The AM Forum

THE AM BULLETIN BOARD => Class E Forum => Topic started by: K8DI on March 20, 2020, 02:58:57 PM

Title: some ideas to be shot down...
Post by: K8DI on March 20, 2020, 02:58:57 PM
Well, I have a lot of extra time, since my day job depends on construction schedules that have been paused. I can't really get on the radio with this time, because I have a cold /sore throat that won't go away and makes talking painful.  So I'm gonna build something.

But I was musing about class D topologies.  People have made H bridge and push pull class D transmitters which are line powered without a proper PSU, and are driven with transformers (that also give them the two drive polarities) and themselves drive output transformers (that combine the sections) to provide the RF output. Being isolated by those transformers they can let the RF section sit on the AC line. Some have modulated that with another transformer, sometimes a power transformer, driven by an audio amp. 

So starting from there:

Say I have a typical mid 80's commercial grade dual channel audio amp. It has dual power supplies and totally separate channels.
Say I gut one channel's audio parts, keeping its power supply, heat sink, and fan.
Say I build up a class D RF amp with input and output transformers (DC isolated type circuit) on that heat sink.
Say I connect the power transformer in full wave instead of as a bridge, so it is single ended, say 80-85 volts.
Then I connect the FET sources together, take that point, and connect the other amp channel's speaker output to it.

The other channel can swing audio +/- 85 volts.
At +85, no current will flow through the RF FETs because both ends will be the same potential -- 100% negative mod.
At -85, the voltage across the FETs would be 170, 100% positive mod.
No audio (dead carrier), that amp output, having a very low impedance/high damping factor, will be at 0 volts. Then the FETs will see 85 volts and make carrier.

The issues I see right now are
  bypassing the RF across the audio amp output, need a good RF cap
  dead carrier/no audio will be putting a large constant current through the audio output transistors/they will run hot

The pro here is I have at least ten boat anchor commercial dual mono amps in a pile to be recycled. There's no cost to me (other than the RF parts) if I turn one into smoldering ash.

So who thinks this might work? Who thinks I am nuts??


Title: Re: some ideas to be shot down...
Post by: Opcom on March 23, 2020, 10:35:43 AM
Never shoot down an idea, I prefer to say.

Agree with repurposing dual channel amps, especially with dual power supplies, if the amps are separate and one can be removed to make room for the RF section.

Will there be enough room for the coils, etc? If not, and you have identical amps, you may be able to 'stack' the cabinets, cutting a hole in the lower one's top and top one's base, to make a taller unit that could still look good. This used to be easier when things were rackmount and had definite removable top and bottom panels.

Alternately, gut and convert one to RF, and use another identical unit for the audio and power supplies, and with the spare room left by removal of one audio channel in the 'audio' unit, and both audio channels and power supplies in the "RF" unit, gain space for all the control circuits, fans, etc.

only mechanical suggestions here. No experience with class D, E, etc..

Title: Re: some ideas to be shot down...
Post by: w8khk on March 23, 2020, 12:45:07 PM
No, Ed, I certainly don't think you are crazy.  I personally take great pleasure in re-purposing obsolete or surplus devices, rather than just stripping them for parts and starting anew.  But when I have a cold and/or sore throat, I lose all ambition to be creative.  You are ahead of me on that score!

A couple suggestions, if I may:

First, since you are going to nuke one of the amplifiers anyway, why not do a hopefully non-destructive test on that side before starting to disassemble?  Test your assumption that the DC current through the audio output may or may not cause a failure.  Load that side with the DC supply in series with a load resistor similar to the impedance of the RF stage you plan to add.  Hang a scope probe across the resistor and critically examine the output waveform under those conditions.
Run it with no audio, then moderate to full-strap audio, and see if that audio amplifier holds up.  If it does not, consider feeding DC to your RF stage through a modulation reactor, capacitively coupled to the audio output of the remaining amplifier.

One thing to consider when driving a load with a DC offset:  Look at the schematic of the audio amplifier to determine whether a DC load will upset any feedback loops, possibly causing a bias shift such that some transistors are close to cutoff and others are closer to saturation.  If this is an issue, perhaps you could change the feedback such that it passes low frequency audio, but does not pass DC.  (This may not be an option if the amplifier requires DC feedback to stabilize the overall operating point.)

Second, you may want to test to be sure the remaining audio stages are stable in the presence of RF, before you build the high-power FET amplifier in the vicinity of the low-level audio stages.  If exposure to RF causes instability or distortion, either shield the RF half from the audio half, or consider Patrick's suggestion of using two boxes to get the job done.  This has the benefit of providing a full shelf of spare parts, and less to recycle.  If totally successful, post your excess amplifiers in the "for sale" section here, with instructions to build a strapping transmitter!

Good luck on your project, sounds like fun.  


Title: Re: some ideas to be shot down...
Post by: K8DI on March 23, 2020, 01:58:21 PM

Thanks for the comments. I will have to go connect a power supply to the output of one and see what it does.

Then I thought, I wish I still had an old Crown DC300 laying around....  For those who do not know, DC stood for direct-coupled. The DC300 would amplify DC.  You can hook a battery to the input and turn it into a variable power supply -- turn up the gain, the output voltage goes up.  I could connect a battery and AC coupled audio into the input, have a standing DC offset plus audio on the output, and use the amp as the class D/E RF deck power supply!  Ah, well. I will have to try that if I run across a DC300 someplace....


AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands