|
How many people have spectrum analyzers or access to one?
I have read about spurs created by chips like the AD9850. How does a typical ham checkout one of these devices?
Back in the OLD DAYS (1960's) we used frequency selective voltmeters. They were very much similar to today's direct conversion receiver.
I have an article from the ARRL on building spectrum analyzer that works with an ordinary osciliscope. And I does require a super good scope because what it has to look at is just low frequency information.
I started building one many years ago.
Bob Macklin
K5MYJ
Seattle, Wa
|
Steve,
I worked with the HP gear back in the 70's an 80's. So I am familiar with
it.
A few years ago I joined the HP Test Equipment group and saw the problems
they were having with the digital stuff from the 70's and 80's.
So I started looking at the design of that stuff. Most used bipolar PALS and
PROM devices. These all have a know failure mode and are hard to replace
these days. I did have a programmer for programming them but I sold it to
someone that was restoring Drake radios that also used that technology.
The older analog HP stuff is mostly repairable. The coils and variable
capacitors don't go bad. And most of the tubes and early transistors are
replaceable.
I don't know how much of a problem spurs from an AD9850 might be. I also
don't know about the Silicon Industries devices.
I found the AD9850 modules to be a convenient device to work with. And good
filters my solve some problems.
I've been a ham since 1957 and we never had more than a GC receiver to find
spurs. I still have 3 GC receivers but that takes a lot of time.
Bob Macklin
K5MYJ
Seattle, Wa.
"Real Radios Glow In The Dark"
toggle quoted message
Show quoted text
----- Original Message ----- From: "STEVE" <ratzlaffsteve@...> To: "Bob Macklin" <macklinbob@...> Sent: Thursday, August 08, 2019 8:21 PM Subject: Re: [qrp-tech] Spectrum Analyzers? Hi Bob,
I have a spectrum analyzer; almost always have had one, at least for about
past 20 years. I'm retired from an electronics R&D facility where I used
such stuff on the bench, now use it on my hobby bench. Used to have HP and
Tektronix gear but it's old and eventually quits working. I'm not good at
repairing test equipment so I gradually buy modern gear. I bought this
Siglent SA about a year ago when the old HP SA died. Years ago I built
that W7ZOI SA from that article too, just for fun since I had access to
good SAs at work. I'm spoiled I guess as the homebrew one couldn't do
stuff I was used to doing with commercial SAs.
If you have a decent modern rig with a general coverage receiver in it
(almost all do now) you could use that to hunt for spurs from the cheap
AD9850 DDS and get a relative idea of their strength just from listening
to the level of the spur. The W7ZOI SA would not really be suitable for
seeing such spurs, they're pretty far down though I suppose you could put
a preamp ahead of the SA input. I still have a good HP frequency selective
voltmeter, the HP 3586C. It has a built in tracking generator; very
accurate level meter, 1 Hz tuning, etc. A very fine piece of test gear and
mine hasn't died yet. :)
73,
Steve AA7U in AZ
On 8/8/2019 8:07 PM, Bob Macklin wrote:
How many people have spectrum analyzers or access to one?
I have read about spurs created by chips like the AD9850. How does a
typical ham checkout one of these devices?
Back in the OLD DAYS (1960's) we used frequency selective voltmeters.
They were very much similar to today's direct conversion receiver.
I have an article from the ARRL on building spectrum analyzer that works
with an ordinary osciliscope. And I does require a super good scope
because what it has to look at is just low frequency information.
I started building one many years ago.
Bob Macklin
K5MYJ
Seattle, Wa
|
Software defined receivers are a useful tool. Some of them can see a
wide area of the spectrum at once: about 2 MHz for RTL-type receivers
and nearly 10 MHz for Airspy or SDRplay. The dynamic range of the RTL
receivers is limited and only useful for finding egregious problems,
the higher resolution Airspy and SDRplay are good enough for FCC
compliance testing. You can get software that turns them into a slow
spectrum analyzer by repeatedly retuning the receiver and grabbing
another sample.
SDRs that use a QSD (Softrock, RS-HFIQ, KX2 and KX3, etc) have a much
more limited bandwidth; up to about 150 KHz at a time if you use an
audio card with 192 KHz sampling. They are less useful for broadband
testing, but their excellent low noise performance makes them a great
tool for exploring problems that are near the carrier.
SDRs that use direct sampling (an RF-rate ADC) could in theory do a
spectrum analysis of the entire HF spectrum at once. But it would be a
challenge to build a computer fast enough to do that. And some
incorporate switched analog filters on the input that would make that
less useful.
The spectrum scope features in some of the fancy big name transceivers
can also be useful. The resolution of the display that is built into
the radio is likely to be too low to be useful, but some rigs also let
you connect an external monitor that lets you see more detail about
the signals.
toggle quoted message
Show quoted text
On Thu, Aug 8, 2019 at 11:07 PM Bob Macklin <macklinbob@...> wrote:
How many people have spectrum analyzers or access to one?
I have read about spurs created by chips like the AD9850. How does a typical ham checkout one of these devices?
Back in the OLD DAYS (1960's) we used frequency selective voltmeters. They were very much similar to today's direct conversion receiver.
I have an article from the ARRL on building spectrum analyzer that works with an ordinary osciliscope. And I does require a super good scope because what it has to look at is just low frequency information.
I started building one many years ago.
Bob Macklin
K5MYJ
Seattle, Wa
|
I have tried searching on "AD9850 spur problems" and have not found much of
interest.
How bad is the spur problem with a DDS. Can bandpass filters eliminate them
?
Bob Macklin
K5MYJ
Seattle, Wa.
"Real Radios Glow In The Dark"
toggle quoted message
Show quoted text
----- Original Message ----- From: "Shirley Dulcey KE1L" <mark@...> To: < [email protected]> Sent: Thursday, August 08, 2019 9:28 PM Subject: Re: [qrp-tech] Spectrum Analyzers? Software defined receivers are a useful tool. Some of them can see a
wide area of the spectrum at once: about 2 MHz for RTL-type receivers
and nearly 10 MHz for Airspy or SDRplay. The dynamic range of the RTL
receivers is limited and only useful for finding egregious problems,
the higher resolution Airspy and SDRplay are good enough for FCC
compliance testing. You can get software that turns them into a slow
spectrum analyzer by repeatedly retuning the receiver and grabbing
another sample.
SDRs that use a QSD (Softrock, RS-HFIQ, KX2 and KX3, etc) have a much
more limited bandwidth; up to about 150 KHz at a time if you use an
audio card with 192 KHz sampling. They are less useful for broadband
testing, but their excellent low noise performance makes them a great
tool for exploring problems that are near the carrier.
SDRs that use direct sampling (an RF-rate ADC) could in theory do a
spectrum analysis of the entire HF spectrum at once. But it would be a
challenge to build a computer fast enough to do that. And some
incorporate switched analog filters on the input that would make that
less useful.
The spectrum scope features in some of the fancy big name transceivers
can also be useful. The resolution of the display that is built into
the radio is likely to be too low to be useful, but some rigs also let
you connect an external monitor that lets you see more detail about
the signals.
On Thu, Aug 8, 2019 at 11:07 PM Bob Macklin <macklinbob@...>
wrote:
How many people have spectrum analyzers or access to one?
I have read about spurs created by chips like the AD9850. How does a
typical ham checkout one of these devices?
Back in the OLD DAYS (1960's) we used frequency selective voltmeters.
They were very much similar to today's direct conversion receiver.
I have an article from the ARRL on building spectrum analyzer that works
with an ordinary osciliscope. And I does require a super good scope
because what it has to look at is just low frequency information.
I started building one many years ago.
Bob Macklin
K5MYJ
Seattle, Wa
|
BTW:
My current interest is to use a DDS to replace the LO/VFO in a NorCal40 and
a NorCal Sierra.
Bob Macklin
K5MYJ
Seattle, Wa.
"Real Radios Glow In The Dark"
toggle quoted message
Show quoted text
----- Original Message ----- From: "Shirley Dulcey KE1L" <mark@...> To: < [email protected]> Sent: Thursday, August 08, 2019 9:28 PM Subject: Re: [qrp-tech] Spectrum Analyzers? Software defined receivers are a useful tool. Some of them can see a
wide area of the spectrum at once: about 2 MHz for RTL-type receivers
and nearly 10 MHz for Airspy or SDRplay. The dynamic range of the RTL
receivers is limited and only useful for finding egregious problems,
the higher resolution Airspy and SDRplay are good enough for FCC
compliance testing. You can get software that turns them into a slow
spectrum analyzer by repeatedly retuning the receiver and grabbing
another sample.
SDRs that use a QSD (Softrock, RS-HFIQ, KX2 and KX3, etc) have a much
more limited bandwidth; up to about 150 KHz at a time if you use an
audio card with 192 KHz sampling. They are less useful for broadband
testing, but their excellent low noise performance makes them a great
tool for exploring problems that are near the carrier.
SDRs that use direct sampling (an RF-rate ADC) could in theory do a
spectrum analysis of the entire HF spectrum at once. But it would be a
challenge to build a computer fast enough to do that. And some
incorporate switched analog filters on the input that would make that
less useful.
The spectrum scope features in some of the fancy big name transceivers
can also be useful. The resolution of the display that is built into
the radio is likely to be too low to be useful, but some rigs also let
you connect an external monitor that lets you see more detail about
the signals.
On Thu, Aug 8, 2019 at 11:07 PM Bob Macklin <macklinbob@...>
wrote:
How many people have spectrum analyzers or access to one?
I have read about spurs created by chips like the AD9850. How does a
typical ham checkout one of these devices?
Back in the OLD DAYS (1960's) we used frequency selective voltmeters.
They were very much similar to today's direct conversion receiver.
I have an article from the ARRL on building spectrum analyzer that works
with an ordinary osciliscope. And I does require a super good scope
because what it has to look at is just low frequency information.
I started building one many years ago.
Bob Macklin
K5MYJ
Seattle, Wa
|
Recently using VNWA:
in spectrum analyzer mode measuring RF amplifier parameters including
noise figure and OIP3.
Some results in:
Required an external preamp for Noise figure measurements. An
inexpensive way to have a quality antenna analyzer, two port VNA, and an
OK spectrum analyzer in one little box.
John KN5L
|
I have both HP8658B and Rigol DSA815t.
Testing the 9850 output is easy, pump it into the SA and look at the spectra.
However... you will find spurs. The 9850 is a very old part and the number
of bits used for the sine conversion are not enough to insure the waveform
is without jumps or skips and as a result there will be odd outputs that are
typically about 60db down from the set frequency. That parameter is in the
data sheet as SFDR (spur free dynamic range) and varies with frequency set.
Better parts like the 9851 and later are better for this.
FYI the AD9850 is easily 20++ years old! The only reason its on people
radar is it suddenly got cheap about 5-6 years ago when a source in
china sold off a large number of them because the reconstruction filter
was incorrectly done (greater SFDR). Plus the Arduino make it easier
to use.
Note filters unless very narrow (kilohertz wide) will not take out the
spurs as they are often in band.
The statement that you found little suggests your google foo is broke.
Allison
--
Please reply on the group, no private emails we as a group get to share info that way.
|
While looking at the DG8SAQ-VNWA3 on ebay, I came across a low cost device called the nano VNA for around $64.00. It appears to have decent features for the price in a small package including display. Anyone use this? Just search for Nano VNA on eBay
Thanks,
-Steve K1RF
toggle quoted message
Show quoted text
-----Original Message----- From: John KN5L Sent: Friday, August 09, 2019 5:55 AM To: [email protected]Subject: Re: [qrp-tech] Spectrum Analyzers? Recently using VNWA: in spectrum analyzer mode measuring RF amplifier parameters including noise figure and OIP3. Some results in: Required an external preamp for Noise figure measurements. An inexpensive way to have a quality antenna analyzer, two port VNA, and an OK spectrum analyzer in one little box. John KN5L
|
I use both the VNWA 3 from DG8SAQ and love my old 400Mhz Marconi 2382.. All good for xtal and ladder filter design..
Have uploaded various jpg's to the Yahoo DG8SAQ group showing the filter responses of SSB/CW 10 pole tscebychev 4.915Mhz filters...
73 Ken G3UDA
|
What does "LOW COST" mean?
Bob Macklin
K5MYJ
Seattle, Wa.
"Real Radios Glow In The Dark"
toggle quoted message
Show quoted text
----- Original Message ----- From: "John KN5L" <john@...> To: < [email protected]> Sent: Friday, August 09, 2019 2:55 AM Subject: Re: [qrp-tech] Spectrum Analyzers? Recently using VNWA:
in spectrum analyzer mode measuring RF amplifier parameters including
noise figure and OIP3.
Some results in:
Required an external preamp for Noise figure measurements. An
inexpensive way to have a quality antenna analyzer, two port VNA, and an
OK spectrum analyzer in one little box.
John KN5L
|
What should I use to GOOGLE for AD9850 spurs?
Bob Macklin
K5MYJ
Seattle, Wa.
"Real Radios Glow In The Dark"
toggle quoted message
Show quoted text
----- Original Message ----- From: "ajparent1/kb1gmx" <kb1gmx@...> To: < [email protected]> Sent: Friday, August 09, 2019 3:49 AM Subject: Re: [qrp-tech] Spectrum Analyzers? I have both HP8658B and Rigol DSA815t. Testing the 9850 output is easy, pump it into the SA and look at the spectra. However... you will find spurs. The 9850 is a very old part and the number of bits used for the sine conversion are not enough to insure the waveform is without jumps or skips and as a result there will be odd outputs that are typically about 60db down from the set frequency. That parameter is in the data sheet as SFDR (spur free dynamic range) and varies with frequency set. Better parts like the 9851 and later are better for this. FYI the AD9850 is easily 20++ years old! The only reason its on people radar is it suddenly got cheap about 5-6 years ago when a source in china sold off a large number of them because the reconstruction filter was incorrectly done (greater SFDR). Plus the Arduino make it easier to use. Note filters unless very narrow (kilohertz wide) will not take out the spurs as they are often in band. The statement that you found little suggests your google foo is broke. Allison -- Please reply on the group, no private emails we as a group get to share info that way.
|
Per sdr-kits web page:
VNWA 3EC (recommended for S11 and S21 with single sweep) = USD $525.78
John KN5L
toggle quoted message
Show quoted text
On 8/9/19 8:47 AM, Bob Macklin wrote: What does "LOW COST" mean?
Bob Macklin
K5MYJ
Seattle, Wa.
"Real Radios Glow In The Dark"
----- Original Message -----
From: "John KN5L" <john@...>
To: <[email protected]>
Sent: Friday, August 09, 2019 2:55 AM
Subject: Re: [qrp-tech] Spectrum Analyzers?
Recently using VNWA:
in spectrum analyzer mode measuring RF amplifier parameters including
noise figure and OIP3.
Some results in:
Required an external preamp for Noise figure measurements. An
inexpensive way to have a quality antenna analyzer, two port VNA, and an
OK spectrum analyzer in one little box.
John KN5L
|
A spectrum analyzer would be a great instrument for a serious homebrewer,
but they're expensive - like $1,500 or so although you can go higher or
maybe lower with some effort. I almost jumped on a used one for about $500
not too long ago but someone beat me to it.
If I were going to try to homebrew one, I'd take a serious look at Ashar
Farhan's SPECAN. I don't think it would be too expensive, but the homebrew
shielding requirements might be a challenge for my building skills. A web
search will show results that various hams have achieved.
As for spurs - you want to use it for the VFO in one of your QRP kits? I'd
say go for it.
Those spurs aren't necessarily going to be harmonically related and may be
both above and below the fundamental, so an LPF isn't going to eliminate
many of the low and/or close in ones. But they aren't typically too
strong.
In the manual for AmQRP's DDS-60 (AD9851 based), there were SA shots from
K8IQY showing possibly one spur only 40 dB down and the rest looked like
about 50 dB down or more. These shots were taken at 10, 30 and 60 MHz.
73-
Nick, WA5BDU
toggle quoted message
Show quoted text
On Thu, Aug 8, 2019 at 10:07 PM Bob Macklin <macklinbob@...> wrote: How many people have spectrum analyzers or access to one?
I have read about spurs created by chips like the AD9850. How does a
typical ham checkout one of these devices?
|
Bob,
You should try one of these(under $4USD):
3 outputs so you could do LO/VFO and BFO
73,
Gary
WB6OGD
toggle quoted message
Show quoted text
On 8/8/2019 9:40 PM, Bob Macklin wrote: BTW:
My current interest is to use a DDS to replace the LO/VFO in a NorCal40 and
a NorCal Sierra.
Bob Macklin
K5MYJ
Seattle, Wa.
"Real Radios Glow In The Dark"
----- Original Message -----
From: "Shirley Dulcey KE1L" <mark@...>
To: <[email protected]>
Sent: Thursday, August 08, 2019 9:28 PM
Subject: Re: [qrp-tech] Spectrum Analyzers?
Software defined receivers are a useful tool. Some of them can see a
wide area of the spectrum at once: about 2 MHz for RTL-type receivers
and nearly 10 MHz for Airspy or SDRplay. The dynamic range of the RTL
receivers is limited and only useful for finding egregious problems,
the higher resolution Airspy and SDRplay are good enough for FCC
compliance testing. You can get software that turns them into a slow
spectrum analyzer by repeatedly retuning the receiver and grabbing
another sample.
SDRs that use a QSD (Softrock, RS-HFIQ, KX2 and KX3, etc) have a much
more limited bandwidth; up to about 150 KHz at a time if you use an
audio card with 192 KHz sampling. They are less useful for broadband
testing, but their excellent low noise performance makes them a great
tool for exploring problems that are near the carrier.
SDRs that use direct sampling (an RF-rate ADC) could in theory do a
spectrum analysis of the entire HF spectrum at once. But it would be a
challenge to build a computer fast enough to do that. And some
incorporate switched analog filters on the input that would make that
less useful.
The spectrum scope features in some of the fancy big name transceivers
can also be useful. The resolution of the display that is built into
the radio is likely to be too low to be useful, but some rigs also let
you connect an external monitor that lets you see more detail about
the signals.
On Thu, Aug 8, 2019 at 11:07 PM Bob Macklin <macklinbob@...>
wrote:
How many people have spectrum analyzers or access to one?
I have read about spurs created by chips like the AD9850. How does a
typical ham checkout one of these devices?
Back in the OLD DAYS (1960's) we used frequency selective voltmeters.
They were very much similar to today's direct conversion receiver.
I have an article from the ARRL on building spectrum analyzer that works
with an ordinary osciliscope. And I does require a super good scope
because what it has to look at is just low frequency information.
I started building one many years ago.
Bob Macklin
K5MYJ
Seattle, Wa
|
That's beyond my budget!
Bob Macklin
K5MYJ
Seattle, Wa.
"Real Radios Glow In The Dark"
toggle quoted message
Show quoted text
----- Original Message ----- From: "John KN5L" <john@...> To: < [email protected]> Sent: Friday, August 09, 2019 7:03 AM Subject: Re: [qrp-tech] Spectrum Analyzers? Per sdr-kits web page:
VNWA 3EC (recommended for S11 and S21 with single sweep) = USD $525.78
John KN5L
On 8/9/19 8:47 AM, Bob Macklin wrote:
What does "LOW COST" mean?
Bob Macklin
K5MYJ
Seattle, Wa.
"Real Radios Glow In The Dark"
----- Original Message -----
From: "John KN5L" <john@...>
To: <[email protected]>
Sent: Friday, August 09, 2019 2:55 AM
Subject: Re: [qrp-tech] Spectrum Analyzers?
Recently using VNWA:
in spectrum analyzer mode measuring RF amplifier parameters including
noise figure and OIP3.
Some results in:
Required an external preamp for Noise figure measurements. An
inexpensive way to have a quality antenna analyzer, two port VNA, and an
OK spectrum analyzer in one little box.
John KN5L
|
I have two of the QRP Labs SI5351 modules. One is the Arduino shield module.
I've been thinking of assembling it.
Bob Macklin
K5MYJ
Seattle, Wa.
"Real Radios Glow In The Dark"
toggle quoted message
Show quoted text
----- Original Message ----- From: "wb6ogd" <garywinblad@...> To: < [email protected]> Sent: Friday, August 09, 2019 7:29 AM Subject: Re: [qrp-tech] Spectrum Analyzers? Bob,
You should try one of these(under $4USD):
3 outputs so you could do LO/VFO and BFO
73,
Gary
WB6OGD
On 8/8/2019 9:40 PM, Bob Macklin wrote:
BTW:
My current interest is to use a DDS to replace the LO/VFO in a NorCal40
and
a NorCal Sierra.
Bob Macklin
K5MYJ
Seattle, Wa.
"Real Radios Glow In The Dark"
----- Original Message -----
From: "Shirley Dulcey KE1L" <mark@...>
To: <[email protected]>
Sent: Thursday, August 08, 2019 9:28 PM
Subject: Re: [qrp-tech] Spectrum Analyzers?
Software defined receivers are a useful tool. Some of them can see a
wide area of the spectrum at once: about 2 MHz for RTL-type receivers
and nearly 10 MHz for Airspy or SDRplay. The dynamic range of the RTL
receivers is limited and only useful for finding egregious problems,
the higher resolution Airspy and SDRplay are good enough for FCC
compliance testing. You can get software that turns them into a slow
spectrum analyzer by repeatedly retuning the receiver and grabbing
another sample.
SDRs that use a QSD (Softrock, RS-HFIQ, KX2 and KX3, etc) have a much
more limited bandwidth; up to about 150 KHz at a time if you use an
audio card with 192 KHz sampling. They are less useful for broadband
testing, but their excellent low noise performance makes them a great
tool for exploring problems that are near the carrier.
SDRs that use direct sampling (an RF-rate ADC) could in theory do a
spectrum analysis of the entire HF spectrum at once. But it would be a
challenge to build a computer fast enough to do that. And some
incorporate switched analog filters on the input that would make that
less useful.
The spectrum scope features in some of the fancy big name transceivers
can also be useful. The resolution of the display that is built into
the radio is likely to be too low to be useful, but some rigs also let
you connect an external monitor that lets you see more detail about
the signals.
On Thu, Aug 8, 2019 at 11:07 PM Bob Macklin <macklinbob@...>
wrote:
How many people have spectrum analyzers or access to one?
I have read about spurs created by chips like the AD9850. How does a
typical ham checkout one of these devices?
Back in the OLD DAYS (1960's) we used frequency selective voltmeters.
They were very much similar to today's direct conversion receiver.
I have an article from the ARRL on building spectrum analyzer that
works
with an ordinary osciliscope. And I does require a super good scope
because what it has to look at is just low frequency information.
I started building one many years ago.
Bob Macklin
K5MYJ
Seattle, Wa
|
For some reason, cannot find this topic online to check whether anybody has noted that there is some pretty functional spectrum software available for sdrplay v1,2. There are some problems with it, but I was able to use it to test my ubitx for compliance. There is also a noise source that is recommended that is <$15 on ebay.
73
Mark
AJ6CU
|
Nick,
The spurs from 9850 are not directly related to harmonics or frequency settings
( they are related but the math will make your eyes bleed).
Generally they are from D to A artifacts like each step is not monotonic to the next,
or the phase accumulator has reached a frequency where all the D/A bits are not
usable and the steps needed to approximate a sine wave are few. The D/A for
the 9850 is 10 bits (1024 steps) and at some frequencies it maybe as low as 6bits
or 64 steps. For the later an approximation of a sine wave is going to be looking
more like a rough set of unequal steps. That is the source of the limits to SFDR.
Chips like the SI570 and SI5351 approach the frequency generation differently
and are generally cleaner if one ignore the square wave output (harmonics).
The harmonics are easily filtered.
In general for modern radio projects use AD9851 or newer (and more expensive)
parts if SFDR is an issue (high performance radios) or SI5351 or better yet si570.
The Si5351 is generally good enough for a wide range of radio projects that need
one to 3 oscillator outputs. If your doing a high performance radio then SI570 or the
much later exotic AD DDS parts are likely the cup of tea.
Allison
--
Please reply on the group, no private emails we as a group get to share info that way.
|
As to gear to measure the output of a DDS or other RF source.
DVB-t with upconverter, generally cheap and with care and step attenuator
measuring to -60dBc is not difficult.
If you need more than that a downconverter from signal to audio base band
and using spectran or other PC sound card spectrum analysis software
is also a super cheap route.
Cheap is under 50-60$US.
Allison
--
Please reply on the group, no private emails we as a group get to share info that way.
|
The AD9850 modules seem to be the most common.
The AD9850 makes square waves as well as sine waves. Are the square waves
OK. The SI570/SI5351 only make square waves.
The way I am reading this the AD9850 problems lay in the D/A.
I can work with the square waves. I'm not trying to do "DC TO DAYLIGHT".
Just a narrow range for a VFO. Maybe a hundred KHz. Maybe 350KHZ. 20M or 15M
LO.
Bob Macklin
K5MYJ
Seattle, Wa.
"Real Radios Glow In The Dark"
toggle quoted message
Show quoted text
----- Original Message ----- From: "ajparent1/kb1gmx" <kb1gmx@...> To: < [email protected]> Sent: Friday, August 09, 2019 8:58 AM Subject: Re: [qrp-tech] Spectrum Analyzers? Nick, The spurs from 9850 are not directly related to harmonics or frequency settings ( they are related but the math will make your eyes bleed). Generally they are from D to A artifacts like each step is not monotonic to the next, or the phase accumulator has reached a frequency where all the D/A bits are not usable and the steps needed to approximate a sine wave are few. The D/A for the 9850 is 10 bits (1024 steps) and at some frequencies it maybe as low as 6bits or 64 steps. For the later an approximation of a sine wave is going to be looking more like a rough set of unequal steps. That is the source of the limits to SFDR. Chips like the SI570 and SI5351 approach the frequency generation differently and are generally cleaner if one ignore the square wave output (harmonics). The harmonics are easily filtered. In general for modern radio projects use AD9851 or newer (and more expensive) parts if SFDR is an issue (high performance radios) or SI5351 or better yet si570. The Si5351 is generally good enough for a wide range of radio projects that need one to 3 oscillator outputs. If your doing a high performance radio then SI570 or the much later exotic AD DDS parts are likely the cup of tea. Allison -- Please reply on the group, no private emails we as a group get to share info that way.
|
Shirley Dulcey KE1L
Nick Kennedy
wb6ogd