Post-July 2016 - a revised 6 Meter 1KW RF Deck
This RF deck has the same performance as the original done on TC350 substrate, but is easier to build. The footprint is a bit larger (4 by 6.5"), and it uses the same PC board as a newly developed HF-6m rf deck, but optimized for 6m. The board material is .078 thick FR4; here are the advantages of this design over the prior one:
A simpler heat spreader is used, cut from .375" flat copper plate
The heat spreader no longer requires a slot (for mounting the LDMOS) to be machined into the copper
heat spreader has a bit more mass ( it weighs more than 3 pounds)
The heat spreader does not require tapped mounting holes; instead, it uses 12 pass-through holes designed to make the rf deck easier to construct and mount to a heat sink. Your Aluminum heat sink is now the only piece requiring tapped mounting holes.
Overall height has been reduced
The schematic diagram for this new version is here
If you are building this RF deck from a kit I supplied, here are the assembly instructions:
Begin with the output board, and install all of the capacitors to their
positions shown here
Cut two pieces of RG316 coax to a length of 11 inches (28cm). Pair them as shown, and wind them through one of the large ferrite cores, passing the paired coax 3 times through the center of the core from left to right.
By using two pieces of identical 50 ohm coax like this in parallel, we are
duplicating the characteristics of 25 ohm coax. At 50 MHz, this RG316
Teflon coax can handle more than 3 times the power it will actually need to
handle in this circuit.
Remove 1/2" insulation from each of the 4 ends. Flare the shields back,
leaving about 5mm of the shields unflared; tin the unflared shields at the
location of the red arrow.
Trim the flared parts of the shields away with a pair of diagonal cutters; this is best done by holding the cutting blades parallel to the coax as the flared portions are trimmed away.
Taking care not to cut into the center conductors (which weakens them), remove all but 3mm of insulation from the ends as shown in the photo, and tin the center conductors.
Prepare the second transformer in the same way as the first.
Install the first transformer by soldering both cables at one end to the traces as shown, center conductors to the bottom trace, shields to the top, making certain the insulated part of the center conductors is all that bridges the gap.
the other ends of the transformer across the vertical gap between the bottom two
traces, center conductors to the right side, shields to the left.
the second transformer to the right side in mirror image style in the same
manner as the first (see photo).
Make the bifilar drain choke; locate the two #14 tinned magnet wires, straighten them out, and pass them
through one of the larger ferrite cores as shown here. One should extend out the
end of the core by 1.75 inches, the other by 1.25 inches.
Bend these wires coming out of the top away from you over the top of the core to secure their length.
Pull the other end of the wires toward you, and fold up against the side of
the core. Pass the ends of these wires through the top of the core and through
the bottom. Place the remaining turns to the right of the first one.
the ends of these wires in the same direction as the other ends, and the choke
Form the short leads as you see them in the photo, trim excess length from the ends (if necessary), and solder to the two inner traces.
form, trim and solder the longer wires to the outer traces.
Construct the output balun
The RG402 coax supplied should be trimmed to a length of 36.5 cm (the extra 2 inches left over are provided for practicing how to safely trim it.
Practice these next steps on the short piece of RG402 left over from trimming; proceed with trimming the balun ends only when you feel you've mastered the technique.
Remove 1/2 inch (12.5mm) insulation from each end, exposing the outer conductor. Roll the coax under a knife blade to 'score' it, at 5mm away from the insulation, taking care not to cut completely through the shield (outer).
a pliers, bend the shield end back and forth, breaking it away at the scored
location. Once it is broken, it can be removed with a pair of diagonal cutters,
or just pulled away from the end of the cable. This needs to be done at both
ends of the balun.
both ends, remove all but 2 or 3mm of insulation from around the center
conductor, taking care not to cut into the center conductor itself. The balun is
now ready to be installed.
the coax into 3 loops, and set up the ends as shown here.
with the leftmost lead as seen in the photo, form the center conductor so it will lie flat on the
output board trace
in plane with the outer, and solder it in place across the gap as shown; center
conductor soldered to the left side, outer soldered to the right.
In this last step, solder the remaining end in place at the output pads; outer conductor to the ground trace first, then bend the center conductor down onto its trace and solder it in place.
The output board is finished.
Continuing now with assembling the input board; install all components (except T1) as shown in their locations below. The component locations for RG1,2,3, C5, rtb1, C6 and rtb2 are not used in this design. The R13 designator is next to the pad with the two 68pf chip capacitors and the 120 nh inductor. All other markings are correct.
7.5 inches (190mm) of RG316 and remove 5mm of insulation from the center,
exposing the shield in that location. Tin the shield there.
2.25 inches (57mm) insulation from each end, exposing the shield.
the shield in place with your left hand, push the right end of the shield to the
left, forming a mushroom bulge about 5mm away from the insulation on the left.
Tin the 5mm of shield at the spot referenced by the red arrow.
a pair of diagonal cutters with the cutting blades parallel to the coax, trim
away the mushroom bulge and slide the loose shield off of and away from the
insulated center conductor.
the coax into the smaller ferrite core in criss-cross fashion as shown.
the transformer as shown, and pass the coax on the left across the core to the
right, then through the core from right to left. Solder the exposed shield
to the left gate trace, taking care not to bridge the gap between the traces.
the coax on the right across the core to the left, then through the core from
left to right. Solder the exposed shield to the trace on the right, taking care
not to bridge the gap between the traces.
The wire on the right (in this photo) will go to the R13 pad, and the one on the left to the ground pad near C3.
Trim the ends to the correct lengths as in the photo...taking care not to cut into the center conductor, trim 4mm of insulation away from each end and tin them. Solder them into place.
Solder the coax center tap to the trace below it (one end of R10 is also connected to this trace).
The input board is finished.
Your heat sink should be drilled/tapped for 40-40 machine screws (or your metric equivalent) using the pattern in this template.
Position your copper spreader (with LDMOS attached) over the drilled pattern in your heat sink.
The recommended way to attach your LDMOS to the copper spreader is to flow-solder it as you see it pictured here. Machined spreaders with LDMOS attached using this process are available on the parts page here.
You can use screws to mount your LDMOS, but the heat transfer and rf
grounding will be inferior.
Slide the boards into place under the transistor tabs.
Secure the board and spreader to the heat sink with 4-40 x 5/8 machine screws (or your metric equivalent) using flat washers under the screw heads. Once the boards are in proper position, tighten the screws.
Solder the 4 transistor tabs to the PC boards.
Turn on the 50v main supply voltage, but not the bias; there should be no current drawn
Turn on the bias and note the idling current drawn from the 50v supply. Adjust IDQ for 2 amps. Note: the current drawn by the bias supply (usually12v) is not what you are measuring here...you must measure the idling current (IDQ) the LDMOS draws from the 50v supply.
Shut off the power supply, and remove current limiting.