B&W 5100B: Negative Carrier Shift/Asymmetry?

[KK7UV]

Using the standard practice of adjusting the oscilloscope to show +1 and -1 divisions around the centerline for a carrier, when I modulate with either voice or a sine wave, I see the negative peaks/troughs hitting the center line, but the positive peaks are well below the expected +2 and -2  division lines.  Is this asymmetry called negative carrier shift?   How do I solve it?

I tested all component values in the modulator, and tried swapping the plate leads.

Steve

[w7fox]

This can be caused by insufficient drive to the final.  I would check this first.

[Opcom]

then, final screen volts would be good to check.

[KD1SH]

  You might also want to take a look - with the proper oscilloscope probe to do it safely - at the modulated B+, before the plate choke, to see if it's asymmetrical.
  Others might have different definitions, but my definition of negative carrier shift is when the entire carrier shifts downward, like what you might see with a B+ supply that sags under load.
  Asymmetry is the bane of AM, and the potential causes are legion. It can start at the microphone (try switching the element's leads, if the microphone's configuration allows it); at the lower levels of amplification; even with your own personal voice characteristics. None of us speak in sine waves - voice is extremely complex - and our vocal cords don't come with an offset adjustment.  I built myself a W3AM all-pass-filter, but I can't say it did much for me.
  Running sine sweeps can tell you a lot, but use discretion: you can hose your mod transformer by hammering it hard with sine - it was designed for the sort of duty cycle that comes from the syllabic nature of voice.
  You might have already done this, but I suggest you make use of this forum's search function; there's a trove of topics related to this, with many points of view.

[WD5JKO]

Hi all,

    This topic needs to embrace a few things. First, some of the BA designs out there just do not modulate properly...as designed.

You need to separate the big parts. Here is a list:

1.) The Power Supply. It is common that Rigs with a common B+ supply will have B+ sag under modulation. Even rigs of higher power, have a common power cord, and line voltage sag exists with modulation. Rigs with two 5R4's in the HV supply will droop if the tubes are not strong.

2.)The RF PA needs to linearly modulate from 0-100% and preferably beyond. Setting up for a Trapezoid scope pattern (a whole topic by itself) should reveal straight walls up to the limit of the modulator. All too often the RF PA is not linear at all, and there are issues with the G1 bias, grid current, Screen modulation percentage, and more. The modulator can be distorted badly, but the Trapezoid can still show linear modulation.

3.) The modulator itself should be clean and linear. If the modulated PA is linear with the Trapezoid, then you can drive the modulator with a Triangle waveform, and then reconfigure the scope to examine the RF modulated envelope (use about a 1ms/sec sweep). The Triangle uses less power to fully modulate when compared with a sine wave.

4 Scopes ; For 100 watt class rigs, a 100X AC compensated probe is ideal. Most are rated for around 2KV. Amazon and Ebay have them for around $25.
For higher power rigs, with a high B+ like 2500v you need a 1000X AC compensated probe. These can be expensive. Many would build up an AC compensated voltage divider rated for the task.

Jim
Wd5JKO