Hi Roger, a couple of thoughts come to mind. The 400 mA idle current is
problematic. This would not cause an oscillation but it allows the tubes to
operate pretty much at maximum gain and that can make the taming process more
difficult. In addition, the plate dissipation rating is being exceeded
before there is even any rf drive applied. There may be thermal run away if
the screen voltage is not stiff enough. The negative screen current will try
to charge the screen voltage upward and depending on the screen voltage
source impedance can rise as high as the plate voltage.
Under these conditions the tube gain is way greater than that expected.
The purpose of neutralization is to purposely create a feedback path from
plate to grid that is of opposite phase in respect to the normal internal
grid to plate feedback path. In addition, the amplitude is supposed to be
such that the internal feedback path is effectively canceled.
You have the perfect test equipment to make the test and adjustment. This
is not related to any parasitic suppression as that is a separate issue.
While I don't have direct experience with this aspect of using a 4CX250, I
have tested other larger tetrode tube amplifiers. They showed about 40 dB
input to output isolation when an rf signal was connected to the input and
measured at the output. Also, as a check. the signal can also be applied to
the
output and measured at the input. Of course this is done with the
amplifier completely cold and at the expected operating frequency.
I would expect the neutralization null to be fairly narrow band but that
will depend on the degree of neutralization needed to make the null happen.
Once you get to this point, you can then start looking at possible
parasitic oscillation issues. The 4CX250 family of tubes are quite small
physically so it would seem that they would be a lot less prone to parasitics
than a
larger glass tube that inherently has more inductance in the internals.
Actually, I am thinking your root problem is more likely due to a run away
issue as discussed above.
The idea of swamping the grid is a valid method of reducing the gain but if
that does not help, that again points to the run away scenario.
Please let us know how you make out.
73,
Gerald K5GW
In a message dated 7/25/2015 3:20:45 P.M. Central Daylight Time,
ai7rogerroger@gmail.com writes:
I've been building a 6m amp from scratch based on quad 4x150a's for more
than 3 years. I've never built any type of rf producing device before and
unbeknownst to me at the time, chose a project that was way over my head.
I've read and studied volumes of material written on 6m amps and tetrodes.
The power supply was also built from scratch and is sound. It has been
successfully tested on a 2m rf deck running a pair of 4cx250b's in push
pull. HV: 1680vdc, Screen: 360vdc, Bias: ~ -50vdc, when not jumping off the
bench idle current 400ma. Though not currrently designed for individual
adjustment and not perfect, the tubes have been selected and are very close
to matching. I'm at the point of irradicating runaway, oscillations and/or
parasitics. I have access to quality test equipment and am fairly
knowledgeable with its use. I'm conducting a cold test using a tracking
generator with spectrum analyzer to observe the output while injecting a
signal into the grids. This test is suggested in Care and Feeding Chap
5.5.2 (Feeding the Grid Circuit). This states: "The neutralization can then
be adjusted for minimum feed through." One question of many: 1) Is the
author suggesting that the tester knows the frequency of the parasitic and
that is the frequency being injected and neutralized? Continuing. I have
indications of runaway, in that, when the amp is keyed without excitation,
excessive grid, plate and negative screen current is drawn. Some of the
fixes I've tried include raising the cathodes off ground by as little as 10
and as much as 33 ohms; various configurations of parasitic suppressors,
grid swamping with various values of resistors (all non inductive) and
different methods of neutralization.
I have no intention of dropping this project, but would really appreciate
any advice or suggestions beyond that to rein this thing in.
Thanks so much for your time.
73, Roger
AI7RR
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|