Marshall 1987 and 1959 RI-- surprise inside

Discussion in 'Amps/Cabs Tech Corner: Amplifier, Cab & Speakers' started by gtrnstuff, Dec 18, 2005.


  1. gtrnstuff

    gtrnstuff Member

    Messages:
    2,215
    Joined:
    Oct 19, 2005
    Location:
    Middle TN
    I always thought my RI Marshall 50w and 100w sounded quite different from each other. After all these years I'm actually taking a closer look.

    The 50w uses a 1k cathode resistor for V2a, no .68 bypass cap. Otherwise pretty normal, has a sticker that says "5881" but came with EL-34.

    The 100w has an 820/.68 on V1a cathode(!?!), and an 820/330 like you'd expect on V2b. On the 470k mixer resistor for channel 1, a 470p in parallel with a 120p. On V2a, an 820, this time with the .68 bypass cap.
    At least I know why the 100w seemed to have more headroom--that V1a cathode thing, and the more aggressive mids. Now, was that a goof or were they really copying some great old head that they liked?
     
  2. gtrnstuff

    gtrnstuff Member

    Messages:
    2,215
    Joined:
    Oct 19, 2005
    Location:
    Middle TN
    D'oh. I realized later I had said that backwards. All things equal, a smaller cathode resistor would actually *increase* gain at that stage, wouldn't it? Headroom coming from something else, then.
     
  3. donnyjaguar

    donnyjaguar Member

    Messages:
    4,234
    Joined:
    Nov 4, 2005
    Location:
    Canada
    Actually, a lower value Kathode resistor would decrease gain. There is a rule of thumb that works with any ampliifier device:

    More current = less gain = less distortion = more heat :)

    For the hi-fi folks you can add also = better sound, but for the musical instrument crowd it becomes debatable. That said, if the voltage supplying the tube is higher, it is also possible that the cathode resistor will be a higher value so that the tube is sitting at the same idle conduction point of perhaps 1mA, which is where Fender amp's seem to stay near. Makes sense, because this puts the tube in a good linear region and doesn't tax it too heavily either hence it should last a long time.

    All this said, the arbitrary bypass capacitor value will definitely affect the sound depending upon the K resistor value. For 820R its -3dB point is 2.8kHz and for 1k it is 2.3kHz.

    DJ
     
  4. gtrnstuff

    gtrnstuff Member

    Messages:
    2,215
    Joined:
    Oct 19, 2005
    Location:
    Middle TN
    Thanks for the tutorial. My electronics knowledge is very limited, mostly from staring at schematics and parts and swapping out components within generally accepted parameters. So I was right the first time, and didn't know why;-0
    Still curious about the amps' deviation from Marshall's schematics for late Plexi circuits.
     
  5. Wakarusa

    Wakarusa Member

    Messages:
    1,477
    Joined:
    May 7, 2004
    Location:
    Newnan, GA
    All else being equal, 1K vs 820 isn't a huge difference. Regarding the discussion on smaller Rk = lower gain, that's not really the right answer. The combination of tube type, plate voltage and plate load are the significant bits (that and the cathode bypass cap Ck, but more on that in a second).

    The cathode resistor sets the bias point for the stage and is usually selected after the other stuff is determined. The target is to bias the stage Class A. Vary too far from this and the stage will clip asymmetrically. Just guessing, but if the plate loads are the same, I'd look for different B+ between the two amps.

    The bigger difference is the presence or absence of Ck. An unbypassed cathode will give a stage with lower gain than with a bypassed cathode. The 0.68uF actually results in a partially bypassed cathode -- i.e. some frequencies get the boost from bypassing, others do not (fully bypassed for audio work would have Ck in the 20-30uF range). So... the 0.68uF Ck gives the stage a high freq boost (or a low freq cut, depending on how you want to look at it).
     
  6. gtrnstuff

    gtrnstuff Member

    Messages:
    2,215
    Joined:
    Oct 19, 2005
    Location:
    Middle TN
    Thanks Todd! I haven't measured all the voltages yet, but the last time I checked pin 3 on the EL-34s, the 100w was 447vdc and the 50w was 468vdc. So if the B+ trend continues down the line, the Rk affecting Channel 1 of each amp is tending to compensate for that, right? Of course Channel 2 is 820/330u on both amps. Something tells me Marshall wasn't so concerned so much about the type of clipping as the amount.

    I also got this response on the Weber board:
    Date: December 19, 2005 at 13:36:35
    From: Bob Clear
    Subject: Re: Marshall 1987 and 1959 RI-- surprise inside

    The story as told by Ritchie Fliegler was that he and Mitch Colby scoured the country for the 'golden' 100W Marshall. They found two they liked - one belonged to Steve Melkesithian at Angela, the other to Warren DiMartini.

    Supposedly they sounded nearly identical but Warren's was one of the 550V+ versions and was unreliable with current (at the time) EL34s, but Steve's was ~480V and was solid. So they took Steve's and measured everything, resulting in that weird bypass cap buildup on the bright channel.
     

Share This Page