5E3 1/4 power switch?

Discussion in 'Shock Brother's DIY Amps' started by Uncle Daddy, Jun 13, 2018.

  1. timewaster1700

    timewaster1700 Tele-Meister

    Posts:
    317
    Joined:
    Feb 17, 2016
    Location:
    Utah
    Ok I did the math for anyone that's interested. My guess is I made it too complicated but here we go:

    Here's my makeshift circuit diagram without the switch for simplicity:

    >------^^R2=16ohm^^----^^R3=8ohm spkr^^------>GND
    | |
    |_________^^R1=12 ohm ^^________________|

    Here's my math:
    We want an equation such that the speaker power is some fraction of the total power: P3 = (1/n)*Ptotal
    where we can define n to be 4 for quarter power, 9 for 9th power, etc.

    P3 = (1/n)*(P1+P2+P3)

    We can rearrange this to be: 1/n = P3/(P1+P2+P3) ;

    P1 = V*I1 = V^2/R1

    P2 = V2*I2 = V2^2/R2

    V2 = (R2*V)/(R2 + 8)
    therefore,

    P2 = (R2*V^2)/(R2+8)^2 ;

    P3 = V3^2/8

    V3 = (8*V)/(R2+8)^2
    therefore,

    P3 = (8*V^2)/(R2+8)^2;

    Now if we plug P1, P2, and P3 into our original equation we get:

    1/n = [(8*V^2)/(R2+8)^2] / [(V^2/R1) + (R2*V^2)/(R2+8)^2 + (8*V^2)/(R2+8)^2]

    The voltage terms cancel out and we can multiple the numerator and denominator by (R2+8)^2 to get:

    1/n = 8/[(R2+8)^2/R1 + R2 +8]

    If we take the reciprocal of both sides we get:

    n = (R2+8)^2/(8*R1) + R2/8 + 1

    Now lets solve for R1 in terms of R2 and n:

    (n - R2/8 - 1) = (R2+8)^2/(8*R1)
    then
    8*R1 = (R2+8)^2/(n-R2/8-1)

    which simplifies to our final equation:

    ||| R1 = (R2+8)^2/(8*n-R2-8) |||


    If we want 9th power (n=9) and we make R2 = 16 as @FenderLover suggested we can enter those numbers into this equation to solve for R1 and we get:

    R1 = (16+8)^2/(8*9-16-8) = 12



    One last thing though you also have to keep the total resistive load on the output transformer 8 ohms so you have to also fulfill a second equation at the same time:



    8 ohms = 1/[1/R1 = 1/(R2+8)]
    which is easier to think about like this:
    1/8 = 1/R1 + 1/(R2+8)


    Maybe the best procedure would be find a bunch of R1 and R2 values that work for the
    above equation then plug them into this equation to get your fractional power value:


    1/n = 8/[(R2+8)^2/R1 + R2 +8]

     
    Last edited: May 3, 2019
  2. timewaster1700

    timewaster1700 Tele-Meister

    Posts:
    317
    Joined:
    Feb 17, 2016
    Location:
    Utah
    Ok I wrote a little LabVIEW program with the above two equations and got the following resistor options:
    upload_2019-5-3_0-40-17.png

    You can see the standard 16 ohm and 8 ohm for 1/4 power and then @FenderLover 's 1/9th power suggestion of 12 ohm and 16 ohm.
    Looks like 1/25th power can be had with a 10 ohm and 32 ohm resistor.

    I tried up to 1000 different combinations of R1 mixed with 1000 combinations of R2 and these are the only ones that nail the 8 ohms exactly.

    If I loosten the tolerance for the load to be between 7.968 ohms and 8.032 ohms it opens up the possibilities a little more:

    upload_2019-5-3_0-42-5.png

    I can expand the tolerances if someone is interested. I have no idea how close the impedance of the load needs to be to the ideal 8 ohms. I'm guessing I could open up the tolerances and be fine.
     

    Attached Files:

    Last edited: May 3, 2019
    BigDaddy23 and mcentee2 like this.
  3. timewaster1700

    timewaster1700 Tele-Meister

    Posts:
    317
    Joined:
    Feb 17, 2016
    Location:
    Utah
    Here's another run with the tolerances opened slightly and a 4th column that shows the total impediance the output transformer will connect to using the resistor combination and an 8 ohm speaker assuming everything was ideal and perfect values (which of course is not true).

    upload_2019-5-3_0-43-26.png
     
    Last edited: May 3, 2019
    BigDaddy23 and mcentee2 like this.
  4. mcentee2

    mcentee2 Tele-Meister

    Posts:
    387
    Joined:
    Feb 18, 2011
    Location:
    York
    From what I have read speaker rating is "nominal" anyway given its variance with an AC signal/frequency.

    I don't know how tolerant it is though, but I suspect an "ohm" either way won't break anything.

    Maybe pick out the pairs with common values, easily purchased?
     
  5. timewaster1700

    timewaster1700 Tele-Meister

    Posts:
    317
    Joined:
    Feb 17, 2016
    Location:
    Utah
    I opened the tolerances a bit more
    upload_2019-5-11_16-25-0.png
    Looks like 10 and 33 (both available from mouser) would give 1/25th power. I'm tempted to try this and then the 1/9th power and compare.
     
  6. timewaster1700

    timewaster1700 Tele-Meister

    Posts:
    317
    Joined:
    Feb 17, 2016
    Location:
    Utah
    Here's my hand sketch for an attenuator wiring diagram that uses an ON/OFF (SPDT) switch and a LOW/HIGH ATTENUATION (DPDT) switch. I think I'll give this a shot.
    Attenuator wire diagram.jpeg
     
    robrob likes this.
  7. Uncle Daddy

    Uncle Daddy Tele-Holic

    Posts:
    781
    Joined:
    Sep 26, 2015
    Location:
    Maldon, England
    Good luck fitting that lot into a 5e3 chassis
     
  8. timewaster1700

    timewaster1700 Tele-Meister

    Posts:
    317
    Joined:
    Feb 17, 2016
    Location:
    Utah
    I should have mentioned the plan was to put it in a pedal box. Either the small A size Hammond or the larger B whichever works.
     
    robrob and Uncle Daddy like this.
IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.


  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.