Modulated Audio Stage for Ricebox/Amp Combination

[N1BCG]

There seems to be a lot of interest in better quality audio than can be achieved with many of today's rigs. Most are designed for that punchy DXing sound, but often times something a bit cleaner and more natural is preferred for ragchews with friends.

That led to this somewhat half-baked idea of creating an outboard modulated stage that could be switched in between a rig and a linear amplifier. The idea is to drive it with an unmodulated carrier of just a couple of watts as well as an audio input. The output would be set to whatever the amplifier needs, which can be accomplished with relatively low power. I have two audio tubes in PP and a single RF amp in mind. Screen modulation simplifies this even more since the output requirements are so modest.

Has anyone tried this? Ironically, I own neither a solid state rig nor an amplifier, but if I did, I'd rather try something like this than drill and blast an expensive transceiver.

[kb3ouk]

Haven't tried it but I did consider it at one point. Even had some of the parts for it. The idea was going to be a single 1625 (or an 807 would work, but the 1625 I already had) driven by my Icom 718. To modulate it, I was gonna use a 100 watt audio amp I already had, using an old power transformer I had as a modulation transformer. Planned on running it at about 20 to 25 watts. All I really needed was parts for the output tuning and a power supply, but then the linear I had gave it up and I decided to change directions, now I'm thinking of going solid state class E.

[KD6VXI]

Yes.   I have exactly what your talking about.

I used a mosfet stage,  100 Watt pep.   Stripped the carrier operated relay out of a CB amplifier.   Built tuned input.

Modulate the stage with a pulse width modulator.   Takes about 2 watts of carrier,  and I get about 100 watts of hi fi audio out.

When you unkey,  12 volts is switched off to the pdm, and the relay goes back to shorting the two antenna Jacks.

I basically built a modulated linear.   But it works good,  and has some bodacious loud and clean audio.

--Shane
KD6VXI

[Pete, WA2CWA]

Using the transceiver as a VFO/driver to drive an outboard modulated RF amplifier has been done.  Don't some of the Class E users do this now? However, there are a number of modern rigs that now provide access directly to the internal modulator, by-passing all the audio tailoring that normally hangs between the mike input and the modulator of the transceiver. Even with some of the older modern rigs, inputting your audio directly to the balanced modulator, also by-passing all the audio tailoring,  is not that difficult to do.

Hi-Fi sounding radios are highly over-rated to have/use given the atmospheric static, adjacent frequency interference, possible fading, and other atmospheric/interference anomalies that you can normally find on the amateur bands. If I can understand what you're saying, loud and punchy works for me even on AM. At least I won't doze off especially if it's an old buzzard transmission. 

[WD8BIL]

Interesting. The IC 718 can be made to sound really good by simply feeding line level audio into the acc jack. A lineeear is all that's needed after that. Why re-invent the wheel?
 

[ka1tdq]

That's exactly what I do.  I use a ricebox as a VFO delivering 10 watts of CW carrier to an external RF deck.  The audio drive comes from a car audio amp.  The single MOSFET operating in class C delivers 23 watts carrier, but I knock it down a little before it goes to the linear. 

I don't want to start another discussion about positive peak modulation >100%, but it does sound good.

Jon
KA1TDQ

[W9ZSL]

Could you plate modulate a linear?

[N1BCG]

Well, plate modulating a linear changes several things in an interesting way. If you have a modulator that can provide at least half your desired carrier power then the bias could be changed on the amplifier from linear operation to the vastly more efficient class C. The only reason for low efficiency linear operation is to amplify a fully modulated carrier.

That idea certainly opens up the discussion!

[KD6VXI]

Yes,  you can plate modulate a linear,  but you have to bias it class c.   Otherwise,  the needle swings,  but no audio is mixed onto the carrier.

You can also cathode modulate a linear.   

You can also s reel modulate a tetrode linear.

An interesting aside.   If your going to do this with a bipolar stage,  you need to modulate both a driver and final stage.   If going the mosfet route,  a single stage will be enough.

If anyone wants pix,  I'll upload some of the rig I built.

--Shane
KD6VXI

[kb3ouk]

Interesting. The IC 718 can be made to sound really good by simply feeding line level audio into the acc jack. A lineeear is all that's needed after that. Why re-invent the wheel?
 

Well in my case, I could've done that and it would've worked just as well, but I wanted to be different. The other reason was the 718 I have I believe has a set of somewhat weak finals in it and struggles to put out anything over 60 watts pep, and 15 watts wouldn't drive the amp I had very well. In that case, it would've been easier and cheaper to reinvent the wheel than replace the finals in the 718, plus without the amp I would've had a pretty decent pw rig.

[Opcom]

Plate modulating a linear amp that is running at its normal voltages can cause electrical overstress of the amplifier's output section. It is the same result, if it is set to class C, the electrical overstress.

What I mean is that something store-bought designed to run at 2000V as a linear amp is just not made for higher voltages. I can see the sparks flying from (someone else's) NCL2000 or SB-220 when presented with 4500V peaks.

A sensible thing to do is to reduce the amp's plate voltage to half and then plate-modulate from there. Check the ratings of the RF tank and plate bypass caps, etc.

/\/\/\

Assuming this has been done and the amp is safe from overstress and will be kept linear, feeding its input with a 100% modulated signal and then modulating its plate voltage 100% with an identical in phase copy of that audio signal would cause a huge increase in modulation because the (stereotypical) carrier would be 1/8 the PEP instead of 1/4.

No extra power would be had over the linear amp's usual rating but it looks like the modulation would be expanded in the positive direction. I don't claim it is a good thing or suggest it. The diode detectors and AGC in the good old receivers, as has already been beat to death, don't like >150% positive mod.


Variations of this scheme of increasing modulation to silly levels have surfaced in interesting arrangements where the plate voltage was not modulated.
A cathode driven 813 linear amp has added to its input a damper tube voltage doubler to feed a positive audio envelope voltage to the screen grid. Interestingly the cathode loads the driver on the negative half cycle of RF and the screen rectifier loads the driver on the positive half cycle of RF. An RF Amp made this way is called, in the low vernacular, a "modulator". The schematic for that mono-band amp example is around here somewhere.

[w4bfs]

Variations of this scheme of increasing modulation to silly levels have surfaced in interesting arrangements where the plate voltage was not modulated.
A cathode driven 813 linear amp has added to its input a damper tube voltage doubler to feed a positive audio envelope voltage to the screen grid. Interestingly the cathode loads the driver on the negative half cycle of RF and the screen rectifier loads the driver on the positive half cycle of RF. An RF Amp made this way is called, in the low vernacular, a "modulator". The schematic for that mono-band amp example is around here somewhere.

sounds somewhat like the G2DAF amp ckt only it was grid driven rather than grounded grid ... I would like to see a schizmatic of this 'cause I have friends in low places where the beer chases my blues away and i'll be ok ...

[N1BCG]

I can see the sparks flying from (someone else's) NCL2000 or SB-220 when presented with 4500V peaks.

Indeed, and not an insignificant point for experimenters emboldened by this thread. Screen modulation would take advantage of the linear design of the amp and would eliminate the need for a large modulator.

[Opcom]

I have searched as best I can and can't find the schematic. It could be the G2DAF and I have misspoke, or maybe it is actually on paper, in which case who knows where. Anyway 300K+ files on the computer and it could be named badly

[KD0HUX]

Yes,  you can plate modulate a linear,  but you have to bias it class c.   Otherwise,  the needle swings,  but no audio is mixed onto the carrier.

You can also cathode modulate a linear.   

You can also s reel modulate a tetrode linear.

An interesting aside.   If your going to do this with a bipolar stage,  you need to modulate both a driver and final stage.   If going the mosfet route,  a single stage will be enough.

If anyone wants pix,  I'll upload some of the rig I built.

--Shane   :)Yes please post some pix
KD6VXI

[Pete, WA2CWA]

Yes,  you can plate modulate a linear,  but you have to bias it class c.   Otherwise,  the needle swings,  but no audio is mixed onto the carrier.

You can also cathode modulate a linear.  

You can also s reel modulate a tetrode linear.

An interesting aside.   If your going to do this with a bipolar stage,  you need to modulate both a driver and final stage.   If going the mosfet route,  a single stage will be enough.

If anyone wants pix,  I'll upload some of the rig I built.

--Shane   :)Yes please post some pix
KD6VXI

When you quote someone and respond, watch where the quote [ ] brackets are, especially the trailing [/] quote bracket. When your response is within the quote, it's sometimes difficult to tell who is saying what.

Yes,  you can plate modulate a linear,  but you have to bias it class c.   Otherwise,  the needle swings,  but no audio is mixed onto the carrier.

You can also cathode modulate a linear.  

You can also s reel modulate a tetrode linear.

An interesting aside.   If your going to do this with a bipolar stage,  you need to modulate both a driver and final stage.   If going the mosfet route,  a single stage will be enough.

If anyone wants pix,  I'll upload some of the rig I built.

KD6VXI

--Shane   :)Yes please post some pix

[KD6VXI]

Will do,  tomorrow.   Rig is in shop,  I'm in bed :-)

--Shane
KD6VXI

[W9ZSL]

I was just curious but mainly still wonder if I can use the Kenwood 440 in CW mode as an exciter because it does have variable output and an 813 doesn't need much drive.  It would save me the trouble of building a separate exciter.

[N1BCG]

Plate modulation as you originally mentioned would work for a triode but would require the amp to be derated with a lower plate supply voltage. With an 813 (tetrode) based amp, screen modulation would be easiest and require less drastic mods and significantly less modulator power.

That seems to be the general summary of the posts on this topic.

[KD6VXI]

Sorry,  got side tracked.   

Years ago,  my Homebrew cable from ts440 to sb220 went in wrong,  1 pin off.   Shoving 110 DC into ptt.   No,  wasn't pretty,  but,  after going on 7 years learned enough digital to find the problem and repair it.   So didn't get pictures until this morning.

So,  crappy transmitter porn here we go....

[KD6VXI]

When I get to the pc,  I'll label a little better.   But,  that's an aftermarket rf stage contained in the heat sink on back.

From input,  rf is routed to switch.   On rx,  relay just shorts ant to radio side.   Hit rf in with above a qtr Watt and it will switch.   This activates a few things.

Rf in is put through tuned input, matching 50 ohms to approx 10 ohm input.   Without this,  my exciter would heat up and power drifted.   I also noted modulation or something of the incoming signal with no tuned input,  morphing to distortion on output.   

The rf sense circuit also switches a pulse width modulator on.   The PWM runs on 13 volts,  and I can use up to 200 vdc on the output switching fets.   Output switches are floating, and I used 600 volt 14 gauge wire on the torroids.  Eventually,  I'm planning on replacing the mosfets with some 1.5 kv units,  and trying to direct plate modulate a gi-7....  We shall see.

Pwm filter is closely copied from another design I found online.   Why reinvent the wheel.   Just tweaked the output cap value to get it for the load impedance of the amp.

Npl and anti alias filter is close to the one Steve designed years ago,  but working on single ended voltage.

I'm currently using 22 to 24 volts on the output,  and can run  from 2 watts of carrier (that's the input drive level)  to about 30, and see pep values of up to 130.  With npl engaged,  I can make crazy stupid mod percentages.

Oh,  the rf amp is class d.  Very efficient setup.   It's a 10 meter amplifier,  but runs decently from 10 to 12 meters.

I'm planning on more rf boards for more bands.   Have some ARF devices here for other bands.   Just got sidetracked on my ts440 vfo lol.

--Shane
KD6VXI

[Opcom]

sounds somewhat like the G2DAF amp ckt only it was grid driven rather than grounded grid ... I would like to see a schizmatic of this 'cause I have friends in low places where the beer chases my blues away and i'll be ok ...

Yes it is that one or much like it. He also made some bigger amps with the same principle. All grid driven. I still think I have seen a GG version though, powering the screen from the rectified envelope. I remember because the unique claim was to have made use of both half cycles of the RF drive. No joy on that schematic, but did find the G2DAF files.

If nothing else hams always find interesting ways to do stuff. - this topic for instance!