Adding ALC and overdrive protection to the MJF ALS-500M 500 watt amplifier

The MFJ ALS-500M is a (nominally) 500 watt amplifier that was produced by MFJ capable of covering from 160 meters through 15 meters - and 12/10 meters if so-equipped.

If you own an ALS-500M, you may have soon realized that it is a bit awkward to use:  If you are using it with a 100 watt radio, driving it at full power will not only cause it to be badly overdriven - causing terrible on-air distortion on an SSB signal - and (very) likely causing damage to the amplifier itself by throwing about twice as much power at it as it needs to work properly.

To understand the problem, an ALS-500M was powered from a variable-voltage supply with known-accurate wattmeters - one on the input to measure drive power and another to measure the output power, the amplifier itself terminated in a known-good 50 ohm load.  This same test set-up included a known-accurate DC ammeter as it was determined that the ALS-500M's own ammeter wasn't particularly accurate.

With this set-up, the characteristics of a friend's ALS-500M were measured on most amateur bands in terms of input and output power, at both 14.5 and 12.5 volts from the power supply.  (The voltage at the amplifier was lower than this due to resistance of the factory-supplied power cable.)

Freq (kHz)	PWR In	PWR Out	DC Voltage	DC Current
1825	5	65	14.5	18
	8	180	14.5	31
	22	300	14.5	43
	30	410	14.5	52
	45	525	14.5	60
	60	600	14.5	63.5
	6	82	12.5	21
	22	300	12.5	43
	30	450	12.5	56
	45	480	12.5	59
	65	500	12.5	63
1975	6	90	14.5	23
	20	350		45
	33	480		56
	45	560		63
	60	600		66.5
	33	425	12.5	53
3650	6	80	14.5	23
	20	290		47
	36	425		60
	50	500		66
	62	500		70
3850	6	78		22
	19	300		45
	33	410		57
	50	450		61
	63	490		67
	33	370	12.5	54
5371	6	95	14.5	27
	18	340		52
	34	480		64
	50	525		71
	34	380	12.5	60
7050	4	85	14.5	22
	18	350		47
	30	400		53
	52	460		59
	33	350	12.5	50
7250	5	87	14.5	23
	18	350		46
	33	410		53
	52	480		58
	33	350	12.5	49
14225	4	95	14.5	26
	18	310		47
	34	380		53
	52	400		56
	33	290	12.5	47.5
18100	5	75	14.5	22
	20	210		37
	35	260		41
	55	300		44
	35	200	12.5	36
21250	6	95	14.5	23
	24	210		35
	37	275		38
	56	280		39
	37	200	12.5	33
28345	5	68	14.5	25
	22	275		50
	34	325		55
	52	390		62
	34	300	12.5	53

From this data we can determine several things:

• There are no instances where more than 50 watts drive is useful.  If you were to graph the input versus output power in the above chart you would see that the curve "flattens" by the time you get to about 50 watts drive meaning that further increases of input power do not result in the same proportion of increase in output power.  It is at this point that the amplifier is becoming very non-linear and severe distortion of SSB and AM signals will result if one attempts to drive it to still-higher output power.
  
• While described as a "500 watt" amplifier, this is clearly optimistic. While barely capable of about 600 watts at 160 meters, the maximum usable "clean" (non-distorted) output drops to about 400 watts at the highest band, 10 meters.  This effect is due to physics:  The transistors in the amplifier are simply less capable at higher frequencies.
• The power output is lower with a 12.5 volt supply than a 14.5 volt supply.  This is also due to physics and clearly specified in the manual:  You'll get 25-100 watts less output at the lower voltage, depending on the frequency and drive power.

Too much power is NOT a good thing!

The ALS-500M manual clearly warns against driving with too much power for the reasons mentioned above, but in addition to producing a bad-sounding signal on the air, feeding too much power to the amplifier (more than about 60 watts) is significantly exceeding the specifications of the (expensive!) transistors and it will dramatically increase heating of the components:  On this test amplifier, even briefly driving it at 65 watts caused the input power resistors to overheat slightly, resulting in an obvious smell.

Why no ALC?

Since at least the 1960s both amateur transmitters and amplifiers have included an ALC (Automatic Level Control).  In the transmitter, this circuit monitors the transmitter's output power and if it exceeds the pre-set threshold (e.g. 100 watts for a radio rated at 100 watts) it will send a signal back to reduce the output power.  RF amplifiers have a similar circuit:  It detects the amount of RF being output and sends a negative voltage back to the transmitter driving it.  If the output of the amplifier gets too high, this voltage causes the transmitter to reduce its drive power.

In both cases this circuit does two important things:

1. Prevents excess drive to the amplifier(s), which prevents distortion of the transmitted signal.
2. Preventing damage.  All amplifiers have electrical and thermal limits above which they may be damaged and/or their operational lifetime may be dramatically shorted. 

Despite most commercially-produced amplifiers having an ALC circuit since the '60s, the ALS-500M does not - which is all the more confusing as this circuit is not complicated at all:  Having this circuit would help in the prevention of grossly overdriving the ALS-500M and having bad signals on-air and it may have saved many ALS-500M's from damage.

Adding an ALC circuit

As it made sense to do so, an ALC circuit was added to my friend's ALS-500M:

Figure 1:
This is the schematic of the ALC circuit.  It develops a negative voltage related to the RF output
power that is fed into the transmitter driving it.  When properly adjusted, this feedback loop will
limit the maximum drive to the amplifier, reducing the probability of distortion and damage.
Click on the image for a larger version.

 The circuit is quite simple - consisting of just TEN components including the output jack.  Here's how it works:

• J1 is the existing "RF Out" jack on the ALS-500M, an SO-239.
  
• Resistor R1, attached to the RF Out connector, samples the transmit power.
• Resistor R2 - with R1 - form a voltage divider:  At 500 watts into 50 ohms, there would be 447 peak-to-peak volts on the RF Output, but it is divided to 34 volts peak-to-peak at the junction.
• Capacitor C2 couples the RF to diodes D1 and D2, blocking DC.
• Diode D1 clips the positive-going voltage and together with D2, forms a voltage doubler circuit.
• Capacitor C3 filters the output of diode D2 - a negative voltage -  removing residual RF.
• Potentiometer R3 allows adjustment of the produced ALC voltage so that the proper threshold may be set for the transmitter being used to drive it.
  
• Capacitor C4 further filters any RF from the ALC line.
• J2 is a phono ("RCA") jack used to connect the ALC voltage to the driving transmitter.

Frequency compensating capacitor C1

Capacitor C1 requires more explanation.  Real-world components aren't like their "ideal" theoretical counterparts and resistor R1 is no exception:  Even though it is a "resistor", it has some capacitance - albeit small - plus there is some stray capacitive coupling between the center pin of the RF Out connector and the nearby components.  Because of this, at higher frequencies, some RF energy "leaks" around R1, causing more voltage to appear at the junction between it and R2:  This higher voltage would cause more AGC voltage for a given power level and in testing, while 500 watts produced about -37 volts at the top of R3 on 80 meters, it took only about 150 watts to produce that much voltage on 10 meters.

Figure 2:
The as-built ALC circuit built atop the RF OUT connector,
using it for component support.  To the right of the large
resistor is the ALC adjustment potentiomteter and jack below.
Click on the image for a larger version.

We actually want this to occur for the simple reason that the ALS-500M cannot output the same "maximum" power on each band - this level decreasing as frequency goes up - but as we can see from the table, whereas we could "safely" output about 400 watts at 80 meters at 12.5 volts, we'd probably want no more than 325 watts or so at 10 meters, so our ALC output voltage should be the same at those two power levels.

Capacitor C1 "compensates" for this:  Being a capacitor, it has lower impedance at increasing frequency and we can select its value to give us about the same ALC voltage at 400 watts on 80 meters as 325 watts would on 10 meters.  For the ALS-500M and our circuit, a value of 6.8pF turned out to be about right - but this would vary with components:  A variable capacitor (somewhere in the 2-15pF range) would allow easy adjustment of this compensation.

Figure 3:
The rear panel of the modified ALS-500M.  The ALC adjust
potentiometer is between the RF OUT and DC IN connectors
with the added "ALC OUT" jack below.
Click on the image for a larger version.

Construction

Figure 2 shows how the ALC circuit was constructed.  In the top-center, we see the "RF Out" connector and R1, the 12k resistor.  Clustered around R1 - and using the ground lug (plus an added lug) on the "RF In" connector we see the other components.  Just to the right of center - between the RF Out connector and the DC connector we see R3, a 10k potentiometer and below it - partly obscured by R3 - is the "ALC Out" jack.

Figure 3 shows the rear panel of the amplifier - the "ALC ADJ" potentiometer (R3) near the top and the Phono (RCA) plug below it - both labeled.  Looking at the label of the ALC ADJ control, you will notice that the label shows that rotating it counter-clockwise will result in "minimum" power - but this corresponds with maximum ALC voltage.  While this may seem counter-intuitive, remember that the the more negative the ALC voltage, the more it will try to reduce the output power of the transmitter - but if the potentiometer were turned fully clockwise (no ALC voltage at all) it would be the same as disabling the ALC altogether.

In testing with an Icom IC-7300, setting the ALC control to "Min" resulted in no more than about 80 watts out of the amplifier, no matter the "RF Output" setting on the radio and this indicated that the ALC was doing its job.  Setting the ALC control for about 425 watts at 80 meters resulted in about 325 watts on 10 meters, maximum - both within the "linear" and safe range of the amplifier.

ALC Overshoot and other anomalies

In many radios, ALC isn't perfect:  There will be a slight lag in many radios between the appearance of the ALC voltage and the radio's cutting back in transmit power - some of this being due to the radio itself having "ALC Overshoot" and some being due to the ALC voltage from the amplifier being a bit slow to respond.  What this means is that it is possible for the radio to briefly output WAY more power than expected for a brief instant before throttling back.

On the air, this can cause a burst of amplifier overdrive at the beginning of words/syllables - often showing up as a "popping" (or "clicking" on CW during key-down) and over time, this burst of high power could damage the transistors and other components in the amplifier.  What this means is that you SHOULD NOT rely entirely on the ALC to limit the output power - you should, at the very least, turn down your transmit power to about 50 watts or so even if you have the ALC.

Some radios have another problem:  They can do ALC overshoot even without an external amplifier - briefly driving their own amplifier  to much higher than expected power.  Some radios - even if you turn the power down - rely on feedback from their built-in wattmeter and will briefly output higher than the desired output power.  Both of these mean that you could still end up with a somewhat "dirty" signal on the air even if you believe you have taken steps to prevent it.

Overdrive protection:  Adding a 3dB pad.

While adding ALC to the ALS-500M is a "no-brainer", it would be easy to forget to connect the ALC - or your radio and amplifier combination could still cause the "popping" or "clicking" from brief overdrive conditions even if you turn down your power and/or connect the ALC.  To prevent this, it would be a very good idea to prevent too much power from ever reaching the amplifier circuitry itself.

As you can see from the chart above, there is never a frequency or band combination where more than 50 watts drive would yield clean output power.  What this means is that we could lose half of the drive power of a 100 watt radio and still push the amplifier to its useful limit - and protect its expensive transistors and other circuitry against an accidental "oops" should we accidentally overdrive it.

Figure 4: 
The 3dB "Pi" resistive attenuator.
Click on the image for a larger version.

The addition of a 3dB attenuator would accomplish this, soaking up half the transmit power before it gets to the amplifier allowing 100 watts of output..  The easiest place to install this attenuator would be on the input of the amplifier - but this would also affect the receive signal by about 1/2 "S" Unit:  If your S-meter reads well above S1 on even the quietest band, you won't "miss" any signals by doing so:  A 100 watt 3dB "pad" can be found commercially and on the surplus market if you look carefully.  The other down-side of having a 3dB pad inline would be that if your turn the amplifier off, you are still losing half of the transmit power.

A technically "better" solution would be to place the 3dB attenuator right on the input of the RF power amplifier circuit, inside the amplifier.  Doing so avoids placing this loss in the receive path and it will also not affect the transmit signal when the amplifier is turned off.  Figure 3 shows this attenuator schematically.

These resistors must, collectively, be able to dissipate 50 watts of power and rather than trying to assemble a large mass of lower-wattage resistors, we can use thin or thick film power resistors in transistor-like package which may be bolted to a heat sink. For the ALS-500M, there is a flat area on the rear panel that is next to amplifier deck and large enough to accommodate these resistors and dissipate the power dropped.  Examples of suitable resistors include:

• 18 Ohms, 100 watts:  Bourns (Riedon) PF2472-18RF1  (DigiKey P/N:  696-PF2472-18RF1-ND - link)
• 300 ohm, 100 watts:  Bourns (Riedon) PF2472-300RF1  (DigiKey P/N:  696-PF2472-300RF1-ND - link)

Figure 5:
The three resistors comprising the attenuator, mounted on
rear panel of the amplifier for heat sinking.  The white
RG-316 coax from the T/R switch comes in from the left
while that going to the amplifier input goes to the right.
Thermal paste is used under the resistors' tables to enhance
thermal conductivity to the case.
Click on the image for a larger version.

These particular resistors are "Thin Film" and their construction is such that while not intended specifically for RF applications, they work perfectly well at HF and into VHF for a non-critical application like this - plus they are relatively cheap!  These resistors have metal heat sinks, but these are isolated from the resistors within, having only a few 10s of pF of capacitance coupling between the internals of the resistor and the ground when bolted to the case.

The coaxial cable from the T/R switch to the input of the amplifier will need to be extended (carefully splicing the two together, minimizing the length of the ground/shield connections) to reach the resistors when mounted on the rear panel:  RG-316 PTFE coaxial cable was used for this and the short jumper that connected from the output of the attenuator, back to the input of the amplifier.

Figure 5 shows the attenuator, mounted to the back panel of the radio with a small amount of thermal compound:  The RF power from the T/R switch enters from the left and one leg of 300 resistor R1 is connected directly to the shield of that piece of coaxial cable.  The center conductor then connects to the junction of it and R2.  On the other side of R3, the process is repeated, the shield of the coax tied to the shield as well:  This cable then connects to the input of the amplifier module.  Between R1 (on the left) and R3 (on the right) is a piece of 12 AWG (2mm) wire that connects between the shields of the "in" and "out" coaxial cable.

Final results:

While it might seem wasteful to throw away half of the drive power, doing so protects the power amplifier from being damaged by overdriving when one inevitably forgets to reduce the output from the transmitter.  It also protects those that might be listening on the air to a badly distorted signal:  Adding the ALC circuit is, I believe, a necessary addition as this helps prevent even mild overdriving of the amplifier that is still possible under some conditions - even with the added attenuation.

* * * * *

This page stolen from ka7oei.blogspot.com

[END]