The Belcher - Working pulse unit for Linear Amp Tune-Up - Pictures

[K1JJ]

Here's the finished product in service. It's working FB on the rigs.  This pulser was designed by Mike, N5RLR. It generates a ~65 hz pulse at a duty cycle of about 10%.    Injected into the audio of an ssb rig (or AM rig) will simulate full peak power at low average power. Great for tuning up the amplifier without heat stress, etc.

I also used it to test my amplifer for bandwidth on ssb. Into a dummyload, it is a good reference for bandwidth checks when changes are made.

Notice the last picture of the lonely pulse.  This pulse get filtered into a damped sine wave pulse on the ssb rig. I listened on the receiver and the test pulse is only about 3kc wide, just like a normal voice. However, on AM it will be very wide and should be tuned into a dummy load.

The third picture of the ssb filtered pulse should show just ONE waveform - the double waveform is a picture freeze error.

The bottom line is this unit will enable the amplifier to generate peak power of say, 1500w on the wattmeter and scope, while the average power is only about 150w.

Thanks again to Mike for an excellent design and schematic!

Here's the schematic and pics.

73,
Tom, K1JJ

[K1JJ]

Picture of the raw pulse and the schematic below:

[AB2EZ]

Tom

Neat circuit!

As an alternative:

Play the attached MP3 file into your audio chain from your computer or your MP3 player.

The waveform is repetitive... with each cycle consisting of (see attached .jpg):

A) A fast rise / fast fall square-shaped positive-going pulse of amplitude A and duration 2ms, followed by

B) 2ms of silence, followed by

C) A fast rise / fast fall square-shaped negative-going pulse of amplitude -A/2, and duration 4ms,  followed by

D) 192 ms of of silence prior to repeating.

The repetition rate is 1/(200 ms) = 5 Hz, the duty cycle is very low (~ 6ms / 200 ms = 3%); and the pulses simulate voice peaks (positive peaks larger than negative peaks).

It works great for loading up linear amplifiers, and for testing modulated transmitters.

Best regards
Stu

[kb3ouk]

where did you come up with a name like that?
Shelby

[K1JJ]

where did you come up with a name like that?
Shelby

Yep, "The Belcher" was named after your Tron belch recording since they sound similar...  Tnx for the suggestion.


Stu:  I heard you on 75M earlier describing your 10hz generator.  Yep, that sure sounds like a squarewave pulse in the .wav file.  The positive peak idea is a nice function that wud be helpful.  I'm all set with the box I have now - guess they all do the same thang more or less.  I have it hooked up so I simply toggle two switches and the Belcher does its job. Real cornvienent for fast tune-ups.

BTW, you were very strapping in here at the late afternoon peak condix - like against the pin.

T

[N5RLR]

I'm honored, Tom.  And, feel a little like a dad again.

And Shelby, congrats on your inspiring the name of this new tool in the box.

[Rob K2CU]

Tom,

Good idea to create a "standard" audio source for tunning up transmitters/amps, but into a dummy load only. I'd be leary about putting the signal into an antenna as it falls into the category of being either pulse modulation (not allowed in HF) or even type B emissions (damped oscialltions) also a no no anywhere. Damped oscillations are what spark gap used to produce!

In the past, I have been involved with various voice communication systems. We have used a variety of voice simulators. One used a "pink" noise generator that was amplitude modulated at 1 or 2 Hz to give sylabic punches. Another used a Votrax IC and a digital word sequencer to speak standard utterences, words, or even sentences that had been selected to produce a repeatable voice band peak to average ratio. With that system thery were often nonsense words. With a ROM  for data it was a battery powered stand alone unit that required no computer to operate....laptops did not yet exist! Still have some of those IC's...

A better solution would be a tone source with either selectable or adjustable frequency...perhaps 300Hz, 1000Hz, and 2000Hz, that would jump the amplitude of every tenth cycle by a factor that was also adjustable up to perhaps 10x the normal value.

If there is enough interest, I'll design one.

[K1JJ]

Very good ideas - as usual from you, Bob...


I've set up my antennas so they are close to 50 ohms near the operating freqs I use, so fortunately I can tune up into a dummy load and be close enuff when the antenna is switched on. Or I just tweak it a little afterwards.


Yes, your proposed circuit sounds great. As long as it's simple and doesn't have tons of IC's or LSI chips, I'll build it - otherwise I'll stick with the one I have now.  Having a unit that is "street legal" wud be a plus for sure...

In fact, if you cud add in a 2-tone feature without complicating it too much, that wud be great - to have guys with their SDR spec analyzers check others' linears using ssb as the exciter.   I wonder about an IMD tone test for AM?  Would it require just one tone to give the same relative test parameters?


T

[Rob K2CU]

It's great that you have your antennas tuned correctly and not depend upon a 'tuna to fix it!

I would think about the IMD issue for AM in three scenarios:

1. Low level followed by linya.
2. High level plate modulated class c toob.
3. Class E with PWM modulator.

The low level bit would be akin to SSB where two tone is often used. But, with DSB AM you have three tones. So, single audio tone here.

High level is intrinsicly prone to distortion as it is a non linear process. You have probable distortion in the big audio amp to start with, and then the distortion in the modulation process itself. You could use either a single tone for the modulator stage and sample output of same for harmonic distortion. Any one of several audio analysis programs, like Spectrum Lab, would show harmonic distortion products. As far as reception by an SDR or low  distorion detector using audio analysis program would want the use of a single tone here as well. Again, you have the issue of reports were the receiver used produces harmonics. You need a synchro detector, product detector or other low distortion detector there...no diode type detectors.

And for Class E PWM transmitters a single tone would also work for all the same reasons.

Hey, the idea was to have a simple audio oscillator that produced a peak to average function, say ten to one.  You start with a clean sinewave oscillator. Clean means less than 1% THD. You square it to then drive a Johnson counter, like a CD4017. One of it's outputs drive a fet switch to change the ampltude of the otherwise passed through audio tone. use something like a CD4066 swtich and maybe an opamp output driver.. The tone oscillator could be a piece of gear you already have, like a Wavetek....

looks like CD4017, CD4066, and a quad opamp, like a TL084, and a few resistors and caps. I will try and work on a complete design over the weekend. I want it to be inexpensive to make as a weekend project. Should cost under $15. cost depends on fancy box etc.

K2CU- Rob.

[N2DTS]

The flex has 2 tone, triangle and saw tooth wave tests built in, what are they supposed to look like on the scope?
Are they supposed to look like 2 tones, triangles and saw tooths?

I really should read up on tuning up ssb rigs and amps....

Brett

[N3DRB The Derb]

Stu, tnx. I now have a good reason for moving the oldest and slowest mac in the house out to the shack...play my Itunes library and tune up teh amp. 

[w4bfs]

where did you come up with a name like that?
Shelby

Yep, "The Belcher" was named after your Tron belch recording since they sound similar...   Tnx for the suggestion.

so thats what Tim is doing in the middle of his transmissions .... worlds fastest tuneup ?!?