The AM Forum
March 28, 2024, 07:42:53 PM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: [1] 2   Go Down
  Print  
Author Topic: Amplitude Modulating a DDS VFO  (Read 16919 times)
0 Members and 2 Guests are viewing this topic.
W2PHL
Contributing
Member
*
Offline Offline

Posts: 146


Phil


WWW
« on: March 17, 2017, 10:30:19 AM »

Gang,
Over the past few years I've been having fun experimenting with cheap DDS signal generator boards from overseas. The AD9850 chip used in many of them can make quality HiFi AM. After getting many "what did you say you're running?" questions, I thought I'd document some information just in case any of youz guys are interested in experimenting.  The board I focus on in the attached article is the best one I've found so far.  You can use it to build a standalone AM transmitter with a cannibalized PA deck from an old HF rig or you can modify an existing HF rig by unplugging the low level transmit RF cable that goes to the PA stage and replacing it with the output from the modulated VFO. It's important to note that by replacing a transceiver's internal TX signal with the VFO's signal, the rig's ALC and SWR fold-back circuits will not function leaving the final amplifier vulnerable to destruction by over driving or output load problems. Destruction is the operative word.  Shocked
I made a video so you can see and hear how Hi the Fi is.

https://www.youtube.com/watch?v=qziatWoF-co

Phil

Logged

I think therefore AM
Steve - K4HX
Administrator
Member

Offline Offline

Posts: 2722



« Reply #1 on: March 17, 2017, 11:52:56 AM »

I've heard Phil on this rig. It sounds fantastic.
Logged
M0VRF
Member

Offline Offline

Posts: 172


« Reply #2 on: March 17, 2017, 03:08:06 PM »

www.analog.com/media/en/technical-documentation/application-notes/AN-423.pdf

Is the 'Original' article.

Just generate low A.M. and use a linear to amplify.

Nothing new....However look how linear the DDS is, wow, that's linear!
Logged
W2PHL
Contributing
Member
*
Offline Offline

Posts: 146


Phil


WWW
« Reply #3 on: March 17, 2017, 04:42:52 PM »

Thanks Steve! I hope other hams find the information useful.
Logged

I think therefore AM
W1AEX
Un-smug-a-licious
Contributing
Member
*
Offline Offline

Posts: 1488


Apache Labs SDR


WWW
« Reply #4 on: March 20, 2017, 11:16:44 AM »

As always Phil, your experiments are much cooler than anything else I have seen! Very nicely done!

Rob
Logged

One thing I'm certain of is that there is too much certainty in the world.
KF7WWW
Member

Offline Offline

Posts: 91


« Reply #5 on: March 20, 2017, 12:29:19 PM »

I ordered one of those DDS's. This seems like a cool little project.. I have an amplifier from some sort of rig that has multiple gain stages. So I believe it should work very well..
Logged
W2PHL
Contributing
Member
*
Offline Offline

Posts: 146


Phil


WWW
« Reply #6 on: March 20, 2017, 06:36:32 PM »

I've also used one of those DDS modules to build an external VFO for a Tempo 2020. Here's a link to the enclosure:
http://www.ebay.com/itm/262638765798?_trksid=p2060353.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

It's a nice size box for a VFO. The top is thin aluminum and the bottom is sheet metal.
Attached is a pic of the Tempo 2020 VFO project.  I'm waiting on a bezel....and maybe a better way to label the buttons...yikes!


Phil


* tempo2020vfo.jpg (2193.32 KB, 3264x2448 - viewed 551 times.)
Logged

I think therefore AM
WB2CAU
Member

Offline Offline

Posts: 342


« Reply #7 on: March 20, 2017, 09:43:40 PM »

I ordered one also.  Next I'd like to think about constructing an intermediate linear amplifier to bring the level up to something capable of driving the SB-220.  Less than 10 watts carrier should do it.  Question is... should it be vacuum tube based or solid state?  

I've never built a solid state RF amplifier.  This might be the time to think about trying something new.  

Eric
Logged

"Life is tough, but it's tougher if you're stupid." -- John Wayne
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #8 on: March 21, 2017, 10:45:19 AM »

I ordered one also.  Next I'd like to think about constructing an intermediate linear amplifier to bring the level up to something capable of driving the SB-220.  Less than 10 watts carrier should do it.  Question is... should it be vacuum tube based or solid state?  

I've never built a solid state RF amplifier.  This might the time to think about trying something new.  

Eric

Hi Eric,

Some ideas:

First, building a high power SS amplifier that is CLEAN, that is, -30dB 3rd order or better is difficult. Frank, WA1GFZ spent over a year testing and modifying his big homebrew SS linear amp and found that he needed to optimize the core and transformers extensively to break into the "clean" arena. He wrote some good articles for QEX on the subject.

Personally, I have built a few SS amps, even using the expensive MRF 150 transistors. I failed in the first attempt, as it was -25 dB 3rd IMD or worse. The second attempt was better and I currently use it.

Bottom line is unless we are talking "Pure Signal" software to clean it up -  be prepared to optimize the transformers out of the mainstream ways. Tubes may be the better choice.

The second alternative is to build up a pure class A amplifier using a big transistor, maybe a version of the MRF-150 family. That's an easy 10 watts of -40dB 3rd... clean drive for your SB-220 amp. (and 40+ watts of heat)


That said, if you only need 10 watts, running barefoot you are gaining a 10 dB "clean signal" advantage over the regular 100 watt station, so nobody will really notice if you are a little dirty.  (until you drive an amplifier with it) Bear in mind that the average, modern open loop SS ricebox is usually about -29dB 3rd or worse.

As a two-tube self-contained 100 W tube project: - Running a single tube designed for linear service, set up properly in class AB1, will give an easy -30 to -35dB 3rd IMD. Pretty clean. Ten dB in IMD improvement makes a huge difference in cleanliness. Make it a two-stage class A  grid driven - driving an AB1 GG amplifier and you may be able to drive it directly with the VFO.  Anything between 10 watts to 100W carrier is easily do-able with tubes. The cool thang about this design is if you make both the driver and final grid driven, the driver and final are close to each other on the same chassis so that you can add some RF negative feedback to make it -40 to -45 3rd  - an ultra clean overall system!  Grounded gird configuration NFB can also be done with a single tube too thru a filament/cathode winding.


And then, a single tube driver designed for linear service (not a class C designed tube) running class A should make a great 10 watt driver.  You will be running about 30 watts of heat to generate 10 watts carrier out, but very clean.

The key thing to remember is that your driver should be at least 10 dB cleaner than the final amplifier - so that it will not degrade and pull down the final amplifier IMD numbers significantly. IE, a -40 dB 3rd IMD driver is adequate for a -30dB IMD  final amplifier so that the final will drop from only -30dB to -29dB 3rd IMD.

Your 3-500Zs SB-220 may be good for at least -35 dB 3rd, so you will need to really pay attention to details to get a -45dB driver or better.

I assume the VFO modulation system is super clean, but it will need to be verified first or all the other work is for not.

T
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
W2PHL
Contributing
Member
*
Offline Offline

Posts: 146


Phil


WWW
« Reply #9 on: March 21, 2017, 08:07:41 PM »

Here are some non laboratory grade test results. The two tone IMD test was generated on a laptop running Spectralab: 700Hz and 1900Hz tones @ 95%modulation and viewed on a Flex 1500 (10dB per division). One pic is the DDS output barefoot the other is at 50 watts PEP using a Kenwood TS450s PA deck. The harmonic spectrum was measured on a Agilent E6380A communications test set.

Looks like 3rd order IMD is about -33 barefoot and -30ish with the TS450s amp @ 50 watts PEP
 
 
Phil


* dds output.jpg (155.97 KB, 1280x960 - viewed 444 times.)

* 50watt PEP.jpg (202.03 KB, 1280x960 - viewed 469 times.)

* -0dBMm harmonics.jpg (175.16 KB, 1280x961 - viewed 441 times.)
Logged

I think therefore AM
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #10 on: March 21, 2017, 08:46:15 PM »

Here are some non laboratory grade test results. The two tone IMD test was generated on a laptop running Spectralab: 700Hz and 1900Hz tones @ 95%modulation and viewed on a Flex 1500 (10dB per division). One pic is the DDS output barefoot the other is at 50 watts PEP using a Kenwood TS450s PA deck. The harmonic spectrum was measured on a Agilent E6380A communications test set.

Looks like 3rd order IMD is about -33 barefoot and -30ish with the TS450s amp @ 50 watts PEP
 
 
Phil

Very good, Phil -

-33 dB 3rd IMD as a driver is OK.   If you drive it into an amplifier chain, the chain will never be better than -33... and slightly worse than the amplifier itself. You will not have any splatter problems on the band with it as long as you can keep the whole chain's output  cleaner than -30...  Most ricebox amplifier modules and 3-500Z amps will accomplish this.  However, the cheap SS amps will not make the grade and are usually -25 dB 3rd or worse.

Of course, there are SS low level drivers much cleaner than the DDS. High end test generators usually have better numbers running AM. I tap off of my FT-1000D pre-driver as a source for my linear amp chain. It has 3rd IMD of about -70 dB 3rd ! at that point. The FT-1000D power amplifier deteriorates to -30dB after that.  I use this 70 dB 3rd source into a lab amp, to the 3CX-350J, etc.

I have a similar DDS that I use as a VFO/driver source for my 4-1000A plate modulated rig. Running as a VFO it had some spurs up and down the band that were about -60 dB down bare bones. I was able to cure this adding another -30 dB simply with the existing input tuned circuits to the 4X1 rig. The 4X1 grid tuning Q had enough selection to kill the spurs.

Looking at your spectrum I see a blip down about 3877 or so... is that a spur that stays without modulation? Maybe not.   I worry about them because if my signal were 60 over at a local location, then the spur would be S9 with modulation - enough to bother or at least annoy an adjacent QSO... :-)  So if spurs become a problem, try a sharp tuned circuit after the DDS and see what it does.

T
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
W2PHL
Contributing
Member
*
Offline Offline

Posts: 146


Phil


WWW
« Reply #11 on: March 22, 2017, 10:08:36 AM »

Interesting comments Tom! That blip seems to be a PowerSDR/Flex anomaly. It disappeared after a restart of the SDR. Yet another reason for me to buy a real spectrum analyzer. ha!  73

 Phil
Logged

I think therefore AM
Steve - K4HX
Administrator
Member

Offline Offline

Posts: 2722



« Reply #12 on: March 22, 2017, 11:26:41 PM »

Eric: One possibility,

http://www.qrpkits.com/10wamp.html


Logged
WB2CAU
Member

Offline Offline

Posts: 342


« Reply #13 on: March 23, 2017, 06:06:10 PM »


Thanks Steve.  It would be a great possibility if there were more band options.  

Although, since SSB and CW are mentioned as the modes, it's probably a 10W PEP or 2.5W AM carrier, a bit short of what I would need to drive the SB-220. 
Logged

"Life is tough, but it's tougher if you're stupid." -- John Wayne
W2PHL
Contributing
Member
*
Offline Offline

Posts: 146


Phil


WWW
« Reply #14 on: March 24, 2017, 10:29:19 AM »

Eric,

[/quote]
 it's probably a 10W PEP or 2.5W AM carrier, a bit short of what I would need to drive the SB-220.  
[/quote]

Although fun to do, building a kit or building an amplifier from scratch is more effort than necessary. If you can find one, using a PA deck from an old HF rig is a much easier way to attack the problem.  The five transistor types that use a single 2SC1971 as a pre-driver (1st stage) will work great with this DDS. No additional input gain is needed for driving the PA deck to 100 watts PEP. Hook up the band filter relays and you are good to go.

Phil
Logged

I think therefore AM
WB2CAU
Member

Offline Offline

Posts: 342


« Reply #15 on: March 24, 2017, 10:50:08 AM »

I ordered one also.  Next I'd like to think about constructing an intermediate linear amplifier to bring the level up to something capable of driving the SB-220.  Less than 10 watts carrier should do it.  Question is... should it be vacuum tube based or solid state?  

I've never built a solid state RF amplifier.  This might the time to think about trying something new.  

Eric

Hi Eric,

Some ideas:

First, building a high power SS amplifier that is CLEAN, that is, -30dB 3rd order or better is difficult. Frank, WA1GFZ spent over a year testing and modifying his big homebrew SS linear amp and found that he needed to optimize the core and transformers extensively to break into the "clean" arena. He wrote some good articles for QEX on the subject.

Personally, I have built a few SS amps, even using the expensive MRF 150 transistors. I failed in the first attempt, as it was -25 dB 3rd IMD or worse. The second attempt was better and I currently use it.

Bottom line is unless we are talking "Pure Signal" software to clean it up -  be prepared to optimize the transformers out of the mainstream ways. Tubes may be the better choice.

The second alternative is to build up a pure class A amplifier using a big transistor, maybe a version of the MRF-150 family. That's an easy 10 watts of -40dB 3rd... clean drive for your SB-220 amp. (and 40+ watts of heat)


That said, if you only need 10 watts, running barefoot you are gaining a 10 dB "clean signal" advantage over the regular 100 watt station, so nobody will really notice if you are a little dirty.  (until you drive an amplifier with it) Bear in mind that the average, modern open loop SS ricebox is usually about -29dB 3rd or worse.

As a two-tube self-contained 100 W tube project: - Running a single tube designed for linear service, set up properly in class AB1, will give an easy -30 to -35dB 3rd IMD. Pretty clean. Ten dB in IMD improvement makes a huge difference in cleanliness. Make it a two-stage class A  grid driven - driving an AB1 GG amplifier and you may be able to drive it directly with the VFO.  Anything between 10 watts to 100W carrier is easily do-able with tubes. The cool thang about this design is if you make both the driver and final grid driven, the driver and final are close to each other on the same chassis so that you can add some RF negative feedback to make it -40 to -45 3rd  - an ultra clean overall system!  Grounded gird configuration NFB can also be done with a single tube too thru a filament/cathode winding.


And then, a single tube driver designed for linear service (not a class C designed tube) running class A should make a great 10 watt driver.  You will be running about 30 watts of heat to generate 10 watts carrier out, but very clean.

The key thing to remember is that your driver should be at least 10 dB cleaner than the final amplifier - so that it will not degrade and pull down the final amplifier IMD numbers significantly. IE, a -40 dB 3rd IMD driver is adequate for a -30dB IMD  final amplifier so that the final will drop from only -30dB to -29dB 3rd IMD.

Your 3-500Zs SB-220 may be good for at least -35 dB 3rd, so you will need to really pay attention to details to get a -45dB driver or better.

I assume the VFO modulation system is super clean, but it will need to be verified first or all the other work is for not.

T

Thanks for all the well thought out advice, Tom.  You make some very excellent points but I'm looking for a simple, low cost, add-on amplifier to go along with the bargain basement DDS.  

I'm not an RF engineer (by any stretch) so I can't just whip up a solid state RF amplifier out of my head and expect it to work.  Everything I build from scratch is a rip-off of an existing design with my own adaptions if necessary.  I'm looking either for detailed project instructions, a non-SMT kit or an assembled board.  

As far as the vacuum tube option, what exactly is a "single tube driver designed for linear service (not a class C designed tube)" in this power range?  Does such a tube really exist?  Do you have any particular tube in mind?  I think a 6146 is a class C designed tube, isn't it?  Although a 6146 would just about fit in the correct power range, a 6146 type tube is not a choice for grounded grid since the suppressor grid is tied to the cathode.  I would suspect a TV sweep tube might be a better choice but it was not designed for linear RF service.  

Eric
Logged

"Life is tough, but it's tougher if you're stupid." -- John Wayne
WB2CAU
Member

Offline Offline

Posts: 342


« Reply #16 on: March 24, 2017, 10:57:14 AM »

Eric,

it's probably a 10W PEP or 2.5W AM carrier, a bit short of what I would need to drive the SB-220.  
[/quote]

Although fun to do, building a kit or building an amplifier from scratch is more effort than necessary. If you can find one, using a PA deck from an old HF rig is a much easier way to attack the problem.  The five transistor types that use a single 2SC1971 as a pre-driver (1st stage) will work great with this DDS. No additional input gain is needed for driving the PA deck to 100 watts PEP. Hook up the band filter relays and you are good to go.

Phil
[/quote]

Phil, your advice is well taken, but I think you just got lucky with having the PA deck of an old HF rig at your disposal.  This is not exactly something easily found lying around most ham shacks, except yours :-). 
Logged

"Life is tough, but it's tougher if you're stupid." -- John Wayne
Steve - K4HX
Administrator
Member

Offline Offline

Posts: 2722



« Reply #17 on: March 24, 2017, 11:00:38 AM »

You could probably buy the hulk of a crapped out FT-757 or similar aged Kenwood or Icom for less than you could build an amp.
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #18 on: March 24, 2017, 12:33:50 PM »

As far as the vacuum tube option, what exactly is a "single tube driver designed for linear service (not a class C designed tube)" in this power range?  Does such a tube really exist?  Do you have any particular tube in mind?  I think a 6146 is a class C designed tube, isn't it?  Although a 6146 would just about fit in the correct power range, a 6146 type tube is not a choice for grounded grid since the suppressor grid is tied to the cathode.  I would suspect a TV sweep tube might be a better choice but it was not designed for linear RF service.  
Eric

Hi Eric -

What I mean by a tube designed for linear service is the series of high-mu, directly heated cathode tubes designed for grounded grid service, like the 3-500Z, 8877, 3CX-3000A7, YC-156, etc.  These tubes came after the popular 813, 4-400A, 4-1000A and other tubes designed to run grid-driven in class C. (And class B service for modulators)  However, the class C series of tubes can be made as clean and even cleaner than the "linear" tubes by adding a loop of negative feedback.  There are anomalies like the 813 and 4-1000A that do extremely well in GG too.  But the 3-500Z, etc., are naturals and so easy to use (less circuitry) and get great results in linear service.

There are other good GG tubes of lower power available.

My intention for this discussion was to give a "heads up" that yes, we can add a cheapo SS amp to the VFO, but if we amplify it further with a big 3-500Z amplifier, etc., watch out for IMD problems caused by a marginal driver. (creating overall less than -30 dB 3rd IMD)  Many of the small SS amplifier QRP kits and even many SS ham commercial amps on the market do not get better than -25dB 3rd, if that, and what happens is after a few bad reports of splatter we end up shelving the rig in fear of making too many enemies... :-)  Notice that almost ALL of the ham solid state  commercial manufacturers never publish their 3rd order IMD products. It takes a call to the factory and reluctantly you may get an engineer there to admit the specs... maybe.  Hopefully things have improved with the new devices on the market and better transformer designs for those $10K SS ham amps. But the little kit, QRP and CB-style amps for $50 are a red flag unless we keep them at QRP levels.

The driver (VFO and first RF IPA amplifier = 10 watts) is absolutely critical to achieving a clean overall signal once a bigger, clean tube amplifier is added resulting in over 100 watts output.

Steve's suggestion of using an older ricebox SS amplifier module will work FB keeping us on the clean edge of -30dB 3rd.  After all, there are plenty of riceboxes successfully driving 3-500Zs on the air. Once we fall into the -20's, the band splatter problems usually start.  We simply must do a 2-tone test on our rigs (complete chain) to know for sure.

T
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
KC2ZFA
Member

Offline Offline

Posts: 441



« Reply #19 on: March 24, 2017, 01:35:02 PM »

solution !

http://online.sfsu.edu/hl/tmc2.html
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #20 on: March 24, 2017, 01:59:05 PM »


Agreed!  I owned one a few years back. The first thing I did was increase the internal RF negative feedback until the IMD improved to -50 dB 3rd. Only 100 mW input drive required for a couple of hundred watts out carrier. It's one of the best made amplifiers you can buy.  Mil-spec, heavy duty.

I built an external power supply and installed PTT, etc. What an amp for the VFO project.  The only problem is the VFO is -33 dB 3rd while the amp is 10-15 dB cleaner...  Generally, the opposite relationship is desirable.


T
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
n1zpy
Member

Offline Offline

Posts: 16


« Reply #21 on: March 25, 2017, 11:31:10 AM »

Thinking about building one of these.  I have an old ft757 for parts. What will I need for a mic?  All I have right now is some old cb base mics(d-104 both styles and some turner plus threes)  Will I need to add a mic preamp? 
Logged
W2PHL
Contributing
Member
*
Offline Offline

Posts: 146


Phil


WWW
« Reply #22 on: March 26, 2017, 11:56:00 AM »

The FT 757GX RF deck will work OK fine with this VFO, but it isn't plug and play. Any RF deck or amplifier will require modifications (switching of bias, band and TX/RX relays etc...) to make it work. As for the audio,  I'm using line level from an Inovonics 222. I haven't tried using an amplified microphone directly to the VFO. I would be hesitant to go that route anyway as there would be no control over peak limiting and bandwidth. There are some excellent articles here on AMfone and on the web that discuss audio processing for AM.

Phil
Logged

I think therefore AM
n1zpy
Member

Offline Offline

Posts: 16


« Reply #23 on: March 26, 2017, 01:20:52 PM »

I don't see an Inovonics 222 in my future but would a BEHRINGER Ultragain Pro MIC2200 and a condeser mic work ok as a starting point?
Logged
W2PHL
Contributing
Member
*
Offline Offline

Posts: 146


Phil


WWW
« Reply #24 on: March 27, 2017, 08:42:16 AM »

As a starting point? It will work but far from ideally.

Looking at the specs, the Mic2200 is a just microphone preamp with EQ. Without peak limiting it will be difficult to prevent over-modulation and distortion.

A common modification for many HF rigs is to feed audio directly to the balanced modulator. Because all of the transceiver's audio processing (gain/compression/limiting) is bypassed, a broadcast style audio chain is required. Same scenario with modulating this DDS VFO.

In that silly YouTube demo I made, the audio to the transmitter is being processed by Breakaway Broadcast. Without that software processor the transmitted audio would sound bad...really bad! If it were on the air, it would cause splatter up and down the band!!!

Logged

I think therefore AM
Pages: [1] 2   Go Up
  Print  
 
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone © 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.064 seconds with 18 queries.