Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

donate to eham
   Home   Help Search  
Pages: [1] 2 3 Next   Go Down
  Print  
Author Topic: noise from DDS VFO  (Read 10552 times)
AJ4SN
Member

Posts: 67




Ignore
« on: June 11, 2017, 09:04:47 AM »

I just built a DDS VFO to replace the analog VFO in my homebrew radio. It's pretty noisy. With a dummy load connected to the receiver, there are numerous buzzes and tones present. The DDS VFO is built in a separate shielded box. I am using the VFO design of AD7C that incorporates an Arduino microcontroller and an AD9850 module. Anyone have suggestions for improving the noise performance of the VFO?

Thanks,

Stan
aj4sn
Logged
AC7CW
Member

Posts: 961




Ignore
« Reply #1 on: June 11, 2017, 03:26:46 PM »

Perhaps adding a lpf between the vfo and the radio would help? Ground noise is a possibility, could play with the wiring between the vfo and the radio and see if the noise changes. Overloading or mismatching the input stage in the radio is something to investigate maybe....
Logged

Novice 1958, 20WPM Extra now... (and get off my lawn)
KL7CW
Member

Posts: 255




Ignore
« Reply #2 on: June 11, 2017, 03:52:29 PM »

It is not at all unusual to hear various DDS "noises" on a receiver connected to a dummy load.  Fortunately often these signals are below the band noise especially on the lower HF bands.  So if a birdie is say an S2 on 80 meters and the band noise is S3...usually not a problem.  If the reverse is true you need to do some trouble shooting.  So hook up an antenna and see if band noise obscures the birdie.  A BPF or LPF for the VFO frequency may help.  Double check your cables and shielding.  You could even try running your VFO off of a separate (battery) supply.  Perhaps others can give you some more ideas to try.
          Rick  KL7CW
Logged
WW7KE
Member

Posts: 576




Ignore
« Reply #3 on: June 11, 2017, 03:58:26 PM »

Perhaps adding a lpf between the vfo and the radio would help?

Many switching mixers require square waves to drive them, so an LPF won't work in those cases.
Logged

He speaks fluent PSK31...  One QSO with him earns you 5BDXCC...  His Wouff Hong has two Wouffs... Hiram Percy Maxim called HIM "The Old Man..."  He is... The Most Interesting Ham In The World!
VK3YE
Member

Posts: 227


WWW

Ignore
« Reply #4 on: June 11, 2017, 04:53:39 PM »

It's important to work out how the noise is getting in.  

Is it getting through the RF output from the VFO, through badly decoupled power connections or direct pick-up?  

I discuss this and how I sorted my problem out at

https://www.youtube.com/watch?v=zOGzQDwFOag

Even if your home location has a lot of RF noise that overcomes anything internally generated it's still good to fix the problem.  

Then you can operate portable from remote locations and fully enjoy their RF quietness.

 https://www.youtube.com/watch?v=9n_QYo6yyd0

« Last Edit: June 11, 2017, 04:57:32 PM by VK3YE » Logged

Peter VK3YE

TOP-SELLING RADIO BOOKS: 'Minimum QRP', 'Hand-carried QRP antennas' & 'Getting back into Amateur Radio'. 
Paperback and ebook editions. See http://home.alphalink.com.au/~parkerp/vk3yebooks.htm or search titles in Amazon.
AJ4SN
Member

Posts: 67




Ignore
« Reply #5 on: June 11, 2017, 05:01:49 PM »

Thanks so much for all the suggestions. I do have a LPF included with the design and, by coincidence, the VFO runs from a separate power supply. I will definitely look into the grounding and shielding situation. I may try to shield the LPF which is in the same enclosure as the rest of the DDS VFO. Peter, I did watch your videos. They were very informative. Thanks for taking the time to make them. I am curious about the impact of the VFO output level on the spurious emissions. I inserted a step attenuator between the VFO output and the radio, and dropping the drive level somewhat does help. Is there an appropriate drive level that minimizes spurs? I am using a minicircuits TAK-3H mixer, and I try to keep the drive level as close to the 17dBm as possible. I understood that was the best way to have a strong mixer. Am I off base here?
Logged
VK3YE
Member

Posts: 227


WWW

Ignore
« Reply #6 on: June 11, 2017, 05:09:06 PM »

I'm not that familiar with that mixer but excessive drive could cause problems.  

A lot of literature on diode mixers emphasise the importance of proper termination and good impedance matching over a wide range of frequencies. Discussed extensively in Experimental Methods for RF Design etc.

Also while it's no substitute for removing the noise in the first place an audio amplifier chain that has a limited audio bandpass (ie 300 - 3000 Hz) with a roll off below and above can reduce the effect of hum etc.
Logged

Peter VK3YE

TOP-SELLING RADIO BOOKS: 'Minimum QRP', 'Hand-carried QRP antennas' & 'Getting back into Amateur Radio'. 
Paperback and ebook editions. See http://home.alphalink.com.au/~parkerp/vk3yebooks.htm or search titles in Amazon.
VK4FFAB
Member

Posts: 412




Ignore
« Reply #7 on: June 11, 2017, 11:48:10 PM »

Thanks so much for all the suggestions. I do have a LPF included with the design and, by coincidence, the VFO runs from a separate power supply. I will definitely look into the grounding and shielding situation. I may try to shield the LPF which is in the same enclosure as the rest of the DDS VFO. Peter, I did watch your videos. They were very informative. Thanks for taking the time to make them. I am curious about the impact of the VFO output level on the spurious emissions. I inserted a step attenuator between the VFO output and the radio, and dropping the drive level somewhat does help. Is there an appropriate drive level that minimizes spurs? I am using a minicircuits TAK-3H mixer, and I try to keep the drive level as close to the 17dBm as possible. I understood that was the best way to have a strong mixer. Am I off base here?

The AD9850 should not be putting out 17dbm, it should be putting out -6dbm and then have a amp and buffer stage following it to bring the signal up beyond what your mixer requires. 17dbm is 5v p-p the sine wave output of the AD9850 is not much more than 300mV p-p.
Logged
N3QE
Member

Posts: 4874




Ignore
« Reply #8 on: June 12, 2017, 06:30:08 AM »

Constructing a good DDS VFO for receive use is in fact very difficult. Spurs from the DDS must be filtered out by tracking a tracking PLL and bandpass filters. NONE of this is present in the AD7C design and the majority of the buzzes and whines you are hearing are the result of the DDS spurs you would fully expect to hear by using a raw unfiltered DDS as the LO.

I think the AD7C design would be usable for a QRP transmitter where some tuned circuits in the amplifier chain helped clean up the worst of the spurs, at least the ones that were completely out-of-band. You would still be left with a number of in-band spurs that could only be cleaned up by a tight tracking PLL and even more filtering. Above the QRP levels these would be absolutely necessary to not pollute the whole band.
Logged
K9MB
Member

Posts: 36




Ignore
« Reply #9 on: June 12, 2017, 07:42:56 AM »

N3QE, very good points.
I am just getting parts together to build one of these things, but was thinking last night that the thing has to be very noisy and that I need to build up a nice stable and "quiet" VCO and phase lock it to the DDS, with appropriate filtering to eliminate jitter and spurs. Might make sense to put the pristine VCO in a separate compartment from DDS.

Close in noise may still be present, though reduced by loop filter, but spurs away from the channel can be pared off, I think.

I am surprised that this issue has not been addressed. I am new to this kind of PLL.

I used divide by N traditional loops with down mixers and always used a solid crystal oscillator for first conversion in my vhf and uhf weak signal receiver designs.

At HF  where noise figure in the front end is of no importance, the bigger issue is spurious out of bandpass signals that cause noise and heterodynes as one tunes across a band already full of outside intereferences.
A tracking vco should serve nicely there.
Logged
N3QE
Member

Posts: 4874




Ignore
« Reply #10 on: June 12, 2017, 09:07:58 AM »

I am surprised that this issue has not been addressed.

Again, at the hobbyist QRP transmitter level, I think the AD9850 with just the usual transmit chain bandpass filtering may be acceptable. But only at QRP levels. Above QRP levels, the in-band spurs will quickly earn a bad reputation among neighboring hams. I think most homebrew users never get past the QRP stage.

DDS's are very often important parts of modern ham synthesizers. And synthesizer noise (not just DDS but also simpler PLL) took DECADES for the major ham manufacturers to address adequately for good receive performance. So this has been largely addressed but the devil is in the details. Even today, the best synthesizers still produce noticeable spurs and phase noise that can be measured as adversely affecting receive performance.

Real-world ham rigs that use a DDS as part of the synthesizer, use a DDS running at a very high clock (sometimes bigger than 1GHz) for fine frequency control feeding a mixer that is part of a PLL loop that also uses more conventional divide-by-n synthesizers for the coarse steps. The DDS is run over a rather limited output frequency range and has superb tight bandpass filtering after it to make sure all the spurs are down far enough. This is in addition to the cleanup provided by the PLL loop and divide-by-n factor. Often there are multiple PLL loops in real ham synthesizers, not just one, to provide even further cleanup while still offering frequency agility.

Good synthesizer designs to research, would be the Elecraft K3 synthesizer and the Ten-Tec Eagle synthesizer. Even then, for example, Elecraft recently upgraded its synthesizer to take down even further close-in phase noise that was mearuably affecting rig performance. So there is always room for improvement even with the very best the industry has developed.
« Last Edit: June 12, 2017, 09:13:41 AM by N3QE » Logged
AC7CW
Member

Posts: 961




Ignore
« Reply #11 on: June 12, 2017, 01:10:56 PM »

Perhaps adding a lpf between the vfo and the radio would help?

Many switching mixers require square waves to drive them, so an LPF won't work in those cases.

The VFO generates a sine wave by calculating the value. It's stepped iow, and always requires some filtering to have a cleaner waveform. We really don't have much to go on in this question so i tossed out some general idea of what I'd look for first...
Logged

Novice 1958, 20WPM Extra now... (and get off my lawn)
AJ4SN
Member

Posts: 67




Ignore
« Reply #12 on: June 12, 2017, 04:54:21 PM »

Here's a little more information on the DDS VFO. I did include a feedback amplifier and a low pass filter taken right from Experimental Methods in RF Design (page 6.39 revised first edition). That enabled me to raise the output level to 15.9 dBm. I use this 5-5.5 MHz signal to drive my main station receiver which has a 9 MHz IF. It's a homebrew unit that uses a minicircuits TAK-3H mixer. A dual gate MOSFET serves as the RF amplifier in the receiver. The mixer is followed by W7ZOI's "hycas" IF and a 2.4 kHz 8-pole crystal filter as well as an 8-pole 500 Hz filter for CW. Another TAK-3H serves as the product detector. The receiver has been in use for several years, and it performs well with its analog VFO; however, the tuning is just a little too fast for CW, and I thought that the DDS VFO would be a nice improvement. It sounds like the DDS project may end up as another signal generator for the shack (I can always use another one). I just didn't expect it to produce so much noise! Apparently, a DDS system that is comparable to the analog VFO in noise performance is relatively complex, and I might be fighting above my weight class trying to build one! I really appreciate your feedback and suggestions, though. Thanks for taking the time to help. 

73,

Stan
Logged
G4AON
Member

Posts: 1019




Ignore
« Reply #13 on: June 13, 2017, 07:56:58 AM »

A couple of points. The drive level to a mixer should be measured into a 50 Ohm termination and not measured with an oscilloscope across the mixer osc input pin, which can give misleading results.

I use an SDR Kits Si570 based synth in a few projects. They are relatively quiet and low cost. If you get one, ask for firmware version 4.21 as the current listed one (4.20) removes the IF offset when you engage dial lock! 4.21 works OK. The output of an Si570 is a square wave of about +10dBm, with a 3dB attenuator it drives a level 7 mixer just fine.

http://sdr-kits.net

73 Dave
Logged
YL3GND
Member

Posts: 55




Ignore
« Reply #14 on: June 13, 2017, 02:38:32 PM »

"mixer is followed by "hycas" IF and crystal filter"...really? Sounds pretty bad composition. Mixer with an IIP3 around +23dBm and OIP3 around +29dBm is followed by something so crappy as "hycas" amplifier.. Or there is a typo and crystal filter is in between?

Egmont
Logged
Pages: [1] 2 3 Next   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!