2204 Cold clip to Symmetrical

Discussion in 'Shock Brother's DIY Amps' started by XTRXTR, Jul 16, 2020.

  1. XTRXTR

    XTRXTR TDPRI Member

    Posts:
    21
    Joined:
    Jul 13, 2020
    Location:
    Northern Colorado
    I am planning a rebuild of my 2204. The old board (actually a 2550 board) will be yanked, cleaned, and either saved or sold, not sure yet. Before the amp is rebuilt using terminal strips I want to experiment with a few mods. I will approach them one at a time as I experiment for sound. The ultimate goal is happiness and I have Covid-19 blues as all gigs are currently canceled. A great time to seek tone and have fun.

    https://drtube.com/schematics/marshall/2204prem.gif
    [​IMG]

    My first order of business is to figure out how to get a symmetrical clip from V1a-b. I'll read about it if I can get a reference but I am open to an explanation here if anyone has done this already. The thing is I don't know what symmetrical clipping sounds like since I have only used JCM 800/2204 amps my whole career. Currently the scope shows clipping at low gain only on the positive half after emerging from the High intensity input and through the V1b and then V1a stages.

    What values should I change to the cathode(s) to get a symmetrical clip? I would like to set this up as a switch from Symmetrical to Asymmetrical so I can be certain I only want one or both. It is possible that both could be wanted depending on the type of music and style being played, right? I say that with all the ignorance and naivety of having only heard the cold clipper my whole life.

    I read all the mods possible on RobRobinette dot com (this is my first post and I don't know if I can hard link that or not) and plan to implement many of them experimentally and permanently, in fact a PPIMV type 3 is already done but can be set back as before to test other mods. Switching the bright cap in the tone stack, switching NFB, tube buffered FX loop, hum balance bias circuit and more. I want versatility to play any style and dial it in.

    Thanks
    X
     
    Last edited: Jul 16, 2020
    andrewRneumann likes this.
  2. andrewRneumann

    andrewRneumann Tele-Meister

    Posts:
    291
    Joined:
    Mar 22, 2020
    Location:
    Cincinnati, OH, USA
    Welcome aboard @XTRXTR !

    Disclaimer: I have no experience on a JCM800.

    Have you considered temporarily wiring in a 10k linear taper pot in the place of that v1a cathode resistor? Find what you like, measure the pot resistance, and replace it with a resistor.

    There has been some talk recently here about that v1a and whether it is actually a true cold-clipper. The issue at hand was whether v1b could drive v1a hard enough if v1a was unbypassed. I don’t really know either way, but I bring it up to say you might also have to experiment with a cathode bypass capacitor on v1a to get the clipping you seek.

    Good Luck!
     
    XTRXTR likes this.
  3. Dacious

    Dacious Poster Extraordinaire

    Posts:
    8,508
    Joined:
    Mar 16, 2003
    Location:
    Godzone
    I have a 4210, 50 watt combo version of this amp. There's a few mods. If you just pull the diodes out it just turns the dirt channel into a mega clean. To turn it into a tube-onlu dirt channel you have to do a few component mods.

    https://www.schematichell.com/modifications/4210_mod.html
     
  4. XTRXTR

    XTRXTR TDPRI Member

    Posts:
    21
    Joined:
    Jul 13, 2020
    Location:
    Northern Colorado
    Thanks @andrewRneumann the pot idea is good. I will likely add a small resistor to the ground lug side to be sure I don't get too low of resistance by accident. But I should be able to send a signal through and watch the output for gain change and symmetry if achieved. Something tells me it won't be that simple. :twisted:

    Just a side note here. I always thought it interesting that Marshall named V1a and V1b differently depending on which year or drawing you look at for the 2204. Just so we can be on the same page in the future perhaps we should refer to low and high input rather than V1a or V1b. Low will always refer to the same half of V1 as will High. From the signal perspective the AC signal on the High input triode will always pass through the Low input triode as it exits V1 output. Where as the Low input triode signal will never pass through the high input triode.

    There is also the 100pf cap between the High input triode plate and cathode, shown and not shown, I have it in my amp.

    This side note is not meant to be a giant monkey wrench rather a clarification o_O

    [​IMG]
    [​IMG]
     
    Last edited: Jul 16, 2020
    andrewRneumann likes this.
  5. andrewRneumann

    andrewRneumann Tele-Meister

    Posts:
    291
    Joined:
    Mar 22, 2020
    Location:
    Cincinnati, OH, USA
    I understand the concern but having done the experiment myself I’m confident you won’t need the extra resistor.

    Having no cathode resistance will not hurt the valve or your amp. The 100k load resistance remains intact and keeps your valve from over-dissipating even when the grid and cathode are both at zero volts.

    I show 171mW with a 280V HT+ and 100k load resistor. Well within the 12AX7’s 1 watt limit.

    D37F0507-5478-4C5C-8AFE-1895A4BD2E75.jpeg
     
    XTRXTR likes this.
  6. andrewRneumann

    andrewRneumann Tele-Meister

    Posts:
    291
    Joined:
    Mar 22, 2020
    Location:
    Cincinnati, OH, USA
    I shouldn’t have made such a definitive statement on an Internet forum. Anytime someone is giving you advice you have to make sure it passes the sanity test. I have been wrong in the past and I will be wrong again. So... I want to say that any time you open an amp and start making changes you have a risk of hurting your amp or even yourself. Be careful, know the risks, and understand what you are doing.

    Good luck!
     
    XTRXTR likes this.
  7. XTRXTR

    XTRXTR TDPRI Member

    Posts:
    21
    Joined:
    Jul 13, 2020
    Location:
    Northern Colorado
    @andrewRneumann After reading the discussion you mentioned concerning the cold clipper and the referenced Valvewizard common gain cathode information, section 1.18, http://www.valvewizard.co.uk/Common_Gain_Stage.pdf
    It seems I may have to simply make the second stage like the first stage with the same cathode bypass cap and resistor Ck and Rk and I should get a symmetrical output. I have seen schematics of amps that do this...I think. I'll add a triple ON ON ON or a On Off On switch to it and see what happens. Maybe a Cap switch with the Pot in parallel so I can dial in a preferred gain and sound, if any. I'll clip in the values until I find what I want. I'll post results after. Thanks for the direction to get started.

    I may have questions as I go, stay tuned.
    X
     
  8. Bitsleftover

    Bitsleftover Tele-Meister

    Age:
    47
    Posts:
    469
    Joined:
    Oct 17, 2016
    Location:
    Lincolnshire UK
    I’m a bit confused by your concept.
    The 2550 jubilee is a very different amp to a 2204 JCM800.
    Is it the 2550 you’ve been playing most of your life?
    Because as far as I can recall. The Jub has a very clean “clean” channel. A symmetrical diode clipped “rhythm clip”channel and an asymmetric diode clipped “lead boost”.
    You could probably get a feel for the difference just by setting your lead boost up to be as similar as possible to your rhythm clip channel by way of grit and switch between the two.
    Or am I missing the point?
    BTW. I’d kill to get my old black Jubilee back. It’s the one that got away.
    If it’s just that you want to mod a stock 2204 design, then just go with 100k plate resistors and 1k5 cathode resistors or V1 a and b. That’ll be about in the middle. People used to turn that cold clipper into a standard gain stage by doing that to get a little extra gain out of the front end at the cost of losing the asymmetric effect.

    Sent from my iPhone using Tapatalk
     
    XTRXTR likes this.
  9. XTRXTR

    XTRXTR TDPRI Member

    Posts:
    21
    Joined:
    Jul 13, 2020
    Location:
    Northern Colorado
    I did not like the 25th anniversary Jubilee and 16 years ago I wanted a rack mountable chassis for an amp. I got tired of lifting all my equipment. I designed the chassis and had it custom made for me. I then took all the iron out of the Jubilee and dismantled most of the board but reused some of the board traces for the 2204 circuit instead. Its the diamond plate faced one in my avatar - the blue one I built from scratch, also a 2204 circuit.

    16 years later in the middle of the Covid Pandemic, the gigs are gone, the amp has hiss and hum issues its time for new caps anyway I decided it was time to give the amp a proper internal rebuild. And now, I want some additional mods, NFB select, tone stack select, tube buffered fx loop, and want to experiment to find my ultimate amp so it is versatile for any style of music. I am inspired by RobRob's mods so one at a time. However, this first thing I am doing with V1 I want to be clean and symmetrical to get more Power tube break up but switchable if possible to get the 2204 Marshall sound as well. I added the type 3 PPIMV which has insane fuzz at low volume and now the front end needs to be tamed. Once I find the mods that work for me I am rebuilding/re-soldering and rearranging the power and bias supply section along with the rest of the amp. It is going to have a new wiring layout optimized for silence at idle and sweet tube sounds. I want Dire Straits clean, Hendrix/SRV Blues, AC/DC amp, EVH and Satriani heat. I know that is a high aim but if I get 67%-70% I'll still be happier.

    For now V1 symmetry. You say standard gain stage o_O I'm interested - Do you have a schematic you can refer me to? And Thank you;)
     
  10. Bitsleftover

    Bitsleftover Tele-Meister

    Age:
    47
    Posts:
    469
    Joined:
    Oct 17, 2016
    Location:
    Lincolnshire UK
    The way I understand it, a 12AX7 triode with a 100k plate resistor like you have, is biased roughly mid way with a 1k5 cathode resistor resulting in symmetrical clipping. Reduce down to around 820r for more distortion, up to about 2k7 for a bit less. But still fairly symmetrical.
    The 10k you have on your cold clipper biases the stage so one half clips and the other always stays clean resulting in less dirt but more harmonically rich tone due to the clean half of the wave.
    I’d try a switched 1k8 resistor in parallel with with your 10k cathode resistor. That will give around 1k5 total and set your bias somewhere near mid way. You’ll be able to flip back n forth to hear the difference and confirm with your scope.
    I think you’ll end up with more distortion for a given gain knob position though. And a more buzzy distortion at that. Which may be what you didn’t like about the jubilee. (Which is sort of the opposite of what you’re trying to achieve). But it will be an interesting and fun experiment and you won’t know till you try.
    I’m sure there was a similar thread a couple of months ago and it worked out well. [​IMG]


    Sent from my iPhone using Tapatalk
     
    XTRXTR and andrewRneumann like this.
  11. XTRXTR

    XTRXTR TDPRI Member

    Posts:
    21
    Joined:
    Jul 13, 2020
    Location:
    Northern Colorado
    I'mma go try this tonight and see what happens.:) @Bitsleftover Thanks for the help.

    X
     
  12. XTRXTR

    XTRXTR TDPRI Member

    Posts:
    21
    Joined:
    Jul 13, 2020
    Location:
    Northern Colorado
    So I did get more gain from V1 but I also got some kind of oscillation only on the negative half of the 1kHz signal. Following the idea above, 1.8K Ohm switched in parallel to the 10K. I had more clean gain Vp-p up to a point, about 0-4 on the pre-amp volume (higher than non switched) but then a fuzzy oscillation worm:eek:started to show on the neg half as I went past 4 and it got bigger as I went to 10. I changed the 12ax7 out just in case it was microphonic or had a sympathetic resonant frequency. The same thing happened with the new tube however.

    When tracing to V2 one half of the cycle was clamped; looked like a half wave diode non filtered signal. I wonder now if its V2 that I need to be working on to allow a full sine wave to the PI.

    Can I jumper from V1 to PI temporarily to see a sine wave actually make it to the Power tubes? I'll wait for guidance from here before doing that. :cool:
     
    andrewRneumann likes this.
  13. Tom Kamphuys

    Tom Kamphuys Tele-Meister

    Age:
    40
    Posts:
    471
    Joined:
    Sep 18, 2018
    Location:
    The Netherlands
    https://www.tdpri.com/threads/preamp-signal-through-my-rr800v.1032352/

    The cold clipper seems to be more of a low gain symmetrical stage than a cold clipper.

    The next (warm) stage clips asymmetrical in a way you describe.

    The unbypassed cathode resistor changes the 'static anode characteristic' (load line plot). All sources I found only mention a gain difference (Valvewizard, Kuehnel, robrob), but essential knowledge from 1945 seems to have made it to today in the mind of @Ten Over.
     
    XTRXTR likes this.
  14. andrewRneumann

    andrewRneumann Tele-Meister

    Posts:
    291
    Joined:
    Mar 22, 2020
    Location:
    Cincinnati, OH, USA
    Hello again! Sounds like you are having fun and learning a lot about your amp. I think we all might benefit if you posted some images of your scope traces.

    I’m not sure from your description what you are experiencing after V1. Keep in mind that the valve has very low impedance to portions of grid signals that are positive (relative to cathode). So you can basically see the positive side clipping “before” the signal hits the tube. (I hate using “before” and “after” because I like to think of the entire circuit as acting in a concert at the same time.)

    Example: Gain stage has about 3Vp-p headroom and is biased warmly at -1V. If I hit the grid with a 10Vp-p signal and measure the signal on the grid, I will see clipping on the positive side at +1V and the entire negative side intact down to -6V. On the output side, I will see the full clipped version at 3Vp-p.

    edit: Clipped output wave would be 3Vp-p times the gain of the stage.
     
    Last edited: Jul 19, 2020
    XTRXTR likes this.
  15. XTRXTR

    XTRXTR TDPRI Member

    Posts:
    21
    Joined:
    Jul 13, 2020
    Location:
    Northern Colorado
    It seems @NSB_Chris and I are or were on the same path to enlightenment. @Ten Over ('s) simplified tube illustration seems to be right on point. The more positive AC signal amplitude on the grid of V2 the more it behaves like a diode and clamps. I think if you look at it from the electron flow perspective it is much easier to understand. As electrons flow from cathode to anode the anode becomes more negative and the grid is the control of that flow. Once the grid is more positive than the bias point the electrons are attracted to the grid and don't make it to the anode, clamping the gain of the positive half to the grid and allowing the negative to increase. So the output on V2 plate looks like a positive half wave diode since it is inverted from the input on the grid. I'm not sure if i am expressing this idea right but that is how my brain is visualizing it.

    So, if this is the case, and it looks as though the 2204 and the RR800 are exactly the same except for the single input, I won't get what I thought I was going to. The output gain from V1 is already high enough to clamp V2. I was looking for a capability that is not possible in the 2204 as it is. Which brings me back to this, can I bypass V2 and go to V3 the PI, Im guessing the affect is similar. The PI will be clipped for similar reasons and the power tubes will see a square wave or worse maybe a notched square or a clipped sawtooth "MWMW" type of wave.

    @andrewRneumann Your example agrees with what Tom Kamphuys points out and what I have seen in my trace. I will try to take pictures of the scope traces when I go down into my lair to play foolish with the universe's mysteries...I'll stop at nothing!! Except for where the boundaries have already been found apparently by Jim Marshall.

    There is still room for Rob Robinette's mods. I'll see about getting those pics for completeness. Maybe less gain from V1 will give me more control - there I go again - that is what the Low input is for. Maybe a switch on the High input that allows me to bypass the high input and go directly to the low input without unplugging my cable - I'll have to look at the schematic. :twisted:

    X
     
  16. NSB_Chris

    NSB_Chris Tele-Meister

    Age:
    55
    Posts:
    266
    Joined:
    Sep 2, 2017
    Location:
    New Smyrna Beach Florida
    That is what I have done on a few Fender clones. I did it mostly to save space on a crowded faceplate.

    input pic.JPG
     
    XTRXTR likes this.
  17. XTRXTR

    XTRXTR TDPRI Member

    Posts:
    21
    Joined:
    Jul 13, 2020
    Location:
    Northern Colorado
    @andrewRneumann @Bitsleftover
    Okay Pics of the Fuzzy Worm I mentioned above
    0-1 PV Probe A is input waveform Probe B is on cold clipper plate below input waveform

    1-2 as I turn up PV

    5-6 on the PV dial

    Cranked

    This is on V2A Grid and pretty much rest of amp cranked PV

    I think I'm gonna remove the PPIMV mod and add the RobRob three way Cold Clipper switch and play guitar for a while, I'll post about it soon.

    X
     
    andrewRneumann likes this.
  18. mrriggs

    mrriggs TDPRI Member

    Age:
    40
    Posts:
    46
    Joined:
    May 24, 2020
    Location:
    Vancouver, WA
    I had the same parasitic oscillation in the power tubes of my last build. 'Grid stopper' resistors fixed it.
     
    andrewRneumann and XTRXTR like this.
  19. XTRXTR

    XTRXTR TDPRI Member

    Posts:
    21
    Joined:
    Jul 13, 2020
    Location:
    Northern Colorado
    Them grid stoppers are important. I added those to all three of my 2204 amps including the 20th anniversary series white tolex head. That was another interesting missing resistor on some of the Marshall schematics.
    Not shown:[​IMG]
    Shown:
    [​IMG]
    Anyway, I am not sure what type of oscillation this was but it only showed when the switch was closed to include the 1.8k resistor in parallel with the 10k. The general take away for me was that 1.5k cathode resistance on the Cold Clipper is too hot to operate with stability on this specific amp. There are many long and crossed wires in this amp as it was my first DIY and I am using a board meant for the 2550. Some of the distances between points are quite extended and can easily be the cause of induced noise and oscillations. When I add mods I am trying to keep wire lengths as short as possible to avoid creating issues. This amp is an ongoing project and will be rebuilt after I experiment with the mods I want to include. Then I will optimize a new layout etc and rebuild the guts.

    So I reversed the PPIMV mod, for now, I might add one in addition to the Pre PI MV so I can tame that PI distortion...IDK, we'll see. I bought extra bits and pieces just to experiment with many different mods. I went back to original values and I added the 3-way Cold Clipper Switch per RobRob's page of awesome Mods. I'm gonna jam on some guitar for a few hours and then post about my experience at a later time. Since this thread was started with focus on the Cold Clipper I want to continue that until I have settled on mods concerning V1 on the 2204.
     
    andrewRneumann likes this.
  20. mrriggs

    mrriggs TDPRI Member

    Age:
    40
    Posts:
    46
    Joined:
    May 24, 2020
    Location:
    Vancouver, WA
    Grid stoppers aren't just for power tubes. Did you add one to the pre-amp tube that was oscillating? With the resistor right at the socket pin, it should kill that oscillation, even if there is a long lead feeding it.

    If symmetrical clipping is your goal then it may be easier to achieve with a simple diode "hard" clipper.

    [​IMG]

    The following scope trace shows why I put quotes around "hard". Tube diodes can clip a whole lot softer than solid-state diodes.

    [​IMG]
     
    FenderLover likes this.
IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.