ADF5355 Signal Source


ed.murphy31@...
 

Hello All,


Has anyone tried one of these ? Looks like they will not release the source code which would be useful.


http://www.ebay.co.uk/itm/ADF5355-Module-Touch-Color-Screen-LCD-Sweep-RF-Signal-Source-VCO-Synthesizer-PLL-/272780479899?hash=item3f82fbcd9b:g:B7IAAOSw3ChZerDQ


Ed GM3SBC



Derek Kozel <derek.kozel@...>
 

I've used the ADF4351 version of the same. It works but is a bit fiddly. The problem is that sometimes you cannot tell if it has tuned or not, so without a spectrum analyzer or frequency counter it can cause confusion.

Here is a C++ driver for the ADF 5355 synth released under GPL license. If anyone has questions I'd be happy to help with any of the ADF435x, ADF535x, or LMX259x synths.
https://github.com/EttusResearch/uhd/blob/maint/host/lib/usrp/common/adf5355.cpp

On Mon, Aug 21, 2017 at 2:54 PM, ed.murphy31@... [ukmicrowaves] <ukmicrowaves@...> wrote:
 

Hello All,


Has anyone tried one of these ? Looks like they will not release the source code which would be useful.


http://www.ebay.co.uk/itm/ADF5355-Module-Touch-Color-Screen-LCD-Sweep-RF-Signal-Source-VCO-Synthesizer-PLL-/272780479899?hash=item3f82fbcd9b:g:B7IAAOSw3ChZerDQ


Ed GM3SBC




ed.murphy31@...
 

Thanks Derek,

I may have a go with the ADF5355 version. 

I have not had a chance to check this chip out but can you tell me if the 10KHz increment is the limit for this device. 

Ed GM3SBC


Andy G4JNT
 

If you look at the data sheet you'll see it is advertised as a fractional N device with 38 bit modulus.   Delving into the data sheet quickly suggests a 24 bit Fractional register and 14 bit auxilliary Fract one.   So programming is a bit more complex thean teh ones I've used to date.

But even a 24 bit Fract value , and assuming a 10MHz Fcomp will give you  0.6Hz on the VCO, and even finer than that when divied down.

''bout time I looked at this chip - are there any (recommended and trustworthy) Ebay suppliers of a PCB without the controller - like there were with the ADF4351 devices.

10kHz step you mention is probably just an artefact of one persons implementation of a controller.

Andy  G4JNT

On 23 August 2017 at 14:34, ed.murphy31@... [ukmicrowaves] <ukmicrowaves@...> wrote:
 

Thanks Derek,


I may have a go with the ADF5355 version. 

I have not had a chance to check this chip out but can you tell me if the 10KHz increment is the limit for this device. 

Ed GM3SBC



Barry Chambers
 

Andy

Both  'ACE and I have purchased good ones from this seller

http://www.ebay.co.uk/itm/ADF5355-phase-locked-loop-RF-output-54M-to-13-6G-Development-Board-PLL-VCO/272655869682?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2057872.m2749.l2649

73

Barry, G8AGN



On 23/08/2017 14:45, Andy Talbot andy.g4jnt@... [ukmicrowaves] wrote:

 
If you look at the data sheet you'll see it is advertised as a fractional N device with 38 bit modulus.   Delving into the data sheet quickly suggests a 24 bit Fractional register and 14 bit auxilliary Fract one.   So programming is a bit more complex thean teh ones I've used to date.

But even a 24 bit Fract value , and assuming a 10MHz Fcomp will give you  0.6Hz on the VCO, and even finer than that when divied down.

''bout time I looked at this chip - are there any (recommended and trustworthy) Ebay suppliers of a PCB without the controller - like there were with the ADF4351 devices.

10kHz step you mention is probably just an artefact of one persons implementation of a controller.

Andy  G4JNT

On 23 August 2017 at 14:34, ed.murphy31@... [ukmicrowaves] <ukmicrowaves@...> wrote:
 

Thanks Derek,


I may have a go with the ADF5355 version. 

I have not had a chance to check this chip out but can you tell me if the 10KHz increment is the limit for this device. 

Ed GM3SBC



Derek Kozel <derek.kozel@...>
 

In the system I work with we avoid the sub hertz tuning and do the final compensation digitally, but it can do it.

Any Fractional N synthesizer is going to produce spurs related to the divided reference (phase detector frequency) and the output frequency. If you are only dealing with narrowband signals then it won't be too much of an issue, but if you put a panadapter style receiver inline you may see some spurs. I would recommend trying to keep it in Integer N mode and moving your IF to compensate the last few MHz. Also on transmit you'll end up with spurs all over the spectrum on the output so having a bandpass filter before the final amplification/antenna would be important.

On Wed, Aug 23, 2017 at 3:45 PM, Andy Talbot andy.g4jnt@... [ukmicrowaves] <ukmicrowaves@...> wrote:
 

If you look at the data sheet you'll see it is advertised as a fractional N device with 38 bit modulus.   Delving into the data sheet quickly suggests a 24 bit Fractional register and 14 bit auxilliary Fract one.   So programming is a bit more complex thean teh ones I've used to date.

But even a 24 bit Fract value , and assuming a 10MHz Fcomp will give you  0.6Hz on the VCO, and even finer than that when divied down.

''bout time I looked at this chip - are there any (recommended and trustworthy) Ebay suppliers of a PCB without the controller - like there were with the ADF4351 devices.

10kHz step you mention is probably just an artefact of one persons implementation of a controller.

Andy  G4JNT

On 23 August 2017 at 14:34, ed.murphy31@... [ukmicrowaves] <ukmicrowaves@...> wrote:
 

Thanks Derek,


I may have a go with the ADF5355 version. 

I have not had a chance to check this chip out but can you tell me if the 10KHz increment is the limit for this device. 

Ed GM3SBC




ed.murphy31@...
 

Thanks Andy,

I have been so busy recently on my Globemaster SDR project that I sadly have missed out on all the good microwave activities going on.

However, I would like to make amends and keep my finger on the pulse. This chip cannot be that bad?

I have a very nice 12 to 24 GHz doubler IC that may be good for driving from the ADF5355. 

OK, if there is anyone around to tackle this it has to be you. I know you are up for it ! 

It would be good to get your professional opinion on just how well it does perform.  

Sounds like a good winter project.

Regards 

Ed GM3SBC


ed.murphy31@...
 

Thanks David,

All understood. 

Ed.
GM3SBC


Neil Smith G4DBN
 

I bought a complete board back in April and it appears to be a real chip.  Ebay item number 122308406711.  Not seen any bare boards for sale yet though.  This board is sort-of-ok, but the lead lengths to the capacitors on the power leads leaves a lot to be desired, so I wouldn't recommend it for anything where noise and spurs are critical.

The calculations for fixed frequencies are simple enough.  I haven't really played with it other than using it to produce test signals on 5.7 and 10GHz, but the frequency step only really matters if you want fast stepping.  It is trivial enough to switch to any frequency if you have a few milliseconds between steps to program all the registers.  The output at 10GHz from this board is rather low, it actually gets better at 12GHz.  I haven't tried fiddling with the terminating inductors yet though

Neil G4DBN


On 23/08/2017 14:45, Andy Talbot andy.g4jnt@... [ukmicrowaves] wrote:
 

''bout time I looked at this chip - are there any (recommended and trustworthy) Ebay suppliers of a PCB without the controller - like there were with the ADF4351 devices.



Andy G4JNT
 

That's the sort of thing - no on board controller, so I can do my own thing.

Trouble with all fully integrated synths is that the internal VCO has to go though a calibration phase, which requires it to go out of lock for a bit.   How oftyen this is done really ought to be under operator contoll - but in the LMX2541 is automatic on the F register being programmed.  Which makes frequency agile systems a real complication


'jnt



On 23 August 2017 at 22:07, Neil neil@... [ukmicrowaves] <ukmicrowaves@...> wrote:
 

I bought a complete board back in April and it appears to be a real chip.  Ebay item number 122308406711.  Not seen any bare boards for sale yet though.  This board is sort-of-ok, but the lead lengths to the capacitors on the power leads leaves a lot to be desired, so I wouldn't recommend it for anything where noise and spurs are critical.

The calculations for fixed frequencies are simple enough.  I haven't really played with it other than using it to produce test signals on 5.7 and 10GHz, but the frequency step only really matters if you want fast stepping.  It is trivial enough to switch to any frequency if you have a few milliseconds between steps to program all the registers.  The output at 10GHz from this board is rather low, it actually gets better at 12GHz.  I haven't tried fiddling with the terminating inductors yet though

Neil G4DBN


On 23/08/2017 14:45, Andy Talbot andy.g4jnt@... [ukmicrowaves] wrote:
 

''bout time I looked at this chip - are there any (recommended and trustworthy) Ebay suppliers of a PCB without the controller - like there were with the ADF4351 devices.




Neil Smith G4DBN
 

I think that you can disable the auto-calibrate on the 5355 if you are not moving out of range of the current VCO and the current selected sub-band (they use four VCOs, each with 256 overlapping bands....

"Write to Register 0 to enact (by default) the VCO automatic calibration, and to choose the appropriate VCO and VCO subband.

Write 1 to the AC1 bit (Bit DB21) to enable the automatic calibration, which is the recommended mode of operation.

Set the AC1 bit (Bit DB21) to 0 to disable the automatic calibration, which leaves the ADF5355 in the same band it was already in when Register 0 is updated.

Disable the automatic calibration only for fixed frequency applications, phase adjust applications, or very small (<10 kHz) frequency jumps.

Toggling automatic calibration (AUTOCAL) is also required when changing frequency. See the Frequency Update Sequence section for more information."

Looks like it is safe to disable the autocal for small changes, but not sure what that last sentence actually means though.  The word "also" seems a bit odd.

Neil G4DBN

On 23/08/2017 22:12, Andy Talbot andy.g4jnt@... [ukmicrowaves] wrote:
 

Trouble with all fully integrated synths is that the internal VCO has to go though a calibration phase, which requires it to go out of lock for a bit.   How oftyen this is done really ought to be under operator contoll - but in the LMX2541 is automatic on the F register being programmed.  Which makes frequency agile systems a real complication



Andy G4JNT
 

Tnx.  Have just ordered one

Andy


On 23 August 2017 at 16:11, Barry Chambers b.chambers@... [ukmicrowaves] <ukmicrowaves@...> wrote:
 

Andy

Both  'ACE and I have purchased good ones from this seller

http://www.ebay.co.uk/itm/ADF5355-phase-locked-loop-RF-output-54M-to-13-6G-Development-Board-PLL-VCO/272655869682?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2057872.m2749.l2649

73

Barry, G8AGN



On 23/08/2017 14:45, Andy Talbot andy.g4jnt@... [ukmicrowaves] wrote:
 
If you look at the data sheet you'll see it is advertised as a fractional N device with 38 bit modulus.   Delving into the data sheet quickly suggests a 24 bit Fractional register and 14 bit auxilliary Fract one.   So programming is a bit more complex thean teh ones I've used to date.

But even a 24 bit Fract value , and assuming a 10MHz Fcomp will give you  0.6Hz on the VCO, and even finer than that when divied down.

''bout time I looked at this chip - are there any (recommended and trustworthy) Ebay suppliers of a PCB without the controller - like there were with the ADF4351 devices.

10kHz step you mention is probably just an artefact of one persons implementation of a controller.

Andy  G4JNT

On 23 August 2017 at 14:34, ed.murphy31@... [ukmicrowaves] <ukmicrowaves@...> wrote:
 

Thanks Derek,


I may have a go with the ADF5355 version. 

I have not had a chance to check this chip out but can you tell me if the 10KHz increment is the limit for this device. 

Ed GM3SBC




Brian Flynn GM8BJF
 

I have been using one of the "green PCBs" with the ADF5355. I agree with Neil the the layout is possibly not quite ideal. It is also on double sided 1.6 mm FR4 so the there will be some losses and impedance mismatches. I think the "black" boards may be a better bet as I believe they are on a thinner substrate. Looking at the markings on the board they look as if they come from the same stable as the "black" ADF4351 boards. They also use a higher reference frequency.(125 MHz as opposed to 26 Mhz) I attach a zipfile of some checks I did on my board. I am using it to build a personal beacon on 24 GHz.

73s
Brian GM8BJF

--


Andy G4JNT
 

Reverting to the original question, having now looked at the data sheet.

The way the two Fractional modulus registers are arranged, it is possible to get up to 38 bits of resolution

So if you used a 10MHz Fcomp, generating at 10GHz that would allow 0.0023Hz steps.
So the resolution is good enough to directly generate WSPR with its 1.464Hz tone shift at 10GHz (although why anyone would ever contemplate WSPR on 10GHz....!)

Andy  G4JNT

On 23 August 2017 at 14:34, ed.murphy31@... [ukmicrowaves] <ukmicrowaves@...> wrote:
 

Thanks Derek,


I may have a go with the ADF5355 version. 

I have not had a chance to check this chip out but can you tell me if the 10KHz increment is the limit for this device. 

Ed GM3SBC



Neil Smith G4DBN
 

Just checking the AD config tool which I use to find the reg values.  If you set the step size to an arbitrary 100 milliHertz at 10.368MHz, the error per tone from the calculated register values is under 100 microHertz when using 100MHz ref and 50MHz PFD.

Neil G4DBN




On 24/08/2017 13:37, Andy Talbot andy.g4jnt@... [ukmicrowaves] wrote:
 
Reverting to the original question, having now looked at the data sheet.

The way the two Fractional modulus registers are arranged, it is possible to get up to 38 bits of resolution

So if you used a 10MHz Fcomp, generating at 10GHz that would allow 0.0023Hz steps.
So the resolution is good enough to directly generate WSPR with its 1.464Hz tone shift at 10GHz (although why anyone would ever contemplate WSPR on 10GHz....!)

Andy  G4JNT

On 23 August 2017 at 14:34, ed.murphy31@... [ukmicrowaves] <ukmicrowaves@...> wrote:
 

Thanks Derek,


I may have a go with the ADF5355 version. 

I have not had a chance to check this chip out but can you tell me if the 10KHz increment is the limit for this device. 

Ed GM3SBC



DougF VK4OE
 

Greetings to all readers!

Many folk will have become familiar in recent years with the 'ZLPLL' 4GHz PLL board designed and sold by Wayne, ZL2BKC.  They are rather good.....provided, of course, that the reference signal is clean.

Well, Wayne has also been on the scent of the ADF5355 trail and has produced his first (small) batch of what he now calls the '13G ZLPLL'.  Different to the earlier edition, the new design requires an off-board reference, but like the earlier version it's fully programmable and it works well way up past 13 GHz.  For a digitally generated microwave PLL at these sorts of frequencies, its PN seems

To generate a 13+ GHz output signal the chip uses a doubler that's internal to the chip so the PLL 'fundamental' is also present in the output spectrum, the desired output needing filtering if the F/2 component is a problem.

I have one of Wayne's early batch of 13G ZLPLLs here with me in the UK and I propose, for possible comparison purposes, to have it with me at the Crawley RT next month.  Or I may be able to show it off to some other folk between now and then.....  (I do also have with me a very good quality 10 MHz OCXO that is appropriately interfaced to the 13G ZLPLL, so it can be a plug-and-play comparison.)

It should be interesting to compare the output signal from Wayne's design, one that uses a reliably genuine chip set, with the Chinese ones that some folk are obviously acquiring!

Cheers and 73,

--Doug Friend, M/VK4OE


Mary Echternacht <scootiebug02@...>
 

Can anyone give me the email address for Wayne? AC0PY Richard USA


On ‎Thursday‎, ‎August‎ ‎24‎, ‎2017‎ ‎03‎:‎19‎:‎48‎ ‎PM‎ ‎EDT, DougF uwaves@... [ukmicrowaves] wrote:


 

Greetings to all readers!

Many folk will have become familiar in recent years with the 'ZLPLL'
4GHz PLL board designed and sold by Wayne, ZL2BKC.  They are rather
good.....provided, of course, that the reference signal is clean.

Well, Wayne has also been on the scent of the ADF5355 trail and has
produced his first (small) batch of what he now calls the '13G ZLPLL'. 
Different to the earlier edition, the new design requires an off-board
reference, but like the earlier version it's fully programmable and it
works well way up past 13 GHz.  For a digitally generated microwave PLL
at these sorts of frequencies, its PN seems

To generate a 13+ GHz output signal the chip uses a doubler that's
internal to the chip so the PLL 'fundamental' is also present in the
output spectrum, the desired output needing filtering if the F/2
component is a problem.

I have one of Wayne's early batch of 13G ZLPLLs here with me in the UK
and I propose, for possible comparison purposes, to have it with me at
the Crawley RT next month.  Or I may be able to show it off to some
other folk between now and then.....  (I do also have with me a very
good quality 10 MHz OCXO that is appropriately interfaced to the 13G
ZLPLL, so it can be a plug-and-play comparison.)

It should be interesting to compare the output signal from Wayne's
design, one that uses a reliably genuine chip set, with the Chinese ones
that some folk are obviously acquiring!

Cheers and 73,

--Doug Friend, M/VK4OE


Bob Lockley VK6KW
 

Hi Richard,

Try:

https://zl2bkc.com/store/

 

73,

Bob VK6KW

 

From: ukmicrowaves@... [mailto:ukmicrowaves@...]
Sent: Friday, August 25, 2017 6:17 AM
To: DougF uwaves@... [ukmicrowaves]
Subject: Re: [ukmicrowaves] ADF5355 Signal Source

 

 

Can anyone give me the email address for Wayne? AC0PY Richard USA

 

 

On ‎Thursday‎, ‎August‎ ‎24‎, ‎2017‎ ‎03‎:‎19‎:‎48‎ ‎PM‎ ‎EDT, DougF uwaves@... [ukmicrowaves] <ukmicrowaves@...> wrote:

 

 

 

Greetings to all readers!

Many folk will have become familiar in recent years with the 'ZLPLL'
4GHz PLL board designed and sold by Wayne, ZL2BKC.  They are rather
good.....provided, of course, that the reference signal is clean.

Well, Wayne has also been on the scent of the ADF5355 trail and has
produced his first (small) batch of what he now calls the '13G ZLPLL'. 
Different to the earlier edition, the new design requires an off-board
reference, but like the earlier version it's fully programmable and it
works well way up past 13 GHz.  For a digitally generated microwave PLL
at these sorts of frequencies, its PN seems

To generate a 13+ GHz output signal the chip uses a doubler that's
internal to the chip so the PLL 'fundamental' is also present in the
output spectrum, the desired output needing filtering if the F/2
component is a problem.

I have one of Wayne's early batch of 13G ZLPLLs here with me in the UK
and I propose, for possible comparison purposes, to have it with me at
the Crawley RT next month.  Or I may be able to show it off to some
other folk between now and then.....  (I do also have with me a very
good quality 10 MHz OCXO that is appropriately interfaced to the 13G
ZLPLL, so it can be a plug-and-play comparison.)

It should be interesting to compare the output signal from Wayne's
design, one that uses a reliably genuine chip set, with the Chinese ones
that some folk are obviously acquiring!

Cheers and 73,

--Doug Friend, M/VK4OE


Mary Echternacht <scootiebug02@...>
 

Thank you, I just bought one of his boards. 73 AC0PY Richard



From: 'Bob VK6KW' rlockley@... [ukmicrowaves]
To: "ukmicrowaves@..."
Sent: ‎Saturday‎, ‎August‎ ‎26‎, ‎2017‎ ‎02‎:‎32‎:‎30‎ ‎PM‎ ‎EDT
Subject: RE: [ukmicrowaves] ADF5355 Signal Source

 

Hi Richard,

Try:

https://zl2bkc.com/store/

 

73,

Bob VK6KW

 

From: ukmicrowaves@... [mailto:ukmicrowaves@...]
Sent: Friday, August 25, 2017 6:17 AM
To: DougF uwaves@... [ukmicrowaves]
Subject: Re: [ukmicrowaves] ADF5355 Signal Source

 

 

Can anyone give me the email address for Wayne? AC0PY Richard USA

 

 

On ‎Thursday‎, ‎August‎ ‎24‎, ‎2017‎ ‎03‎:‎19‎:‎48‎ ‎PM‎ ‎EDT, DougF uwaves@... [ukmicrowaves] <ukmicrowaves@...> wrote:

 

 

 

Greetings to all readers!

Many folk will have become familiar in recent years with the 'ZLPLL'
4GHz PLL board designed and sold by Wayne, ZL2BKC.  They are rather
good.....provided, of course, that the reference signal is clean.

Well, Wayne has also been on the scent of the ADF5355 trail and has
produced his first (small) batch of what he now calls the '13G ZLPLL'. 
Different to the earlier edition, the new design requires an off-board
reference, but like the earlier version it's fully programmable and it
works well way up past 13 GHz.  For a digitally generated microwave PLL
at these sorts of frequencies, its PN seems

To generate a 13+ GHz output signal the chip uses a doubler that's
internal to the chip so the PLL 'fundamental' is also present in the
output spectrum, the desired output needing filtering if the F/2
component is a problem.

I have one of Wayne's early batch of 13G ZLPLLs here with me in the UK
and I propose, for possible comparison purposes, to have it with me at
the Crawley RT next month.  Or I may be able to show it off to some
other folk between now and then.....  (I do also have with me a very
good quality 10 MHz OCXO that is appropriately interfaced to the 13G
ZLPLL, so it can be a plug-and-play comparison.)

It should be interesting to compare the output signal from Wayne's
design, one that uses a reliably genuine chip set, with the Chinese ones
that some folk are obviously acquiring!

Cheers and 73,

--Doug Friend, M/VK4OE