ALC for use with a modulation monitor (e.g., REA)

[AB2EZ]

I have finished the design and construction of a small unit that serves as an automatic level control between the sensor and the main unit of my REA modulation monitor.

As the input signal level to this circuit (from the sensor) varies over a +/- 6dB range, with a time constant of 1 second or longer, the output signal level changes very little. Thus, one does not have to tweak the "set carrier level" control.

This circuit is transparent to normal audio frequency variations of the envelope... but tracks out slowly varying changes in the overall (carrier) level.

One could use a larger capacitor at the output (e.g., 1000uF instead of 220uF) if one wants to further reduce the frequency at which the variable gain control circuitry will begin to track out the modulation.  The circuit used here should be good to, at least, below 20 Hz.

http://mysite.verizon.net/sdp2/id18.html

Best regards
Stu

[WA1GFZ]

Stu,
You gave pretty high modulation levels from my V2-CDC. When I check it on a scope the rig barely hits 100%. I think you are seeing a transient and the monitor reacts to the transient as a modulation peak. This could be AGC reaction time letting a transient slip through.
I've been meaning to send you a message that sumptin didn't add up. My high voltage is pretty stiff now changes maybe 30 volts on peaks so may have allowed a bit of a transient peak but I think your numbers were around 130%. I couldn't see it on a scope although didn't set up a trigger to catch the transient. gfz

[AB2EZ]

Frank

Hi!

Your observation refers to the use of the modulation monitor to attempt to measure the modulation levels (positive and negative) of a received signal. In that case, I am using the 455kHz if output of my 75A-4 to drive an external envelope detector... which, in turn, drives the modulation monitor.

You are correct... this is a much trickier measurement to make than measuring the modulation of your own transmitted signal with a local off-air sensor connected to your transmitter's output.

The difficulties include: rapid fading of the received signal, noise and interference on the received signal, inadequate bandwidth of the receiver's rf/if chain, and the nature of your receiver's AGC control methodology.

a. Fading

If the received signal that you are trying to measure fades down, the AGC circuit inside your receiver will turn up the gain of the receiver's rf/if chain to compensate. When the received signal you are trying to measure fades back up again, the AGC circutry takes time to re-adjust... and, until it re-adjusts, the positive peaks that are displayed on the modulation monitor will be higher (perhaps much higher) than the true positive modulation peaks on the received signal. Fast fading makes this particularly problematic. Selective fading makes things much worse.

b. Noise and interference

Noise and interference add to desired, received signal at the output on your receiver; and, therefore, diisturb the modulation monitor's readings.

c. Bandwidth of the receiver's rf/if chain

If the bandwidth of the receiver's if chain (e.g., the selected if filter in the 75A-4) is too narrow, the positive peaks (and, in some cases, the negative peaks) of the incoming rf signal may be reduced... thus giving a relatively low measured value for the true modulation peaks of the (other guy's) signal. I.e., depending upon the amount of compression the other guy is using, and the bandwidth of his modulation chain (including both his audio chain, and his transmitter/modulator) his actual transmitted positive peaks may be significantly higher than the positive peaks that come out of your receiver's rf/if chain.

d. The nature of your receiver's AGC control methodology

It is best, for measuring the modulation index of a received AM signal, to use a receiver whose AGC control circuit reacts to the averagevalue of the rectified if output signal. However, many receivers, including the 75A-4 use an AGC control methodlogy that reacts to the peakof the received signal... including the modulation peaks of an AM signal. Depending upon the attack and decay time constants, that type of AGC circuit will tend to track out slowly varying components of the received signal's modulation... and, in general, make it diffcult to measure the modulation levels of the received signal. When I use my 75A-4 to measure other people's modulation indices, I turn off the 75A-4's internal peak-detector-based AGC (a.k.a. AVC), and I inject my own, externally generated AGC control signal... which is proportional to the average value of the 75A-4's i.f. output signal

Bottom line:

All of the above means that is is usually difficult to measure the outer guy's modulation peaks, with any degree of precision/accuracy... unless: his signal is well above background noise and interference levels, unless the fading is slow, unless you open up the bandwidth of your receiver to at least +/-  4kHz, and unless you turn off your receiver's peak-detector-based AGC control circuit.

Best regards
Stu

[WA1GFZ]

YUP, best way is when you were listening to Nick with an audio generator.
Real peak is very hard to catch. I just wanted to let you know there was quite an error. Maybe manual gain control is best....fc

[AB2EZ]

Frank

Agreed... but, again...

All of this has to do with the use of the modulation monitor to measure the modulation of a received signal ... which is not what it was designed to do (i.e., an "off label use")

None of these problems/issues come up when using the modulation monitor as it was intended to be used... i.e., to measure the modulation of your own, transmitted signal.

Best regards
Stu

[WA1GFZ]

I agree Stu, But it would be nice to make your system work....

[steve_qix]

Well, I guess I'll jump in, too !

I've been using my Modulation Monitor with my so-called HI FI receiver for a while now, and I've had some experiences similar to those of Stu, but I also get very good results under stable band conditions.  I should point out, I am using a different AGC circuit, however I do not believe this necessarily makes any real difference. The function of my circuit is the same as the one Stu is using - to keep the carrier input level fed to the modulation monitor steady under receive fade conditions.

I get VERY accurate readings IF there is no selective fading.  I've observed many, many stations under all conditions.  Selective fading is the biggest problem with respect to getting good readings, because of course the modulation level relative to the carrier changes under selective fading conditions.

But, with reasonably steady signals, the measurement accuracy of received signals is quite good.  I've actually run experiments with other folks who are also using the REA modulation monitor connected to their transmitter, and their level of modulation as measured in their own stations with the monitor directly connected agrees closely with that I see over the air, with the monitor connected to the receiver - again, when no selective fading is happening.  In fact, I did this very thing this evening.

It is quite interesting to scan the broadcast band with the mod monitor - receiver combo and observe what they are doing.  There's a lot of over modulation out there [in the negative direction], and out of phase AM stations !!!  Many stations only modulate about 50-70 % - and this may be on purpose, to reduce the effects of selective fading at night.

Now, a few other notes:  I have characterized the receiver and the [external] AGC controlled RF pickup using triangle waves and by sweeping the audio range using a very low distortion test transmitter.  The end-to-end performance is excellent - triangle waves are undistorted.  The AGC circuit has no effect on the audio quality to 20hZ.  At 10hZ, the AGC is starting to cause waveform modification, due to a small amount of audio starting to show up in the AGC.  This is a compromise between reasonable response time (about .5 second), and good audio performance of the AGC. I figured 20Hz was probably a reasonable low frequency :-)

Interesting stuff, for sure !

Regards,

Steve

[WA1GFZ]

Steve,
It will be very hard to deal with fading and tracking gain. The problem as I see it is the attack time of the AGC too slow making modulation look higher than it is. Again hard to make right without some delay in the system or range gating so to speak. Don't look at modulation till gain is right.
Range gating is a radar trick to improve signal to noise ratio and reduce jamming. The receiver is setting at high gain but gated off until the return pulse is expected. Just before it arrives the gate is opened. This could be as easy as a monitor on the IF output. Don't open the path to the mod monitor until the average level is right. Again easy to say but hard to do.
I ran into the same problem designing the noiseblanker in my homebrew rx. I put a delay 400 ns in the main channel to give me time to detect noise and generate a blanking pulse.
Your old 813 pdm rig had so much modulation that I had to disable the blanker.
I'm interested in how well you guys deal with this. I burned out on it after awhile.

[John K5PRO]

Steve (QIX),
Do you have any figure for accuracy for the REA mod monitor? Broadcast monitors (Belar and TFT at least) typically quote +/-2% for their meter,when using a 400 Hz sinewave tone at 100%. Peak flashers are said to be quite accurate for TFT and not spec'd at all for Belar AMM2.

A local 1 kW AMer is looking for a monitor, and as you know, FCC has dropped the requirement for stations owning them. Also, I doubt the FCC type accepts them anymore.
The probable reason that so many stations are far off nominal, is that they aren't monitoring anymore. FCC will still bust them if they confirm that they are overmodulating during an inspection.

73

[steve_qix]

Steve (QIX),
Do you have any figure for accuracy for the REA mod monitor? Broadcast monitors (Belar and TFT at least) typically quote +/-2% for their meter,when using a 400 Hz sinewave tone at 100%. Peak flashers are said to be quite accurate for TFT and not spec'd at all for Belar AMM2.

A local 1 kW AMer is looking for a monitor, and as you know, FCC has dropped the requirement for stations owning them. Also, I doubt the FCC type accepts them anymore.
The probable reason that so many stations are far off nominal, is that they aren't monitoring anymore. FCC will still bust them if they confirm that they are overmodulating during an inspection.

73

Hi John,

The actual accuracy of the AMM-HF1 mod monitor as shipped from the factory is between 1 and 2 %.  The calibration is pefrormed at 100% modulation with a 400hZ sine wave.  The accuracy is further verified at 90% and 95% modulation, and the positive peak accuracy is checked at 150% positive.

Since the AMM-HF1 is peak reading, as opposed to quasi-peak, the shape of the modulating waveform (triangle, complex, dynamic, etc.) doesn't change the accuracy.  I find this feature very desirable for AM, where there are absolute peak limits (particularly in the negative direction).  I want to know the peaks, not the quasi-peaks.  This is probably the biggest advantage the AMM-HF1 has over analog metered equipment.

I found it interesting that the FCC doesn't require type approved equipment for monitoring.  The AMM-HF1 was going to be submitted for FCC type approval, but there's nothing about it in the rules anymore !  None of the competitive, new equipment (Innovonics, Belar) seems to have an FCC number either.  It appears as if much of the equipment is "self certified" by the manufacturer.

Broadcast stations have purchased and are using the AMM-HF1 - replacing existing, working equipment!  I think the main draw is the concurrent display of positive and negative peaks, so you get an immediate display of the asymmetry in the modulation.  I have to admit, that's what I really like, too.  It's so nice to sit there and see, one above the other, your positive and negative peaks layed out for you - and with peak hold.  The modulation is right there, the good, the bad and the ugly :-)  (and some of the stuff I see out there is pretty ugly     )

Talk later and Regards,

Steve