The AM Forum

Hi!

I'll provide more details in a later post... but I just finished using a test signal that is perfect for tuning up an AM transmitter or a linear amplifier. I produced it using the latest version of Spectrum Laboratory... and I'm amazed at how well it works.

Using the features of Spectrum Laboratory, you can easily construct a periodic audio signal with an arbitrary shape... and then Spectrum Laboratory will "play" it on the output of your sound card... as well as displaying the audio spectrum of the signal. You can vary the period (1/repetition frequency) to suit your needs.

I'm doing this using a separate sound card on my computer, which is not the sound card I am using for running Flex Radio's Power SDR software. Any plain vanilla sound card will be fine for generating the test signal.

The waveform I am currently using is shown in the attached .jpg slide. I adjusted it for a repetition frequency of 1kHz. Thus the audio spectrum consists of discrete tones which are multiples of 1 kHz.

Be careful not to put this signal out on the air. Use it with a dummy load. The harmonics of the audio signal will fill up whatever bandwidth your transmitter can pass. In the case of my externally modulated 6L6 transmitter, I was amazed at how closely the envelope of the rf output of the transmitter tracks the audio waveform. I'll post some pictures of the actual audio input and rf output waveforms in a later post or on my web site.

This test waveform allows me (for example) to adjust the loading on my linear amplifier to accommodate the positive peaks I want to put out from my driver transmitter.

For those who already know how the get to the audio waveform generator programming screen in Spectrum Laboratory... the formula to paste into the arbitrary waveform generator is

(x<0.2)-0.67*((x>0.2)&&(x<.5))

I assume that if I record a .wav file of this tone, you could just play that into your audio chain.

I observed that the best way to run the test was with the audio processing completely bypassed, except for wideband gain.

Best regards
Stu

Stu,  I am not familiar with your software.    If it can generate an audio bandwidth limited white noise, that would work also.    Generating a frequency hopping audio signal fast enough produces the same effect.   I used to use an HP digital function generator to test the spectrum of a sonar-radar combination (called RASS) using this method.    The sound card is much cheaper than the $6K for the HP unit.

Hi Stu,
       Could you post a .wav or .mp3 of the waveform? I'd like to give it a try. I've been using everything from white noise to The Who for doing A/B comparisons and the old yayyy fourrr when tuning the amp. I have also used a sound card generated two tone pulse that seemed to work ok. Thanks, Phil

Phil

Can I send the .wav file (10 seconds of audio, 2MB) to you... and ask you to post it?

I tested it by playing it on my Microsoft Media Player directly into the audio amplifier that modulates my transmitter. The waveform is pretty much the same as when I play it directly from the Spectrum Laboratory audio signal generator.

What e-mail address should I send it to?

Stu

I think you can simply add it as an attachment to your next post. Maybe the file is too big? Otherwise you can send it to w2phl@aol.com. Thank You!

Phil

Looks like it was okay (not too big).

I attached it to my previous post.

Stu

Looks pretty kewl on Microsoft OsillyScope...

This waveform is even better. The repetition rate is 100 Hz. The positive pulse occupies 2% of the total duration of one period, and the negative pulse occupies 4% of the total duration of one period. Thus, the stress on the modulator and the transmitter is much less... and more representative of real voice peaks.

Stu

I hope to have time this weekend to do some testing. Thanks again Stu.

Attached, below, is a photograph... taken from my oscilloscope... of a) the audio input to my audio power amplifier (blue) and b) the rf output from my home-brew legal-limit plate-modulated rf amplifier (yellow) ...using AM Test Waveform 2.wav (posted above)

1. I am using the .wav file I posted, and playing it from Windows Media Player into the input of the audio power amplifier. The blue waveform is (as stated above) the actual audio waveform at the input to the audio power amplifier.

2. The plate-modulated RF amplifier was (in this test) putting out 350 watts of carrier into a dummy load.

3. The plate modulated RF amplifier is modulated in a modified Heising configuration. The modulation transformer is an Antek toroidal power transformer, with a 1kVA rating and a step up ratio of approximately 24:1 (actually two off-the-shelf Antek power transformers in series, with step up ratios of a) 67.5:115, and b) 115:1600, respectively)

4. I am driving the plate modulated RF amplifier in grounded grid configuration with my Johnson Ranger. The Ranger is externally modulated in modified Heising configuration. The modulation transformer for the Ranger is an Antek toroidal filament transformer (backward connected) with a step up ratio of approximately 19:1 (12:230). The Ranger is plate modulated with approximately 30% negative peaks to produce a modulated input on the cathodes of the GS-35b tubes in the RF power amplifier.

Note:

A. The rf output (from an rf sniffer) of the RF power amplifier is faithfully following the modulation waveform from 100% negative peaks to over 150% positive peaks (confirmed by one of Steve's modulation monitors... as well as the scope)

B. The horizontal scale is 200us per division ... which demonstrates the very wide and flat bandwidth of all of the Antek toroidal transformers.

Best regards
Stu   

You can "write" an arbitrary waveform with Goldwave....it's been my only digital audio processor for ages! 

Another great signal would be a fast sweep, "Bronx Cheer" signal, because of what it sounds like on the air. :)

(Again, don't actually use this on the air. :)

eric

Eric

Next time I find myself entertaining a group of six year old boys... I'll keep your suggestion in mind.

For tuning up AM transmitters... I'll stick with the waveform I have posted.

Stu