1. Win a Broadcaster or one of 3 Teles! The annual Supporting Member Giveaway is on. To enter Click Here. To see all the prizes and full details Click Here. To view the thread about the giveaway Click Here.

Continuing saga of plexi clone build with harsh distortion

Discussion in 'Shock Brother's DIY Amps' started by itsGiusto, Oct 20, 2020.

  1. _Steve

    _Steve TDPRI Member

    Age:
    40
    Posts:
    82
    Joined:
    Jul 14, 2020
    Location:
    Los Angeles
    OK theres a different issue here - not oscillations. You should be rocking hard by 4-5. 8 should be screaming crazy hair-metal distortion. Did you take any pics of your scope as you were tracing it with the 82Hz signal? I want to compare the amplitudes. Somewhere its not right. We should be able to find this.
     
  2. itsGiusto

    itsGiusto Tele-Holic

    Posts:
    616
    Joined:
    Dec 31, 2017
    Location:
    Massachusetts
    I didn't, but I can go take some. What settings do you want to see? And what trace points in the circuit?

    One circuit variation which I think can affect how early the onset of distortion is, are the voltage dropping resistors right after the choke. Some variations have as little as 16.4k (like mine) which results in less distortion, and some have as much as 20k, which results in more distortion. Which does yours have?

    Also, I noted that your B+ is much higher than mine. The voltage on my plates is like 410v
     
    Last edited: Oct 30, 2020
  3. _Steve

    _Steve TDPRI Member

    Age:
    40
    Posts:
    82
    Joined:
    Jul 14, 2020
    Location:
    Los Angeles
    Can you take the traces at the same spots as me? I labelled each picture with the location. I used an 82Hz signal @ 500mV Peak (ie 1V p-p). Make sure all the amp knobs settings are the same too (especially the presence at 0). If your scope has a Measure function turn it on.

    We'll just see where we (hopefully) see any differences in amplitude and narrow it down from there.

    Also can you please measure the B+ voltages at each node as well as each triode's cathode pin voltage.

    I dont think its the dropping resistors. But once we get there we can compare - not by my amp right now. Im hopeful we can nail this.
     
    itsGiusto likes this.
  4. itsGiusto

    itsGiusto Tele-Holic

    Posts:
    616
    Joined:
    Dec 31, 2017
    Location:
    Massachusetts
    Okay, sounds good, but I'm actually not 100% sure what the terms B+1, B+2, B+3, B+4, and B+5 refer to. I've only ever heard of B+. Is it just each step along the power section? Like, after the diodes, it's B+, or B+1, then after the choke it's B+2, then after the dropping resistors it's B+3, etc?

    Also, you said your power section (and possibly your power transformer) is the 100w plexi version. Does that mean we should expect your voltages to be different than mine, which is the 50w plexi version?
     
    Last edited: Oct 30, 2020
  5. _Steve

    _Steve TDPRI Member

    Age:
    40
    Posts:
    82
    Joined:
    Jul 14, 2020
    Location:
    Los Angeles
    Yup B+ is exactly that. Take note of where each B+ section goes to (eg B+2 -> Phase Inverter etc)

    Yes my power transformer is different so voltages will be a little different - not a whole lot though and it doesn't make a huge difference except to the cork sniffers ;) Tell me what PT you have and i'll look up what its voltages should be, just to sure.
     
  6. itsGiusto

    itsGiusto Tele-Holic

    Posts:
    616
    Joined:
    Dec 31, 2017
    Location:
    Massachusetts
    K, I'm using this one: http://www.classictone.net/40-18054.html
    For the circuit, I'm using the orange secondaries, which are 625v total (312.5 each).
     
  7. trxx

    trxx Tele-Afflicted

    Age:
    42
    Posts:
    1,563
    Joined:
    Dec 6, 2019
    Location:
    Between reality and imagination
    My technical knowledge on amps is still pretty limited, but I have a Ceriatone Plexi 51 (which can supposedly be switched to 67 specs), a strat, a meager recording setup, and an old scope and signal gen that I never really got around to making use of. If I can be of any help for comparing some things, let me know. I could probably learn a thing or three from checking some things out.

    What looks strange to me in your video titled 'Oscilloscope Marshall 3' is that the notching is only happening on the positive half cycles. Maybe some positive DC is being passed to the output which that higher frequency signal is riding?
     
    Last edited: Oct 31, 2020
  8. dan40

    dan40 Friend of Leo's

    Posts:
    2,041
    Joined:
    Aug 19, 2015
    Location:
    Richmond Va
    The late 60's and early 70's models usually ran a b+ of 420v or so. These had 16-20k of preamp dropper resistance to achieve the proper plate voltage on the preamp tubes. In the late 70's when the supply of reliable el34's was drying up, Marshall lowered the b+ voltage to 370-380v for a few years. These models usually show up with just a 10k resistor to keep the preamp voltages at the same levels. When the JCM 800 line came along the b+ was once again raised above 400v. The lower b+ models definitely broke up a bit earlier and felt a bit spongier than the higher plate voltage models. The 72-73 models had the higher voltages but Marshall made some slight changes to the preamp and the nfb circuit. These small changes made these models sound a bit more dirty than the earlier models. The changeover from a 500pf bright cap on the bright volume to a 5000pf cap somewhere in the late 60's also brought the overdrive on a bit earlier on the volume dial.

    Just a few small changes in these preamp circuits can really make a difference in how the amp reacts. Marshall slowly changed the circuits from shared cathode arrangements/tube rectified circuits to split cathode/ss rectified throughout the 60's as players were wanting more volume and more distortion from these amps. As the years went they got brighter and more aggressive sounding to meet the demand of the times.
     
  9. dan40

    dan40 Friend of Leo's

    Posts:
    2,041
    Joined:
    Aug 19, 2015
    Location:
    Richmond Va
    It will work provided the b+ is not extremely high and the bias circuit will provide the correct range. 6v6's will cause the b+ to go up because they don't draw as much current as el34's so this usually works better in lower b+ amps.

    I mentioned earlier that many of these old Marshall circuits will exhibit ghosting at higher volume levels. Raising the mains filtering usually helps but you can't go very high with a tube rectifier. Pushing the amp to 10 on the volume knob will only make the ghosting issue worse and it's possible that the attenuator is also creating some artifacts as it interacts with the OT. It would be nice if you could try a high volume test without the attenuator to see if the issue lessens any.
     
  10. itsGiusto

    itsGiusto Tele-Holic

    Posts:
    616
    Joined:
    Dec 31, 2017
    Location:
    Massachusetts
    Interesting. Would it be safe for me to put the 6v6 tubes in and turn it on, and then if I see that the plate voltage is getting too high, I can quickly turn it off? Or is even a few seconds of exposure to "too high" voltage dangerous to me, the amp, or the tubes? What voltage would I see on the plates that would necessitate me deciding that I'd need zeners or resistors to drop the voltage to be compatible with 6v6s?

    Unfortunately, I've tried bumping all filter caps to 50uf, and the ghosting sounds the same, at least at high volume.

    Could you think of a way I could test the proposal that the attenuator is causing the problem without a high-volume test? I could probe both pre and post attenuator, but I don't quite know what to look for on the oscilloscope, what the ghosting looks like.
     
    Last edited: Oct 31, 2020
  11. elpico

    elpico Tele-Afflicted

    Posts:
    1,178
    Joined:
    Sep 14, 2011
    Location:
    Vancouver BC
    The B+ will rise if you switch to 6V6 due to reduced current draw. If it's already over 400Vdc with EL34 you'll probably need some pretty stout 6V6 like the JJs because it could hit 430, 440 etc.
     
    dan40 and itsGiusto like this.
  12. dan40

    dan40 Friend of Leo's

    Posts:
    2,041
    Joined:
    Aug 19, 2015
    Location:
    Richmond Va
    Like elpico mentioned, it will rise up a bit with 6v6's so it's really best done with lower voltage amps. If you want to try it for a quick test, remove the el34's, dial up the highest negative bias voltage on pin 5 that your pot will give you and set your impedance selector for half of what the speaker or cab is rated at. Power up the amp and do some quick measurements to see if your bias is in the correct range. You will be fine doing this for a few minutes as long as you don't see any redplating occurring. If the bias circuit won't allow a high enough negative voltage to bring the dissipation down you can swap the bias resistor for a different value. Many of these 6v6 plexis are running a plate voltage of 430v or so but with a stout 6v6 like JJ's you should be fine. The amp will still be loud but it will breakup earlier on the dial. Most of the distortion in these Plexi type circuits comes from the phase inverter and power tubes. These amps need to be run loud to achieve that great Marshall crunch that many players love. The master volume amps that came out in 76-77 used a cascaded preamp to achieve that crunch at lower volume settings so you may want to consider converting your board over to that if volume is an issue for you.

    I can't think of a way to test this without actually cranking the amp up without the attenuator. The power tubes, OT and attenuator all work together when the amp is being pushed. I would think that you would really need to hear the results through your chosen speaker to see if the amp sounds different without the attenuator. I have also read in the past that certain attenuators work better with certain power amps so it may be a good idea to try a different type if you have one available.
     
  13. Nickfl

    Nickfl Friend of Leo's

    Posts:
    2,399
    Joined:
    May 24, 2016
    Location:
    Florida
    Do you live in an apartment or have you already had neighbors complain? If so I see your issue, but if not I think you might be surprised what you can get away with, especially if its just for 5 min to test it without the attenuator and it isn't at unreasonable hours...

    A cranked 50watt amp is devastatingly loud to you in your bedroom, but when you factor in walls and distance, to your neighbor in their living room it probably isn't any louder than your lawn mower is. YMMV depending on construction, distance between houses and touchiness of neighbors.
     
    dan40 likes this.
  14. dan40

    dan40 Friend of Leo's

    Posts:
    2,041
    Joined:
    Aug 19, 2015
    Location:
    Richmond Va
    I also wanted to suggest that if you are set on having a Marshall Plexi type sound with volumes set on 10, you may want to convert the circuit over the lead spec. If you built to the JTM 50 schematic, your preamp is passing a ton of low end to the power amp and driving the snot out of the OT with the channels jumped and volumes maxed. With the OT saturated and then squashed down through an attenuator, you may be hearing artifacts of that through your speaker cab. Marshall slowly converted these circuits over to lead spec to cut a lot of the low end from the preamp and let a lot more highs pass through. These few changes let the amps achieve a much crunchier sound and also tightened up the low end. Playing with the NFB circuit may help you a bit. If you are on the 8 ohm tap now, switch the wire to the 16 ohm and see if it helps with the issue you are having. If you can't switch to a higher tap you can always reduce the NFB resistor value down. The extra NFB will help to tighten up the power amp a bit and may help to alleviate some of the issue.
     
  15. itsGiusto

    itsGiusto Tele-Holic

    Posts:
    616
    Joined:
    Dec 31, 2017
    Location:
    Massachusetts
    I think I have a pair of Mullards lying around. I saw the datasheet seems to indicate that it's good up to 450 to 470v.

    Nah, I mostly just want to make sure that it's period correct, and sounds like its supposed to. I've never been into MV amps, which is why an attenuator is my go-to most of the time.

    I'm not set on getting the "Plexi on 10" sound, I just want to be able to play it at the point where you have the ability to get crunch if you roll up your volume knob, and get clean if you roll it down. That's where I've always played amps. It's just that for this amp, it seems like that point is, like, 8, and then I get those ghost notes. I'm definitely going to look into what @_Steve was saying that maybe there's some problem causing me to not get crunch as early as I should. The next opportunity I get, I'll get pictures of the oscilloscope trace and compare to his pictures.

    That said, I am interested in the "lead-spec" sound too. If I ever get this thing to work, I wanted to try out adding switches to split the cathode and swap to lead tonestack.

    Which circuit variations are you specifically thinking of? Like I mentioned above, I want to put the lead tonestack and split cathode on a switch. I know Hendrix used to get his plexi amps of this spec modded to the lead tonestack, so maybe that really helps in achieving higher gain?

    One mod which I'm less likely to be able to put on a switch (and still maintain good lead dress) is changing the post-phase-inverter coupling caps be switchable to be 22nf instead of 100nf. However, I was told that that particular change, while it does affect the amount of bass sent to the power tubes, isn't that drastic of a sonic difference.

    I actually have the NFB tapped from the 16 ohm tap, not the selector switch, so it's constant for different impedance speakers. I experimented with different amounts of NFB, and definitely the problem got worse when there was less NFB.
     
  16. itsGiusto

    itsGiusto Tele-Holic

    Posts:
    616
    Joined:
    Dec 31, 2017
    Location:
    Massachusetts
    I have had the neighbor complain in the past. Our houses are really close together. However, she has been gone for a while during the pandemic, so maybe now would be a good chance to try it out.

    But I also feel like, how could I even turn it up for myself? It's gonna be so loud I'll be in pain from it. Like maybe it'd work if I could play the amp from another room, but I don't really have a set-up that would allow that.
     
  17. _Steve

    _Steve TDPRI Member

    Age:
    40
    Posts:
    82
    Joined:
    Jul 14, 2020
    Location:
    Los Angeles
    Was the NFB on the JTM50 off the 16ohm tap? It's always been on the 8 ohm tap for every plexi-ish amp i've looked at. What sized resistor are you using with that? I think you mentioned before that you have tried with and without the NFB connected right?
     
  18. itsGiusto

    itsGiusto Tele-Holic

    Posts:
    616
    Joined:
    Dec 31, 2017
    Location:
    Massachusetts
    My understanding is that the JTM45 and early EL34 and solid state-rectified plexis had NFB with a 27k resistor off the 16 ohm tap. As @dan40 hinted at, some of them were also done with a 27k resistor tapped off the selector switch, meaning you got variable NFB depending on what impedance the selector switch was on. Of course, as we all know, nothing was 100% consistent back in the day, so there will always be variations therein. But it wasn't until around 1968 that plexis started being released more consistently with either a 47k NFB resistor, or being tapped off the 8ohm tap.

    From RobRob's site:
    https://robrobinette.com/Generic_Tube_Amp_Mods.htm#3-Way_Negative_Feedback_Switch
    Yeah, I've tried both with and without NFB connected. I'm currently using 16 ohm tap, 27k resistor. Definitely gets the least of the ghost-notes this way, compared with other NFB configurations.

    I should revise my previous statement after doing some more testing and listening. 5 does get some light to medium crunch. 8 is where heavy distortion kicks in on mine.

    I'll be sure to take those pictures and measurements shortly!
     
    Last edited: Nov 1, 2020
  19. dan40

    dan40 Friend of Leo's

    Posts:
    2,041
    Joined:
    Aug 19, 2015
    Location:
    Richmond Va
    With a 27k and using the 16 ohm tap you got plenty of NFB working for you. Marshall raised the value to 47k in most amps and up to 100k in the 72 50 watters which lowered the NFB level and made them extra aggressive. Yours should be good with the setup you are running. Raising the resistor value will help the amp to breakup a bit earlier but it will also be a bit looser and may make your issue worse. Cutting some of the low end from the circuit will also help it to hold together better at higher volumes. Lowering all of the coupling caps or just the last two between the PI and power amp will remove some of the lower frequencies and help the OT to cope better.

    You never mentioned what pickups you are using with this. Jimi had to run these amps painfully loud with the singlecoils in his Strats to get the overdriven sound that he was after. A good set of medium output humbuckers should be pushing the amp into a good crunch about halfway up the dial though.
     
  20. Nickfl

    Nickfl Friend of Leo's

    Posts:
    2,399
    Joined:
    May 24, 2016
    Location:
    Florida
    Earplugs. If you're worried about being not able to hear the harsh distortion artifact clearly with earplugs in, just record it too. You can also get "musician earplugs" which are supposed to protect your ears without changing the sound, though in my experience they do make it sound a bit different in actual practice.
     
    dan40 and Lowerleftcoast like this.
IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.