Amplifier ALC Theory questions and help

Discussion in 'Amateur Radio Amplifiers' started by N2RDQ, Jun 29, 2018.

ad: L-HROutlet
ad: l-rl
ad: L-MFJ
ad: Left-3
ad: Left-2
ad: MessiPaoloni-1
ad: Subscribe
  1. N2RDQ

    N2RDQ Ham Member QRZ Page

    I'm currently working on a dentron MLA-2500, one of those weekend lazy kind of projects.

    The dentron has an alc circuit but the voltages are way to high for my icom rig, so, I want to replace the ALC circuit with something that would be more compatible. even if I voltage divided it, or used a zener to limit the voltage, the idea of connecting my transceiver up to something capable of -100 volts or more is not a risk I care to take. Id rather make something appropriate.

    with that in mind, i am trying to find good writeups on how an alc in a tube amplifier should behave and the proper interactions between the radio and the amp.

    unfortunately, I am finding some conflicting writeups and various alc designs which don't mesh with each other on how they would behave.

    also, I have yet to find a curve of the alc output versus drive examples.
    from what I gather though, as a rough theory, it should have a fairly fast attack and a slow decay. but beyond that, its not clear at what point getting close to an amplifiers peak desired output should the alc voltage start to increase. is it also a time domain driven though, the longer the drive is too high, the voltage keeps increasing until it reaches max or the level drops to an acceptable leveel?? not sure...

    id love to learn more about this topic, with some accurate information. the goal at the end of the day would be to design my own replacement ALC circuit.
     
  2. G3YRO

    G3YRO Ham Member QRZ Page

    A.L.C. was developed as a system to prevent overdriving of the Valve P.A.s in SSB Transmitters . . . the whole idea was to limit the peak drive power so that the Valve(s) stayed within the linear part of their curve . . . this is why it was called Automatic Linearity Control.

    The best circuits detected the Grid Current drawn by the Valves, and when this reached a certain limit that would cause distortion, a DC voltage was detected and fed back to an earlier stage (usually in the I.F. amp) to reduce the drive.

    Confusingly, when typing the ALC acronym in full, a lot of rig handbooks have made the mistake of thinking it stands for the more widespread use of the term ALC as used in millions of tape recorders etc - "Automatic Level Control". But that is not really what ALC is - it's not there to control your Level (that's what a Compressor or Processor is for) - it's there to maintain Linearity of your PA.

    Obviously, the same concept can be continued to an external high power Linear Amplifier, and the detected control voltage fed back to your transmitter's existing ALC control circuit.

    Why do you dismiss the idea of a simple resistor divider to reduce the voltage from your Amp? How is it unsafe? If you, for example, had a 10k resistor feeding a 1k resistor to ground, that 100V would only ever reach 10V. That is the only way to do it properly - you want the DC voltage fed back to your rig to change proportionally, so just reduce it that way.

    And the Time Constant circuit will already be in your Transmitter - you don't want to add any extra delay. (if anything, rigs with Transistor PAs already have a much too slow trailing edge on the time constant - that's because they use it to adjust the actual power output of the rig, which isn't what proper ALC was designed to do)

    If you want to improve the ALC circuit in your Amplifier, then I would suggest designing a circuit that actually detects excessive Grid Current. Very few Amps seem to do this, even though it's the best way to do it, as it doesn't just look at the drive level, it takes into account whether the Amp is properly loaded to maintain linearity too.

    I might have done this to my own Amps . . . except I never use ALC from the Amp! I always use a Compressor on SSB, so make sure I set the drive power so I'm not overdriving. On CW I just set the carrier level from my rig for the power I want.

    Roger G3YRO
     
    Last edited: Jun 29, 2018
    WD4IGX and K9AXN like this.
  3. KA9JLM

    KA9JLM Ham Member QRZ Page

    The amps ALC voltage is normally proportional to the amps RF output power.

    If the radio has a way to Limit / Set power out level, I would not even bother using the amp alc output.

    Just watch the drive level, And never exceed the amps input rating or max grid current.

    Not connecting the extra wire also gives one less path for amp RF to get into. :rolleyes:

    What model Icom are you using to drive your MLA-2500 ?

    Have Fun.
     
    Last edited: Jun 29, 2018
    AI3V and AC0OB like this.
  4. K6BSU

    K6BSU Ham Member QRZ Page

    The amp circuits I have studied use a simple diode peak detector at the amp input circuit. This provides the same ALC signal regardless of the power output or tuning of the amp. For that reason, I do not use ALC. That only requires careful attention to not over driving the amp, which is easy enough when the driving tx has an "RF Output" level control.
     
  5. WB2WIK

    WB2WIK Platinum Subscriber Platinum Subscriber QRZ Page

    That used to be true decades ago, but modern amplifiers use much more sophisticated ALC circuitry than that. Even the inexpensive tabletop AL-80B does this (see below) and I've found it not only very effective and adding zero distortion products, but also compatible with every reasonably modern (past 30 years) HF transceiver sold that has an ALC input port:
    ALC1.PNG
    ALC2.PNG
     
    WB1E and AG5DB like this.
  6. WB2WIK

    WB2WIK Platinum Subscriber Platinum Subscriber QRZ Page

    The ALC in the MLA-2500 is rudimentary and a bit basic, but its output level is adjustable all the way down to "0 V" if you want to simply turn it down; it will still work to some degree.
    MLA-2500 ALC.PNG

    See R10? One end of it is at ground potential, so turning R10 all the way down will produce 0V no matter how much drive you have.:)

    Updating the circuit with something more modern and effective is a noble idea and I might try it; although this amplifier really only needs ~65W PEP max drive for full output power and I might consider changing R1 from its original value (100 Ohms) to 50 Ohms, 50W non-inductive to further limit drive. Even with lighter drive, you could exceed the Ig1 dissipation rating of the 8875s if you don't fully LOAD the amplifier. UnderLOADing is really bad for tubes having limited grid dissipation. I'd be watching output power and GRID current while tuning, and not care much about PLATE current.
     
  7. N2RDQ

    N2RDQ Ham Member QRZ Page

    ic-7300 alc voltage 0 to -4v
     
  8. KA9JLM

    KA9JLM Ham Member QRZ Page

    For that radio, I would just set the radio power out level, Then no ALC voltage from the amp would be needed to control the amp from being over driven.

    If you do use ALC for that radio and amp, The control will be very touchy, Maybe useless.

    Have Fun.
     
    Last edited: Jun 29, 2018
  9. N2RDQ

    N2RDQ Ham Member QRZ Page

    great info all around...it helped a lot, but still a bit unclear on how the alc is supposed to react.

    so.... the attack and delay should be in the transceiver... ok that's something didn't understand before, and that makes sense.

    so if i understand the al-80 write-up, its not a linear increase on output as the drive increases / grid current increases, but, there is a threshold, and at that point, the output goes negative. still, I am not understanding if then its a linear curve starting at a threshold grid current level, or, if its more switch-like, where you reach a certain current level and you are going from 0v to -4v immediately??

    the exsisting dentron circut looks like its astright protression that mirrors the inut rf.

    why not a voltage divider.. that's a legit question, and it comes down to paranoia .... the solid-state circuitry in the transceiver is not very forgiving, a little glitch in the pot and bang, you have hv into the alc input. probably less of a possibility of that with a voltage divider, but, the idea of isolation from the HV appeals to me. over paranoid perhaps??

    maybe a more simplistic solution for just isolation would be an optoisolator circuit and leave the original design alone.

    so, is the proper way to drive an alc circuit on a tube amplifier to detect the input power, or, to be monitoring the grid current. judging from the al-80 write-up, id think grid current is the right way to go if i am going to bother building my own circuit anyway.

    going with the grid current concept for a moment...Since i am modifying the grid current metering circuit to increase the metering resolution (its currently horrid, a 1A scale, its basically unusable, making it a 100ma scale), I can pretty easily then also monitor the grid current with a circuit and drive the alc output, and at the same time, make it compatible with various voltage schemes used by modern radios very safely. basically, let the grid current meter voltage drive an opamp similar to what I see in the al-80 schematic . .. if something goes terribly wrong, I'll blow the alc circuit in the amp and not the radio .


    also, i ran across an interesting write-up by WA8AJN http://www.robkalmeijer.nl/techniek/electronica/radiotechniek/hambladen/hr/1984/08/page40/index.html
    where he did a modification on the alc circuit to make it driven by grid current... however, he also modified the grid circuit to self-bias as the grid current increases so it would self-limit to around 50-60ma total. ... to me, that particular idea might make it non-linear, but i think his idea was to basically reduce the need or eliminate the need to have alc or basically have it internally limiting... which to me would just add compression.

    not to get off topic, but since changing the input impedance was mentioned...he did further mods for the input impedance, but, I am likewise going to experiment in that area. (I have dentrons tuned input addon, and am aware of the loading issues dentron has)
     
  10. N2RDQ

    N2RDQ Ham Member QRZ Page

    oh i maybe misread the ciruit....how it looks to me is the alc adjust sets the level at which alc - voltage will start to apper on the alc output.
    the amount of alc voltage would then be controlled by the rectified rf voltage above the potential of the alc adjust voltage and the rectified. the alc output voltage would be whatever the rectified output voltage above the alc adjust voltage increasing in proportion to the rf input increase.

    did I understand the circuit wrong?
     

Share This Page