k4kyv
Contributing Member
Don
Member
Offline
Posts: 10037
|
|
« on: December 27, 2009, 03:55:30 PM » |
|
I was working on my 500Ω line amplifier that goes between the peak limiter and transmitter. The thing has taken flaky spells in the past, with inexplicable distortion and cutting out, intermittent of course. So this week I put it on the bench with scope, signal generator, DVM, etc to see what is going on.
At first I was getting some rather bizarre results. The thing is very simple. A power supply, line to push-pull grids input transformer, pair of small receiving triodes (have used 6J5's, 37's and 6C5's over the 30-year life of the thing), push-pull plates to line low level output transformer. Only a couple of capacitors and resistors other than power supply components. The audio voltage to the grids of the tubes was unbalanced, but I could reverse the unbalance by reversing the polarity of the audio line to the input, even though the signal generator has 500/600Ω balanced output. The scope showed some very strange patterns at the output into a resistive load. Finally, after much hair pulling and frustration, I figured out that the pushpull stage was self-oscillating at about 21 kHz, but the pattern on the scope still made no sense. Then I found that shorting out the input transformer secondary or the output transformer primary, and finally both, had very little effect on the oscillation. The reason for the bizarre waveform at the output was that the oscillation was common mode to both tubes. It made no difference whether the plates, grids, cathodes or all three were shorted with clip leads; the oscillation was still there.
Then I noticed that the leads from the input transformer to grids were crossing over in the vicinity of the leads from plates to the output transformer. I replaced the grid leads with shielded wire and re-routed the plate leads to keep them well away from the grid leads, and the oscillation went away and now the thing works perfectly.
Lead length and placement can be critical not just in rf amplifier stages, especially if you are dealing with broadcast quality audio transformers and other components with inherent frequency response to 20 kHz or thereabouts.
|