Quality Audio Modulation Monitor circuit design?

[K1JJ]

I'm looking to build up a high quality audio off-air mod monitor -  just the audio detector portion with headphones needed.

Take a look at the fifteen parts after the RF input to the last 220 pf capacitor in this modulation monitor by WA1QIX:  

http://amfone.net/21stAM/mod_monitor.gif

I like the idea that the diode is biased vs: no bias.  I already have a simple diode without bias driving a pair of headphones. It sounds OK, but I thought I could use a better circuit.  

I found I needed a good amount of drive and used a tapped resistor off the 50 ohm line.  And maybe an audio stage to drive the headphones will make it cleaner with less loading on the diode.  This simple fifteeen part circuit could be built easily, but wanted to get some feedback first.

I'm not sure if the three pots need to be trimmers or mounted on the panel for easy access...  maybe the det bias pot is a trimmer.

Any opinions?


BTW, I see this circuit is dated 2002.  Is there a more recent schematic available... is this the latest design used in the no-longer-made REA unit?

T

[K1JJ]

Here's an updated circuit (2008) by WA1QIX that will do the job.  Look about 2/3's down the thread.


http://amfone.net/Amforum/index.php?topic=15743.0


I'll use an audio amplifier after the detector  -  should be all set now.


hmmm....  where do I find a 33 mH choke?  Wind some turns on a ferrite core?

T

[WBear2GCR]

Seems unlikely it is a critical value. Probably you can find some chokes that look like resistors in some bit of old gear, or in some junk box? Inductors in series work too...

Worst case you can order them online from Digi-key or Mouser, but the shipping will be 10x the cost of the parts.

You could wind them with magnet wire on ferrite ah 'spose.

                     _-_-bear

[Pete, WA2CWA]

I've used W1AEX's circuit with good success.

[Steve - K4HX]

Weren't the 33 mH units used widely back in the day by the phone companies? Seemed to be all over the fests in years past.

[steve_qix]

Hi Thom,

Let's see... that circuit is not even close to the REA modulation monitor circuit.

But, to help with your problem - the idea is to have a good low distortion detector followed by a good filter, and then a VERY good audio amplifier.  Somewhere in the circuit, you will need to have a phase reversal available for the headphones.  This is ULTRA critical.

Below is the circuit from the REA modulation monitor that is used to drive headphones or output to an external audio amp with tone controls, etc.



You need a buffer amplifier between the detector output and the headphone amp input.  This can be an op-amp circuit.  

I have a neat little circuit that applies AGC (very slow AGC) to the detector output and buffers the detector, and can feed the headphone amp.  I'll see if I can find the circuit.  Using this, the signal fed to your headphones will be constant no matter what power level you are running (within the limits of the detector, of course).  You can vary the power over an approximate 10x range (50 watts to 500 watts, 200 watts to 2000 watts, etc.) without damaging the detector using the REA low distortion detector circuit.

Regards,

Steve

[K1JJ]

Thanks for posting that, Steve.

Well, I just finished your 2008 circuit using the 2 diodes and the double filter and ran some tests.  


http://amfone.net/Amforum/index.php?topic=15743.0  (2/3rd's way down the thread)

It works very FB!  With my old 1 diode mod mon, I used to hear some slight spitting on my highs, but this design is smooth as silk.  I plan to add an audio amp to it.


BTW, instead of 33 mH chokes, I used 2.2K resisrtors in the filters.  Think that is OK?

TNX..

T

[K1JJ]

Steve,

Yes, let's see that AGC circuit.  I'll give it a try. Sounds like a worthwhile bell to add.

I notice that the detector seems in cutoff (class C ) until it gets a good carrier to work with. Is there a better way to bias it on for weaker carriers or does this need to be done simply with more carrier - with a smaller resistor value giving a bigger RF sample?

T

[steve_qix]

The particular detector circuit shown here is designed to work with 3V peak (6V P-P) of carrier RF input at the diodes, at minimum.  That is very low power - much less than a watt.  What is the final load on the detector output itself?  The detector is designed to look into 22k or thereabouts at the far end.  The entire load should be around 10k.



When properly terminated, it is a very low distortion detector.  You *must* have the RF choke from the input to ground also.  Without that, the input will float at DC and cause distortion.

[steve_qix]

R-Atten shown - 24k - is for a 50 to 500 watt RF pickup unit.  If you need lower power, use a lower value of R-Atten, or for higher power use a higher R-Atten.  I don't know why I put a 1% resistor in the schematic.  The value is not critical.  I suppose I put it there because I am actually using a 1% resistor, but such a precision resistor is definitely not required!

I actually made one of these detectors with a switch for different R-Atten values.  Note:  This particular detector was not used above 40 meters.  At higher frequencies the capacitance in the switch could allow coupling "around" the resistors, which would not be good.

[K1JJ]

Yep, that's the exact one I built and it sounds very FB!

I added a second cascaded filter like the first, as you suggested.

I just built a buffer and phase reversal for it and that works well too.  So, into an audio amp - and we're all set.

Question : I used 2.2K resistors  instead of the 33 mH chokes for the filter. Is that OK?

T

[steve_qix]

Yep, that's the exact one I built and it sounds very FB!

I added a second cascaded filter like the first, as you suggested.

I just built a buffer and phase reversal for it and that works well too.  So, into an audio amp - and we're all set.

Question : I used 2.2K resistors  instead of the 33 mH chokes for the filter. Is that OK?

T

The resistors are probably an acceptable substitute for what you're doing.  The inductances (there are two of them in the mod monitor filter) add another 2 poles to the filter, further reducing the RF component.  At short wave frequencies, the resistors should afford sufficient attenuation, looking into the capacitors.

At 550kHz (the mod monitor was designed to work down to the broadcast band), the inductors are necessary.

[W3RSW]

Sorry for delayed response.

Weren't the 33 mH units used widely back in the day by the phone companies? Seemed to be all over the fests in years past.

The ones I've seen mostly are 88mh center tapped and yes, there were scads of CW tone filters, audio bandpass and RTTY filters, you name it using them.  I still have four or five floating around the shack somewhere.

[Steve - K4HX]

Ya. 88 is it. Poor memory on my part. Thanks for the info Rick.

[K1JJ]

Got the unit working, but a little confused about the circuit performance...

These questions are directed to Steve/ QIX, but anyone else please feel free to comment. We are basing it on the last circuit posted using two diodes...

When I modulate over 100% negative, I see flat topping on the audio output. This is to be expected, even with this dual diode circuit, correct?  I imagine we would need a sync detector to see a rounded waveform when modulating over 100%.   The flat topping follows the phase reversal switch as expected.

I am using a balanced modulator so when over 100% negative, the little wavelets fill in and the resultant RF waveform is still rounded. Even so, the detected audio waveform should still flat top, regardless, just like in a diode receiver, correct?


When I was experimenting with the circuit, thinking I could improve this flattopping, I disconnected the lower diode from the circuit. It had no effect on the waveform at all. So, the output audio waveform seems to remain the same with the simple single upper diode alone with the bottom circuitry out of the circuit.  Just what does the bottom diode do?  Does it add a fixed bias to the anode of the top diode?

I will post some pics of the new unit in a day or so.

Tnx.

T

[steve_qix]

Got the unit working, but a little confused about the circuit performance...

These questions are directed to Steve/ QIX, but anyone else please feel free to comment. We are basing it on the last circuit posted using two diodes...

When I modulate over 100% negative, I see flat topping on the audio output. This is to be expected, even with this dual diode circuit, correct?  I imagine we would need a sync detector to see a rounded waveform when modulating over 100%.   The flat topping follows the phase reversal switch as expected.

Flat topping or flat "bottoming".  I would expect the top of the waveform to be unaffected, even if you're modulating 100% negative. 

I am using a balanced modulator so when over 100% negative, the little wavelets fill in and the resultant RF waveform is still rounded. Even so, the detected audio waveform should still flat top, regardless, just like in a diode receiver, correct?

Well, if the transmitter is putting out RF, the detector should detect it, regardless.  The output may be distorted, but not cut off.  Definitely not flat-topped.

When I was experimenting with the circuit, thinking I could improve this flattopping, I disconnected the lower diode from the circuit. It had no effect on the waveform at all. So, the output audio waveform seems to remain the same with the simple single upper diode alone with the bottom circuitry out of the circuit.  Just what does the bottom diode do?  Does it add a fixed bias to the anode of the top diode?

The lower diode is there simply to present a balanced load to the input, and also to protect the top diode from unusually high peak reverse voltages that will occur if there is no load at all on the negative side of the waveform.

I will post some pics of the new unit in a day or so.

Tnx.

T

Ok, we'll definitely check them out!  I don't have any direct experience running the detector with an "overmodulated" balanced modulator, but with overmodulated standard RF amplifiers, the results are completely predictable, and the output wavefrom from the detector is an accurate representation of modulated signal.  I have run extensive tests in this area.

Regards,

Steve

[K1JJ]

OK on all, Steve.

Yes, it is "flat-bottoming" when the balanced modulator goes past 100% negative. So this is a normal thing from what you are saying.  The top peaks (opposite the flat bottoms) seem to have no practical limit.

I have four different RF atten resistor values I can switch in to cover 10 watts to 1500 W. I added the PTT relay for external receiver audio input  as well as a phase rev switch and buffer.   Made up a 7 watt audio amp inside.  Painted white with black lettering, of course.  Cool unit.

I'll run some more tests with both tones and voice and see what I get.   Yes, it is an odd situation with the bal mod able to go beyond negative 100%.  With the class E rig, the neg peak limiter wud prevent it.  

When you are hitting your neg peak limiter hard  on the class E rig, can you see (tapped off the detector into the scope)  and hear in the headphones, the bottom flattening produced by the limiter clipping?


T

[K1JJ]

A few pics of the new, completed AM Detector unit in the QSO section:

http://amfone.net/Amforum/index.php?topic=33293.0

[steve_qix]

When you are hitting your neg peak limiter hard  on the class E rig, can you see (tapped off the detector into the scope)  and hear in the headphones, the bottom flattening produced by the limiter clipping?

T

Yes, it is absolutely noticable on all monitors.  The mod monitor "brickwalls" at about 92% neg and the scope display shows a distinct "minimum carrier" when the negative peak limiter is brought into conduction.  It is also apparent in the headphones as well, although the clipping has to be reasonably heavy before one can really hear it.

Regards,

Steve