| Easy And Accurate Output Tuning (without an oscilloscope!) |
The AMT5000 manual briefly mentions a tuning procedure that calls for adjusting for maximum amplifier drain current then increasing the tuning by one turn. While this will generally work in most cases, it's not very precise. Class E efficiency takes some effort and an oscilloscope to achieve and that might be more than the casual Part 15 broadcaster is willing or able to do.
The AMT5000 is an extremely capable transmitter and is arguably the most powerful Part 15 AM transmitter given its Class E output amplifier compared to the more common Class C design used in other transmitters. Since all Part 15 AM transmitters are restricted to 100mW power input, efficiency will determine the output and the AMT5000 has a potential 15% or more advantage. HOWEVER, and this is important, there are several factors that need to be considered before expectation levels soar.
1) Absolutely nobody will notice a 15% change in signal strength unless they are using sensitive test equipment. It would literally take an increase in output power far beyond the limitations of the Part 15 rules for the casual listener to recongize a gain in signal strength or coverage area. Noticeable changes will more likely come from antenna improvements and that's where the most attention is needed.
2) Getting Class E efficiency in the upper 90% range takes careful adjustment, far more than tuning for peak current then "adding an extra turn" on a trimmer. While following the directions in the AMT5000 manual will produce a highly competitive signal, it's not optimal, and that's where the line is drawn between the casual user who wants to broadcast to radios around the house and the competitive user who wants to boast about having one of the most powerful Part 15 transmitters.
The AMT5000 Challenges - In looking at the design of the AMT5000, it's clearly a Class E capable transmitter but there are four critical isues that can distort the Class E waveform, reduce efficiency, prevent harmonic filtering, and restrict positive modulation peaks.
1) The IRLL014N R.F. output FET should be seeing a load with low reactance yet it is connected to an inductor (the toroid) in the output circuit without a series capacitor to cancel the inductive reactance at the operating frequency. Reactive loads cause the R.F. voltage at the drain of the FET to rise far above the D.C. voltage. The value of the D.C. blocking cap C4 is too large for this purpose so another capacitor is needed that can be adjusted.
2) Another cause of excessively high R.F. drain voltage is insufficient capacitance between the drain and source leads of the FET to prevent excessive voltage peaks due to the "flyback" effect of the inductor in the tuned circuit. While the IRLL014N has internal drain-source capacitance (Cds) it's typically not enough for the full range of operating frequency and load impedance combinations. Interestingly, two jumper-selectable capacitors are included in the AMT5000 for this purpose yet the manual doesn't go into any detail about their use.
3) Unlike in the AMT3000, there is no provision for measuring relative output power, and it is critically important to be able to compare the output and input power in order to tune for maximum efficiency.
4) While the toroid (L1) is only used for matching to a 3 meter wire antenna and can be defeated by placing a jumper at S2, one end remains connected to the circuit which causes issues when matching to other loads.
Solutions! - The first step is to be able to measure the relative R.F. output. Incredibly, many field strength meters designed for CB radio use work very well.
If you don't have one, you could build the simple circuit shown below.
Circuit notes: The component values aren't critical. The capacitors can be anywhere from 100 to 1000pF and the diodes can be 1N914, 1N34, or any small signal type. The meter, however, should be analog (not digital) and as sensitive as possible. A VOM set for a low scale such as 2.5 Volts would work well. You could even use a dedicated meter as long as it can read less than 500uA full scale. I've found that a 100uA full-scale meter is ideal. A ground connection might not be needed.
The input would depend on whether you were using a 3 meter wire antenna or an external matching circuit (described below) to connect to a 50Z coaxial cable. If you use a 3 meter wire attached directly to the transmitter then a short whip antenna or wire a foot or two away should be used as the input. If you have an external load where the transmitter is connected to a coax then the input can be attached directly to the output of the tuning circuit where the coax connects.
R.F. FET Drain/Source Loading Capacitance (S14 & S15)
The next section shows the importance of the S14 and S15 jumper positions that are used to add damping capacitance to the IRLL014N R.F. output FET for a proper Class E waveform and to provide a noticeable increase in positive peak headroom. Users of external processing set for asymmetric clipping will appreciate the dramatic difference, particularly if you have a modulation monitor that shows negative and positive modulation. Another benefit is the reduction of harmonic energy resulting in more power on the desired frequency.
It's important to keep in mind that this is intended only for users who want to optimize their transmitter. It is not necessary to do this if you are happy with the functionality of your AMT5000 as it is. Additional components might be needed, and if you built your transmitter as a kit, you have the skills to do these modifications. And, the obligatory statement, proceed at your own risk.
3 Meter Antenna (Efficiency: >86% )
The AMT5000 output matching circuit is already in a series L/C configuration so if you have the relative R.F. output meter circuit described above then you're almost done. Several tests between 1500 and 1700 kHz have shown that the load impedance on the IRLL014N R.F. FET will be fairly low so a lot of damping capacitance will be needed to limit the peak R.F. voltage at the drain. This is done by placing a jumper at S14 to add 470pF.
Use C1 to tune the transmitter for peak output as seen on the relative output meter which should be located a foot or two away from the antenna depending on how long the antenna is that you connected to the meter circuit. You might have to move it further away if the meter reading gets too high.
It is likely that you will notice that the meter reading will change not only as you adjust C1 but also as you move your arm near the antenna. This is due to a detuning effect where you are actually adding capacitance to the antenna by being near it. You will find that adding a bit more capacitance (past peak) then slowly moving away will help get the meter to peak when you're not closeby. Adding very small amounts of capacitance and moving away repeatedly will allow you to find just the right setting so that the relative output power meter is peaked when you're not near the antenna. Interestingly, this happens when tuning ALL part 15 AM transmitters with a directly connected antenna.
Once the relative output meter peaks to its highest reading when you're away from the antenna, adjust R1 according to the manual for setting the power input to 100mW using the chart on page 41 of the manual. And as expected, you'll have to back away in order to read the voltage (test points T1 and T2) and current (testpoints T2 and T3). It's challenging, but once you get the tuning right then your transmitter will be operating at a MUCH higher output and efficiency than if you just followed the tuning instructions in the manual!
It is likely that the voltage and current combination will be around 1.9V at testpoints T1 and T2, and 0.526V measured between T2 and T3 (52.6mA). This represents a 36Z load to the FET.
50 Ohm Coax - Easiest Way (Efficiency: ~90% )
A basic series L/C circuit provides the simplest means to couple R.F. from a Class E amplifier to a low impedance load. While there are only two components, each serves an important function. The inductor helps convert the squarewave output of the transmitter to a clean sinewave while the capacitor cancels the reactance of the inductor so that together they appear as close to a low, non-reactive resistance as possible. A series L/C circuit also functions as a band-pass filter which reduces unwanted harmonic energy.
You can make either the inductor or capacitor variable, but either way, keep in mind that all connections carry R.F. and none can be grounded. This means that insulated mounting hardware should be used if you plan to mount these in a metal box. The inductor and capacitor values are somewhat flexible but need to resonate at your operating frequency. It's also important to have enough inductance so that the output is a clean sinewave.
Since there is no provision for impedance matching, the AMT5000 will see a low impedance load close to the value of what you have connected. This means that a bit of damping capacitance will need to be added, and since the load impedance is low, you'll probably need 470pF which is done by placing a jumper on S14.
Disable the built-in inductor L1 by placing a jumper at S2. You also need to put a jumper at S3 to provide a termination for the unconnected end of the inductor which is at an extremely high R.F. impedance.
Connect the relative power meter circuit to the output of the series L/C tuning circuit and adjust the tuning of that circuit for a peak reading on the meter. Use the power setting chart on page 41 in the manual to set R1 for 100mW input. With a 50 Ohm resistor as a load, the transmitter power control R1 will likely end up being set for something close to 2.4V @ 0.0417mA (0.417 Volts measured between T2 and T3).
50 Ohm Coax - Better Way (Efficiency: >95% )
Unlike the matching scheme for a 3 meter antenna, coupling to a 50Z load such as a coax is best handled with an external "T-match" circuit for more flexibility. It also takes advantage of a higher FET load impedance of 203Z when operating at 4.5V @ 22.2mA. A higher load impedance means that the damping capacitance can be reduced which increases efficiency even more but the AMT5000 only has options for adding 470pF (S14) or 1000pF (S15). Either is too much although 470pF can be used at a slight cost in efficiency.
There are several options here. One is to install a 120pF mica capacitor in place of either of the above capacitors and use a jumper to add it. Another is to install a 120pF mica capacitor under the board between the drain and source connections of the FET. I can't think of any configuration where this capacitance wouldn't be needed. The third is to add a connector to each lead of a 150pF mica capacitor so that it can be used as the jumper at S14 (150pF and 470pF in series = 114pF). The leads should be kept to less than 1" in length.
The next step is to build the T-match circuit:
The size of the inductor will depend on the operating frequency but the easiest type to use is an airwound inductor that uses small clips to attach the wires. Notes:
- L1 and L2 can be the same long coil or two separate coils.
- C1 is used to add capacitance to C2 since a lot is needed in low impedance circuits.
- L1 and the two capacitors are a series resonant circuit at the operating frequency.
- L2 is used for loading.
Disable the built-in inductor L1 by placing a jumper at S2. You also need to put a jumper at S3 to provide a termination for the unconnected end of the inductor which is at an extremely high R.F. impedance. The T-match should be connected between the transmitter output and a 50 Ohm resistor (for adjustment). Connect the relative power meter circuit to the output of the T-matching circuit then adjust the variable capacitor for a peak reading. Using the power control R1, set the drain voltage (T1 and T2) to 4.5 Volts then check the drain current (T2 and T3). It's likely that the tap on L1/L2 or the value of L2 will need to be changed in order to get a drain current of 22.2ma (0.222 Volts at T2 and T3).
You may need to keep adjusting R1 for 4.5 volts with each change in current but eventually you'll have the 4.5V @ 22.2mA combination you need. At that point, tweak the T-match tuning for a peak reading on the relative output meter.
50 Ohm Coax - Best Way (Efficiency: 97% )
Incredibly, this matching circuit provides even greater efficiency although we're pretty much getting to the maximum that can be expected from even a Class E transmitter. Minimizing circuit resistance is critical in getting every last percent of efficiency so good connections and the use of short thick leads is important.
This configuration is unique in several ways...
- The single roller inductor is used both as a tank circuit and for impedance matching.
- An output voltage monitor circuit is not needed since the inductor is adjusted for proper FET current.
- The FET loading of 203Z reduces the need for more damping capacitance which increases efficiency.
- The use of a variable inductor for matching adds more inductance when its needed for lower frequencies.
The roller inductor was able to easily match operating frequencies from 1300-1700kHz to a 50Z load with no changes in capacitor values. Keep in mind that most roller inductor shafts are part of the circuit so a large insulated knob should be used. Notes:
- L1 is a 30uH roller inductor although any around that value should work.
- C1 is a 2000pF capacitor used to cancel the reactance of the inductor.
- C2 is a 250pF capacitor added across J15. This avoids soldering or making changes to the board.
Disable the built-in inductor L1 by placing a jumper at S2. You also need to put a jumper at S3 to provide a termination for the unconnected end of the inductor which is at an extremely high R.F. impedance. The series match should be connected between the transmitter output and a 50 Ohm resistor (for adjustment). Set the power control R1 for a drain voltage (T1 and T2) to 4.5 Volts then check the drain current (T2 and T3). Adjust the inductor to get a drain current of 22.2ma (0.222 Volts at T2 and T3). Add inductance to lower the current or use less inductance to raise the current.
You may need to keep adjusting R1 for 4.5 volts with each change in current but eventually you'll have the 4.5V @ 22.2mA combination you need. At that point, tweak the T-match tuning for a peak reading on the relative output meter.
|
