Got a question for the amp builders/those in the know...
I have a Palomar Elite 400 which, although it works great, sounds a bit
"scratchy" on SSB. I have clipped the "bias" leads off the input
transformers (the ones with the ferrite beads that go from the top of
the
transformers down to the board), and this made the amp sound better, but
it SEVERLY limited its output. Now the amp only does around 210 watts,
as opposed to the 450+ it was doing before...
The question is this... is there a way to get some of that output back
without sacrificing the quality of the output signal?
YES
Would it help if I added a ferrite bead or two to the "wires" and
re-soldered them back to the tops of the input transformers?
NO
I think what's happening is, with the wires/beads in place, the amp is
closer to class C than to AB, and from what I can see on the schematic I
have, the "wires" from the transformers, with the ferrite beads on them,
act as AC resistance to ground. Dragging the bias closer to ground seems
to increase the output on AM, but at the expense of quality of signal.
I'd like to be able to use this amp on both AM and SSB, without the lost
quality of output. I know I might take a power hit, but I'd like to
think
the hit would be less than half of what the amp does normally...
The gain will actually increase
when the proper bias is applied.
I'm one of those who believes increased quantity does not imply
decreased
quality, so any help would be appreciated.
Here's the deal: the
bias is *still* being grounded out even though you cut the
leads that appear to be the only thing stopping the bias
from being applied.
On the *bottom* of the board there is a factory solder bridge
that grounds out the bias. Remove that bridge and the bias circuit will
operate.
Once the bias circuit is restored, the maximum output will
be the same as it was before you cut the grounding leads.
The problem is that the small signal gain will increase--which will
make the amp less stable. The high instability is the reason the
factory grounded out the bias.
Fortunately, those amps respond well to the addition
of negative feedback. The newer ones that use the SD1446
transistors end up really nice when two mods are done: reduce the
bias, and add feedback loops. Here's how: Increase the value of the
big bias resistor. As I remember, the stock one is a 35 ohm
10 watt unit. Upping it to 50 ohms will bring the amp closer to
Class B. That will improve stability and reliability without any
sacrifice in audio quality on sideband. The bias resistor
will run cooler too. To add negative feedback, put a .1 uF 25 volt disc
cap in series with a 33 ohm 2 watt carbon resistor and connect them
between the collector and base of each of the 4 power
transistors.
RF Parts in San Marcos CA sells a Ready to install Feedback kit with the parts and instructions on how to install them.
The vertical
transformer 400s have some good features. They use the large R-10
style relays, the board is supported by rails all the way down
both sides but is allowed to expand and contract with temperature
changes because it's not screwed down, and the heat-sink/chassis design
is first rate. The price is excellent too. The problem is in that there
were a lot of corners cut in the design in order to keep the production
cost low. Those compromises result in a unit that is unusable in
certain situations and is problematic in many other situations.
Fortunately, with some relatively simple mods, the amp's true potential
is realized. In other words, when the mods are done, the amp ranks right
up there with the trusted names in production amps.
Give the mods a try, I think you'll be pleasantly surprised.
The vertical transformer 400s are diamonds in the rough.
A couple of updates: The feedback loop resistors can be
any non-inductive type. The best are the blue or green
"flameproof" units. I believe they are metal film. Also,
the value of the bias should be no more than .6 volts
(600 mV). 400 to 500 mV is actually the best. The
lower the better for maximum stability. On the other
hand, anything under 400 mV and the SSB distortion
starts to become audible.
This applies to all diode biased amps (just about all
of the biased CB amps use diode bias so listen up):
The diode bias circuit used in most biased CB amps
is just a voltage divider. There is a big resistor and
a diode in series between the 13.8 volt supply and
ground. The actual bias voltage exists at the junction
where the big resistor and diode meet. The diode is
really just acting as a resistor.
The problem with diode bias is that the diode sets the
bias value too high. That increases the likely-hood of
instability and lowers the efficiency. The benefit behind
using a diode is that the diode is *supposed* to track
the transistor temperature and decrease the bias as
the transistors get hotter. The problem is that in today's
amps the diodes are rarely ever put in thermal contact
with the heat sink close enough to the transistors to where
the diode will ever decrease the bias. In a nutshell, with
a diode the bias is too much to begin with, and any thermal
decrease is too little too late. Anyone reading this that
has a diode biased amp (any amp claimed to be Class AB),
a clip lead, a multi-meter, and soldering equipment should
test their amp to see what the bias actually is. If it's over
.6 volts (600 mV) the bias circuit should be modified. To
check the bias voltage key the amp with no input signal
(attach the clip lead to the un-banded side of the diode
that is next to the transmit relay and ground the other end)
and measure the DC voltage on the bases of the power
transistors. Do so with the amp cold (room temperature).
The DC voltage on the transistor bases is the bias voltage.
With diode bias it's likely to be well above .6 volts. Don't
be surprised if it's above .7 volts--especially if your amp
shows signs of instability (high output SWR even though
the antenna is matched properly, fuzzy sound, runs hot,
squealing, won't unkey all the time, makes tons of TVI, etc.).
If the bias is above .6 volts and you want to do something
about it, you can either increase the value of the big resistor
or replace the diode with a resistor. The cheaper way is to
replace the diode. If you decide to do it that way be super,
super careful because if you use a resistor with too high a
value the bias will be too high and you risk damaging the
power transistors. The resistor that replaces the diode
should be a 1/2 watt unit. The value will be based on what
the value of the big resistor is and the value of the power
transistor base resistors. It can be figured mathematically,
but it's just as easy (and more fun) to experiment. Get some
resistors with values in the 1 to 2 ohm range and try them.
Start low and work your way up until the bias is in the .4 to
.5 range. For example Texas Star doesn't use diodes; for
one pair of transistors the big resistor is 25 ohms and the
small one (that goes in place of the diode) is 1.6 ohms.
That results in a bias voltage that's still a little high but not
as bad as if they used a diode.
While you're at it, check the value of the power transistor
base to ground resistors. If they are more than 10 ohms
(black-brown-black) replacing them with 10 ohm units will
increase the stability. Check the value of the feedback
loop resistors too. Each power transistor should have a
feedback loop consisting of a capacitor, a resistor, and
sometimes an inductor in series connected between the
collector and base. The resistor controls the amount of
feedback. Most use 100 ohms. Decreasing the resistor
value increases the feedback and makes the amps output
signal cleaner and makes the amp more stable. If the amp
doesn't have feedback loops add them. Use .1 uF 25 volt
ceramic disk caps and 33 ohm 2 watt non-inductive resistors.
If the amp has feedback loops but isn't as stable as you'd
like, consider increasing the feedback (decreasing the
feedback loop resistor value).
