A 6V6 Plexi with quite a few twists, and videos!

[SnickSound]

Taking a chance sharing this here, hopefully some people find it interesting!

I took the 6V6 Plexi concept "further" by adding switchable cathode circuits in the preamp (so I can go between JTM45 split cathode to a more Superlead sound), a Cascade mode with optional cold bias on 2nd stage (we'll call this the JCM800 mode), a fully bypassable Tube FX Loop that also adds a Master Volume, a cathode biased power amp (should make around 15W-ish, perfect to crank in bars) with a trick half-power option, and tube rectification because why not.

All built from scratch on a turret board.

I'm no professional and my amps tend to look like crap, but they always sound amazing :)

Schematic: https://www.youtube.com/watch?v=f3sjo6LCrSE&list=PLqC_WjyYCKZHSq071HgWoueuDdyoHHHWm&index=1
Layout: https://www.youtube.com/watch?v=IkQ9Km3o6E8&list=PLqC_WjyYCKZHSq071HgWoueuDdyoHHHWm&index=2
Making the turret board: https://www.youtube.com/watch?v=gv-lEHHwVxo&index=3&list=PLqC_WjyYCKZHSq071HgWoueuDdyoHHHWm
Populating the board: https://www.youtube.com/watch?v=gwDj1PdzChE&list=PLqC_WjyYCKZHSq071HgWoueuDdyoHHHWm&index=4
Drilling the front panel: https://www.youtube.com/watch?v=bAp6MOxxCYY&index=5&list=PLqC_WjyYCKZHSq071HgWoueuDdyoHHHWm
Drilling the rest of the chassis part 1: https://www.youtube.com/watch?v=TMVye5ZxeMU&list=PLqC_WjyYCKZHSq071HgWoueuDdyoHHHWm&index=6

I'm "vlogging" this not to show people how to do it, but rather to show that it's not rocket science if you take your time and do your research first. And also that it requires fairly simple and common tools. Basically, to inspire those who might be on the fence about building an amp.

Now I'm going back to the basement, going to try the headphone trick for OT placement :)

[silverfox]

Count me in and interested in the schematic...

silverfox.

[tubenit]

Thanks for sharing your innovative idea!  Looking forward to seeing how this turns out for you.

I did something sorta similar (at some level) a few yrs ago using Marshallish topology (not component values)  with 6V6 cathode biased & an active FX. However, I added a PPIMV.  It is played regularly by different guitarists and I've heard some pretty nice compliments on the amp's tone.  I also added a switchable overdrive channel.

Here's a copy of what I did with this one.

With respect, Tubenit

[12AX7]

I'm really curious about several details in that tone stack, but i won't ask about them all because it's too much. But i WOULD like to ask one thing.....what does that 500pf across the bass pot do?

Thanks for sharing your innovative idea!  Looking forward to seeing how this turns out for you.

I did something sorta similar (at some level) a few yrs ago using Marshallish topology (not component values)  with 6V6 cathode biased & an active FX. However, I added a PPIMV.  It is played regularly by different guitarists and I've heard some pretty nice compliments on the amp's tone.  I also added a switchable overdrive channel.

Here's a copy of what I did with this one.

With respect, Tubenit

[Willabe]

I'm really curious about several details in that tone stack, but i won't ask about them all because it's too much.

If you feel it's too much for this thread, then just start a new thread with your questions on that TS.   

[pompeiisneaks]

Snick, Hey I'm from reddit, pompeiisneaks there too... I dig the videos, been watching them as you go.  Interesting, and I'm learning, I still know very little about the marshall style/topology, keep them coming!

~Phil

[punkykatt]

I also would like a schematic.  Thank you.

[SnickSound]

I also would like a schematic.  Thank you.

I'm still realizing little mistakes (or required tweaks) as I go along, like some resistors to prevent the half-power switch from popping.
As soon as I feel confident sharing, I will definitely do.

In the meantime, here I go doing an amateur job of drilling the chassis (live and learn!)

!Ep 6

!Ep 7

[SnickSound]

I went ahead and made the work-in-progress schematic available on the project page if anyone's curious:
http://www.snicksound.com/projects/6v6ultimateplexi/

This is yet untested, so don't blame me if you attempt it and it doesn't work :D



And a few new videos are up too :)

https://www.youtube.com/watch?v=qefdJF8--1Q&t=39s&index=8&list=PLqC_WjyYCKZHSq071HgWoueuDdyoHHHWm

https://youtu.be/1uIhcfrc924

https://www.youtube.com/watch?v=ihED629rZc8&index=10&list=PLqC_WjyYCKZHSq071HgWoueuDdyoHHHWm

https://www.youtube.com/watch?v=qGMbitu2_0g&index=11&list=PLqC_WjyYCKZHSq071HgWoueuDdyoHHHWm

[MFowler]

After building over 125 amps so far I still laugh at all the mistakes I've made building amps which are the same ones you note in your videos so bravo to you for showing everyone what we encounter building our amps.


Since step bits in metric are hard to find and SAE step bits are too large a hole for Belden sockets I use 18.6mm hole punch for the Belden 9 pin and 30mm for the Belden 8 pin sockets, much easier.  I use the step bits for the through holes for transformer wire and use plastic through hole protectors instead of rubber grommets better for me.


I prefer keps nuts instead of nylock nuts but that is just my preference.


Have fun that's what this is all about.


Mark

[SnickSound]

After building over 125 amps so far I still laugh at all the mistakes I've made building amps which are the same ones you note in your videos so bravo to you for showing everyone what we encounter building our amps.

Since step bits in metric are hard to find and SAE step bits are too large a hole for Belden sockets I use 18.6mm hole punch for the Belden 9 pin and 30mm for the Belden 8 pin sockets, much easier.  I use the step bits for the through holes for transformer wire and use plastic through hole protectors instead of rubber grommets better for me.

I prefer keps nuts instead of nylock nuts but that is just my preference.

This is my... 9th build, and every build has its own set of challenges. But there's also always small little details that "I should know better". I think right now what's working against me is that I've built just enough amp that I'm starting to do some parts by habit, like in this case wiring the two input jacks straight to the tube socket, completely forgetting I had a Cascade switch to go through first.

And since the last video, I've been wiring the front panels and now I'm doubting some of my decisions (there's a LOT of flying leads), so I might have to re-evaluate things a little. I had spent a lot of times during design and conception to bring it down to just 4 preamp switches (not counting FX loop) that would give me the whole "Marshall hall of fame", but it's starting to look like a rat's nest. So I'll probably keep it simple at first and build upon a solid foundation. Using relays to keep the switching points closer to the circuit they're switching might have been a better idea, might even have allowed footswitching between the modes. Perhaps in a future iteration.

Here's the funny story about the KEPS vs nylon lock nuts: I was certain I was putting KEPS nuts in my bag at the hardware store (they sell these in bulk), turns out they were nylon lock nuts. Having used nylon lock nuts with no issues in the past, I went ahead with those. I did use the KEPS for the long screws holding the board though, because a nylon lock nut would have lost all it's "power" by the time it got to the bottom.

Did all the local hardware store and was never able to find KEPS of nylon in 4-40 size though, so Loc-Tite it is. But I'll be ordering a bunch of online.

[SnickSound]

Preview of how it sounds right now :)

https://www.youtube.com/watch?v=tF9a8yhygy4

[Shack]

Awesome thread, and sounds good.....thank you for sharing

[SnickSound]

Did some tweaking this morning, trying to get rid of the mud, fix the FX loop volume drop, get more clean range

https://www.youtube.com/watch?v=Jn6mz9ROHDY

[SnickSound]

I think it's starting to sound good, I'm also starting to realize that a FX Loop is kind of pointless on such an amp... Have to run the power amp clean for it to be useful, which defeats the purpose of a low wattage design. Might have to build a dry/wet rig after all.

[silverfox]

Overall this has been a great project. You've got the video portion included and that alone takes quite a bit of work. In regards to the amp: At time it sounds very good and other times it still sounds muddy, to me. I suspect this may be due to a mismatch between the preamp and power amp section. I personally have had this problem so it sounds at times like that type of mismatch. Turning down the MV seems to bring the stages back to a match. Other than that, great job. At the start of the series you mention, not an expert. Don't sell yourself short. You're doing a great job.

Have you posted the final schematic? Thanks for the contributions,

silverfox.

[SnickSound]

A little update :)

Video Demo

Schematic:
http://www.snicksound.com/projects/6v6ultimateplexi/#schematic

(Previously linked to schematic directly but there was an error in it, it's safer to link to the project page so you always get the most up to date version)

[fixitz]

Here's a dumb question but something I've never been able to totally pin down. What is it exactly that makes a Plexi a Plexi, as opposed to any other circuit? What's special in the circuit that defines Plexi?

[tubenit]

THANKS for sharing your success and your adventure in amp building!  The amp sound really great.

With respect, Tubenit

[SnickSound]

Here's a dumb question but something I've never been able to totally pin down. What is it exactly that makes a Plexi a Plexi, as opposed to any other circuit? What's special in the circuit that defines Plexi?

This is a great question, as Plexi refers to a family of amps more than an actual amp. Of course, the actual definition is a Marshall with a plexiglass front panel (hence the name), but symbolically it refers to Marshall's modification to the Tweed Bassman (5F6a) circuit.

So, while the JTM45 was almost a perfect clone of a 5F6a (save for using different tubes and having more negative feedback), what we call a Plexi (the 1959 and 1987 circuits, 100W and 50W respectively) typically had split cathodes (on the JTM45 and 5F6a, both input stages share a cathode resistor and cap) with a bigger resistor and smaller cap on the lead channel cathode, a cap on the mixer resistor and a bigger cap (500pF vs 100pF) on the Bright volume. This makes the Normal and Bright (now Lead) channels much more different sounding, and allows getting a lot of different textures by mixing them together in various amount. The bright channel has a lot more high mids that gives that signature Marshall grind. And negative feedback was reduced compared to JTM45 (also tone stack lets more mids through)
Also, the 1959 and 1987 are diode rectified.

Mine has some JTM45 bits though: it's tube rectified, has the smaller mixing resistors, higher negative feedback, JTM45 spec tone stack, and only a 47pF treble cap on the Bright volume. Via the 3 front panel switches though I can make it pretty aggressive.

The 3 switches are (from right to left)
- Cascade: Sends the Bright channel's output into the Normal channel
- Bright: bottom position = 47pF on the volume, middle position = add a JCM800 treble peaking circuit (470k bypassed by 470pF) before the volume, top position = add a 510pF cap on the volume.
- Cold bias: bottom position = normal Plexi, middle position = normal channel now has a 8k2 cathode resistor and no cap (for use with cascade mode for JCM800 config), top position = same as middle but adds a 500pF treble cap on volume knob (to simulate fixed voltage divider of JCM800)

EDIT: For those who may not be familiar with the 5F6a, here's what I "think" makes a Plexi sound the way a Plexi does:
- As already mentionned, the Lead channel has a LOT of high-mids boosting built in. By mixing it with the fat full-range Normal channel, you get very interesting textures
- The DC-coupled cathode follower stage (a defining feature of the Tweed Bassman) compressed the signal in a very warm way (increases 2nd order harmonics) due to its inherently flawed nature. It warms up what would otherwise be a fairly harsh amp
- The Bass-Middle-Treble tone stack has a specific response that is part of the sound. That said, there are various iterations of it, but in general the Marshall version has a moderate mid-scoop and lots of emphasis on the highs (for that "grind" sound)
- Long-Tail Phase Inverter has a lot of available swing that allows driving the power amp very hard

Of course, a big part of the Marshall sound comes from playing it through a closed back cab loaded with Celestion drivers. In the video, that's an old UK made V30 in a sealed 1X12 cab.

[pompeiisneaks]

As I mentioned on the youtube video comments, I really dig this sound, I need to build a plexi soon :)

[fixitz]

Appreciate the info SnickSound. That was a pretty detailed explanation, beforehand if someone had asked me what I thought defined a Plexi I probably would have just gone with the tone stack more than anything else. I'm glad nobody asked me.... ;-)

[kagliostro]

Thanks for Sharing

The amp really sounds awesome

Is you project exactly reproducible following the schematic you posted and the layout on the video ?

Can you share the Layout ?

Franco

[SnickSound]

Thanks for Sharing

The amp really sounds awesome

Is you project exactly reproducible following the schematic you posted and the layout on the video ?

Can you share the Layout ?

Franco

Hi Franco,

In theory anyone could build this amp based on the schematic on my website (which I've updated: http://www.snicksound.com/projects/6v6ultimateplexi/#schematic)

At it's core it's a Marshall 1987 circuit, with a tube rectifier thrown in and the fixed bias supply replaced with a big resistor and cap. The front panel mini switches are a b**** to wire though, and some shielded wire has to be used because it gets quite sensitive in cascaded mode. I'm feeding both input grids with shielded wire and that seems to be enough, also have a small cap across the bright channel's plate load.

I will eventually share the layout yes, just have a lot on my plate right now. I might forego the FX Loop though because it's just not a good implementation at the moment. If you get it to attenuate enough to not overdrive pedals, then you don't have enough make up gain to drive the power amp hard (which in some way is not an issue since driving the power amp hard renders the loop kind of pointless, but that's another discussion). So probably easier to build it sans loop, and if needed add like a Metro Lossless loop kit between the treble wiper and the master volume, as I've done on my other amps. Sure, these are silicon based, but so are the pedals you're putting in the loop, so hardly matters.

Or maybe I'll share both schematics and layout with and without loop and people can decide for themselves what to do with those :)

The great thing is that the Marshall Plexi circuit is a platform that can be taken in all sorts of directions! Mark Huss already did the preliminary work to make it into a 20-25W 6V6 powered version, I just went further and made that tube rectified and cathode biased, then added switches to go from old school to JCM800. 15W is still a lot though, but you can swap in a 5Y3GT rectifier and switch to 6K6GT power tubes (they like to see a 10-12K primary though) with a 400+ ohm cathode resistor and bam you're making single-digit power now.

In my case, a 5Y3GT also makes it low enough to run EL84s with the right sockets and cathode resistor. 5Y3 is rated for only 20uF though, and mine runs 32uF but I do have series resistance (those two 47R resistors before the rectifier) which protect the rectifier tube.

[kagliostro]

Many thanks for the answer and all the interesting info

Some 5Y3 datasheet show exactly 32uF as first cap max value, so you are on the ballpark with it (other people use the 32uF value with that tube and without issues)

5Y3 data


About the FXLoop, can I permit to me to suggest to give a try this one ?


Note that the pot is a single shaft pot and the connections are reversed each other so when one signal decreases the other increases


I and a friend planned this 5E3 using that FXLoop with success

click on the image to see in full resolution (or save it on your PC)


Franco

[SnickSound]

This is pretty much the loop I used except without the parallel option (if your pedals can't go 100% wet it can cause phasing issues). Also, I went from 2k2 to 10k on the first tube's cathode. It's really subtle, but somehow I like it better this way (it technically allows more voltage swing but also limits gain so it's not that big of a change).

The loop is fine for my use, just have to remember you can't drive the power amp as hard while using it (which would increase distortion post-loop, making the loop sort of pointless anyway), but I would hate for someone else to build this amp and realize this too late

[kagliostro]

I don't remember the reason at all, but I don't think that R19 (on your schematic) has an adequate value, as I say, didn't remember why, but 1M seems to me too high (10k - 25k will be a better choiche)

Also R29 and R30 values seems to require a bit of investigation, did you experimented very much with that values ?

Franco

[SnickSound]

I don't remember the reason at all, but I don't think that R19 (on your schematic) has an adequate value, as I say, didn't remember why, but 1M seems to me too high (10k - 25k will be a better choiche)

Also R29 and R30 values seems to require a bit of investigation, did you experimented very much with that values ?

Franco

You're looking at the untested preliminary schematic, I'll update that post to link to the project page where you can always find the most up to date schematic:
http://www.snicksound.com/projects/6v6ultimateplexi/#schematic
(Just realized there are two R29s on that one, will fix that)

But to answer your questions:
R19 is basically unnecessary, it just gives a path to ground for the capacitor before it if both Send and Return jacks were left open (e.g. leads connected but nothing on the other side). In such a scenario, it can prevent a pop when you actually connect the other side of the cables.

R29 and R30 (voltage divider to attenuate before the loop) I did experiment with, at some point I was trying to clip the loop driver triode to get some more preamp drive so I removed R29 altogether and… it didn't clip. So I have it this way right now, no attenuation before the loop. R30 is currently a 2M2 but I will lower that to a 470k to keep the load post-tone stack the same when engaging the loop (since the MV is now always on and is a 500k pot)

Without R29, it's possible to turn the Send control up (that too is not on the old schematic) and drive the amp even harder but... it doesn't sound very good (probably just clips the return stage more than anything).

[kagliostro]

OK, I was looking to the posted schematic, the old one ....

Looking to the new link, be patient, don't beat me, I would try a 10k pot in place of that 100k pot (VR6)

Franco

[SnickSound]

OK, I was looking to the posted schematic, the old one ....

Looking to the new link, be patient, don't beat me, I would try a 10k pot in place of that 100k pot (VR6)

Franco

Hehe, no beating! I'm always open to suggestions from internet friends.

Come to think of it, it's quite possible I already have a 10k in there, I'll have to check.

If I remember correctly, this is about keeping the output impedance of the driving circuit low, a noble cause. That said, my thinking so far was that since I'm driving pedals anyway (which typically have at least a 470k input impedance), it wasn't that much of an issue. But it's an easy thing to try indeed.

[kagliostro]

Yes, impedance is the concern

I don't know why, but this idea come to me

  

Franco