What's the Real Deal on speaker impedance mismatch?

Discussion in 'Amp Tech Center' started by SSL9000J, Apr 30, 2018.

  1. SSL9000J

    SSL9000J Tele-Meister

    Age:
    48
    Posts:
    232
    Joined:
    Dec 18, 2016
    Location:
    Atlanta, GA
    Found this Princeton mod on Fenderguru.com

    Output transformer with multiple impedances (built-in attenuator)


    This mod requires the knowledge of how to replace a transformer in your amp. This mod will give you different speaker outputs impedances converting the the external speaker jack to an independent output.

    The original OT is sized for an 8 Ohm speaker impedance. It will make sure the power tubes “see” the correct impedance and can operate at optimal conditions in terms of clean headroom potential and frequency response. If you had an OT designed for a 2 ohm speaker load the tubes would operate at their best with a 2 ohm speaker load. If you use a 8 ohm speaker with a 2 ohm OT, the tubes are pushed harder since they see a bigger load than expected. Both volume and clean headroom is reduced significantly, and the tone is not significantly changed. This effect is similar to an attenuator, which is very nice when you want to achieve the amp’s sweet spot at lower volumes.

    The article goes on to describe how to install an OP transformer with multiple secondary windings for different levels of attenuation. I've read about the proverbial Rule of Thumb regarding impedance mismatching; that it's relatively safe if one doesn't overdo it. But what are the dangers of running a pair of 6V6s into a 6.6 kohm primary output transformer and driving an 8 ohm speaker with the 2 ohm secondary tap? It seems way outside the safe operating parameters. (This particular example is a mod for silverface Princeton Reverb. The intention being more breakup/distortion with less volume. IE sounding dimed at bedroom levels.) What sort of wear & tear would this place on the output tubes & transformer?

    Thanks!
     
  2. Wyatt

    Wyatt Tele-Afflicted

    Posts:
    1,129
    Joined:
    Nov 3, 2004
    Fenderguru.com has always been a prime example of "a little learning is a dangerous thing." Woven within various technical details is a LOT of assumption, armchair electronics, and internet myth on that site. The whole site should be avoided or taken with a grain of salt.

    What they overlook is the flyback voltage that results from a higher than expected load. This voltage works its way backward though the OT and gets amplified to several hundreds or even thousands of volts, which can result in tube failure or more serious issues. The flyback voltage is a factor of the overall AC output voltage and the level of impedance mismatch.

    On solid-state amps, which behave linearly to the load, it is perfectly acceptable to increase speaker impedance to decrease output. On tube amps, just buy an attenuator.
     
    Last edited: Apr 30, 2018
  3. Wally

    Wally Telefied Ad Free Member

    Posts:
    35,313
    Joined:
    Mar 17, 2003
    Location:
    Lubbock, TX
    Wyatt, is it not so that a mismatch to the high side...say an 8 ohm load on a 4 ohm OT....is harder on the OT secondary while a mismatch to the low side puts the stress on the power tubes?
     
    SSL9000J likes this.
  4. Wyatt

    Wyatt Tele-Afflicted

    Posts:
    1,129
    Joined:
    Nov 3, 2004
    I think that may be fair.

    If using a smaller than spec'ed load, the tubes will bias hotter and tube life will be affected. But the amp should stay within safe operating conditions. That's why Fender (and others) always wired those speaker jacks up in parallel, they wanted the load to drop, not go up. Output (voltage swing) drops either way according the tube tube data.

    But flyback voltage is the reasons why many '70's 100-watt JMP's have replacement OTs.
     
    SSL9000J and Wally like this.
  5. Inglese

    Inglese Tele-Meister

    Age:
    40
    Posts:
    107
    Joined:
    Mar 7, 2018
    Location:
    Italy
    Changing the load or the transformer ratio at the end changes the impedance seen by the output tubes.
    Changing the impedance changes the load line slop on the plate curves plot.
    And from the graph you can figure out what's happening.
    It's sad but not everything can be made simple and clear to everyone just by using simple words like "pushing hard", "sweet spot" or similar.
     
    GorgeousTones, SSL9000J and Wally like this.
  6. robrob

    robrob Poster Extraordinaire Ad Free Member

    Posts:
    7,784
    Joined:
    Dec 29, 2012
    Location:
    United States
  7. Old Tele man

    Old Tele man Friend of Leo's

    Posts:
    3,941
    Joined:
    May 10, 2017
    Location:
    Once banned always banned
    Last edited: Apr 30, 2018
    SSL9000J and robrob like this.
  8. SSL9000J

    SSL9000J Tele-Meister

    Age:
    48
    Posts:
    232
    Joined:
    Dec 18, 2016
    Location:
    Atlanta, GA
    Sounds like too many guitar players, not enough technicians. (And quite possibly no engineers.) This forum, on the other hand, is like Mensa for guitar amplifiers. So, as I suspected, the best practice is to just match impedance the way nature intended and look elsewhere to make modifications.
    Many thanks to all!
     
  9. philosofriend

    philosofriend Tele-Holic

    Posts:
    983
    Joined:
    Oct 28, 2015
    Location:
    Kalamazoo
    My experience has been that on well-made equipment, fudging an impedance by half or twice is safe. Any more that that is like running your car's engine over the tachometer redline. It might work, but it is your own fault if you break it.

    On off brand equipment, it sounds crummy no matter what you do and the equipment might be borderline blowing up even at the rated impedances.
     
  10. Old Tele man

    Old Tele man Friend of Leo's

    Posts:
    3,941
    Joined:
    May 10, 2017
    Location:
    Once banned always banned
  11. clintj

    clintj Friend of Leo's

    Posts:
    4,951
    Joined:
    Apr 4, 2015
    Location:
    Idaho
    Let's make an example, shall we?

    Let's say the amp in question idles at 400V, and under a given signal swings +/- 100V on the plates with a proper speaker load. IOW, the plates vary between 300V and 500V.

    When you mismatch a load, the impedance seen by the power tubes changes. In the case of our theoretical amp, driving an 8 ohm speaker with a 2 ohm OT is a factor of four.

    Ok, so what? Let's find out. Now for the same given signal, the plate voltage varies 4 times as much: +/- 400V. On the low end of the cycle, the tube is out of headroom and hits 0V. At the upper end, it now is peaking at 800V. Something will give out eventually, be it the tube arcing, the OT insulation breaking down, etc, somewhere there's a weak point and that kind of voltage will find it.
     
    SSL9000J likes this.
  12. waparker4

    waparker4 Doctor of Teleocity Ad Free Member

    Posts:
    18,996
    Joined:
    Nov 9, 2011
    Location:
    Philadelphia, PA
    Personally I believe it is unsafe to do that to your amp. But also I don't think it makes any sense at all to spend so much time and money replacing the output transformer of your amp to make it a sketchy attenuator, when you can buy a real attenuator or buy or build an L-Pad for under 50 bucks.
     
    SSL9000J likes this.
  13. radiocaster

    radiocaster Poster Extraordinaire

    Posts:
    6,856
    Joined:
    Aug 18, 2015
    Location:
    europe
    Is this much more likely to happen with higher wattage amps?
     
  14. SSL9000J

    SSL9000J Tele-Meister

    Age:
    48
    Posts:
    232
    Joined:
    Dec 18, 2016
    Location:
    Atlanta, GA
    I was about to ask a dumb question about the added impedance of an attenuator, then remembered this:
    https://robrobinette.com/Generic_Tube_Amp_Mods.htm#Quarter_Power_Switch
    Never mind.
    "I see!" said the blind man, (as he picked up his hammer and saw.)
     
  15. Wyatt

    Wyatt Tele-Afflicted

    Posts:
    1,129
    Joined:
    Nov 3, 2004
    Flyback voltage? Yes. Higher volumes at higher voltages and/or higher mismatches. With more AC voltage feeding the mismatch, you would get more flyback voltage.

    Just as an aside, I've never really been worried about most 1:2 or 2:1 mismatches, but it is something that should be assessed on an amp-to-amp basis. I don't like stressing small OT's in vitnage or small amps and I don't like to worry about flyback voltage in 100-watt amps, the gray area is somewhere inbetween. That said, all my setups are matched.
     
    Last edited: May 1, 2018
  16. waparker4

    waparker4 Doctor of Teleocity Ad Free Member

    Posts:
    18,996
    Joined:
    Nov 9, 2011
    Location:
    Philadelphia, PA
    That's an L-pad implemented with fixed resistors.

    [​IMG]
    [​IMG]


    I built one in a small project box from an SWR L-pad that I bought online for $10.

    https://www.parts-express.com/speaker-l-pad-attenuator-50w-mono-3-8-shaft-8-ohm--260-252
     
    radiocaster likes this.
  17. Old Tele man

    Old Tele man Friend of Leo's

    Posts:
    3,941
    Joined:
    May 10, 2017
    Location:
    Once banned always banned
    You are aware that an adjustable L-pad is actually a dual potentiometer configuration (NOT a single potentiometer) with the two "wipers" connected together so that the resistance on one side (input) varies differentialy from the resistance of the other side (output), such that Zin remains relatively constant, while Zout varies with wiper rotation. Purpose is to reduce power (-dB) via resistive-voltage "drop" while maintaining constant Zin impedance:

    FIXED resistors L-pad (NOTE: it's "A/20" for voltage(E) or current(I), but "A/10" for POWER!!!):

    [​IMG]

    Example: with Z = 8 ohm load/speaker @ half-power (-3dB), R1 = 4.0 ohms and R2 = 8.0 ohms.

    ADJUSTABLE resistors L-pad:

    [​IMG]
     
    Last edited: May 1, 2018
    viccortes285 and radiocaster like this.
  18. SSL9000J

    SSL9000J Tele-Meister

    Age:
    48
    Posts:
    232
    Joined:
    Dec 18, 2016
    Location:
    Atlanta, GA
    I was under the impression that in order for an L-pad to work, R1 had to be in series with the speaker and R2 had to be in parallel with both the speaker and R1.

    View attachment 511248
    Using the resistors as a voltage divider would upset the balance, no?

    edit: Disregard; still a series-parallel circuit. Didn't see it previously.
     
    Last edited: May 1, 2018
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.