Home
High power amplifier for 1296
1 KW SSPA for 1.8-54 MHz
A 1.5 KW LPF for 160-6m
1.8 to 54 MHz Dual Directional Detector
1.8 to 54 MHz combiner set
Automatic Transverter Interface
1 KW 6 Meter LDMOS Amplifier
2 Meter 80W All Mode Amplifier
1 KW 2M LDMOS Amplifier
1 KW 222 MHz LDMOS Amplifier
500w 70cm Amplifier
1KW 70cm LDMOS Amplifier
A Big Power Supply for SSPAs
Low Pass Filter/Dual Directional Detector
Sampling RF Power
LED Bar Graph Meter
Amplifier Control Board
LNAs (preamps) and MMICs
LNA Sequencing and Protection
Building UHF Antennas
VHF OCXO
MIcrowave Marker
Crystal Oven Controller
Microwave L.O.
Latching Relay Driver
12 to 28v
Relay Sequencer
High Current DC Switch
L & S Band LNA
Microwave L.O. Filters
PC Board Filters
Using Inexpensive Relays
600w 23cm LDMOS Amplifier
XRF-286 Amplifiers for 23cm
150W 23CM Turn-Key Amplifier
300w 23cm Amplifier
200w 23cm Amplifier
100w 23cm "brick"
100w 23cm Transverter
60w 23 cm Amplifier
23 CM Beacon
23cm Signal Generator
23cm Double Quad
23cm filters
13cm filter
13cm Signal Generator
13cm Transverter
120w 13 cm Amplifier
600w+ 33cm Amplifier
300w 33cm Amplifier
33cm filter
33 cm Crystal Source
33cm Signal Generator
9cm Transverter
Transverter Selector
12 AND 28 volts
Klitzing Amplifiers
IC-910H tweaks
Audio Files
Parts I Can Supply
Current Projects
Links

Comments? email to

A 1 KW SSPA for 1.8-54 MHz

It's been a while since I built an amp for HF; I'd have to go back to the 60's and a pair of the venerable 3-500z tubes that I put together back then. Time to upgrade, so I built this one. A legal-limit version was also built...a few photos of it are near the end of this article.

It can accommodate almost any driving radio, and only requires 2 to 3w drive for 1kw out, a potent companion for my Elecraft KX3 and the other QRP rig I own, the FT817. But because it has a configurable input attenuator, it's also a good match for one of the 100w radios used here, and lately that's the FT991.

The cabinet is a bit deeper than the VHF amplifiers to make room for the large low pass filter board (the cabinet measures 12w by 6h by 16d), but still fits into the operating position with ease.

Details on the construction and performance of the RF deck are at the end of this article.

The rear panel has all the usual connections; input/output, power supply, attenuator configuration jumpers, ALC and PTT.

In addition to these is a band select interface connector, which allows the driving radio to control the band switch function in the amplifier. When used, the front panel band switch is disconnected, and the amplifier band select slaves to the radio.

At the time this photo was taken, I hadn't finished the interface electronics yet, hence the jumper plugged in to enable the front panel band switch.

The inside of the cabinet is packed full of the essentials; the RF deck, low pass filter assembly, control circuits, antenna relays and so forth.

It got a bit crowded at the back panel area due to all of the ferrite cores used as RF chokes for external connections, but it all fit OK.

The 3 cable connectors on the left side are for disconnecting the front panel, which is the next photo.

The 3 disconnects allow one to remove the front panel if the amplifier should ever require service, allowing easy access to the rest of the components, (next photo).

The low pass filter board is also removable...the cable disconnect is tucked under the board, which covers up a few other components mounted underneath it.

It's mounted on 2-inch spacers to make room for those parts, and is positioned in the air stream of the cooling fans, located at the back of the heat sink. Air flows out the front of the heat sink fins, across the filter board, and vents through side cut-outs in the cabinet cover.

Shown before most wiring was in place are some of the power resistors and other parts; the two boards on the right are dual directional detectors.

The one at top right is used to sense whether the band switch is in the correct position (the wrong position can cause damage); if it isn't, it locks out the RF deck (via the control board). The one below it measures forward and reflected power for the bar graph power meters, and also provides the high SWR lockout signal to the control board.

If you will be switching the band select manually, this might be a good idea to talk about the best way to set that first dual detector board up to sense whether the band switch is in a potentially damaging position. Click here for details

The input and output relays are mounted to the right side of the heat sink, clear of other components.

The left side houses the control board; mounted on the top left side of the heat sink are the high current FET switch and the input attenuator.

The photo below shows how it fits in with the other kw amplifiers in the station (vhf/uhf). Good thing there are no more bands to design for, I'm out of room for amplifiers.

 

But naturally, the story can't possibly end there...if one RF deck will do a KW, why not use two of them and build a full legal-limit version that will loaf along at 1500w?

I built the one shown here for a fellow who wanted just such an amplifier, but also wanted it to match the color scheme of the other radio equipment in the station...so the panels are black anodized with white lettering.

Looking inside from the side, the heavy lifting is done on top of the heat sink, where the two rf decks, output combiner, fet switches and input attenuator are located.

The low pass filter is mounted vertically in front of the heat sink in the air stream of the fans located at the back.

The small board at lower left is the band select interface; this one was to be used with an Elecraft K3; the amplifier follows the K3 band selection automatically using this interface.

And here's a look at the rear panel connections; Anderson power poles for the power connections, and a db15 connector for the K3 band data control. Plugging in the control cable from the K3 automatically disables the front panel band switch on the amplifier.
A couple more photos of the interior:

A description of one of the 1kw RF decks follows:

Please note... as time goes on and ways to improve the amplifier are discovered, useful changes will be posted as they become apparent. Here are some of the changes made to date:

• These are the instructions to use if your RF deck or kit was shipped after June 2017. This applies all changes to the amplifier since the original prototype.
• These are the instructions to use if your RF deck or kit was shipped after April 30, 2017, and applies all changes through June 30th 2017.
• These are the instructions to use if your RF deck or kit was shipped between July 2016 and April 30th 2017,

The information below describes the original version:

This RF deck is nice and compact, only 4" by 6.5".

Note: Beginning in August 2016, a modification to the input circuit was added to eliminate an instability discovered when operating above 20 MHz into certain narrow-band antennas. This change, and a revised schematic are documented here. The current rf deck or rf deck kit (August 2016 and later) includes this change.

The VHF amplifiers using the newer Freescale and NXP devices also generated interest in their use for the HF bands. I've read some discussion in various user groups about why these devices, originally intended for plasma exciters, can work at VHF but not at HF. The explanations for this seemed to lack any credible science, so I decided to experiment away. What I found was these parts worked very well indeed...you just need to provide some degenerative feedback to keep them stable (the gain is very high at HF). So far, I've tested this with the NXP BLF188xr and the Freescale MRFE6VP1K25H, and one plays as well as the other, no circuit changes are required.

I looked over several designs before settling on the NXP engineering example provided to me; it originally covered 3 to 30 MHz, but after considerable fussing with matching transformers and minor modifications, I managed to include 160m and 6m with a kilowatt over the entire range. Some bands are even better (1200+), but on the whole, 1kw is what you can expect when you attempt to cover such a wide range of frequencies.

Here's a graph showing typical performance after passing through the low pass filter also featured on the site (I've built several of these rf decks now, and they all played the same).

As far as being linear, here's the input vs output ratios measured at 14 MHz:

Don't hold me to the very low drive power readings at .2 and .4w, I was using my Elecraft KX3's power settings to measure that, and the accuracy is good, but not perfect. Still, it looks like 1db compression happens at about 85% of full output. In this case, and on this band, that's at about 1062w out (85% of 1250w). By now you've probably noticed that you can drive this to full output with an average of just 2 or 3 watts.

The input return loss (SWR) is shown below (less than 1.5 to 1 across the entire range).

Two of these rf decks, properly combined, will loaf along at full legal limit (1.5 kw). The article for the combiner set is located here.

The amplifier is a push-pull design, using a dual-LDMOS device; by nature, this type of amplifier does a good job of suppressing even harmonics, but the odd harmonics are monsters, so be sure to use a good low pass filter. Just look at this spectrum analyzer display of the unfiltered output on 40 meters:

After passing through the LPF, the spectrum is much more reasonable, and in compliance with FCC regs.

The PC board material used in the design is not FR4; it's Arlon TC-350, a material that has excellent thermal conductivity (very helpful in drawing heat away from components).

Here's the schematic for the amp:

And a Bill of Materials (BOM)

The LDMOS is flow-soldered to a 3x5x1/2 inch copper heat spreader; the spreader is then secured to the heat sink with machine screws, thermal paste used at the joint. I used 1/2" spacers around the edges of the spreader so the board would not overhang the copper; that way I was able to use my standard KW copper spreaders I stock here for the VHF amps, which are smaller than the HF board by 1/2 inch all the way around.

If you are building this project from a kit I supplied, the assembly guide is here

template for drilling/tapping your heat sink

Flow-solder attachment is the recommended way to fasten the LDMOS to the copper; here's a video on how to do it. If you are hesitant to do this yourself, that service is available...but it's pretty straightforward, and not at all difficult to do.

For those who enjoy building their own, blank boards only, or kits and assembled/tested RF decks are available on the parts page (kits already contain the PC boards). The kits will save you some $, so you'll have to weigh the value of your time against the cost of the assembled versions, though there is also some comfort in knowing the major parts are working correctly and are ready to go. Here is a list of the major assemblies offered there that will be required to duplicate the amplifier shown at the top of the page:

• 1.8-54 MHz Kilowatt RF Deck, kit or assembled/tested (if building from the basic kit, the items in the indented sub-list will also be needed)
  • Machined copper spreader, if ordering the basic kit
  • board spacer set, if ordering the basic kit
  • LDMOS transistor, if ordering the basic kit
  • flow-solder service for LDMOS, if needed
• machined heat sink, drilled/tapped for RF deck above
• 6-segment Low pass filter, kit or assembled/tuned/tested
• two dual directional detectors, kits or assembled/tuned/tested
• two tri-color bar graph displays, kits or assembled/tested
• 48v high current FET switch, kit or assembled/tested
• control board, kit or assembled/tested
• SPDT High power output relay board
• DPDT Mid power input relay board
• ALC board, assembled/tested
• input attenuator - 10 db recommended for 50w driver, 13db for 100w driver

The cabinet consists of the front and rear panels and a floor plate I have made by Front Panel Express (FPE, www.frontpanelexpress.com); they have design software you can download for free, which allows you to design custom panels and order them through their system. Their CNC machining process does all the hole cutting and engraving of those parts, and the panels can be ordered in a powder-coat or anodized finish.

The wrap-around cover and internal support brackets and tie-ins I make here; If you have your own sheet metal tools (shear and bending brake), you can save some $ there. The Panels for the 2-deck unit are larger and more expensive (about $335 from FPE).

One other decision to make on the 2-deck cabinet is the cover fitting flush and behind the front and rear panels, or with a 1/8 inch over-hang (shadow cabinet cover). I prefer the latter, but some do prefer the former; it can be done either way.

If you do want to go the Front Panel Express way, a block diagram of the amplifiers, and my design files for the panels mentioned above can be found in this folder.

The following two lists are the full bill of materials (BOM) for the single deck amplifier, and following that, for the 2-deck amplifier; the auto-band-select interface is omitted in these lists, as this is clearly an optional component, and will vary with the type of radio you use with your amplifier. For Yeasu and others with BCD encoded outputs, the board offered by Unified Microsystems is a good choice. For Flex SDR radios, a program called DDUTIL and a different interface card is necessary; for Apache (ANAN) SDR's, no interface is required, these radios already have control outputs that can operate the amplifier's band select relays directly.

This first BOM is for the single-deck amplifier (1kw):

The following BOM is for the 2-deck amplifier (1500w +):