Not sure if my Plexi build has problems

Discussion in 'Shock Brother's DIY Amps' started by itsGiusto, Oct 26, 2019.

  1. itsGiusto

    itsGiusto Tele-Meister

    Posts:
    274
    Joined:
    Dec 31, 2017
    Location:
    Massachusetts
    Hey, I just completed my build of a Ceriatone 67' Plexi Lead 50 (basically a Marshall plexi specced from 1967). I biased the tubes to get them sitting between 67% and 69% dissipation at idle.

    The amp was exhibiting some weird problems one of which kind of went away... I'm not sure what to make of it.
    These clips were recorded with the channels jumpered, each channel volume at about 8, with varying tone settings.


    Current "problem" - This problem still exists, but I'm not sure if it's a problem, or just normal for this kind of amp!! ¯\_(ツ)_/
    Sometimes the amp sounds like it's distorting in weird ways, almost like a ring modulator or an octave fuzz. Sometimes when I hold notes, I'll hear weird overtones that modulate
    Here's a clip showing some of the spitty, fuzzy-like sounds. You can hear subtle ring-modulation like things going on:

    The problem also seems more pronounced when I turn my NFB dial down (in place of the NFB resistor, I put a 250k pot in, in series with a limiting 27k resistor). When it's at 27k NFB (from the 16 ohm tap), this problem doesn't really show up, and also the overdrive tone is much less overdriven. When I turn the NFB pot so the NFB is at more like 100k - 120k, the problem is much more present. The tone is more overdriven, but also comes along with these kind of spitty, modulating overtones.

    Note, the problem is a little less pronounced now, after I opened up the amp and made a voltage chart and put it back together again. When I record it, I barely even really hear it on the recording, though it sounds more pronounced when listening in person. Here are more recent clips showing it:



    https://soundcloud.com/itsgiusto/project-10-26-19-3-28-pm

    Is this sort of sound normal for a jumpered Marshall Plexi turned up to 8? Or do I have some sort of unwanted distortion going on?



    Former Problem - This problem seems to have gone away after I opened the amp up and made a voltage chart. ¯\_(ツ)_/
    The volume was varying, especially after I hit a chord. It almost sounded like a compressor with a long attack time pumping - I'd hit the chord, then a half second later, it'd drop in volume. Here's a sound clip showing the problem:

    As you can hear, the longer I waited between playing successive chords, the more pronounced of a "pumping" effect I get.
    You can also hear some weird little crackling sounds sometimes when the volume pumps.




    Here are the voltages I've measured in my amp, with all knobs rolled down to 0, and nothing plugged into the input jacks:

    [​IMG]

    And here are the voltages Ceriatone has published as being normal for the amp:
    [​IMG]

    Some of the differences are accounted for the fact that Ceriatone's is split-cathode in V1, and mine is shared cathode.

    One thing that stands out to me in my build is that pin 8 on V2 is very high. I'm using a JJ short-plate 12ax7, since that seems to be good for cathode-followers.

    Another thing to note: the EL34 plate voltages have dropped since I took these measurements. The next time I went to measure, they were more like 406v.
     
    Last edited: Oct 26, 2019
  2. BigDaddy23

    BigDaddy23 Tele-Meister

    Age:
    45
    Posts:
    264
    Joined:
    Jan 4, 2019
    Location:
    Australia
    Do you have some pics of the circuit? Build and the schematic? This will help.
     
  3. dogmeat

    dogmeat Tele-Afflicted

    Age:
    68
    Posts:
    1,103
    Joined:
    Oct 12, 2017
    Location:
    Alaska
    I had some weirdo tones on an amp I built that had a variable NFB (AB763 type). changing tubes helped some but there were settings that just did not work depending on the volume. a great sound with the master down would be nasty when louder... that kind of thing. this was a few years ago, so sorry, I don't remember all the details
     
  4. itsGiusto

    itsGiusto Tele-Meister

    Posts:
    274
    Joined:
    Dec 31, 2017
    Location:
    Massachusetts
    Yeah, I can try to get a circuit diagram. It'll be hard cause I put so many mods in to make it switchable and get different sounds, switchable cathode, switchable from tube to diode rectifier, etc.

    I'm hesitant to post pics of the build. As result of all my mods, there are wires going everywhere, people are going to flip if they see it. I tried my best to keep it neat, but it isn't really. Also, because I wanted to be able to service the underside of the board as well, I kept some if the wires on one side of the board very long, to give clearance. A real pain point of mine with previous builds and when working on other amps is that if you make the wires as short as possible, you can't get to the underside of the board, making fixes and mods impossible.
     
  5. dan40

    dan40 Tele-Afflicted

    Posts:
    1,355
    Joined:
    Aug 19, 2015
    Location:
    Richmond Va
    It's possible that you have a cold solder joint, possibly around v2 that was causing the volume fluctuations. The act of taking measurements and touching the connections with your meter leads may have helped but the issue will likely return if the joint is truly bad. Another possibility is oscillation caused by the long wire leads that you mention. Just moving a sensitive wire a few mm's can cause the oscillations to start or stop. The wire leading from the board to the presence pot is especially sensitive in these circuits and needs to be kept away from other tone control wires. For maintenance sake, I have gotten in the habit of placing all board jumpers on the topside to make future repair much easier. Everything can then be repaired from the topside and all wire leads can be kept as short as possible.
     
  6. dan40

    dan40 Tele-Afflicted

    Posts:
    1,355
    Joined:
    Aug 19, 2015
    Location:
    Richmond Va
    Positive feedback is also a possibility when starting up a new amp for the first time. It usually shows up as a loud squeal or howl but it can cause other weird issues. You can simply flip the plate leads at pin 3 of each power tube socket or simply disconnect the nfb from the output jack and see if the amp performs properly. With nfb disconnected, it will have slightly more high end and will transition into overdrive much quicker. If the weird symptoms are gone with the nfb disconnected, simply flip the plate leads at pin 3 and reconnect the nfb circuit.
     
  7. Dacious

    Dacious Poster Extraordinaire

    Posts:
    6,963
    Joined:
    Mar 16, 2003
    Location:
    Godzone
    You don't mention where your eq was are set - where's your treble? My limited experience with JCM amps suggests treble settings above 6 (just past noon) along with presence settings will engender weird overtones around the bottom strings. I think your negative feedback mods are possibly contributing to instability - too much Tweed for what is already a OTT preamp. Higher values seem to contribute to more gain it and butsustain but it sound like you have plenty.

    Despite the hype, no-one ever really dimes all the dials on a Plexi unless going for extreme tones like Kravitz' Fly Away - even then, notice he doesn't hold notes and lets the bass do the solo.

    OTOH, guys who've built these say lead dress to avoid parasitic oscillation is critical. So maybe tidy and reroute your leads.
     
    itsGiusto likes this.
  8. itsGiusto

    itsGiusto Tele-Meister

    Posts:
    274
    Joined:
    Dec 31, 2017
    Location:
    Massachusetts
    My settings were varying by clip, really.
    But, yeah, actually, the weird overtones were happening and much more present when my treble was set higher, around 6. Then I turned it down, and it got much less obvious. So maybe it's just normal?

    As far as the NFB mods are concerned, the range I've seen for plexi specs are as follows:
    27k, 16 ohm tap for early plexi
    47k ,8 ohm tap (equivalent to ~70k from 16 ohm tap) for middle years
    100k from 4 ohm tap (equivalent to 200k from 16 ohm tap) for later years
    So my setting of ~100k to 125k from the 16 ohm tap should have still been within plexi ballpark.

    I can try turning down a little bit and see. Maybe I was just pushing it too hard, and putting the treble too high?

    What exactly is "parasitic oscillation"? I wasn't getting any runaway siren-like sounds. Just some weird spitty overtones, that sound a little like a fuzz face or octave fuzz.

    Heh, I really wish I had another similar amp I could compare it to to see if it's sounding like it's supposed to, or if I had a friend nearby who has experience with plexis who could play it and see what they think. I really don't know if there's any problem at all, I just don't have much experience with these amps!
     
    Last edited: Oct 26, 2019
  9. Paul G.

    Paul G. Friend of Leo's

    Posts:
    2,710
    Joined:
    Mar 17, 2003
    Location:
    Rhode Island
    That's not what they sound like.

    I'd shorten and clean up all your leads, this does make a difference. Make sure signal wires are kept away from AC components, only cross them at right angles. The wire to the presence control is critical, keep it away from all the other tone control wires.

    Double-check your layout against the schematic to make sure it's all ok. Make sure the filter caps are properly wired and well soldered.
     
  10. Nickfl

    Nickfl Tele-Afflicted

    Posts:
    1,623
    Joined:
    May 24, 2016
    Location:
    Florida
    You need to post pictures if you want help. There are people here who can help separate benign messy from problem messy and tell you where your problems might be coming from, but not without pictures.

    Those clips definitely don't sound right to me, sounds like some sort of instability in the amp. If its worse when you turn up the treble its probably a high frequency thing. Does your circuit have the stability cap across the PI plates?
     
    BigDaddy23 likes this.
  11. itsGiusto

    itsGiusto Tele-Meister

    Posts:
    274
    Joined:
    Dec 31, 2017
    Location:
    Massachusetts
    Alright, you've convinced me. I'll take some pictures and post them tomorrow. I'll also try to start photo-editing myself an accurate schematic of what my build is.

    My build has a 47pf cap going from one PI plate to the other
     
  12. itsGiusto

    itsGiusto Tele-Meister

    Posts:
    274
    Joined:
    Dec 31, 2017
    Location:
    Massachusetts
    Okay, I've put together a schematic showing my build and my mods:

    [​IMG]

    I'll get out some pics of the actual build tomorrow.
    Thanks all, for the help!
     
    BigDaddy23 likes this.
  13. clintj

    clintj Friend of Leo's

    Posts:
    4,921
    Joined:
    Apr 4, 2015
    Location:
    Idaho
    Parasitic oscillation happens when the closed loop gain at a frequency is greater than 1. That's some big words, let's break it down.

    Every wire carrying AC broadcasts a signal like a radio broadcast antenna does. Any wire in the vicinity can pick up that signal just like a radio does. (Faraday's Law)

    Let's say you have the wire from the plate running too close to the wire from the grid on a preamp tube, and you strum a chord. The plate wire broadcasts a signal, and the grid wire picks it up. The tube will try to do its job and amplify that new signal, and a feedback loop starts. Now because there's capacitance and inductance involved, this doesn't work on all frequencies, just in certain ranges. It also only happens when you're playing a note because the signal is only strong enough to cause this when you hit a note on the guitar, otherwise once the guitar signal fades the feedback stops. The feedback feeds off the main signal like a parasite.

    This is why lead dress requires wires to cross at right angles or be kept separate from each other. Distance and right angles prevent the feedback loop from forming by preventing the wires from "hearing" each other.

    Sent from my Pixel 2 using Tapatalk
     
    D'tar, mexicanyella, corliss1 and 2 others like this.
  14. itsGiusto

    itsGiusto Tele-Meister

    Posts:
    274
    Joined:
    Dec 31, 2017
    Location:
    Massachusetts
    Cool, that's what I thought. But I'm not getting any runaway oscillating sounds, just some weird overtones which sound like it's distorting in some strange way, like intermodulation distortion or crossover distortion. So I'm doubtful that parasitic oscillation is the problem here.
     
  15. itsGiusto

    itsGiusto Tele-Meister

    Posts:
    274
    Joined:
    Dec 31, 2017
    Location:
    Massachusetts
    Okay, here are some pictures of the build. Like I said,it's not neat. Let the blood-bath begin...

    Overall shot of the chassis from the back:

    [​IMG]

    More detailed shot of the board. Note, I moved around some of the typical component placing to make room for some of the mods, alternate tone cap and resistor values, etc.
    [​IMG]

    Shot from the back. Note, the NFB knob is also the switch for changing the values of the PPI coupling caps:
    [​IMG]

    Detailed shot of two switches on the front. One is the power switch, the other is the tube/diode rectifier switch. That switch also changes which bias pot is used:
    [​IMG]

    Shot of prescence knob and bass knob from the back. The bass knob is also the switch that adds the v2 bypass cap in or takes it out of the circuit:
    [​IMG]

    Shot of the middle and treble knobs. The Treble knob is also the switch that changes the tone stack from the bass tone stack to the lead tonestack:
    [​IMG]

    Shot of the channel-volume knobs. The Hi-treble knob has the switch to change the value of the bright cap from 500pf to 4700pf, and the normal channel volume knob has the switch to change V1 from shared to split-cathode.
    [​IMG]

    Shot of the input jacks. I used shielded cable for the runs from the jacks to the board and ground it at the input ground.
    [​IMG]

    Shot of V1. I used shielded cable for the run from the board to V1 input and ground it at the lug:
    [​IMG]

    Shot of V2, the FX loop send and return, and the diode bias pot. I use shielded cable for the cable runs to and from the FX send.
    [​IMG]

    A shot of the speaker impedance selector, the speaker jacks, the tube-rectifier bias pot, and V3:
    [​IMG]

    A shot of the power tube wiring, the bias probe jacks, and the NFB knob. The NFB knob is also the switch for the PPI coupling cap values. The wire from the NFB pot to the prescence knob is shielded, and grounded at the prescence pot.
    [​IMG]

    A shot of the fuses and the capacitors:
    [​IMG]

    A shot of the power input jack and the rectifier tube:
    [​IMG]

    Another shot of the filter cap area:
    [​IMG]

    A shot from the front, showing the knobs:
    [​IMG]

    Shots of the underside of the board. I kept the leads long enough that I could access the underside of the board.
    [​IMG]

    [​IMG]
    [​IMG]

    I know that wire spaghetti, like this, can cause noise to leak in, and crosstalk, but honestly, there isn't much noise in the signal path when I listen to it. I took great care to have power-grounding different from signal path grounding, etc, to have a low noise floor. The problem is not noise, it's distortion, so I'm uncertain if this spaghetti is actually the root cause of the problem. I could be wrong, though, so please let me know. Can noise lead to weird distortions I was getting, stuff that sounds like a spitty octave fuzz?

    I know I can probably make the leads shorter if I don't prioritize accessing the underside of the board, but I refuse to do anything that'll make it difficult to maintain, long-term. So should I move all underside wires to the top of the board so I can shorten the leads? Should I be wrapping the wires around the turret base to solder it there instead? I'm not too familiar with turret wiring practices other than wiring in the top and underside holes.

    Also, I'd be willing to part with some of the mods if it'll neaten things up a bit. The .68 cap on the V2 cathode doesn't have much of a tonal impact. Same for the PPI coupling caps, so I can get rid of those if necessary. But the tone-stack switch really makes a huge difference, as well as the v1 cathode-split, and bright caps so I want to keep those. The NFB pot is a great tonal difference as well, so I'd like to keep that. Maybe if people think there's any reason why it's bad for it to be a pot, I could change it to be a 3-way switch instead?
     
    Last edited: Oct 27, 2019
  16. Diverted

    Diverted Tele-Meister

    Age:
    48
    Posts:
    299
    Joined:
    Nov 11, 2016
    Location:
    Bristol RI


    Great explanation. I formerly didn’t understand this as well. Thanks!
     
  17. dan40

    dan40 Tele-Afflicted

    Posts:
    1,355
    Joined:
    Aug 19, 2015
    Location:
    Richmond Va
    Oscillations can definitely cause the spitty sound that you describe. With the extra wring from the mods you installed, the chances for this to happen are greatly increased. You could try some lead dress adjustment with a chopstick while playing through the amp to see if the problem gets better or worse. A looper pedal helps in this situation so that your hands are free to move about.

    One of the pics shows the wiring to v1 and v2 crossing and overlapping each other. I would try to shorten those up so that they stay away from each other. Removing a few of the switching mods and shortening some of the critical wires should make a notable difference.
     
    Nickfl and itsGiusto like this.
  18. dan40

    dan40 Tele-Afflicted

    Posts:
    1,355
    Joined:
    Aug 19, 2015
    Location:
    Richmond Va
    I forgot to mention that you can shorten all of your wiring to the sockets and pots and still get access to the board underside if you ever need to. Simply remove the potentiometer nuts and face plate, slide the pots back out of the chassis and then you will have the necessary slack to flip the board over and make your repairs.
     
    itsGiusto likes this.
  19. David Barnett

    David Barnett Doctor of Teleocity

    Age:
    63
    Posts:
    11,449
    Joined:
    Mar 17, 2003
    Location:
    The Far-Flung Isles of Langerhans
    Does it do it when you don't jumper the channels?

    I had a small-box, metal-panel JTM50 that would squeal or block if you jumpered the channels and turned 'em both right up. It was fine if only one channel was used at a time. It turned out to be lead dress on the wires to the volume controls.
     
  20. itsGiusto

    itsGiusto Tele-Meister

    Posts:
    274
    Joined:
    Dec 31, 2017
    Location:
    Massachusetts
    Alright, I can work on this in the upcoming days. Which pic shows the V1 and V2 wiring crossing? Which wires are the critical ones to shorten?

    I can't actually pop the pots out, or at least not all of them, because some are soldered directly to the ground bus. I think I'd rather just migrate all my connections to the top of the board, if that's not inadvisable.

    Umm, not sure, I can't remember. I think it seems to be more present on the normal channel, so maybe not if I don't jumper, and just use the hi-treble. I'll have to investigate
     
    Last edited: Oct 28, 2019
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.