The AM Forum

THE AM BULLETIN BOARD => Technical Forum => Topic started by: N2DTS on January 04, 2010, 11:06:21 PM



Title: flex radio musings...
Post by: N2DTS on January 04, 2010, 11:06:21 PM
I had a nice long qso on the weekend with Eric, Jay, and Stu regarding the flex radios.
I have a 5000, Jay has a 3000, Eric and Stu have the old 1000's.
There is a carrier adjustment setting on the flex which sets the carrier level without changing the pep (audio) level.
The flex manual says you can adjust this to run less carrier and more audio, to get say 130% positive modulation.
Eric said its no good, it turns the neg audio peaks around and makes them positive, and injecting a tone into the radio and looking on a scope, and that is exactly what it does. I did the test.
Stu said he could hear it, I could not, but I cant hear much of anything.

I don,t understand what flex would be doing with this adjustment.
You basicly get extra positive pulses where the carrier would normaly be closed off.
I am trying to figure out what the net results of this would be, its not a keep alive carrier, its a pulse/waveform...

With all the things they can do in software, you would think they could just limit the negative modulation without flipping it.
I suppose its better then closing off the carrier, or is it?
Would the signal get wide (splatter) or just have distortion go up?

We also talked about the benefits of modulation over 100% positive, which everyone knows is good, but how much is good before the typical AM detector generates excessive distortion?

Another very odd flex thing, on the 5000 anyway, is turning the preamp on LOWERS the noise level, increases the signal to noise ratio.
That is backwards from every other radio, but flex says that is how it works.
I tested it and its so....

Tom, K1JJ was holding forth on 3885 tonite and working stations all over the US, and I checked the homebrew RX, and the flex in sync detector mode, and normal AM, and with the preamp on and off.
The homebrew has the edge, but the preamp really helps with weak signals on the flex, which seems crazy on 80 meters.

I wish I could have jumped in, but would likely tear up all the TV's in the hood...

Brett




 


Title: Re: flex radio musings...
Post by: KX5JT on January 04, 2010, 11:48:55 PM
Maybe Flex can update the ability to adjust the carrier/audio levels with more options via the software update?  Press them about it! 


Title: Re: flex radio musings...
Post by: WD5JKO on January 04, 2010, 11:53:17 PM


Brett,

  When Gerold Youngblood, AC5OG designed the Flex SDR concept, he was inspired by the old Central Electronics 20a phasing SSB transmitter. He talks about the old 20a, and how he made a SDR version of it. This was back in 2002 in an old QEX article:

www.flex-radio.com/Data/Doc/qex1.pdf

  I have had a passion for these old 20a's, and the effect you see when exceeding 100% negative modulation does cause a receiver diode AM detector fits. To me the received signal sounds really crappy, but switch to Sync-AM or SSB product detector and it sounds just fine.

  Some years ago I was playing around with this phenomena (DSB with reduced carrier), and ways to shape the audio to eliminate the "clicking distortion", to in effect create yet another form of "Super Modulation".

open file, 20aQRO3.doc:
http://pages.prodigy.net/jcandela/CE20AQRO/


  How you might do this with a Flex, I have no clue...but I bet a software routine could accomplish the same.

Jim
WD5JKO


Title: Re: flex radio musings...
Post by: K5UJ on January 05, 2010, 12:48:08 AM

I wish I could have jumped in, but would likely tear up all the TV's in the hood...

I used to worry about that too but went ahead anyway with 300 w. and never had a complaint.  Now that DTV is widely deployed I think it's a nonissue.  In fact, as far as I know, you can pull your TV low pass filter if you ever used one, and throw it away.  Unfortunately the people on the RFI receiving end with all this high tech digital stuff are the hams. 

73

Rob


Title: Re: flex radio musings...
Post by: flintstone mop on January 05, 2010, 10:02:27 AM
What I have noticed with my FLEX 1000 in A.M. mode is that it reacts to the same carrier vs modulation characteristics as a standard SS transceiver. The higher you adjust the carrier the less mod it is capable of. With 20 watts being the limits for carrier. The software cannot do any magic beyond the capabilites of the hardware.
All I need to drive my Titan is 8 watts, for the so-called legal limit, so I am capable of a LOT of modulation. And that's without the processors turned on in the software.
The real magic is the computer processing for extremely dense modulation if you care to process that much. And for the crowded bands, the notch filters and shifting of the I.F., and constant bandwidth adjustments runs circles around my R390A....................but the recovered audio out of the diode load still surpasses the FLEX. Almost all QSO's are now armchair copy.

Fred


Title: Re: flex radio musings...
Post by: KD6VXI on January 05, 2010, 10:35:19 AM
Maybe Flex can update the ability to adjust the carrier/audio levels with more options via the software update?  Press them about it! 

There are 11 meter operators using Flex radios that are capable of well over 100 percent positive mod.

Some of the 11 meter 'techs' have rewritten some of the Flex routines to get what they wanted.

One even has about 1.5 Kc of pull on keydown to simulate the loading on a VFO of a boat anchor transmitter when keyed (he wanted it to bark like a Barker 5100).

I never played with the Flex, so never cared about the software and updates the big AM guys are running, but if I run across it, I'll grab and download it...  Some of them sound REALLY good!

--Shane


Title: Re: flex radio musings...
Post by: WA1GFZ on January 05, 2010, 11:06:05 AM
How do the CBers get Flex to TX out of ham bands.
Flex software guys are not interested in AM so it is set to text book
100%


Title: Re: flex radio musings...
Post by: KF1Z on January 05, 2010, 11:15:40 AM
How do the CBers get Flex to TX out of ham bands.
Flex software guys are not interested in AM so it is set to text book
100%

There's a MARS "patch" for the software that Flex put out to those who prove they are members of MARS. (actually may be a firmware patch?)

Of course, same as all software, it is distributed at will.

It allows "everywhere" transmit capability.

The source code for PowerSdr is available on the Flex-radio website....
Up to version 1.8.x  or something like that.


Title: Re: flex radio musings...
Post by: N2DTS on January 05, 2010, 11:20:48 AM
While the flex works quite good for AM, they could make it a lot better, but I suppose there is little interest in improving AM operation. I will bug them about this and a few other things they could fix for AM.

You can do anything in software, including limiting the neg modulation only, and letting the positive go where it will. Exceeding 100% negitive generates a lot of splatter, I dont know if it SOUNDS so bad.
When the flex exceeds 100% neg modulation, it produces bumps of positive modulation, not sure what that sounds like...

My question was about exceeding 100% POSITIVE modulation, 120% is likely ok, but is 140%?

Also, the carrier control on the flex does not work like the typical ricebox, since you can adjust the carrier from 0 to 25 watts out, while the pep dsb signal stays at the same 100 watt level.
The flex also has a dsb mode button, I wonder what that would sound like with two people running it.

Another very odd thing I noticed while doing the sine wave tests on the flex is the carrier level is not stable with modulation, if I put on a 1000 Hz sinewave, and turn the audio gain up, past about 20% modulation, the carrier level (on the mod monitor) goes up, the more tone, the more it goes up.
I don't think I remember a normal plate modulated rig doing that, at least not as much.

That is different from a ricebox where you get into the alc and under modulation, the carrier can go down, not up....

As far as the TV qrm goes, well, I might also get into phones, but should try it some night, I DO get into the home TV, not sure if its into the dvr or the tv, or both, it might depend on the TX freq, or the channel being watched.
I have cable and the usual hi def LCD tv's....

Brett








Title: Re: flex radio musings...
Post by: N2DTS on January 05, 2010, 11:40:33 AM
As it is, the flex can do whatever percentage of modulation above 100% positive you want.
If I set my carrier control to 50%, I will have 200% positive modulation, 12.5 watts carrier, 100 watts pep.
The only problem is the audio that would normaly be cut off in a normal rig is flipped and turned positive creating 'extra' positive modulation where the closed off carrier would normaly be.
I am trying to figure out what that would do to the sound of the audio.
The CB guys likely dont care much about distortion (from what I hear)!

Think about what a sine wave modulated AM signal on a scope looks like.
If you look at the sine wave of a plate modulated transmitter, imagine turning the modulation up past the point of 100% positive, normaly there would be no carrier and a line there. OK, got that pictured? Now think of the tip of a sine wave starting to poke up where that line is...
I wonder what it would look like on a trapizoid!
Voice is different from a sine wave, with much shorter average peaks, so when they start poking up what would that sound like???
The problem with just turning it up and hearing it is how much is the crazy flex bump, and how much distortion is from the excessive positive modulation?



And what is the result of taking a normal plate modulated rig, and boosting the modulator output in only the positive direction???
My 32V3 runs a phase splitter to drive the modulator grids, which I can and have unbalanced to give more positive modulation, the positive phase just has more gain, or the neg side has less.
Sure its distortion, but what kind, and does it sound bad if done at a reasonable level?

Brett




Maybe Flex can update the ability to adjust the carrier/audio levels with more options via the software update?  Press them about it! 

There are 11 meter operators using Flex radios that are capable of well over 100 percent positive mod.

Some of the 11 meter 'techs' have rewritten some of the Flex routines to get what they wanted.

One even has about 1.5 Kc of pull on keydown to simulate the loading on a VFO of a boat anchor transmitter when keyed (he wanted it to bark like a Barker 5100).

I never played with the Flex, so never cared about the software and updates the big AM guys are running, but if I run across it, I'll grab and download it...  Some of them sound REALLY good!

--Shane



Title: Re: flex radio musings...
Post by: WA1GFZ on January 05, 2010, 12:04:17 PM
If that is true all you need is a negative peak limiter on the audio input.
1 op amp a diode and some resistors


Title: Re: flex radio musings...
Post by: k4kyv on January 05, 2010, 12:23:36 PM
One thing that would be interesting would be for all stations in a QSO to run DSB reduced carrier, and receive with a synchronous detector.  The DSB-RC could be generated with an analogue transmitter or SDR rig.  The idea would be to run double sideband, with just enough carrier to give the synch detector something to lock onto.  The  received signal would sound identical to a full  carrier AM signal with the synch detector, but to receive it without distortion on a conventional receiver would require receiving one of the sidebands in SSB mode.

This is where the ESSB guys are missing the boat.  If they would use synchronous detectors and transmit enough "pilot" carrier, about 20 dB below peak output, that would eliminate the frequency error so that the stations that  do it right would have near broadcast quality audio on SSB.  You  have to have a  very good ear to tune in the suppressed-carrier ESSB signal without frequency error, and it helps if you already know what the other person's voice  sounds like.


Title: Re: flex radio musings...
Post by: K1JJ on January 05, 2010, 12:42:48 PM
Many of the ESSB guys do one thing correctly - they operate on an exact frequency integer, like 3825.00  or even 3825.000 depending on the display readout resolution.  With a modern ricebox or accurate freq counter on the RX, most can be tuned in very very well quickly - no brainer.

Or, when joining in, just axe them what freq they are on and match it on the readout.

T


Title: Re: flex radio musings...
Post by: N2DTS on January 05, 2010, 01:18:29 PM
Well, see, flex missed the boat on that also.
You can do dsb, or AM with little carrier, but in the AM with little carrier, the modulation flips and you would get lots of 'extra' modulation bumps. You cant TX any carrier in the dsb mode.

Then there is the one sideband and carrier mode, I have seen people running that on the flex display.
The KWS-1 was one rig that did it that way. Half the channel noise I suppose, but I never heard a very fi signal out of a transmitter running ssb with carrier. A stock kws-1 sounds nasty!
But I wonder if it has to sound bad like that. The ssb guys could not complain about those wide AM signals with carrier and one sideband.
Would you get half the audio at the far end?
Why cant you have your cake and eat it to???
Who made that rule up?

The flex 5000 has got a .5 ppm clock, so for 99% of ssb tuning, its just point and click and you are right on freq.
And, the ssb signal can sound very good, better than some AM, but there seems to be no inteligent life on ssb.
I had one good ssb qso, talking with a guy on a tug boat in Boston, talking about old AM rigs, flex radios and tug boat stuff...

I dont think ssb is a bad mode at all, its just the people on ssb that ruin it...

Brett






Title: Re: flex radio musings...
Post by: Steve - WB3HUZ on January 05, 2010, 01:21:05 PM
No carrier required. A properly designed Costas loop will lock without the carrier. Extreme freq accuracy would also not be required.

A GPS receiver and a TXCO would provide frequency accuracy, if a carrier were to be used.


Title: Re: flex radio musings...
Post by: N2DTS on January 05, 2010, 01:30:12 PM
Well, I would rather not LIMIT something, as its hard to do cleanly.
Why amplify something a lot and then try to limit it, why amplify it so much in the first place?

You can do peak limiting, or peak clipping, or compression, or compression above some threshold, but I think just not amplifying it so much in the first place might work better.
What is the problem with just reducing the gain in the neg direction, so you get 90% neg mod, at 140% positive?

Brett



If that is true all you need is a negative peak limiter on the audio input.
1 op amp a diode and some resistors


Title: Re: flex radio musings...
Post by: Steve - WB3HUZ on January 05, 2010, 01:37:49 PM
Gain reduction equals limiting. Just two different names for a nonlinear gain curve.


Title: Re: flex radio musings...
Post by: N2DTS on January 05, 2010, 01:42:22 PM
I think its different having something with fixed but lower gain, then it is with something that CHANGES gain over time.
You can hear compression, but you cant hear lower gain except in the volume level?

Brett


Title: Re: flex radio musings...
Post by: Steve - WB3HUZ on January 05, 2010, 01:49:15 PM
What you propose is not "fixed but lower gain." You are proposing to have less gain on the negative cycle than on the positive cycle. This means the gain changes every time the waveform crosses the zero axis. In other words, it does change with time. Further, a discontinuity like this will be very audible.


Title: Re: flex radio musings...
Post by: flintstone mop on January 05, 2010, 01:51:01 PM
It's all going to depend on either the microphone or the voice characteristics of the person speaking to get more positive peaks or asymmetrical modulation. Maybe there are processors out there that can make a person's voice asymmetrical or more positive.....dunno.
Turning on the processors in the FLEX (comp) (cpdr) and playing with the settings really start to schwang the munkey. But make the tx audio fatiguing to listen too.
And pls correct me if I'm out in left field, but whether it's 150% or 200% modulation you will not be able to go beyond the P.E.P. of the hardware.

Fred


Title: Re: flex radio musings...
Post by: K1JJ on January 05, 2010, 02:00:05 PM
What you propose is not "fixed but lower gain." You are proposing to have less gain on the negative cycle than on the positive cycle. This means the gain changes every time the waveform crosses the zero axis. In other words, it does change with time. Further, a discontinuity like this will be very audible.

Steve,

Do you remember the circuit I axed you to design for me maybe 10 years ago?  It was a dual op amp circuit that had more gain in the positive direction than the negative. It could be adjusted to make a symetrical sine wave look asymetrical.

I built it and it worked as intended. I figured it'd be a great way to artifically increase positive peaks.  However, in real voice operation it produced a noticeable distortion, even with small disparity gain. I got reports on the air of intermod, etc, so scrapped it.

It was worth a try... ;)

For what it's worth.

T


Title: Re: flex radio musings...
Post by: N2DTS on January 05, 2010, 02:01:38 PM
No you are not wrong, but consider many only use a very small amount of power to drive an amplifier.
I run my 5000 at under 20% for 150 to 200 watts carrier output.

Theo reticly, I could run things at 150 watts carrier and 1000 watts pep. Or 800 watts pep. Or 600 watts pep.
Barefoot radio only does 100 watts pep, 25 watts carrier or less.

Brett


Title: Re: flex radio musings...
Post by: WB2EMS on January 05, 2010, 02:04:41 PM
Quote
One thing that would be interesting would be for all stations in a QSO to run DSB reduced carrier, and receive with a synchronous detector.  The DSB-RC could be generated with an analogue transmitter or SDR rig.  The idea would be to run double sideband, with just enough carrier to give the synch detector something to lock onto.  The  received signal would sound identical to a full  carrier AM signal with the synch detector, but to receive it without distortion on a conventional receiver would require receiving one of the sidebands in SSB mode.

Years ago when I worked at Harris RF Communications, our transmitter and receiver systems were capable of that. They could run USB, LSB, ISB, 4ISB, full carrier AM or 'pilot carrier' which was -26db down to synch up with the receivers at the far end. On the sites I helped design and build, we typically used that in a 4ISB mode, 4 independent 3 khz sidebands, -26db carrier. Lower Lower sideband, Lower sideband, Upper sideband and Upper upper sideband. Typically 2 voice channels on the top two, and 2 RTTY channels on the bottom two. The matching receiver had 4 ISB capability also (RF550) and 4 audio outputs.

From time to time I've wondered about doing ISB with voice on one sideband and sstv on the other so a qso could be ongoing while the pictures were being exchanged in the background on the other sideband. I bet that wouldn't be that hard to implement in a flex system - might be an interesting market niche for them.  How would the ssb folks handle that - we'd still  be using two sidebands, but be running ssb!  :o

I'm a very new convert to the flex-sdr realm having been curious for a while and having received an SDR-IQ from Santa.  ;D Now I find myself scanning the for sale ads for used gear, or contemplating a flex-1500... I wonder how well 5 watts will drive a Titan 425?


Title: Re: flex radio musings...
Post by: N2DTS on January 05, 2010, 02:08:03 PM
DAMMIT, I was going to patent the idea!
Oh well....

Now I guess I have to add a balance control to the 32v3 phase splitter so I can make sure its balanced.
I notice most hifi tube stuff had a resting current balance control, and some had a level balance also, but no ham gear has these adjustments....

Brett



What you propose is not "fixed but lower gain." You are proposing to have less gain on the negative cycle than on the positive cycle. This means the gain changes every time the waveform crosses the zero axis. In other words, it does change with time. Further, a discontinuity like this will be very audible.

Steve,

Do you remember the circuit you designed for me maybe 10 years ago?  It was an op amp that had more gain in the positive direction than the negative. It could be adjusted to make a symetrical sine wave look asymetrical.

I built it and it worked as intended. I figured it'd be a great way to artifically increase positive peaks.  However, in real voice operation it produced a noticeable distortion, even with small disparity gain. I got reports on the air of intermod, etc, so scrapped it.

For what it's worth.

T


Title: Re: flex radio musings...
Post by: Steve - WB3HUZ on January 05, 2010, 02:11:26 PM
Yes, I recall Tom. The circuit seemed good in theory, but didn't work in practice.

The bigger thing to keep in mind is that negative peak limiting or other magic boxes are usually not needed to run extended positive peak modulation. Just set the polarity of your audio system to take advantage or the natural asymmetry of your voice. I can get 130-150% positive modulation without any special processing. As Fred noted, you just need a transmit system that can produce such modulation.


What you propose is not "fixed but lower gain." You are proposing to have less gain on the negative cycle than on the positive cycle. This means the gain changes every time the waveform crosses the zero axis. In other words, it does change with time. Further, a discontinuity like this will be very audible.

Steve,

Do you remember the circuit you designed for me maybe 10 years ago?  It was a dual op amp circuit that had more gain in the positive direction than the negative. It could be adjusted to make a symetrical sine wave look asymetrical.

I built it and it worked as intended. I figured it'd be a great way to artifically increase positive peaks.  However, in real voice operation it produced a noticeable distortion, even with small disparity gain. I got reports on the air of intermod, etc, so scrapped it.

For what it's worth.

T


Title: Re: flex radio musings...
Post by: N2DTS on January 05, 2010, 02:20:12 PM
Until about 6 months ago, the idea of computer junk in the shack really turned me off, till I got the sdr-iq.
Now I find the sdr stuff the most interesting part of the hobby.
I think they could do wonders with the stuff, but its all in the hands of small companies with limited resources.

I wish I understood the software, if I was 18, I am sure I would be into it, but now it makes my head hurt along with math....

The flex 1500 is going to have much reduced spec's but still likely much better than any vintage and most modern receivers.
I think I will get one to listen on in the den.
If they ever come out....its like Apple starting in the garage....a ways to go...

Hey, maybe I should buy flex radio stock!


Brett


Title: Re: flex radio musings...
Post by: N2DTS on January 05, 2010, 02:27:48 PM
I am not buying it.
If a little is good, more is better! Its the American way.
More food, more beer, more tubes, MORE POWER, it was Scott, an American company who sold 'the worlds most powerful receiver'!
8 cylinders is not enough, we want 12, or 16, or 32.

So what if it distorts a little, we get more!

Brett



Title: Re: flex radio musings...
Post by: KD6VXI on January 05, 2010, 03:05:35 PM
How do the CBers get Flex to TX out of ham bands.
Flex software guys are not interested in AM so it is set to text book
100%

AFAIK, it's "unlockable", to general coverage TX.

As I said, I've never played with them, but I did send an email to a friend who started on the 1000, and recently sold his 3K Flex.

--Shane


Title: Re: flex radio musings...
Post by: KD6VXI on January 05, 2010, 03:09:45 PM
What you propose is not "fixed but lower gain." You are proposing to have less gain on the negative cycle than on the positive cycle. This means the gain changes every time the waveform crosses the zero axis. In other words, it does change with time. Further, a discontinuity like this will be very audible.

Which is called compression.

Limiting implies Brick wall, a la diode peak limiters.  Compression implies a non-linear (usually???) reduction in gain due to an external circumstance (audio signal, overdrive, etc).

At least, that's the way I look at it.

--Shane


Title: Re: flex radio musings...
Post by: KD6VXI on January 05, 2010, 03:15:41 PM
Quote
One thing that would be interesting would be for all stations in a QSO to run DSB reduced carrier, and receive with a synchronous detector.  The DSB-RC could be generated with an analogue transmitter or SDR rig.  The idea would be to run double sideband, with just enough carrier to give the synch detector something to lock onto.  The  received signal would sound identical to a full  carrier AM signal with the synch detector, but to receive it without distortion on a conventional receiver would require receiving one of the sidebands in SSB mode.

Years ago when I worked at Harris RF Communications, our transmitter and receiver systems were capable of that. They could run USB, LSB, ISB, 4ISB, full carrier AM or 'pilot carrier' which was -26db down to synch up with the receivers at the far end. On the sites I helped design and build, we typically used that in a 4ISB mode, 4 independent 3 khz sidebands, -26db carrier. Lower Lower sideband, Lower sideband, Upper sideband and Upper upper sideband. Typically 2 voice channels on the top two, and 2 RTTY channels on the bottom two. The matching receiver had 4 ISB capability also (RF550) and 4 audio outputs.

From time to time I've wondered about doing ISB with voice on one sideband and sstv on the other so a qso could be ongoing while the pictures were being exchanged in the background on the other sideband. I bet that wouldn't be that hard to implement in a flex system - might be an interesting market niche for them.  How would the ssb folks handle that - we'd still  be using two sidebands, but be running ssb!  :o

I'm a very new convert to the flex-sdr realm having been curious for a while and having received an SDR-IQ from Santa.  ;D Now I find myself scanning the for sale ads for used gear, or contemplating a flex-1500... I wonder how well 5 watts will drive a Titan 425?

Myself and another person have done something similar with our "Hi Fi" modulators.

Essentially, it's a Class A series modulator high level modulating a single device, then we feed it into linears to increase the Pout.  I have done a similar design for HIGH level, 16 transistor amplifiers at 12 volts before.  Lots of pass transistors.

BUT, I did find that when using computer injection, it was entirely feasible to FM modulate a 30 Khz tone.  Something inaudible.  It wreaked havoc with the other operators, I could play music, in decent fidelity (limited to 10kc) on an AM channel any amount of KC away.

Boy, it SURE did make the amplifier heat up, though.

I also used negative peak limiting in the modulator, which in effect gave me supermod.  It would make the SB220 have a 75 watt carrier, increasing to 750 PEP.  AVG Pout was in the neighborhood of 500+ watts.  I've uploaded audio clips of it, albeit I was at an S1 to S7 at the RX end....  Depending on mod percentage.

This was 13 devices in the "audio rack", using Reaper or Adobe Audition 3.0 as the audio processors, and both included and purchased plugins. 

Anywho, another musing of "fun" on AM with multiple channels...  I guess my method was SCA, though?

--Shane


Title: Re: flex radio musings...
Post by: w1vtp on January 05, 2010, 04:57:35 PM
<snip>...click and you are right on freq.
And, the ssb signal can sound very good, better than some AM,  <snip>...
I had one good ssb qso, talking with a guy on a tug boat in Boston, talking about old AM rigs, flex radios and tug boat stuff...

I dont think ssb is a bad mode at all, its just the people on ssb that ruin it...

Brett


Eric WB2CAU and I do HIFI SSB (Slopbucket) with some regularity and it sounds just as good as the counterpart in AM.  We both run Flex radios and have calibrated our timebases.  All we have to do is set our xmit BW to whatever we agree on (5 KC for example), set our rigs to the same frequency, set the AGC for "Long" and it's armchair HIFI audio.  I'd upload an example of the audio but I'm not using that computer just now.

Al


Title: Re: flex radio musings...
Post by: K5UJ on January 05, 2010, 06:26:48 PM
Eric WB2CAU and I do HIFI SSB (Slopbucket) with some regularity and it sounds just as good as the counterpart in AM.  We both run Flex radios and have calibrated our timebases.  All we have to do is set our xmit BW to whatever we agree on (5 KC for example), set our rigs to the same frequency, set the AGC for "Long" and it's armchair HIFI audio.  I'd upload an example of the audio but I'm not using that computer just now.

Al
Al, without having heard you I must nevertheless disagree.  It is impossible for SSB to ever sound as good as AM potentially can sound, because a sideband with no carrier is not giving a faithful representation of the baseband tx audio.   What you have is a complex cw signal that varies continually in frequency offset from an imaginary carrier (to represent tx audio frequencies) with volume of the original audio represented by the RF power in the various cw signal frequencies.   This is why analog diode clipping of the audio peaks is okay for AM but it isn't for SSB.  With SSB if you clip the analog audio you wind up txing something like a cw sig with an instantaneous rise (key clicks).   I'm getting into this only to try to illustrate why a SSB sig. doesn't represent audio the way an AM signal does.     Recently, a certain ham radio equipment company has in my opinion, irresponsibly reissued an old speech processor design for SSB that employs clipping by converting the mic audio to RF and filtering it before demodulating it back to AF.  They employ clever wording to make it sound like this is the best thing going of course, ignoring DSP based compression, and multi-band AF processing.  They claim as much as 6 dB increase in average power.  If ur average is 20 watts with a 100 w. SSB rig, that would be almost 80 w.!   It is impossible for that to:  A. Sound good and B. not cause garbage QRM.
</soapbox>

73

Rob


Title: Re: flex radio musings...
Post by: K5UJ on January 05, 2010, 06:40:17 PM

The bigger thing to keep in mind is that negative peak limiting or other magic boxes are usually not needed to run extended positive peak modulation. Just set the polarity of your audio system to take advantage or the natural asymmetry of your voice. I can get 130-150% positive modulation without any special processing. As Fred noted, you just need a transmit system that can produce such modulation.



There are a couple of problems with that however (to my thinking):  1.  Not everyone has a lot of voice asymmetry.   2.  You can run pos. peaks to the moon but as someone, Fred I think, mentioned you eventually run (okay some of us run  ;D) into the limits of the transmitter components.   It's fine to be asymmetric in whatever way you want to do it, but it's also important to build high average power in the sidebands by using tight gain reduction peak limiting so you can compress then drive the final limiter to get dense loud audio that still sounds pleasant.   I would rather have a smaller peak to average than huge peaks with low level vowel sounds.


Title: Re: flex radio musings...
Post by: KD6VXI on January 06, 2010, 03:56:08 AM
The flex 1500 is going to have much reduced spec's but still likely much better than any vintage and most modern receivers.
I think I will get one to listen on in the den.
If they ever come out....its like Apple starting in the garage....a ways to go...

Look at the WonderRadio.  1/2 watt output Flex radio, in kit or assembled form.

I spoke to one of my friends who used the Flex for years on HiFi AM.  His answer, honestly, was "Buy an Omnia Processor".  He LOVES... I mean L.O.V.E.S. his Omnia, and he used that, and the Flex.

I'm still waiting for one of the big boys in the northeast to answer me about his setup....  Although I'm starting to hear more and more musings about using the AM setup as stock, and more people are running to a second computer to do audio processing.


--Shane


--Shane


Title: Re: flex radio musings...
Post by: Steve - WB3HUZ on January 06, 2010, 10:09:43 AM
Different levels of the same thing - both are a nonlinear gain curve.

What you propose is not "fixed but lower gain." You are proposing to have less gain on the negative cycle than on the positive cycle. This means the gain changes every time the waveform crosses the zero axis. In other words, it does change with time. Further, a discontinuity like this will be very audible.

Which is called compression.

Limiting implies Brick wall, a la diode peak limiters.  Compression implies a non-linear (usually???) reduction in gain due to an external circumstance (audio signal, overdrive, etc).

At least, that's the way I look at it.

--Shane



Title: Re: flex radio musings...
Post by: Steve - WB3HUZ on January 06, 2010, 10:24:05 AM
I agree. I was not proposing doing nothing to increase the average level. But if asymmetry already exists, use it to your advantage - don't waste time on circtuits or processing to create it if it already exists. Once that's set correctly, process away to get the average level up.

Just remember, you can never increase the average level of the negative modulation beyond 100 percent, but you can on the positive peaks. Which will be louder, a station with both the positive and negative peaks at 90 percent on average or the station with the negative peaks at 90 perrcent average and the positive peaks at 150 percent average?

Yes, there are limits - how much your transmitter can handle, and how much the receiving end can handle (based on AGC and the detector). I keep my positive peaks below 150 percent, usually in the 120-130 range. Most receivers seem to deal with this OK FB. YMMV.

BTW, you sounded very good on 3885 last night. Hope to hear you again soon.


The bigger thing to keep in mind is that negative peak limiting or other magic boxes are usually not needed to run extended positive peak modulation. Just set the polarity of your audio system to take advantage or the natural asymmetry of your voice. I can get 130-150% positive modulation without any special processing. As Fred noted, you just need a transmit system that can produce such modulation.



There are a couple of problems with that however (to my thinking):  1.  Not everyone has a lot of voice asymmetry.   2.  You can run pos. peaks to the moon but as someone, Fred I think, mentioned you eventually run (okay some of us run  ;D) into the limits of the transmitter components.   It's fine to be asymmetric in whatever way you want to do it, but it's also important to build high average power in the sidebands by using tight gain reduction peak limiting so you can compress then drive the final limiter to get dense loud audio that still sounds pleasant.   I would rather have a smaller peak to average than huge peaks with low level vowel sounds.


Title: Re: flex radio musings...
Post by: K5UJ on January 06, 2010, 12:48:14 PM
I agree Steve, and thanks for the comment on the audio here.  I also run about 130% positive (if the specs on the Inovonics 222 can be relied upon).   I may be out for a few days as I get this buzzard transmission heating problem nailed down and fixed.


Title: Re: flex radio musings...
Post by: Pete, WA2CWA on January 06, 2010, 01:38:40 PM
[
Look at the WonderRadio.  1/2 watt output Flex radio, in kit or assembled form.
--Shane


If you're referring to the one in India, they're out of business.


Title: Re: flex radio musings...
Post by: K3ZS on January 07, 2010, 02:05:04 PM
Referring to K5UJ's comment on RF speech processing,  Angelo W8ERN, an active oldtimer (but not on AM) has the most effective and good sounding standard bandwidth SSB that I have heard on the air.   When people comment on his great audio, he explains that he is using an RF speech processor.    I don't know if he designed it himself or bought it.   By the way, he is one of the designers of the AF-67 when he worked for Multi-Elmac.   



Title: Re: flex radio musings...
Post by: KD6VXI on January 07, 2010, 02:09:53 PM
[
Look at the WonderRadio.  1/2 watt output Flex radio, in kit or assembled form.
--Shane


If you're referring to the one in India, they're out of business.

Well.  Thanks for the bad news.  That's who I was referring to.

Too bad, was a nice little kit.

--Shane


Title: Re: flex radio musings...
Post by: KF1Z on January 07, 2010, 02:26:10 PM
Not only went out of business, but did not deliver radios that were paid for.

And did NOT refund the money!


The radio did not work well at all...
Terrible layout, bad bad construction.

don't waste any money on it it you find one used!



Title: Re: flex radio musings...
Post by: N2DTS on January 07, 2010, 03:07:26 PM
If it seems to good to be true, it is likely a rip off.
Like the 'new' globe king 500...remember those?

Brett


Title: Re: flex radio musings...
Post by: W3RSW on January 07, 2010, 03:34:44 PM
Anyone interested in a Wilsonski 2000A?   SDR, 250MHz DDS, 32Bit FGPA, 1.5kwpep, 400AM  all T/R relays built in, 4 independent receivers, 10Hz to 148Mhz, built in ant. tuner, good for 20:1 all bands.   Oh almost forgot - built in Mac.
Lifetime guarentee,
..no spurs, no gears, no fears...


Title: Re: flex radio musings...
Post by: KF1Z on January 07, 2010, 03:43:41 PM
Anyone interested in a Wilsonski 2000A?   SDR, 250MHz DDS, 32Bit FGPA, 1.5kwpep, 400AM  all T/R relays built in, 4 independent receivers, 10Hz to 148Mhz, built in ant. tuner, good for 20:1 all bands.   Oh almost forgot - built in Mac.
Lifetime guarentee,
..no spurs, no gears, no fears...

Interest and reality may be slightly seperated in this intance...  ;D

(The deal-killer is the built-in Mac)


Title: Re: flex radio musings...
Post by: W3RSW on January 08, 2010, 07:46:35 AM
Oh that was a market specific come-on for Steve.
In the not so fine print you can alternately select a Winnogoes machine, 4Ghz processor , 10gig mem, 1Ghz FSB, 3 ea. 10TGs hard drives, and two 23" monitors.  Uh, ashamed to say the printer is extra.   urk.
AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands