Mosfet Rf
's
Allmode HF band RF Power Amplifier
for HF: 80*, 40, 30, 20, 17 and 15 meterband
MOSFET: Metal-Oxide Semiconductor Field Effect Transistor: MOSFET: Metal Oxide Semiconductor FET: MOTAS: Member Of The Appropriate Sex: MOTD: Message of the Day: MOTOS: Member Of The Opposite Sex: MOTSS: Member Of The Same Sex: MOUS: Microsoft Office User Specialist: MP3: MPEG-1 Audio Layer-3: MPAA: Motion Picture Association of America. IXYS-RF offers N-Channel enhancement mode RF Power MOSFETs that are optimized for RF and high speed switching.
RE-PA30HF5C
By Guy, de ON6MU
RE-PA30HF5B rev1.1b oct/09 5 band switchable replaced/discontinued
Rev2.0 jan/19
Prototype
About the HF amplifier RE-PA30HF5C (prototype)
This project uses a widely available IRF510 MOSFET. This N-Channel enhancement mode silicon gate power field effect transistor is an advanced power MOSFET designed, tested, and guaranteed to withstand a specified level of energy in the breakdown avalanche mode of operation.
MOSFETs operate very differently from bipolar transistors. MOSFETs are voltage-controlled devices and exhibit a very high input impedance at dc, whereas bipolar transistors are current-controlled devices and have a relatively low input impedance. Biasing a MOSFET for linear operation only requires applying a fixed voltage to its gate via a resistor.
The built-in self-regulating actions prevent MOSFETs from being affected by thermal runaway, but still needs some thermal protection (R6). MOSFETs do not require negative feedback to suppress low-frequency gain as is often required with bipolar RF transistors.
I chose the IRF510 because lots of hams use 'em and they're cheap. But they perform a bit less when it comes to constant gain and/or power output across a wide range of frequency bands. I wasn't especially concerned with that, and the advantages outweigh the contra's, so I went with that MOSFET.
Rather than using a 1:4 toroid (which is excellent) to match Q1 impedance to 50 Ohms, I have applied the 'old school' radio valve coupling; impedance matching circuitry between the output and the antenna using a L-filter...Why? FET devices are more closely related to vacuum tubes than are bipolar transistors (and because I do like to do things my way HI). Both vacuum tubes and the FET are controlled by the voltage level of the input rather than the input current. They have three basic terminals, the gate, the source and the drain. These are related and can be compared to the vacuum tube terminals. The ralationship between the two doesn't stop here...The two most important relationships are called the transconductance and output. An advantage of MOSFET devices is that they do not have gate leakage current and MOSFETs do not need input and reverse transconductance.
The amplifier is made to be driven by transmitters in the ½ to 2 watt range. Built-in to the power amplifier is a sensitive (Q2) T-R relay which will switch the unit in and out of the
antenna line. When in receive, the amplifier is bypassed and the antenna feeds directly to the input jack, when you go to transmit, the T-R circuit detects the transmit RF power and automatically switches the power amplifier into the circuit and amplifies the applied RF power. If you decide to run 'barefoot' turning off the AMP it will disable the amplifier and your QRP
transmitter will feed directly through the amplifier without any amplification.
Power is supplied by any 14 to 25 volt (or 2 x 12v battery) DC source with a current draw of 1 to 2 amps depending upon RF power output and applied voltage.
The linear amplifier can be used with QRP SSB/CW/FM/AM/PSK transmitters on any of the amateur bands between 40m...15 meters. 80m with limitations;
The completed amplifier will reward the builder with a clean, more powerful output signal for a QRP rig when radio conditions become marginal.
Band selection
Switching beween bands could be done manually using a rotary switch.
You can build the amplifier for only one band or a combination of any other of the five available bands.
Drive
The input drive can be anything from 0.3watt to 1.5 watt (2 watt is outer max), which will be amplified to +/- 22 watt respectively.
The output varies on the drive power, frequency and the applied voltage.
Power
The power output is not perfectly linear to the input frequency/band. The impedance 50 Ohms match could be solved by using a 1:4 toroid, or as I like to use, the 'old school' radio valve coupling; impedance matching circuitry between the output and the antenna using a L-filter...And, the IRF510 isn't perfect (note: there are also low grade versions of the Mosfet out there which can lower the output power and influence the quality of the signal/waveform).
The N-channel mosfet has an input capacitance thats a bit on the high side and the output capacitance that varies with the cross over frequency. It can be a slight problem when it comes to constant gain and/or power output across a wide range of frequency bands. I wasn't especially concerned with that so I went with this MOSFET anyway. Of course the main issue was the simple design to be able to use one band or even up to five bands if wanted, which always has some compromise in this type of design. This means that there is some fluctuation of the output power par band.
When driven between the optimal range of +/- 1 watt to 1.5 watt the amplifier more than capable to deliver 22 watts +/- 5%. Output power for AM should be set to +/- 50% of max.
Although the design allows you to work in a varied range of voltages, the maximum output is only guarenteed @ 24volts.
I'm sure if you peak the amp close to perfection for one specific band you could get more power. Quality of all components, construction etc also influences the performance.
I used breadboard to build my protoype and some 'dead bug' work. Dead bug prototyping and freeform electronics are a way of building working electronic circuits, by soldering the parts directly together, or through wires instead of the traditional way of using a printed circuit board (PCB.)
Graph: Input/Output Power vs Voltage
Higher power than 2 watts does not improve linearity and could damage the mosfet.
Bias
The power amplifier require biasing for proper RF performance. BIAS has be applied to Q1 to have clean proper and correct SSB modulation using this amplifier. Set P1 so that +/- 80 mA current flows through Q1.
Thermal protection
R6 is a PTC resistor that allows which is used here for thermal protection. As the resistor heats up the resistance increases, which lowers the bias voltage. R6 should be placed near Q2.
Modulation modes
If proper BIAS to Q1 is applied, you can amplify any type of modulated wave.
Output power for AM should be set to +/- 50% of max.
Filter
RF purity and harmonic suppression is done here. Also allowing the FET to be coupled to the antenna system through antenna impedance matching circuitry (C16, Ct1, L2, Ct2, C21, C26, L4, C27). Care is taken at this stage so that no harmonic frequency is generated which will cause interference in adjacent band/harmonics on other bands. This 4-element L-type narrow band-pass filter circuit and a 3 element low-pass Butterworth PI filter for the desired frequency removes out any remaining harmonic signals efficiently.
A picture from my oscilloscope:
RF-sensing
The basic principle of RF-sensing using a relay is clearly drawn in the schematic and pretty much self explaining.
Tip: I would like to recommend to add a mini-switch between C31 and GND if you plan to use it for CW modes. The on-time is to long for continues wave modulation formats.
Input Attenuator
I made provisions to include an RF attenuator consisting out a Pi network of R2, R3/R4, R5 which gives a Forward Attenuation of 3.63 dB and a Input Return Loss of 23.23dB. There are numerous of reasons why I implemented it in this design. It improves overall linearity, achieves some 'protection' and enhances stability of the drive input (being a transmitter, transceiver) and Q2 gate.
Cooling/heatsink
Q2 needs to be mounted isolated from the heat sink. Use proper thermal grease and isolator.
I used an old P3 heat sink, which work just fine.
I mounted a Pentium 3 heatsink on the back of the alu-casing. A square space is cut out of the back of the alu-box to allow Q2 to be screwed onto the heatsink. The heatsink is firmly mounted on the back of the chassis with thermal grease allowing the chassis as extra cooling surface.
Construction considerations
HAMs that are experienced in constructing RF projects will know a number of possibilities to create a good RF design.
Note when using a band switch selector: Because I started from scratch and still was in experimental/design stages I have placed the capacitors/trimmers of each band directly around the switch. This works when short connections are used. You can however solder them directly to the PCB. Always start with the lowest band and set the capacitors to maximum output and work are way up from there. Also set your transceiver to the middle of the (each) band segment..
I mounted a Pentium 3 heatsink on the back of the alu-casing. A square space is cut out of the back of the alu-box to allow Q2 to be screwed onto the heatsink. The heatsink is firmly mounted on the back of the chassis with thermal grease allowing the chassis as extra cooling surface.
One thing on the trimmer capacitors (Ct1x and Ct2x). Do not use plastic trimmers, they will melt and perhaps burn through causing shortening and possible failure of Q2 and who knows what else. Please use air- or ceramic based trimmers.
If you do not have them, then the only way tweaking the amplifier is by trial-and-error, (C16x and C17x).
Use a choke (or a snap-on ferrite bead) at the point where the Vcc wires leave the alu-box.
Try to limit the maximum current to 2 Amp. I placed a 1 Ohm resistor R12 in series to buffer a minimum the maximal current and peak at power up.
Use small 50 Ohm coax between the in- and output of the PCB connections to the SO-239 connectors.
Enclosure Recommendations
To accomplish RF shielding the whole circuit needs to be mounted in an all-metal/aluminum case.
Grounding
To prevent ground loops, spurious oscillations etc. please take attention to:
- decouple the PCB in the chassis (housing)
- all connections and wire leads should be made as short as possible
- a proper PCB layout with enough ground surface ensuring normal ground paths
- the source of Q2 (IRF510) should also be grounded to the chassis as close as possible
-
Specifications RE-PA30HF5C
Allmode: AM/FM/CW/SSB/FSK
Bands: 80m*, 40m, 30m, 20m, 17m, 15m*
Average output RF power: +/- 22W @ 24v
Quality of all components, construction etc also influences the performance. I build it on breadboard and some 'dead bug' work. It's a prototype and always under construction HI.Works great with Yaesu FT-817, Ramsey QRP rigs or any other 0.5-2 watt transmitters
Input power drive: 250mW...2watt top peak max (ideal 1.5watt)All modulation modes
Efficient band-pass type harmonic L-filter + lowpass Butterworth PI filter
Usable voltages: Vcc 13.8 - 25 volts
Average current I: +/- 1.65A @ 24v at full load, minimal 0.6A @ 14v
Built-in T/R relay automatically switches between receive and transmit
VSWR overload resistant (short period of time, not unlimited!)
Multistage band pass and low pass filter for a clean signal
Optional: Manual band switching (if build for more than one band)
The 5-band HF power amplifier 'insides' (this pic shows the version with band switch)
The MOSfet HF-band Amplifier settings
Needlessly to say, but I will say it anyway, before testing anything please be sure to double check every connection. The project should be finished HIHI.
Connect a proper dummy load and a power meter to the output of the amp. Also put a Ampere meter in series with the Vcc, allowing monitoring of the current during the setup.
Set all trimmers (Ct1 & Ct2) half way (in medium setting).
Set P1 to the ground (zero ohms).
Now gently increase the voltage to the amplifier while checking the current till you reach 18 volts. The only current you should see is a the liddle idle current of Q1 (a few milli amps and a a few mA of LED D3 if connected). We do not need the full 24 volts during the tuning/setting stages.
Now gently turn P1 till you get approx. 70...80 mA.
So far so good? Now we need to check if the (Q2) RF-sensing circuit is working properly (Although I would like to recommend to test this before anything, rather than building the entire project and test it. Or at least before mounting the PCB in the alu-box, and without Q2 soldered. The RF-sensing ON-time can vary according to the relay used).
Connect your transceiver (or other drive) to the input, and set it to the lowest power rating of +/- 0.5watt and set your transceiver to the band you designed this amp for and place it CW/FM.
Be sure the dummy load is still at the output of the amp.
Key your transceiver and if all goes well the relay (Re) should power up and you should see the current rise and your power meter should already show an amplification of the RF input power.
Also set your transceiver to the middle of the (each) band segment.
All working as planned? Excellent! Now we need to tune the filter-unit by setting the Ct's to maximum output.
Set the drive power (your transceiver) to +/- 1.5 watts or less
At 14 volts the average amp would be around 0,6 amp,
Current should be around 1.7 Amp +/- max @ 24v (depending on the voltage, input power and output impedance which should be 50 Ohms).
Note if working with a band switch:
Turn the band switch (if used) to the lowest frequency/band, as we start with the lowest band and work are way up from there.
If you build the amplifier for more than one band, next is to repeat the above for each band and setting the Ct capacitor trimmer(s) according to each band respectively.
After the filter is tuned in respect to each band you can increase the voltage to 24 volts. Check everything again, band by band. Could be that you notice a slight difference in the peak output power, do to the capacity of the switch and the filter components. Just re-tune (if needed) each trimmer (Ct1...Ctx) for each band respectively.
The maximum current of the amplifier should never exceed 2 amps.
RF-sensing considerations
The basic principle of RF-sensing using a relay is clearly drawn in the schematic and pretty much self explaining. Q2 (BC338, 2N2222) will conduct when RF energy is applied at the input of the amp (via R10, C29, D3, D4, C30 biasing the base of Q2) hence powering up a RF capable relay. This relay switches between RX and TX with amp. When no Vcc is applied to our amplifier (and so Q2 too) no amplification is done bypassing the amplification. The input is simply re-directed directly to the output (as if your transceiver is connected without an amp). The RF sensing circuit is sensitive enough to react on .5 watt easily.
To allow the amplifier in SSB-modulation some extended PTT time-on the RF-sensing unit (Q2->relay) has to be increased. C31 adds the needed 'breathing' time. In FM/CW/AM/FSK modes a carrier is present and extended PTT time-on of the amplifier isn't needed, hence can be short.
Important: Everything will be within specs if you use RY5W relay, but timing delay (the 'breathing time') can vary on the type of relay used (Ohms resistance value of the relay coil), hence experimentation of C31 is needed.
Although this example of RF-sensing isn't the Worlds most best sollution, it is pretty easy for beginners and effective though. Better would be to drive Q2 from your transceiver (amp drive) as this will switch the amp at the very moment of PTT.
Tip: I would like to recommend to add a mini-switch to disconnect C31 if you plan to use it for CW. The delay is too long for those modes.
Note:
Always use a dummy load for testing and adjusting the amplifier!!!
Remember that this is a prototype.
Parts list 5-band HF power amplifier
Q1: N-Channel IRF510 MOSFET
(with proper heatsink isolated from the mosfet)Q2: NPN BC338/337, 2N2222...
IC1 = 78H05 or 78L05
C1: 100n
C2: 1uF/50v
C3: 1uF/16v
C4: 100n
C5: 2.2uF/50v
C6: 100n
C7: 4.7nF
C8: 4.7nF
C9: 220uF/50v
C10: 100n
C11: 47n
C12: 1uF/16v
C13: 68n
C14: 100n
C15: 100n
C16: all 150...200volt ceramic
15M -> 30p
17M -> 39p
20M -> 68p
30M -> 180p
40M -> 2 x 150pF parallel (or 330)
80M* -> 2 x 220pF parallel (or 470), NOTE: 80m band is stil experimental, power goes only to 14watt. Needs further investigation/experimentationC17: all 150...200volt ceramic
15M -> 100p
17M -> 120p
20M -> 220p
30M -> 240p
40M -> 470p
80M -> 1200p NOTE: 80m band is stil experimental, power goes only to 14watt. Needs further investigation/experimentationC26: 220, ceramic 200v
C27: 100, ceramic 200v
C28: 2n2
C29: 470p
C30: 47n
C31: 68uF/tantalum 16v (determines the ON-time for RF-sensing)
C32: 150n
Ct1: 0...100pF ceramic or air-spaced trimmer
(for 15M to 20M: 0...40pF)
(for 20M to 80M: 10...100pF)Ct2: 0...40pF ceramic or air-spaced trimmer
R1: 47 1/2w
R2: 390 1/2w
R3: 47 1/2w
R4: 47 1/2w
R5: 390 1/2w
R6: 470 PTC
R7: 1k
R8: 10 1/2w
R9: 18k
R10: 1k
R11: 1k
R12: 1 Ohm - 5 watt
R13: 560
P1: 5k potentiometer (BIAS setting Q2)
D1, D3, D4: 1N4148
D2, D5: 1N14001
S1: Toggle switch (ON/OFF-Bypass)
S2: 5-position quality switch (if possible silver plated)
2 x SO239 connectors
Re: RY5W-K relay
F1 = 2 amp slow
Alu-box
Heatsink + thermal grease
Dr1: ferrite core 3mm diameter, 5...8mm long. 40 turns, 0.3mm wire (+/- 5uH)
Dr2: yellow/white toroid of +/- 13mm diamter, +/-55 turns of 0.5mm wire (values between 100...200uH works)
Note: It could be that some experimentation is needed to find the optimum value.Dr3: a ferrite bead with 4 turns of 0.6 mm wire
L1: 29nH; 2 turns, no spacing, 5 mm inside diameter, 0.6mm wire
L2: 1.4uH; 22 turns close together, 1.2mm enameled wire. Inside diameter is 9.5 mm (27mm long)
L4: 410nH; 8 turns close together of 1.2 mm enameled wire. Inside diameter is 6.5mm (10mm long)
.
Coils
All we need to do now is make a few remaining coils that have to be handmade - for that 'old-world craftsmanship' touch!
The wire used for the coils are enameled wire (stripped from any AC transformer).
For 80M band you need an additonal coil in series with L2. Secondary coil for 80M Coil specs: 3.2uH; 15 turns close together, 0.6 mm enameled wire. Inside diameter 9 mm and 14mm long. You could add the numer of turns to L2 making one coil if you would use this amp only for 80M band. All feedback is welcome as this band needs further tweaking with this amp.
Dr1: you need a ferrite core of 3mm diameter and about 5...8mm long. You wind 40 turns up and down the core, with no spacing. Wire used is 0.3mm enameled wire.
Dr2: small yellow/white toroid of +/- 13mm diamter (like those often found in PC switched power spupplies etc.). It has about 55 turns of 0.5mm enameled wire.
Tip: remember to vernish or glue-fix the coils to prevent FM'ing do to vibrations
Note: the caps C16 till C17 may have higher voltage specifications, but no less than 100v.
IRF510 MOSFet specs:
Drain to Source Voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .VDS: 100 V
Drain to Gate Voltage (RGS = 20kW) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .VDGR: 100 V
Continuous Drain Current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .ID: 5.6 A
TC = 100oC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ID: 4 A
Pulsed Drain Current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IDM: 20 A
Gate to Source Voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VGS: ±20 V
Maximum Power Dissipation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . PD: 43 W
Linear Derating Factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.29 W/C°
Single Pulse Avalanche Energy Rating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . EAS: 19 mJ
Operating and Storage Temperature Range . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . TJ, TSTG: -55 to 175 C°
Input Capacitance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . f = 1.0MHz - 135 - pF
Output Capacitance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . COSS - 80 - pF
Reverse-Transfer Capacitance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CRSS - 20 - pF
Internal Drain Inductance LD Measured From the Contact Screw On Tab To Center of Die . . . . . . . . . 3.5nH
Pulse Source to Drain Current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .ISDM - - 20 A
Source to Drain Diode Voltage VSD TJ = 25C°, ISD = 5.6A, VGS = 0V. . . . . . . . . . . . . . . . . . . . . . . 2.5 V
Related
MOSFET specs:
. .
.
LT-Spice simulations
Pictures of users who build the project
This is how John SV1ONK did it:
He made it for the 20-meter band.
Thanks John!
This how Konstantinos SV1ONW made it:
Thanks Konstantinos!
Little note on Antenna's
It's important to use a correct designed antenna according to band you would like to operate, or at least use a good antenna tuner to match the antenna (protecting your transmitter and proventing harmonics/interference...).
A resonant antenna is an absolute requirement for QRP operation, and an amplifier is not a 'band-aid' for a poor antenna system!
We cannot expect good results from low levels of RF output if the power gets wasted in lousy coax, corroded connections, or poor antennas. Several examples can be found on my website and all across the Web. A dipole is always a good alternative (total length = 150 / freq - 5%).
Mosfet Rf Amplifier
Another related project:
..15 & 17 meter band transistor 10 watt amplifier
Remember that transmitting and/or using an power levels higher that your local license permit is illegal without a valid radioamateur license!
Home
Mosfet Rf Amplifier Design