Aiken's Reactive Dummy Load.

Discussion in 'Amps/Cabs Tech Corner: Amplifier, Cab & Speakers' started by James Freeman, Apr 27, 2012.

  1. always-ultra

    always-ultra Member

    Messages:
    4
    Joined:
    May 30, 2011
    Thanks so much for your work!
    Just one question: did you use a non-inductive 8Ohm power resistor? If not, do you know if its inductance matters somehow in the frequency range of a guitar amp?
     
  2. James Freeman

    James Freeman Member

    Messages:
    746
    Joined:
    Sep 3, 2010
    Its a wire-wound power resistor so it will have some inductance, but its so small its insignificant.


    I will have to resurrect the "Power Tube distortion Myth or Reality" thread from the dead.
    Why? you ask.
    I've ordered some 12at7 12ay7 12au7 to play with the Phase Inverter and Power Tube distortion.
    I think I may have to crush some people dreams about "power tube distortion sounds better than preamp".

    My Phase Inverter drives the power tubes into blocking distortion and crossover way before the grid of the PI clips.
    It sounds... well.... bad. Like breaking glass.
    This is because at the PI input the signal is less the 1v and at the output of the pi the signal is more than the bias voltage (-50v in my case)
    of the power tubes hence clipping the Power tubes grids, so I can clearly tell you its power tube distortion.

    I will record for you some actual music from my MP3 -> FX loop return -> PI -> Power tubes -> Dummy -> PC.
    So you can actually hear how power tubes distortion sound like without distorting the PI.

    Plus I will post some Scope pictures of pure 1k sine wave going into the pi and how it will look at the dummy exit.

    stay tuned...
     
    Last edited: May 29, 2012
  3. eddy999

    eddy999 Member

    Messages:
    220
    Joined:
    Apr 1, 2009
    Location:
    Surrey, England
    Sounds great to me, very similar to the mic'd amp alone. There's a big difference between the purely resistive and reactive loads - great job.
     
  4. teemuk

    teemuk Member

    Messages:
    2,942
    Joined:
    Apr 21, 2009
    The whole statement is so broad that it's easy to either crush it or to agree with it, depending on several factors like the amp you're evaluating, individual tastes, etc.

    Yep. Pretty typical for most of the generic tube power amp designs: power tubes are overdriven into clipping and grid conduction waaaayyy before PI. Even swapping the phase inverter tubes usually does not make a super distinctive difference, except for maybe introducing secondary effects like drastically killing open loop gain, which will then alter the damping characteristics if the amp happens to utilise feedback.

    That it does, especially if you plug straight into the power amp. Then again, also a plain preamp stage alone would likewise sound damn horrible, fizzy, buzzy mess and like breaking glass. An overdriven PI would also sound like breaking glass ...and oh, pretty much all clipping distortion sounds like breaking glass if it's not filtered in some way.

    Yes, tube power amps introduce crossover distortion, hard clipping, odd order harmonics, high order harmonics, and in some cases several other distortions that - according to all parroted myths - should sound horrible. Then again, pick any classic puh-pull tube amp and it's main components generating tone will likely be all those "atrocious" things.

    Tube amps are often hyped with odes of "musical" and non-discordant low order even order harmonic distortion, coupled to nice DRAWN pictures portraying sine waves that distort smoothly at both half waves. Ironically, the thing those drawings usually portray inherently means odd order distortion, and about the only things I've seen generating such smooth clipping in real life are the most generic and buzzy sounding plain solid-state diode-based fuzzboxes. Real life circuits that sound nice when they distort usually introduce sharper edges, kinks at least on the rising edge of the waveform, sometimes at crossover regions and maybe at descending edges too, possibly asymmetry and all other stuff that doesn't LOOK aesthetically pleasing. Then again, hype is more about looks so...

    Anyway, add a carefully designed preamp that mangles the signal in several different ways, couple it to power amp than mangles it furthermore, and finally output the whole mess with a speaker that mangles it yet again. Do all this in a proper manner and it will likely sound pretty good. Isolate those components from one another and they likely sound less impressive, if not downright horrible.

    Just the straight power amp out vs. power amp out + cabsim comparator is a pretty good case in point example of this. How many actually employ nothing but a broad bandwidth power amp instead of a heavily-filtering, very narrow bandwidth preamp to a (broad bandwidth) power amp to a (narrow bandwidth) speaker; both preamp and the speaker usually having a totally unlinear response instead of a perfect flat line? Good signal filtering (usually introduced by the preamp and by the speaker) can be THE difference between good and bad "power amp tone".
     
    Last edited: May 29, 2012
  5. James Freeman

    James Freeman Member

    Messages:
    746
    Joined:
    Sep 3, 2010
    teemuk agreed to your whole post.

    My goal now is to drive the PI and not drive the grids of the power tubes.
    Sort of like using PPIMV but without it effecting the negative feedback.
    I will used a lower gain tube (12ay7 or 12au7).

    The output transformer saturation and power tubes distortion sound really bad actually.
    Sort of like a bass heavy messy clipping/buzz the whole time.
    Not very musical.

    I know for sure that I like my Preamp gain tones.
    Now just to make the whole volume range of the amp usable without it turning into a big mess when Master is on 3.

    Who knows, maybe a lower gain PI tube like 12AY7or 12AU7 will suit the high gain beasts like Soldano or Boogie much better.
    One thing for sure, they will have alot more controllable volume and gain (without squealing) range.

    Now for a low gain vintage amp like Marshall 1959 or any fender, if you want
    a gainy sound you will have to use a higher gain PI (12ax7) to drive the power tubes grid into oblivion.

    So the whole quest is finding a good balance between Volume Range + PI clipping AND Power Tube distortion (transformer, crossover, bias shift).
    Which will be between 12ax7 and 12au7.


    A little guidance:

    If you are a High Gain (Preamp) Metal player (OR a super clean Jazz player) and you mike your amp on stage.
    Use a lower gain Phase Inverter tube, you will have more control over the amp.

    If you are a more classic "Crunch" player.
    You want every stage of the amp to crunch.
    A 12ax7 in the PI is a good choice.
     
  6. catalin gramada

    catalin gramada Member

    Messages:
    28
    Joined:
    Mar 23, 2012
    Location:
    Europe
    Hi ,
    started my Aiken's load project / slightly different values.

    [​IMG] catalin
     
  7. James Freeman

    James Freeman Member

    Messages:
    746
    Joined:
    Sep 3, 2010
    Looks good.

    What values do you use and what caps are those?
    Did you install a line out?

    If you are going to install a line-out make sure that you use a cliff jack and Ground-Lift switch.
    A cliff jack is made of plastic and its sleeve will not make contact with the case.
    A ground-Lift is also very important for recording and reamping situations.
    For example: when reamping I leave the ground grounded (Amp and SS poweramp share the same ground).
    But when I record into a PC I have to lift the groud because of alot of noise and hum.

    For me, without the line-out its just an expensive and heavy door stopper.
     
  8. catalin gramada

    catalin gramada Member

    Messages:
    28
    Joined:
    Mar 23, 2012
    Location:
    Europe
    Hello ,

    I used capacitors for engine (Ducati 50uF/400V - Romania made), they are mkp type and works fair well for low freq. aplication.(also resistant and stable in high temperature conditions).(btw.they are industrial caps but are better rated for audio aplication rather than Erse pulse-x audio grade !). I choose 300 uF with Hammond 159zj -10mH/5A (effective 8,35 mH) and 47 ohm resistor - for ca. 100 Hz resonant point. Load resistors are 12 ohm/150 w-a basic rule said to use it for max 1/3 rated power.Top plate is decoupled from the box with 10 mm spacers to have a good flux ventilation - but is still hot like hell. Total DC resistance 6,8 ohm. As for line-out I prefer to use a line transformer, instead a voltage divider- not included in, cause allready have a separate good device for that, so have no care for loops or any grounding problems.
    I spent less than 100 e to build it, box included, and I consider it is not expensive at all. It worth every cent- because it works !
    I will try to put a plot graph reponse and recording soon.
    Thanks Mr. Aiken and all comunity for support.

    Catalin
     
    Last edited: May 31, 2012
  9. Amptone

    Amptone Member

    Messages:
    43
    Joined:
    Jul 30, 2006
    Harry Kolbe, Soundsmith designed the Silent Speaker reactive dummyload product in the mid-80s and I bought it in 1989. I've had info about it for years at: http://amptone.com/kolbesilentspeaker.htm

    The HotPlate shouldn't have been a successful product. Kolbe's The Attenuator and Silent Speaker deserve the credit, in the mid-80s, instead of Alex/Aracom/Faustine/RockCrusher/Ironman and the other "new"-generation reactive loads, which follow in Kolbe's footsteps some 25 years after him.

    My Silent Speaker page has linked to this excellent Kolbe page for years:
    http://www.soundsmith.com/fake.htm
    It shows and clearly explains the goal correctly, and shows the reactance curve which all guitar power attenuator products and dummy loads must show. He shows the curve for:
    o Marshall 4x12 cab
    o Marshall Power Brake
    o THD Hot Plate
    o PowerSoak
    o Resistor

    Please provide the absolute minimum parts list eg no case, no resistive load switch -- only the *absolutely* essential parts. What is the least possible money one needs to spend to buy the heart of the circuit? For example, can the same result be achieved by buying a single resistor and coil? I don't even want to include the price of the jacks; only the heart of the raw innermost circuit. If a resistor is $2, what is the absolute cheapest to build a reactive load -- $5?

    -- Michael Hoffman, Amptone.com: Getting the best guitar tones quietly
    http://amptone.com
     
  10. teemuk

    teemuk Member

    Messages:
    2,942
    Joined:
    Apr 21, 2009
    Could be too late now but it's definitely beneficial to use a real, properly finned heatsink instead of relying on the extreme poor convenction cooling of a mere plate.
     
  11. torquil

    torquil Member

    Messages:
    1,636
    Joined:
    Aug 19, 2009
    Is it possible to design a passive mixing circuitry that allows to continuously interpolate between the reactive load box and the real speaker, thereby creating an adjustable attenuator, while preserving the speaker-like impedance load for the amp also at all intermediate settings?
     
  12. James Freeman

    James Freeman Member

    Messages:
    746
    Joined:
    Sep 3, 2010
    Amptone (Michael).

    First,
    I love your page. I have learned alot in your website through the years.

    To build a reactive dummy load you basically have to have all the parts mentioned in the first page.
    It WILL cost.

    Big Power Resistor 20$
    Small Power Resistors 10$
    Small Inductor 15$
    Big Inductor 40$
    Big Capacitor 50$
    Alum Box 40$
    Jacks & Switches, Pot, Feet, 25$
    Wiring 10$

    Total: +- 200$


    Torquil:

    This is exactly what the Aracom & Weber are doing.

    The Weber does only the high frequency slope with a simple air-inductor
    shaped like a speaker for the marketing gimmick.
    The Aracom does the complete reactive circuitry like I've build.
     
    Last edited: Jun 5, 2012
  13. James Freeman

    James Freeman Member

    Messages:
    746
    Joined:
    Sep 3, 2010
    Amptone:

    Yes.
    Essentially Aiken's Dummy Load IS the "Silent Speaker".

    After carefully reading the article you've posted:
    http://www.soundsmith.com/fake.htm

    I got a very important thing I want to mention here:
    Using the high frequencies inductor like Weber or UA without a damping (or bypass)
    resistor (33 ohm on Aikens build) will "effectively leave the amp unloaded" in those frequencies.
    "Cranked unloaded output stage can create very high voltage inductive peaks (several thousand volts)
    sufficient to cause catastrophic arcing in the tubes and in some cases,
    punch through the output transformer winding insulation ... end of transformer."

    I think its called Flyback voltage.

    Same case as when you match your tube amp with higher speaker load,
    for example: 4ohm Head -> 16ohm Cab = BOOM !!!
    or no load at all, Head -> Infinity = MEGA BOOM!!

    When my new resistors will arrive I'll will match my Cab impedance closer.


    Also, my 12AX7, 12AT7, 12AY7 & 12AU7 have arrived and I did some very thorough demo of each of these tubes in the PI slot.
    I have recorded the clean & Dist channel on all notches from 1 to 10 playing a riff for clean
    and different riff for Dist, on every level for all the different PI tubes.
    Soon to be posted on the "Power Tube Distortion Myth/Reality" thread.

    Great Info guys.

    A punch line: :boxer
    The smarter you are, the less money you spend... or not. :huh
     
  14. catalin gramada

    catalin gramada Member

    Messages:
    28
    Joined:
    Mar 23, 2012
    Location:
    Europe
    Hi
    You're right, the power of resistors is direct proportional with thermic dissipation. For instance my 150w are not rated more than 50 w without a proper sink, and that in 25*C ambient conditions (ca.77*F) and not in hot air closed box- more important as we run "almost DC" on output. The second consideration is relative to its nominal value in overheating condition.
    Catalin
     
    Last edited: Jun 6, 2012
  15. catalin gramada

    catalin gramada Member

    Messages:
    28
    Joined:
    Mar 23, 2012
    Location:
    Europe
    Hello
    maybe I'm wrong but have a Marshall Power Brake schematic with a resonance LC circuit centered at 113 Hz and not in around 350 hz as it was presented in a plot of http://www.soundsmith.com/fake.htm. Also confused to compare diferent impedances loads on the same plot graph. Light me please. Thanks.
     
  16. Amptone

    Amptone Member

    Messages:
    43
    Joined:
    Jul 30, 2006
    Yes.
    The Attenuator - Harry Kolbe, Soundsmith
    http://amptone.com/kolbeattenuator.htm

    -- Michael Hoffman, Amptone.com
     
  17. catalin gramada

    catalin gramada Member

    Messages:
    28
    Joined:
    Mar 23, 2012
    Location:
    Europe
    Hello,
    useful link : http://www.reocities.com/sunsetstrip/studio/2987/speaker.html
    some building considerations: keep an eye on RLC meter when you mount inductors in a case. The inductance will vary with position in metal box and also is relative with distance from another electromagnetic components. sometimes helps just to reverse position up side down. also avoid to mount inductors with a screw through the coil. better to measure the value directly in box, out of circuit of course.
     
    Last edited: Jun 16, 2012
  18. James Freeman

    James Freeman Member

    Messages:
    746
    Joined:
    Sep 3, 2010
    Useful and good info:
    http://www.soundsmith.com/fake2B.html

    The typical 4x12 Resonant frequency is around 110Hz, 2x12 around 120Hz (both closed back).

    I have few question untested that I will have to test for the greater good.
    * Whats happening to the RF frequency as the volume goes up?
    * If speakers are wired in parallel how does it change the inductance?

    If anyone can please plot their 4x12 using REW (Room EQ Wiazrd) or any impedance plotting device/software this would help alot.

    Also if anyone can plot an expensive Reactive Attenuator (Aracom,UA,etc) this wound be a huge step forward.
    If an attenuator company stand behind their product they would share their impedance curve anyway.
    Or is there something to hide? ;)
     
    Last edited: Aug 15, 2012
  19. torquil

    torquil Member

    Messages:
    1,636
    Joined:
    Aug 19, 2009
    That's a very good page. And it contains unusual impedance plots for series/parallel configurations of two unequal speakers in Chart 8. Nice!

    I saved it in case it happens to disappear from the web.
     

Share This Page

  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice