[AMPS] Alpha 77DX Bias Switching

Vincent Fedele Vince.Fedele@VSTTECH.COM
Wed, 26 Mar 1997 10:37:07 -0400

Re: [AMPS] Alpha 77DX Bias Switching

Several comments on the pros and cons of the 77 EBS system have been
included in recent discussions here on the Amps forum, and have provided
interesting and informative feedback.

The original goal of the EBS system, as I understand the history, was
twofold. First, it is an attempt to increase the dynamic range of the
users audio by keeping the linear amplifier cut-off during non-speech
portions of the users transmission, rather than amplifying all of the
background noise - which is typically blower (white) noise, and
therefore not desirable. Second, since the tube is cut off during
non-speech (and non-CW) portions of the cycle, no plate current flows at
all, and thus the effective duty factor is reduced, and the tube runs
with less power dissipation. These are the supported benefits of EBS
systems, and while I agree that these goals may be desirable, they must
be properly implemented in order to achieve the desired results without
introducing any negatives along the way.

A better solution to the dynamic range issue might be a simple noise
gate on the mic input ahead of the mic jack. I have heard a very good
implementation of this method by W2IHY -?- (or perhaps WB2IHY-someone
help me out here) that works wonderfully well.  The idea here is to
simply mute the mic input to the exciter during non-speech portions of
the users transmission. Also, with respect to plate dissipation, the
8877 has plenty--in a properly cooled application, and thus any real
benefit here is minimal--unless proper cooling (i.e., air flow) can not
be achieved.

Dick Byrd, and others have commented on the downsides of the EBS circuit
being correctly stated as a crackling or popping sound between syllables
of speech, and this is indeed a weak point of the standard (i.e., stock)
Alpha 77 biasing system.  Also stated was the fact that it is a
tri-state system, with:
    1)  the tube cut off at 60V; 
    2)  partially on at 20V when the relay (PTT) line is keyed but with
no RF present from the exciter; and 
     3)  final bias of 8.2V which occurs with the presence of about
200mW of RF from the exciter. 

The crackling or popping sound between syllables of speech is heard when
the tube is switched from partial cutoff at 20V abruptly to 8.2V full-on
as RF is detected from the exciter. I believe the spike or crackling
that is being heard in this case is distortion caused by operating the
tube from a non-linear region (20V) to its’ linear, full-bias region
(8.2V). If the threshold of the bias switch is set at the background
noise level, then this switching will occur routinely (with noise)
whenever the PTT is keyed, effectively negating the purpose/function of
the EBS circuit. Allowing the tube to switch into nonlinear bias at any
time when the RF relays are in the transmit position is not good
engineering practice.  

Given the Alpha design as it is, there are several adjustments that must
be performed with a monitor scope, and which may alleviate some or all
of what is being heard. First, use of improperly set up speech
processing may be responsible for processing background and input mic
ckt noise if too much processing is used. In almost all cases, this is
more than enough to provide 200mW of RF which will switch the bias
transistor (i.e., 20V to 8.2 V) on/off with noise and or between

The second adjustment [not provided by Alpha] which would also help is
an RF detector threshold adjustment from 100mW to about 600mW to better
control when the EBS bias transistor turns on.  This along with proper
adjustment of mic gain (and/or PROC setup) will preclude unwanted random
EBS switching of the EBS with noise at the threshold--by moving the
threshold above the noise level so that it is tripped with speech,
rather than background noise.

It should be a trivial change to simply disable the EBS, and cause the
biasing to revert to the traditional on/off system. However, this can
not be modified without taking into account the QSK sequencing system
that the EBS is interconnected with. For example, the 77 uses
pre/post-RF detectors on the input relay which prevents the T/R relays
from operating if RF is present from the exciter while the amp is in the
standby mode, and also will prevent the amp from dropping from operate
to stby mode if RF is present and the T/R line is dropped. These analog
circuits are used to create the correct ‘logic’ to protect the relays
from hot-switching and they work very well.

I will post shortly how to disable the EBS while retaining all of the
QSK, keying and hot-switching protection already inherent in the 77

Hope this has been helpful.

Vincent Fedele

FAQ on WWW:               http://www.contesting.com/ampfaq.html
Submissions:              amps@contesting.com
Administrative requests:  amps-REQUEST@contesting.com
Problems:                 owner-amps@contesting.com
Sponsored by:             Akorn Access, Inc. & N4VJ / K4AAA