2W High-Gain build into an old SS combo

Discussion in 'Shock Brother's DIY Amps' started by Badside, Feb 28, 2020.

  1. Badside

    Badside Tele-Meister

    Posts:
    433
    Joined:
    Jul 5, 2011
    Location:
    Montreal, Qc, Canada
    So, there's this old Beckemer combo in my basement. It's... not doing well. But it's got a nicely sized (15x9x2.5") thick chassis that would have no trouble holding good iron, and a surprisingly good 10" speaker.

    Intention is to keep the front panel unchanced, meaning working with the existing control layout (still replacing the pots though since they are PCB mounted). It's a 2-channel amp.

    Controls are:
    - Level 1 = Clean volume basically
    - Channel Switch (between Clean and Distortion)
    - Drive = Gain for the Distorted side
    - Level 2 = Volume for the Distorted Side
    - "Contour" = Not sure what it does, but I'll reuse it as a backward Mid control
    - "Bite" switch = makes it more aggressive, so I'll just make it a NFB defeat switch
    - Bass and Treble = Self explanatory

    So therefore I started researching, and I found my inspiration in the Marshall JCM1 (see here: https://guitar-gear.ru/forum/index....attach&attach_rel_module=post&attach_id=40985)

    Except I want one more gain stage than this, because I'm replicating a preamp topology I built last year and that I really like: Marshallesque input stage into non-bypassed warm-biased stage into cold clipper into DC-coupled cathode follower feeding a tone stack with some tweaked values (not as mid strong as Marshall, but not as scooped as Fender).

    So I'm adding a 4th tube which gives me my extra gain stage, and with the extra triode to work with I'm going back to a tube cathode follower instead of the MOSFET. Then a Cathodyne splitter into an ECC99 power amp.

    The NFB is interesting here, it's obtained by wiring the output transformer in reverse polarity, so that you can feed the output back into the input of the Cathodyne. Good old swap the blue and brown wires. This is exactly as Marshall did it.

    The Clean mode will be very early Plexi like. The Distorted mode will be like a JCM800 with a Tubescreamer in front.

    Transformers are chosen based on... I already have them. The OT is a 18W transformer (Hammond 1750PA), so running the 8 ohm speaker into the 4 ohm tap gives me a 16.8K primary impedance, which is right on track.

    Anyway, here's where I'm at so far
     

    Attached Files:

    FenderLover, BB and Ditchgator like this.
  2. Badside

    Badside Tele-Meister

    Posts:
    433
    Joined:
    Jul 5, 2011
    Location:
    Montreal, Qc, Canada
    Working on the layout. Since I'm reusing a preamp topology I've already built, I figured I'd reuse the layout since that amp is dead quiet despite having SLO levels of distortion. It uses a Hoffmanesque approach with B+ and Ground rails at the top. I save space (and money) by using radial filter caps.

    Past the Cathode Follower though, it's a brand new design since that other amp then has a MOSFET FX Loop, a tube recovery stage and LTPI. Instead here, I go straight to a Cathodyne PI.

    So I decided to split this into 3 boards: Preamp, PI and Power amp.
    Cause I'm always changing my mind! This way, I can insert an FX Loop if I so choose, I can also move to EL84 or 6V6 output without ending up with a butchered board like I often do.

    I've added a load resistor and a Slave output (repurposing the Headphone out, and using the Reverb control as a Send Level). My plan for this amp is to not mic it but instead slave it out into my HX Stomp for "post amp" effects and IR cab sim into the PA. This way I can run the power amp as hard as I want and still get clean delays and reverb and a lead boost.

    The layout looks more complicated than it really is because of the channel switch. It's a 3PDT, has to be. First pole (from the right) selects what feeds into the 4th stage: the Level 1 output (taken after the 1st stage), or the output of the 3rd stage (aka, the "Clean" mode bypasses stages 2 and 3). 2nd pole (middle), grounds the input of the 3rd stage when in clean mode (else you'll hear some distorted sound behind your clean, not sure if it's because a 12AX7 envelope is being shared, either way it's quieter when doing this). And 3rd pole bypasses "Level 2" in clean mode (allows you to balance the volume of both channels).

    Again, already built this preamp in a 20W version and it's super quiet so I'm confident in this layout. HOWEVER, that version had relays for channel switching, which allowed me to distribute the switching closer to the affected circuits. That said... if I like this amp, I'll probably end up adding relays anyway (in which case the front panel switch controls the relays in absence of a footswitch). Just want to keep things simple at first.

    This is a work in progress of course. Not a build ready layout!
     

    Attached Files:

    Last edited: Mar 2, 2020
    FRANKENJAG and FenderLover like this.
  3. Badside

    Badside Tele-Meister

    Posts:
    433
    Joined:
    Jul 5, 2011
    Location:
    Montreal, Qc, Canada
    Finally have time to build this.

    I've changed some things around, but this is still "under development" so circuit will get tweaked.

    Installing the turrets on the two boards (preamp and power amp)

    IMG_20200424_161606.jpg

    Populating the boards. Would be cleaner looking with the PSU wires under the board, but I hate not having access to everything once installed. I'll take convenience over better looking guts. This is a small budget project for fun, so these cheap radial caps will do (sooo much cheaper, and they've worked well in other projects so far). Will add a bit of silicone so they don't flop around.

    I want the first stage to have as much voltage as possible (want maximum gain and headroom, later stages are supposed to overdrive) so that's why it has a parallel rail from the Cathode Follower node. Had to do that under then over thing else there wasn't enough room for the 2nd cap. Again, convenience over good looks.

    IMG_20200425_095647.jpg

    View of the donor amp. The reverb tank is shot unfortunately, broken bits inside. Would have been cool to be able to use it but another time

    The power amp board doesn't have PSU caps because I'll be reusing an old dual cap can for the power tube and PI nodes (no screens since it's a dual triode power amp).

    Cathode bias resistor is a place holder, very high value so I can then parallel other resistors to figure out what I really need. There's also no need for a 10W here, but that's what I had. Bypass cap will go in later (to prevent damaging it while messing around with the Rk).

    DC blocking cap for NFB loop also missing, will also play around there to get the right amount of resonance boost.

    IMG_20200425_111215.jpg

    Analyzing options for the layout. The speaker is right against the chassis so have to keep the tubes and transformer back. It is however a 10" speaker in a 15" wide cab so the small PT can go beside it.

    94225283_671548050289399_7469092185242075136_n.jpg
     
    jtcnj and RollingBender like this.
  4. FenderLover

    FenderLover Poster Extraordinaire

    Posts:
    5,397
    Joined:
    Jun 11, 2009
    Location:
    Minnesota
    Cool build. I'm anxious to hear your impressions once it is fired up. There is plenty signal without needing a LTPI. Kudos on the ECC99. People seem to be afraid of using them for some reason.
     
  5. Badside

    Badside Tele-Meister

    Posts:
    433
    Joined:
    Jul 5, 2011
    Location:
    Montreal, Qc, Canada
    Thanks!

    I've never used an ECC99 before, so this will be a first for me. I did build a Firefly 12 years ago with a 12AU7... wasn't a fan (but whole different amp, plus I later got to compare the Hammond 125C OT with the 1750 PA I'm using here and yeah... the 125-series is not great).

    I saw that Marshall was using the ECC99 in the DSL5, I saw it could take over 300V, also saw it's not particularly expensive, so I bought one to try.

    But I'm getting confused... The original DSL5 says "12BH7 or ECC99", the new DSL5C says "12BH7" only. As far as I can tell, only JJ makes the ECC99 and it's not the same thing as a 12BH7 (which they also make), it's significantly taller, takes a lot more voltage and current. However... the max dissipation is the same at 3.5W, but the datasheet linked on TDSL is older and says 5W.

    Basically, I did see reports of ECC99s failing early. Maybe that's cause they were running them at 5W and JJ later realised that was too much?

    Anyway, I figured I can just change to a 12BH7 if it doesn't work. But given I'm working with 16.8k impedance (instead of the usual 22-25k for these), I wanted to get a beefier tube.

    My other option if the ECC99 is a bust, is to go all in and put a pair of EL84s (eh, I'll already have the OT for it!). PT might not like it though, so I'd have to swap that.
     
  6. Badside

    Badside Tele-Meister

    Posts:
    433
    Joined:
    Jul 5, 2011
    Location:
    Montreal, Qc, Canada
    I'm gonna say, as I was wiring this little guy I started having some doubts. So much going on around the channel switch. But it turned out great! No hum, maybe a little more hiss than I'd like. No squealing, even though I forgot the usual small cap from plate to cathode on the 1st stage (common on high gain builds).

    I'm still tweaking things. Circuit is basically a JCM800 with a boost built-in at this point. Except for the tone stack which favors a more scooped tone.

    Anyway, here's a little quick test so far. This is through the original 10" speaker that came with the amp.



    This is what the circuit looks like at the moment

    [​IMG]
     
  7. awasson

    awasson Poster Extraordinaire Gold Supporter

    Age:
    56
    Posts:
    5,427
    Joined:
    Nov 18, 2010
    Location:
    Vancouver
    That sound clip sure is compelling. It sounds great! I was planning on building a small watt amp like a bassman or similar but maybe I’ll follow in your footsteps.
     
  8. Badside

    Badside Tele-Meister

    Posts:
    433
    Joined:
    Jul 5, 2011
    Location:
    Montreal, Qc, Canada
    The higher gain and dual channel aspect makes this a LOT more complicated though. Here's a unfinished gut shot for reference, lots of flying leads.

    There's a few things I could have done better, and I don't like having leads under the board so that makes things messier (I do have some going under the board, but they connect over it). Also, this was a salvage job and the tube sockets are not all the same model, which I realized later and that screwed up my parallel filament wiring plans but... at the end of the day, it works (on first try too)! I was expecting a noisy squealing beast, but even without any snubbing caps (forgot to install them) it's behaving quite nicely! Hum is a non issue even though I never bother with twisting wires for the preamp tubes, DC elevation is all it takes! I do follow Valvewizard's best practice for grounding so it is connected to the chassis only at one end (extreme right, aka close to the input), with the mains ground connection on the left side (you can see it under the power switch). Chassis isn't used for ground returns, nor is there a buss on the back of the pots. Buss is on the boards and laid out to keep every node together (helped by the localized filter caps). The salvaged dual cap can forced me to combine the grounds for the power amp and PI, but that's late enough in the circuit to not be an issue.

    I'll say though, the power supply section (left side) looks particularly bad because I didn't trim any of the wires since I'm not yet 100% sure I'll keep those exact transformers (there's still a chance I move to dual EL84s, just need a new PT and an extra tube socket). Also, the little yellow board is for DC elevation of heaters. I usually do this with a small tag strip but without a heater center tap that was too much going on (had to add the 2 100R resistors for the artificial center tap).

    [​IMG]
     
    thankyouguitar, gigante and awasson like this.
  9. NSB_Chris

    NSB_Chris Tele-Meister

    Age:
    55
    Posts:
    219
    Joined:
    Sep 2, 2017
    Location:
    New Smyrna Beach Florida
    This is a very informative thread! What PT would you select for the EL84 alteration?
    For the EL84 I am guessing we need to keep the B+ low but most of the PT choices have higher secondary voltage ratings. Thoughts?
     
  10. Badside

    Badside Tele-Meister

    Posts:
    433
    Joined:
    Jul 5, 2011
    Location:
    Montreal, Qc, Canada
    So this is where I'm at. "Final"... for now. Anyway, I'm happy and I've reinstalled the chassis in the cabinet.

    Most of the work went into cleaning up the low-end, taming the high-end, and increasing the versatility.

    Treble peaker after V1a was increased to a 1nF cap and 1M resistor. This further reduces the effective gain at low frequency but lets more low mids through. A resistor was added in series with the treble cap on the Gain pot, gives more useable range IMHO (no cap at all sounds way too muddy). Also, did add 2 snubbing caps (V1a and V2a). Not that it needed it, but it cut down a bit on fizz without darkening the tone in a noticeable way.

    But by far the biggest difference was the mid cap. I notice now that on the previous schem I had a 47n mid cap, this was backwards. Was actually a 47n BASS cap and a 22n mid cap. I typically prefer 22 here, and based on Tone Stack Simulator the values were chosen to center the mid dip at around 600Hz, but turns out it's not what the amp wanted. I tried 47n and BAM, all the mud went away.

    Also added a Presence control that is basically a backwards Cut control.

    Anyway. This has very limited volume. High gain preamps require clean power amps to work well, this reaches the limit quickly. It's still enough that my GF complains she can't hear the TV upstairs so I guess that's fine. It sounds exactly the same with the volume down though, so this is great for noodling at home, which was the intention. I'm fairly limited by the cheap MDF cab and 10" speaker, not to mention the baffle attached with small metal brackets. It is what it is. But it's a nice test bed for experimenting with high gain circuits.

    I'll do a proper recording soon.

    [​IMG]
     
    awasson likes this.
  11. Badside

    Badside Tele-Meister

    Posts:
    433
    Joined:
    Jul 5, 2011
    Location:
    Montreal, Qc, Canada
    That is an excellent question, most transformers for EL84 amps seem to assume a tube rectifier so they will put out too much voltage with diodes. Then again, I don't think I've ever seen an amp actually run EL84s at 300V or less (even when deducing the cathode resistor drop).

    So I'd probably get the 290PAZ, which is Hammond's Marshall 18W PT. I'm already using the matching OT (1750PA). Then I'd put a "sag resistor" to compensate for the lack of a tube rectifier (like the Marshall 20W heads).

    Another option would be a 290CAX. A bit more current limited but should be enough, and it starts already 20VDC lower.

    But truth be told, I talked of EL84s because of the limited space here. If I was building this from scratch and not constrainted by an existing chassis and combo cab, I'd go 6V6. In fact, I did just that last year. This is inspired by an amp I've built for a friend last year. Preamp topology is pretty much the same, just voiced differently for the different speaker type, and his has relays for footswitching the channels. Also an FX loop. Killer amp, but much louder than this, harder to use at home. That one uses a Hammond 272DX PT, so just over 400VDC, but JJ 6V6S tubes are rated for 500V.

    Next time I may actually build up the courage to go all the way and build one with 6L6s or EL34s, along with dedicated tone stacks per channel. Took me 10 years to build up the courage to make my first high gain channel switcher :)
     
    awasson likes this.
  12. NSB_Chris

    NSB_Chris Tele-Meister

    Age:
    55
    Posts:
    219
    Joined:
    Sep 2, 2017
    Location:
    New Smyrna Beach Florida
    Thanks! Really interested to see where you go with this. I built a two channel JCM800 preamp clone on a 6V6 power amp section based on Rob Robinette's circuit design and loved it. I am currently building a similar amp with a tube driven effects loop to experiment with that. I don't really get these 20+Watt amps into their power amp distortion zone, so was considering for my next amp trying a single channel version on a EL84 power amp section to see the difference thinking I might at least be closer to power amp contribution. Adding an additional gain stage like you have done and maybe an additional gain shaping pot could be fun to experiment with. If there is nothing to gain from the EL84 or the ECC99 I will stick with the 6V6 as that is pretty forgiving.
     
  13. NSB_Chris

    NSB_Chris Tele-Meister

    Age:
    55
    Posts:
    219
    Joined:
    Sep 2, 2017
    Location:
    New Smyrna Beach Florida
    Similar observations! The data sheets that I have say the max for EL84's is 300V and that just does not look like what is achieved by commercial or DIY amp schematics. I have been struggling to arrive at at an EL34 platform with SS rectification that I have confidence in. It has to be as it seems that almost every modern lower power tube amp these days has EL84's.
     
  14. awasson

    awasson Poster Extraordinaire Gold Supporter

    Age:
    56
    Posts:
    5,427
    Joined:
    Nov 18, 2010
    Location:
    Vancouver
    I built a little clone of the Orange Tiny Terror amp which has EL84’s. It’s very loud. I run it into a pair of 12” Mesa’s.

    I used a Hammond transformer and used a silicon full wave bridge rectifier instead of a pair of diodes. I think B+ ended being at ~260 volts, give or take 5 volts. It should have been less. I put a big choke on it and it tamed it slightly.

    I used a Hammond 261G6 power transformer with a Hammond 159P 10 Henry choke and a Hammond 1650E output transformer. It’s got ultralinear taps which I switched with a relay and it did round out the sound a bit but I just don’t use it so I’ll probably pull that out.
     
    NSB_Chris likes this.
  15. radiocaster

    radiocaster Poster Extraordinaire

    Posts:
    6,869
    Joined:
    Aug 18, 2015
    Location:
    europe
    Reminds me of the Mesa Dual Rectifier preamp, same number of triodes + cathode follower.

    I would maybe have used a triode less than you did for the clean channel.
     
  16. drneilmb

    drneilmb Tele-Meister

    Age:
    113
    Posts:
    143
    Joined:
    Nov 30, 2019
    Location:
    Decorah, Iowa, USA
    I'm considering seriously building this, but I have a question about the schematic. I see SW1b used twice, once to ground out the input of V2 and again to switch Level 2 in and out of the circuit. Did you actually use a 3PDT switch and one of those should be SW1c? Or is one of those switches vestigial and you either left Level 2 in circuit all the time or lived with some noise into V2?

    Can I ask more niggly schematic questions as I come up with them? :)

    -Neil N0FN
     
  17. Badside

    Badside Tele-Meister

    Posts:
    433
    Joined:
    Jul 5, 2011
    Location:
    Montreal, Qc, Canada
    One of my main gigging amps is a 6V6 JCM800 (2204 circuit) with a tube FX loop. Great piece of kit! (My other main gigging amp is a dual EL34 2204, with an SS loop from Metro)

    A JCM800 type preamp driven into power amp overdrive is in fact very fun, but this isn't really it. This one is all about preamp gain. I mean, it can be driven into power amp overdrive, but it doesn't sound that good.
    If you wanted that sound, then RobRob's micro JCM800 might be better because it has the proper PI (at the cost of a gain stage, that's the compromise I had to make here). Also, a PT with a higher voltage but less current handling would behave closer to the real thing (I already had the 269GX, else I would have gotten a 269JX). Also smaller filter caps (again, salvage job so dual 50uF, but in my case that works since I wanted a tight low end).

    Still, overdriving small dual triode power amps will never be quite the same since they're not pentodes. A pair of EL84s is still loud AF. A single one in single ended is more tolerable (but still loud), but that doesn't quite overdrive the same as a push pull section with a phase inverter.

    One semi successful attempt I've had in the past was a pair of 6K6GTs. They are to a 6V6 what a 6V6 is to a 6L6, if you will. Fun fact, I was running them exactly off this set of transformer, although they would prefer a load around 12K.
    They're long of production but they're cheap on the NOS market. They're good for 315V which is what I'm getting here. With a shared 400ohm cathode resistor and a 12k load you're getting just shy of 10W. It was a cool toy but I found it still too loud for home, but occasionally not enough for the stage. Might have been better with the proper load (a Deluxe Reverb OT ran a tap lower might do the trick)
     
    awasson likes this.
  18. Badside

    Badside Tele-Meister

    Posts:
    433
    Joined:
    Jul 5, 2011
    Location:
    Montreal, Qc, Canada
    This ain't no coincidence. The Dual Rec is based on the SLO (the difference is mainly in the FX Loop circuit, and the disconnection of NFB in Modern mode), which itself is basically a JCM800 2203 with an extra gain stage. Like having your overdrive pedal built into the amp. Those Marshall "AFD" circuits are also basically this, except they have the cold clipper as the 2nd stage instead of the 3rd.

    You can see this is a micro Dual Rec, but tweaked to work better into the cheap 10" speaker that came with the salvaged combo. One difference is my 10k Rk on the cold clipper, but I may revert back to 39k (I switched to 10k when I wasn't bypassing the 2nd stage's cathode, but now that I am it's just a ridiculous amount of gain). Also, like I mentionned the Dual Rec doesn't use NFB. I may actually put NFB on a switch for fun, but I am using a small cap in the NFB loop to get the low end resonance of a Mesa.

    I actually tried this circuit with one triode less on the clean channel, didn't quite enjoy it. This approach is more Plexi like. Of course there are myriads of ways to do this.
     
    radiocaster likes this.
  19. Badside

    Badside Tele-Meister

    Posts:
    433
    Joined:
    Jul 5, 2011
    Location:
    Montreal, Qc, Canada
    Ah yes sorry, that's a 3PDT and one should read SW1c!

    Ask away!
     
    awasson and drneilmb like this.
  20. drneilmb

    drneilmb Tele-Meister

    Age:
    113
    Posts:
    143
    Joined:
    Nov 30, 2019
    Location:
    Decorah, Iowa, USA
    Okay, I'm working up a BOM for myself and I like the 269JX PT that you mentioned and the 125B OT for infinite flexibility and the 22.5k:8 ratio that Rob Robinette used with his PP 12AU7s.

    Looking at the schematic, I wonder if you noted which places you what power rating resistors you used. I typically use 1/2W jobbies by default, but I imagine that a couple of the cathodes and the droppers might need to be bigger than that. I see the 5W and 2W annotations on R31 and R37, are you using 1W resistors everywhere else, or mixing smaller resistors some places?

    -Neil
     
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.