Copyright © 2013 American Radio Relay League, Inc.

– All Rights Reserved

Build a Linear
2 Meter 80 W All
Mode Amplifier
This solid state amplifier
will give your low power VHF
transceiver or transverter just
the boost you need.

James Klitzing, W6PQL  An RF-sensing TR switch for remote op- input and output power of the amplifier with
There are many different 2 meter low power eration, as well as a hard key option. the built in attenuator for a 10 W exciter
rigs in use, ranging from handheld transceiv-  TR sequencing to protect the S-AV36 mod- shown in Figure 2 is provided in Table 1. This
ers for FM to older multimode transceivers ule and prevent hot switching of the antenna also shows the current required at 13.8 V dc.
and even the newer all purpose types such as relay. Designing the Amplifier
the Yaesu FT-817 or the Elecraft 2 meter  Indicator LEDs and control switches. The S-AV36 module is pretty easy to use;
transverters. Low power (QRP) operation can aside from RF IN and RF OUT, there are two
 Reverse polarity protection.
be fun, but if you’re like me, you have the oc- power connections; one is for BIAS (this turns
casional need for a bit more power. The inside of the final project is shown in the module on and off), and the other for
Figure 1. Note the simplicity. main DC POWER, 13.5 V nominal at up to
If you have a couple of afternoons to spend
on a project, you can build this 80 W multi- The devil is in the details for the designer, 15 A. Since the input power required to drive
mode amplifier with ease. It’s easy because it though, and it did take a little planning, but it is only about 50 mW, the first thing to do is
uses one of the newer Toshiba modules as the the end result was a small PC board made at design an input attenuator to match the output
heart of the amplifier. The Toshiba S-AV36 home using common hobby tools. Add a few of the driver to the S-AV36. The resistive at-
provides direct 50 Ω input and output imped- interconnecting wires, heat sink, connectors, tenuator (R7, R8 and R9) can adapt the atten-
ances with gain galore — so much gain that switches, a couple of sheet metal parts for uator to drive levels ranging from 1 to 10 W
less than 50 mW can drive it to full output in the enclosure, and that’s about it. A schematic as shown in Table 2. There are some strange
any mode. This design will work with any ex- and parts list is provided in Figure 2. The values there, but these are not terribly critical,
citer providing 1 to 10 W drive, through the
use of a built in attenuator.
My original intent in making this was to have
an amplifier capable of boosting an older
10 W multimode radio up to 80 or 100 W.
I wanted to keep it low in cost and simple (no
preamp or power meters), yet capable of fixed
station or mobile operation in any mode and
operation from the usual nominal 12 V dc
power supply. In this way, the supply that
powers a 100 W HF transceiver can likely
power the amplifier as well.
After absorbing the specs in the data sheet, it
was clear to me that this module could be
driven by almost any low power rig; thinking
about it a bit more, and keeping in mind the
low cost and simplicity requirements, a few
more useful features came to mind, such as:
 An output low-pass filter to comply with
FCC regulations for harmonic and spurious
suppression.
Figure 1 — Inside of the compact amplifier showing the simplicity of the final design. The amplifier
 A low loss antenna relay. module is the black rectangle connected to the right edge of the PC board.

30   May 2013 ARRL, the national association for Amateur Radio® www.arrl.org  
 Copyright © 2013 American Radio Relay League, Inc. – All Rights Reserved

you just have to get within a few ohms to get respecting amplifier module likes seeing an holes for #6-32 screws, and the PCB holes for
the job done. For example, a 23 dB attenuator open circuit while those lazy relay contacts 4-40 screws.
is needed for a 10 W radio. The resistors cho- are moving, even if it only takes 20 ms to
Install all the PC board components except
sen were those readily available from major happen. It’s not good for the module, and just
for the module. The relay should be installed
distributors, so 58 Ω became 56, and 351 be- plain rude. For this reason, the module has to
last, and because the pins will protrude
came 360 (close enough). L5 is not really be kept off while the relay contacts are set-
through the bottom of the board, they should
necessary. Its purpose is to compensate for tling. The other reason is to protect the relay
be cut off flush with the board after soldering.
the stray capacitance of R7 at 2 meters (a contacts from that 80 to 100 W the amplifier
35 W tab-mounted resistor). The input SWR will generate before they finally settle; it While I used a PC board and surface mount
was acceptable without it, but it does make tends to shorten the life of the relay. components, there is no reason leaded com-
the input match almost perfect.  ponents could not be substituted. The PC
C4, D1 and D2 sample the input while C5,
board is still recommended because the
The Low-Pass Filter C6, R1 and R2 provide filtering and some
etched transmission lines going in and out of
Now that the input is taken care of, let’s deal timing, depending on the position of S1. In
the TR the relay will provide minimum loss
with the output. The data sheet says the SSB mode, the circuit provides a delay on
on receive — important since any loss adds
­second harmonic will only be down about switching to RECEIVE similar to VOX, pro-
directly to the receive noise figure.
25 dB, and the third about 30. Not good viding a second or so of delay. In FM mode,
enough for the FCC, so we need an output the switch back to RECEIVE is much quicker,  Make the enclosure parts and two alumi-
­filter that will put us in good graces with at as the delay is not necessary for FM opera- num spacers as shown in the fabrication
least 60 dB total suppression. For that 25 dB tion. The circuit is sensitive, and will trigger drawings available on the QST in Depth web
second harmonic, we need another 35 dB. with less than 1⁄2 W drive. page.1
The filter shown (L1-L4, C12-C14 in Fig- Q1 is the switch that operates the relay. When  A PC board is available from the author, or
ure 2) is a standard pi type, Chebyshev filter the relay is turned on by Q1, it also turns on can be made from the artwork on the QST in
of seven poles. The design provides the re- Q2 (the bias switch) after a short delay. This Depth web page. Mount the board to the heat
quired suppression with very little insertion delay is provided by C9 and R4, and is about sink with four 4-40 screws. The two alumi-
loss at the operating frequency. 50 ms in duration, allowing those relay con- num spacers must be positioned under the
tacts to settle before the module becomes board on either end to elevate the board to a
The Antenna Relay active. convenient height for the module and keep
and Switching Controls
the back side PCB connections at the relay
In the spirit of keeping costs low, a PCB When switching back to RECEIVE, the bias to
pins from shorting against the heat sink.
mount type of DPDT general purpose relay the module is cut off before the relay contacts
was chosen for TR and bypass switching. At open. This fast cutoff is timed by C9 and R5,  Some minor tuning of the low-pass filter
less than $5 in cost, the contacts are rated at and is only about 5 ms in duration. coils can be made at this time. Connect a
8 A. At 2 meters, a bit of reactance is intro- dummy load to the output of the board, and a
Another noteworthy component is D6, the re-
duced by this part, but compensated for by a transmitter and SWR meter to the trace at the
verse-polarity protection diode. This diode’s
small capacitor (C15) in series with its input. input of the filter where the module will con-
purpose is to blow the in-line fuse in the
nect. Apply 12 V across the relay coil to close
The best way to tell the amplifier to switch on power cord if you accidentally connect the
the relay, and spread or compress L1-L4 for
is to use a control line back to the driving power cord backwards (come on, we’ve all
lowest SWR reading. If this is inconvenient to
radio (PTT). If this is unavailable or inconve- done it).
do, the filter can be adjusted after the ampli-
nient, the amplifier has an RF sensing circuit The extra contacts on power connector J3, fier is fully constructed, adjusting for max
that samples a bit of drive from the input con- Pins 3 and 4, provide a means to disable the power at about 50 W output. It’s best to do it
nector to provide the transmit trigger. RF sensing and connect PTT directly to the now, though, and you’ll probably find that
In another little twist; switching from receive driver should the RF sensing be deemed un-
to transmit should be sequenced for two rea- necessary. If just Pin 4 of J3 is grounded by
sons; first, the S-AV36 is tough, but no self- a PTT line, the amplifier will be switched
1www.arrl.org/qst-in-depth
by the PTT, but all the delays will apply. If
Pin 3 is grounded the amplifier the will fol-
low the PTT with only the delays de-
Table 1 signed to protect against hot
switching, as described above. Table 2
Output Power and Current
Required with Resting Building the Amplifier Values of R7, R8 and R9 for
Current at 8 A Different Drive Levels
Here are the recommended steps, in
Drive Output Current (A) sequence, for constructing the Drive Attenuation
Power (W) Power (W) at 13.5 V Power (W) (dB) R7, R9 (Ω) R8 (Ω)
amplifier:
1 12 8.2 1 13.0 79 106
2 29 9.0  Mark the heat sink for drilling by 2 16.0 69 154
3 44 9.5 using the PC board as a template. 3 17.8 65 191
4 53 10.0 You can also position the Toshiba 4 19.0 63 220
5 66 11.0 module and mark its two mounting 5 20.0 61 248
6 74 11.5 6 20.8 60 272
holes; leave a small gap of 2 to 7 21.5 59 295
7 80 12.0
8 85 12.5
3 mm between the module body and 8 22.0 59 313
9 89 12.8 the board for strain relief. 9 22.6 58 335
10 92 13.0 10 23.0 58 351
 Drill and tap the module mounting

QST ® – Devoted entirely to Amateur Radio www.arrl.org May 2013   31

 Copyright © 2013 American Radio Relay League, Inc. – All Rights Reserved

Decimal values of capacitance are in microfarads (µF); others are in picofarads (pF);
Resistances are in ohms; k=1,000, M=1,000,000. C8 C7
10 μF 0.01
C15
100 K1A Toshiba Low Pass Filter
Input 360
J1 S-AV36
Attenuator L1 L2 L3 L4
Driver 1W
C4 U1
28 nH 75 nH 75 nH 28 nH
1 56 R8 56
J2 R7 R9 C12 C13 C14
35 W 1/4 W
Antenna L5 20 24 24
K1B 220 nH
C11 C10
10 μF 0.01

MBRS140T FB1 RF Sense Disable

D4
C1
3 1
1000
K1 FB2 J3 13.5 V FB3
PTT 4 2

C2 D6
D3 1000 C3 VBT3080
BAS16 1000

Bypass
S2

D10 XMIT Amplify

C9 R5 Online D9
10 μF 3.57 k
R11 1k R10
R4
1k
22 k Q2
MMBTA64
D2 R3 D5
MBRS140T R6 510
BAS16 1k
R1 511 k
C6 Q1 SET IDQ R12
D1 C5 D7
10 μF MMBT6427 1k
BAS16 1000 R2 5.1 V
S1
QS1305-Klitzing02 51 k SSB
FM

Figure 2 — Schematic diagram and parts list for the amplifier. For parts supplied by Mouser, you can order all of these by ordering the project list
from their website, or order individual parts if your prefer. www.mouser.com/tools/projectcartsharing.aspx. The access ID code for the project is
c3ad150d1a. RFPARTS (www.rfparts.com) is the supplier for the Toshiba module and coax connectors. Heat Sink USA parts are available at
www.heatsinkusa.com. Artwork for the PC board is provided on the QST in Depth web page, along with fabrication drawings for sheet metal parts.
For those not wishing to make their own board, commercially made boards are available from www.w6pql.com.
C1, C2, C3 — 1000 pF ceramic capacitor FB1, FB2 — Small ferrite bead R6 — 510 Ω SMT resistor
(Mouser S102K29Y5PN6TJ5R). (Mouser 623-2643000701). (Mouser CR1206-FX-5100ELF).
C4 — 1 pF SMT capacitor (Mouser FB3 — Large ferrite bead R7 — 56 Ω, 35 W SMT resistor
12061A1R0CAT2A). (Mouser 623-2643000801). (Mouser PWR263S3556R0F).*
C5 — 1000 pF 1206 (SMT) capacitor J1, J2 — Panel mount SO-239 coax socket. R8 — 360 Ω, 1 W SMT resistor
(Mouser 12065C102KAT2A). J3 — Power connector, 4 Pin cable mount (Mouser RK73B3ATTE361J).*
C6, C8, C9, C11 — 10 µF 1206 (SMT) ceramic (Mouser 38331-8004). R9 — 56 Ω, 1⁄4 W SMT resistor (Mouser
capacitor (Mouser 581-TAJA106M016R). K1 (RL1) — Omron DPDT relay CR1206-FX-56R0ELF).*
C7, C10 — 0.01 µF 1206 (SMT) capacitor (Mouser G2RL-2-DC12). R10, R11 — 1 kΩ, 1⁄4 W metal film resistor
(Mouser VJ1206Y103KXXCW1BC). L1, L4 — Inductor, 28 nH. 4 turns #18 AWG, (Mouser 660-1/4DCT52R1001F).
C12 — 20 pF metal mica capacitor 4 mm inside diameter, 8 mm long. R12 (VR1) — 1 kΩ potentiometer
(Mouser MIN02-20J-F). L2, L3 — Inductor, 75 nH. 7 turns #18 AWG, (Mouser TC33X-2-102E).
C13, C14 — 24 pF metal mica capacitor 4 mm inside diameter, 10 mm long. S1, S2 — SPST miniature toggle switch
(Mouser MIN02-24J-F). L5 — Inductor, 220 nH. (Mouser (Mouser A101SYZQ04).
C15 — 100 pF metal mica capacitor 70-IMC10008ERR22J). U1 — Toshiba S-AV36 amplifier module
(Mouser MIN02-100J-F). Q1 — NPN Darlington transistor (SMT) (RF Parts)
D1, D2, D3 — SMT switching diode (Mouser MMBT6427). Heat sink extrusion 5.375 × 8 × 1.376 inches
(Mouser 512-BAS16). Q2 — PNP Darlington transistor (SMT) (Heat Sink USA A009).
D4, D5 — 1 A surface mount diode (Mouser MMBTA64). Heat sink extrusion 5.375 × 8 × 1.375 inches
(Mouser MBRS140TRPBF). R1 — 511 kΩ 1206 (SMT) resistor (Heat Sink USA A008).
D6 — SMT dual diode, 30 A (Mouser CR1206-FX-5118ELF). In-line fuse holder (Mouser 441-R-332B-GR).
(Mouser VBT3080S-E3/8W). R2 — 51 kΩ SMT resistor RG-316 50 Ω Teflon coax to go from J1 and J2
D7 — 5.1 V SMT Zener diode (Mouser CR1206-FX-5102ELF). to K1.
(Mouser MMBZ5231B-V-GS08). R3 — 1 kΩ 1206 SMT resistor
D9 — 5 mm green LED (Mouser (Mouser CR1206-FX-1001ELF). *These parts provide the input attenuation required
941-C503BGCNCY0C0791). R4 — 22 kΩ SMT resistor for a 10 W input. For other levels of driving power,
see Table 2.
D10 — 5 mm red LED (Mouser (Mouser CR1206-FX-2202ELF).
941-C566CRFSCT0W0BB2). R5 — 3.57 kΩ SMT resistor
F1 — 20 A fuse to fit the fuse holder in the (Mouser CR1206-FX-3571ELF).
power cable called out below.

32   May 2013 ARRL, the national association for Amateur Radio® www.arrl.org  
 Copyright © 2013 American Radio Relay League, Inc. – All Rights Reserved

very little adjustment is necessary. ufacturer makes them this way, as do other just guessing, but I believe the manufacturer
module makers. Do not attempt to sand this makes the footing this way for strain relief in
 Using heat sink compound, mount the
footing flat or otherwise fill with any material order to protect the mechanical bonds inside.
Toshiba module with #6-32 machine screws.
except for heat sink compound. There is still Solder the module wires to the appropriate
Note that the mounting bar of the module is
plenty of contact area for heat transfer. I’m traces on the PC board (cut off the excess
slightly concave; this is not a defect, the man-
wire length if necessary, see Figure 3).
 Mount the connectors, switches and LEDs,
and complete the chassis wiring (see Fig-
ure 4). The LEDs have their 1 kΩ resistors
soldered directly to their leads, with the wire
connected to the other side of the resistor;
heat shrink is used to cover the resistor and
connections. Use solder lugs under the
mounting screws for the connectors on the
rear panel; these are for connecting coax
shields, dc chassis ground, and bypass capa­
citors as shown in Figure 5.
 Make the power cord from #14 AWG wire.
Make certain to use an in-line fuse on the
positive lead, and fuse it for no more than
20 A. If you will be hard keying your ampli-
fier from your radio, jumper Pin 3 of the con-
nector to ground, and carry Pin 4 back to your
keying connection from the radio. The radio’s
PTT relay contacts or other switching must
be capable of sinking 12 V at 50 mA to
ground.
Testing the Amplifier
Once everything is wired and in place, you
can test the amplifier using the following
procedure:
 Connect the output to a suitable wattmeter
and dummy load, and the input to your driv-
ing radio.
Figure 3 — The rear panel is attached using #6-32 machine screws.
The module connections are soldered to the PC board as shown.  Connect the power cord to a power supply

Figure 4 — View of the inside of the front panel before wiring. Figure 5 — Details of the rear panel wiring.

QST ® – Devoted entirely to Amateur Radio www.arrl.org May 2013   33

 Copyright © 2013 American Radio Relay League, Inc. – All Rights Reserved

A Quest for a More Linear Amplifier

Your editor, in common with many casual VHF operators, uses
a transverter with a power output of 10 W. In my case it is an Table A
Elecraft K3 HF and 6 meter transceiver with an internal Intermodulation Response as a
2 meter transverter. This unit has a very low noise receiver with Function of PEP Output
the result that I can hear many more 2 meter stations than can
hear me. Power Output Intermodulation Products (dBc)
In an effort to put my station on an equal footing with stations PEP (W) 3rd 5th 7th 9th
with the frequently encountered 100 W multimode radios, I 55 –27 –44 –46 –56
wanted to find an amplifier that could be added to my station to 75 –24 –42 –49 –53
make up the difference. I tested a number of “brick” amplifiers in 100 –19 –28 –36 –47
the hopes of finding one that would provide the needed power
with appropriate intermod products for use on SSB.1,2 While the
tested units delivered the desired power, and indicated that they
were suitable for SSB, the intermod levels were high enough
that I thought they would cause trouble for other stations during IMD was indeed possible on VHF.3 There was also the very
contests. These amplifiers work fine on FM, their primary use nice 2 meter kilowatt amplifier described by Jim Klitzing in a
design goal, as recent QST article.4 That inspired me to commission Jim to
well as on CW. It put his talents to work designing a lower powered amplifier
is only using them that would meet my operational objectives, be easy to build
140
QS1305-Klitz-A
in SSB operation and relatively inexpensive.
that generates Jim has met my objectives in this very nice package. The
120
the spurious IMD output, as measured in the ARRL Lab were as shown
100
Power Out (W)

signals due to the in Table A, and a plot of the input versus output power in
80
multiple simulta- Figure A. Note that it closely follows the straight line until 75
60 neous frequency to 80 W, at which point the compression is evident.
40 components.
20
There was
an alterna- 1J. Hallas, W1ZR, “Product Review —A Pair of Mirage 2 Meter Amplifiers”
0 tive, a much
1 2 3 4 5 6 7 8 9 10 QST, Aug 2010, p 52.
Power In (W) higher priced, 2J. Hallas, W1ZR, “Product Review — TE Systems 1410G 2 Meter

higher powered, Linear Amplifier,” QST, Jan 2012, p 54.

3J. Hallas, W1ZR, “Product Review — Tokyo Hy-Power Labs HL.
amplifier (now 350VDX 2 Meter Linear Amplifier,” QST, Mar 2012 p 48.
Figure A — Power output versus input, as
measured in the ARRL Lab. A straight line
discontinued) that 4J. Klitzing, W6PQL, “Solid State 1 kW Linear Amplifier for 2 Meters,”

(red) is shown for comparison. showed HF type QST, Oct 2012, p 32.

capable of delivering 13.5 V at up to 15 A. duce lower overall gain and crossover distor- screw. Most of us making our own prototype
tion in SSB, while setting it too high resulted boards at home don’t have the ability to make
Place the AMPLIFY/BYPASS switch in BYPASS in higher gain and saturated output power. At plated through holes the way the commercial
mode. Transmit, and verify that bypass mode 10 A IDQ, for example, the amplifier could be board houses do, so if you make your own
works (most of the driver’s power should pass driven to over 100 W output with about half board for this project, you’ll probably need
through the amplifier to the load). The bypass the drive required at 8 A IDQ. This current to implement a similar solution.
mode insertion loss is only about 0.1 dB. drawn at this level is close to the manufactur-
Turn off the driving radio and put the ampli- er’s absolute maximum ratings for the device,
and really doesn’t make any difference on the ARRL member and Advanced class licensee
fier in AMPLIFY mode. The READY LED James Klitzing, W6PQL, was first licensed in
should illuminate. Jumper PTT to ground, air, so I resisted the temptation to leave it that 1964 as WB6MYC. He has been a precision
and the XMIT LED should also illuminate. way. For all mode versatility, leaving IDQ set measurement specialist for the US Air Force
at 8 A is best. and Hewlett-Packard Company. He retired in
Adjust the IDQ trimmer (VR1) for 8 A. Place 2006 as an engineering manager for Agilent
the amplifier back in BYPASS mode and re- One last note. At a drive level of 10 W, I no- Technologies after 34 years with HP/Agilent. Jim
move the PTT jumper. has always enjoyed building his own equipment
ticed R7 (the 35 W attenuator input resistor) and is active on HF through 3456 MHz.
Turn the radio back on, place the amplifier ran hot. This was due to the inadequate heat You can reach him at 38105 Paseo Padre Ct,
in AMPLIFY mode, and transmit. Perform­- transfer of the PC board I made for the origi- Fremont, CA 94536, or at [email protected].
nal prototype, which has just a few rivets Jim’s website, www.w6pql.com, has all current
ance should be similar to the data shown in updates to this and many other projects.
Table 1. where there should have been multiple plated
through holes surrounding this part. My solu-
I experimented some with various IDQ set- tion was to use a piece of 0.040 inch copper
tings, and concluded that Toshiba must have strip soldered to the ground tab of the resistor. For updates to this article,
designed the module to operate close to I used this to transfer the heat to the heat sink see the QST Feedback page at
Class A. Setting IDQ too low tended to intro- by fastening the other end to it with a #4 www.arrl.org/feedback.

34   May 2013 ARRL, the national association for Amateur Radio® www.arrl.org