Plessey NJ8811 with a VCO of my own design. It worked very well, although I can't now quote any meaninful numbers off the top of my head - it was 35 years ago! The number of people who came-up to me at shows and rallies and mumbled nonsense, clearly parroting what had been said in Radcom as though it were the truth saddened me deeply.

However that's now largely in the past!

The coming of single chip synthesisers with integrated VCOs has made the 'design' of synthesisers much simpler, but poor choice of tuning range, loop parameters, selection of power supplies, sometimes layout and decoupling problems and other pitfalls can wreck their performance. I've looked carefully at a few of the Chinese synth boards and they tend to be pretty bad when looked at critically. The only commercially available synth PCB which I've measured and which produced a sigh of pleasure is that produced by SV1AFN - that guy really knows what he is doing!

73

Chris G4DGU

*The phase noise performance of of oscillators depends on two primary factors: the loaded Q of the resonator in that oscillator, and the noise performance of the maintaining amplifier. (There are a number of subsidary factors which are important, but are marginal to this brief explanation). If an oscillator at 10MHz runs with a resonator loaded Q of, say, 250000, all else being equal, that  multipled to 100MHz that will be equivalent to an oscillator with a resonator loaded Q of 25000. A well designed VHF crystal oscillator can easily have a loaded Q of
100000, resulting in a potential improvement in PN of >12dB. That's
worth fighting for.


Robin Szemeti - G1YFG
 

Hi Paul,

I have some and I did consider that .. in the end, I have decided to go a bit more modern and use an IQ modulator. If I have the phasing right it should produce just the lower sideband, but equally, I might have it 90 degrees out and it won't ... so you'll either hear me on 3456MHz .. or possibly about 1606 MHz above there. I should say the chances are 50/50 either way at this point ;)


On Thu, 9 Jul 2020 at 00:29, PAUL NICKALLS via groups.io <paulnick=btinternet.com@groups.io> wrote:
Hi Robin,

I made a pipecap filter for 3.6214 Ghz It was quite narrowband and had an approx 3dB loss.  You can play with loss and bandwidth by adjusting the probes.  I have a bag of 28mm pipecaps here if you want one. A bit old school but quick and easy.

73
Paul G8AQA

On 08/07/2020 23:43, Robin Szemeti - G1YFG wrote:
I'm doing a board for the Ionica units .. so  144 in/out  .. and then a 947->144 receive convertor, and a LO feed to drive the Ionica receive block to do the 3456 to 947 conversion.    On the transmit side it will do the same trick in reverse and produce 10mw max of 3456 to drive the Ionica PA. I've sourced a SAW filter for the 947 segment, I just need to figure out the best solution for a 3456 filter ...   the 144/947 will be done with active mixers, the MAX2680 / MAX2671, I'm thinking something from the Minicircuits lineup for the 947->3456 upconvertor.

Slapping one of the LMX2451 units onto a PCB would be simple enough, I think JNT has done this already though.  If there is interest, I could do that, but probaby best to let me iron the mistakes out on the 9cm board and then pick up the lump of LMX off that board and slap it onto a PCB. I have an SMT machine in the shed, assembling them would be simple enough.

On Wed, 8 Jul 2020 at 23:12, Paul Randall G3NJV <paulfrandall@...> wrote:
Robin,  Hi, not sure of your intentions but if you get  a pcb made for the lmx2541/ pic etc or purchase chips,  i would be very interested in several boards and chips and happy to share costs. This is such a useful item I imagine a lot of people "could"  be interested in a group buy, so perhaps worth asking?
I really enjoy following your posts,  thanks,  Paul, G3NJV



Sent from my Samsung Galaxy smartphone.



-------- Original message --------
From: Robin Szemeti - G1YFG <robin@...>
Date: 08/07/2020 00:06 (GMT+00:00)
Subject: Re: [UKMicrowaves] LMX2541

I was planning on using one of those 0.28ppm crystals you noted the other day at 10MHz, or possibly the 10MHz reference out of the 70cm linear.


On Tue, 7 Jul 2020 at 21:56, Andy G4JNT <andy.g4jnt@...> wrote:
Need to add ...   providing the loop filter is done right  :-)
With a loop bandwidth of a few tens of kHz then the output will take on the (multiplied) phase noise of the reference.  Use a decent  reference, and you've got yourself a decent uWave  signal; at least out to +/- the loop filter bandwidth.



On Tue, 7 Jul 2020 at 13:05, Robin Szemeti - G1YFG <robin@...> wrote:
Perfect, that's all I need to know :)

On Tue, 7 Jul 2020 at 12:23, Andy G4JNT <andy.g4jnt@...> wrote:
Perfectly OK 
There are one or two UHF beacons using them, generating JTxx modulation by reprogrammong (Reg 2 !!!) in real time


G8BKE (or is it G8ACE) is even using one at 47GHz




On Tue, 7 Jul 2020 at 12:03, Robin Szemeti - G1YFG <robin@...> wrote:
Generally speaking, how good is the phase noise on these things for low noise, SSB stylee applications?

I was thinking of using a couple of these to get one of the Ionica units down to 145 MHz.  I'd just like a "yep, they are great providing the loop filter is done right" ranging to "avoid like plague, only suitable for consumer electronics"  sort of ballpark idea, so I know if I am wasting my time pasting them onto a PCB with some mixers.

--
Best regards,

Robin Szemeti

Redpoint Consulting Limited 

E: robin@... 
T: +44 (0) 1299 405028 
M: +44 (0) 7971 883371 

CONFIDENTIALITY NOTICE 
The information contained in this e-mail is intended only for the
confidential use of the above named recipient. If you are not the
intended recipient or person responsible for delivering it to the
intended recipient, you have received this communication in error 
and must not distribute or copy it. 
Please accept the sender's apologies, notify the sender immediately 
by return e-mail and delete this communication. 

Thank you.