开云体育

ARR Preamplifier Comparisons

I had an brief short amount of spare time to compare (and optimize)
some used ARR receive preamplifiers, models ARR432VDG and
ARR460VDG. Being specified for sale within a 50 MHz design
spacing or distance, one might think they would be a same or a
very similar circuit.

The following described recent observations are essentially the
same as recorded (by me) through the years from a respectable
number of similar preamplifier evaluations and technical conversations.

The ARR432VDG and ARR460VDG appear similar in construction,
with modest variations. However, slight differences in their circuits
do notably affect performance in different S parameter results.

The ARR432VDG (VNA 2-port_2-path) S11 return loss was over
11 dB, which is quite a bit better than many reported values for ARR
preamplifiers of this type. The ARR460VDG value here was just
over 6 dB, which is more in line with measured values reported by
other persons. It's not obvious by casual (visual) observation, why
the ARR432VDG achieves this much better return loss number, but
the devil is in the details... where those details are the value(s) and
type of parts used ins the input matching section/portion of the
circuit.

Gain values for both preamplifier models were above 14dB, but again
the ARR432VDG provided above 16dB to 18dB (gain). The best
gain was not coincident with best return loss as described below.

Moving along...

Of special note, maximum gain does not occur with best return
loss. best return loss does not occur with the lowest noise figure,
and lowest noise figure does not occur with the maximum gain.

In the real world, the above values seem to be relatively close, but
never the same. If you're all about other opinions, I will write that I had
a few different conversations about this subject with Chip Angle,
and he confirmed these observations as valid with his products. Of
course you also have to add stability to balance the preamplifier
design along with all the above mentioned.

The trade of max gain versus best return loss and stability is an
interesting (can be life-long) animal to study. Purest types most often
insist the best match is always the most efficient (return loss). Other
types try to justify their opinions based on more empirical observations
and/or measurements. In the long term scheme of the things, everything
is of course, a mix of trade-offs.

I did measure both the ARR432VDG and ARR460VDG performance
when optimized by the single internal adjustment capacitor, from
about 430 to 470 MHz. It was interesting to note the ARR432VDG
clearly outperformed the ARR460VDG even up in to the commercial
radio band segment approaching 470 MHz. I have seen this trend
in other brands/models preamplifiers, why this occurs, is a whole
additional and probably quite long thread post for another time.

So, if you have the choice... or the luxury of comparing and using
a ARR432VDG or ARR460VDG.... one might consider the ARR432VDG
even though it might not be something viewed or selected at first
glance. The ARR432VDG examples I recently evaluated are better
performing preamplifiers when compared to their ARR460VDG cousins.

Your mileage will of course, vary.


that is all

cheers

Kelvin Scale

So, if you have the choice... or the luxury of comparing and using
a ARR432VDG or ARR460VDG.... one might consider the ARR432VDG
even though it might not be something viewed or selected at first
glance. The ARR432VDG examples I recently evaluated are better
performing preamplifiers when compared to their ARR460VDG cousins.

I've tuned many an ARR preamp. They did make very minor changes during the
life of the VDG series, but if you compare a P432VDG and P460VDG of the same
vintage, they should have the same components, just with different tuning of
the input cap.

As you saw, you can improve the input match somewhat by tuning the input
away from the point of lowest noise figure. At the point of best match, the
NF will be above 1 dB. For repeater service when the preamp is following
filtering, this is often a better trade-off - I'll take the improved match
over the fraction of a dB of noise figure any day.

The input network on the ARR's is about as simple as you can get. The
series trimmer cap and shunt coil on the gate form an LC high-pass network
that coarsely transforms 50 ohms to the high-Z of the gate. You can improve
the match by adding adjustable shunt C in parallel with the inductor.
High-Q caps are necessary to keep the NF low.

As I'm sure you know, you can't test GaAsFET preamps with a VNA at high
power. This is a problem if you have a VNA that has a fixed output level,
or only adjustable over a small range like -10 to +10 dBm. As you start to
approach the compression point, obviously the gain will fall but also the
input match will change. I usually sweep them at -40 dBm input power.

--- Jeff WN3A

At 9/16/2018 07:39 AM, you wrote:

As you saw, you can improve the input match somewhat by tuning the input
away from the point of lowest noise figure.? At the point of best match, the
NF will be above 1 dB.? For repeater service when the preamp is following
filtering, this is often a better trade-off - I'll take the improved match
over the fraction of a dB of noise figure any day.

If the tradeoff of NF is small (0.2-0.3 dB) I'd agree.? The NF spec on the ARR 440 preamp is 0.5 dB, so if it's meeting spec w/o tuning I wouldn't be so keen on degrading the NF by half a dB just for a better input return loss.? I once had a preamp in service that had an input RL of only ~3 dB & in service it performed identically to an identical model that had much better RL.? No doubt the duplexer got detuned a bit by the poor RL, but in this case it wasn't enough to affect actual on-air performance.? My point is that a preamp with poor RL can still perform well, whereas a preamp with poor NF will not.

Certainly the optimal solution would be good input RL without degrading the NF.? In tests I've conducted on the Mini-Circuits PGA-103 I was able to increase the input RL to > 25 dB over a 25 MHz bandwidth with corresponding NF degradation of 0.15 dB.? IMO this is quite acceptable.

Bob NO6B