HoSo56 amp project...design input sought

[Geezer]

I have a new project that I need help with (LOTS of help)  :-*
I have lots of "clean" amps, but not any that really do the overdriven thing....at least not like the following linked video!!  :-[

After watching THIS video of Peter Stroud playing one of his 65Amps "Soho" 18w'ers, I really want to build one!

Of course, I (we  ;) ) must figure out what in the heck it is!

From the "clues" that Peter gave in the demo, I have drawn my "best guess" at what the topology of the amp could be.......see the drawing below. I have it in .gif (for easy viewing) and in .sch form (for others to make corrections/add their input).
I'm just doing a basic "block diagram" right now. I'll work out a preliminary schematic later on......

#1 - EF86 Pentode is the input stage (a no-brainer)
#2 - Blackface Tonestack(??) ....I get this idea from Peter's comment, "we wanted to get that sort-of dipped midrange that allows for a cleaner tone". I don't think a 5F6-A/JTM45 "Tweed" TMB would give that mid-dip(?)
#3 - He states that the Tone circuit is "12AX7 driven"....I supect the 12AX7 is AFTER the tonestack (as shown). Wouldn't a pentode, into a triode, then into the tonestack be too much gain before the stack?? I envision it like the AB763 BF circuit....pentode, then tonestack, then a triode "recovery" stage before the PI.
#4 - The "BUMP" switch/circuit....I'm thinking it bypasses the TMB tonestack, and inserts a 5E3 style Volume/tone circuit....what do you think?

So my main questions (to start) are....
What tonestack?
Where should the tonestack be placed?
Where do the 12AX7 triode stage(s) fit into the picture?
Does the "Bump" circuit seem resonable as I've envisioned it?

This is just a start.....I'll worry about figuring out that really great sounding Master Volume later!

All comments & help encouraged & appreciated!!!

Geez'r

[LooseChange]

I think you nailed it pretty well.  But why the 6 step volume control on the bump channel.  Why not just a pot?  Something special going on?

[bigsbybender]

Does Dan Boul still post over to 18watt.com?   He'd surely know (although maybe he won't give it all up since he does make a living making and selling those things)..... I haven't been to 18watt since the forum "cleansing" but he was a regular contributor, he seemed like a nice guy on the internet.


j. 

[Geezer]

why the 6 step volume control on the bump channel.  Why not just a pot?  Something special going on?

I wondered the same thing. Maybe they just wanted to make it a "boost" ("Bump" up) and not have the possiblility of actually reducing the volume, so they limited the scope of the control so that the range is only the same amount or more volume....less volume is not possible. A selector switch would make that fairly easy to accomplish....Just a guess(?)

Does Dan Boul still post over to 18watt.com?   He'd surely know (although maybe he won't give it all up since he does make a living making and selling those things)..... I haven't been to 18watt since the forum "cleansing" but he was a regular contributor, he seemed like a nice guy on the internet.

I've already gone over to 18watt & looked around. Dan readily acknowledges that he learned pretty much everything he knows from thoses guys. I even found the post where he announces the official incorporation of 65Amps! But it seems any posts he made about real innovations or ideas are gone with the "cleansing"  :-X .
I'll look around some more (@18w) & maybe even pose some of the questions over there.

Thanks for the responses!  G

[tubenit]

DaGeezer,

If you're looking for an idea with a boost accenting mids ................... you might consider this?

I used a heavy duty full sized DPDT when switching tone stacks. However this accomplishes a similar feature with a PAB boost that adds a cap when engaged to allow more mids thru. Plus the mid pot on this can get either mid scooped or a pretty full mid tone.

I added a PAB to the Little Wing Reverb and another mini-toggle SPDT (center off) to switch a 820p or .002 cap (paralleling the treble cap).  With either of those caps paralleled and the PAB on ......... it gets some excellent overdrive!

The thing with this is that you can use a mini-toggle DPDT.

With respect, Tubenit

[SirElwood]

Can you really drive a "standard" tonestack after EF86? I believe there was some "pentode high impedance" thing involved in this.

[LooseChange]

It's funny how quickly we can all come up with what might be the circuit.  We are all such a talented bunch of folks. :)

[Blind Lemon]

Can you really drive a "standard" tonestack after EF86? I believe there was some "pentode high impedance" thing involved in this.

Thats just what I was thinking....you got an extra triode why not use it for a cathode follower before the TS. Or use a 6U8A (pentode/triode) for the gain/cathode section on the front and parellel the 12AX7 after the stack.

my 2 cents

BL

[6G6]

Blind Lemon had teh same thought I did and posted as soon as I open the reply box!
I think the EF86 would have enough gain to drive the tone stack, but it might benefit
from a CF between it and the TS.
Maybe that is what he meant by "the TS is driven by a 12AX7".

[Heinz]

A lot of their ads (e.g. http://radicaltone.com/65_amps_soho.php) say: "65 - SoHo 20 WATT (2xEL84), Tube rectified, EF86-12AX7 Cathode Follower with BumpTM circuit and Defeatable Master Volume"
That's where the unemployed triode works ...

The tube line-up is 1xEF86, 2x12AX7, 2xEL84 (http://www.proguitarshop.com/index.php?CategoryID=316) so it must be close to tubenit's sketch. The bump control obviously switches volume/level and tone at the same time, so there is probably some cap switching involves, which explains why they didn't use a pot.

There's another demo of the SOHO on youtube (http://de.youtube.com/watch?v=4vqk9AOOjdg) that might be of interest. Maybe someone with a trained ear can figure out more details by listening.

[Blind Lemon]

"Exactly" 6G6. I went over and listened/watched the clip (very nice BTW) and for some reason I get the impression the "Bump Circuit" is not a parallel TS. :-\
BL

[topbrent]

The Dr Z Route 66 has a ef86 with treble and bass controls, ltpi.

[Heinz]

I may be wrong, but I would guess that a triode gain stage after the EF86 would make it very difficult to produce a clean sound, due to the high output level of the pentode. You would either need a voltage divider or couple the cathode follower directly to the EF86 and use the second triode after the tone stack, so the topology would be:
EF86 - 12ax7/CF - TS - 12ax7/gain - PI - power section

Could the bump circuit then be a switchable cathode resistor/cap-combination for that post-TS triode gain stage (apart from disengaging the TS)?

[macula56]

I have been thinking about building something like this in a VJ chassis. I was either going to use da Geezer's schemo for converting a VJ to 5E7 with 2 EL84's or Tubenit's Nightclub Guy schemo as a starting point. This looks like it has elements of both. The EF86 frontend and the CT. This looks like a very cool little amp with lots of tonal possibilities. Thanks guys, as always. Jmac

[Dynaflow]

 Here's they're PR on the amp on they're site, perhaps give a few more clues:

65 – SoHo (20 watts) – Designed to be a super-functional amp for working players and enthusiasts who desire widely varied tones in one amp. The SoHo can transition seamlessly from the best of American tones to the sweetest British tones via the “Bump” function and NOT switching channels. The Bump will radically shift the tone-stack functionality achieving astonishing renditions of dozens of excellent, natural sounds. In addition to the “Bump” function, there is a defeatable Master Volume that is revolutionary in that is slopes the tone stack as you turn the Master down, giving the Master a very natural sound and ultra-low volumes. The Master can be turned on for “Home” mode and completely defeated for “Gig” mode. This amp is a must for studio players, small club gigs and enthusiasts who need to play at reduced volumes.

Regards,

Dyna

[SirElwood]

Could the bump circuit then be a switchable cathode resistor/cap-combination for that post-TS triode gain stage (apart from disengaging the TS)?

This is what I was thinking... Something like tubenits fat/full/bright switch, but with more options/different values.

[Geezer]

Could the bump circuit then be a switchable cathode resistor/cap-combination for that post-TS triode gain stage (apart from disengaging the TS)?


This is what I was thinking... Something like tubenits fat/full/bright switch, but with more options/different values

Except.....Peter states in the video that the Bump feature disables the TB tonestack...... :-\

[Geezer]

I may be wrong, but I would guess that a triode gain stage after the EF86 would make it very difficult to produce a clean sound, due to the high output level of the pentode. You would either need a voltage divider or couple the cathode follower directly to the EF86 and use the second triode after the tone stack, so the topology would be:
EF86 - 12ax7/CF - TS - 12ax7/gain - PI - power section

I'm thinking Heinz has hit pretty close to it here....(?)

[tubenit]

couple the cathode follower directly to the EF86 and use the second triode after the tone stack

What would that look like?  How would you couple a 12Ax7 triode to a EF-86???

What would you change in this schematic to make that work??

With respect, Tubenit

[Geezer]

What would you change in this schematic to make that work??

Nothing.....that sure looks "direct coupled" to me....

May someone else will say otherwise(?)

[sluckey]

There's a lot of dc on the cathode of that cathode follower. Need to block that dc.


EDIT: I changed the GIF/SCH schematic  (2 posts prior) to match Sluckey's critique. Thanks, Sluckey!

With respect, Tubenit

[Heinz]

I just watched the proguitarshop.com video again and they say that as you back off the bump level you have less bass and less gain. That sounds like a switchable cathode bypass cap to me.

The bump tone seems to be a simple treble filter, but I just can't figure out how that could be connected without needing another switch. Anyone have an idea?

EDIT: the volume pot is in the wrong place. It remains effective even with the bump active, so it must be right before the triode gain stage's grid (after the left side of the switch)
EDIT#2: also forgot the coupling cap sluckey mentioned

[Heinz]

Ok, here's the correction of the previous mistakes.

Peter Stroud also talk about a special master volume that preserves the tone as the volume is turned down. As the tone tends to get fizzy at lower volumes, I guess that this feature is a variable treble filter that rolls off treble in proportion to the volume. This could easily be done with a dual pot as drawn. Will this work as expected?

[tubenit]

Heinz,

GREAT ideas (as usual) ........... keep em coming!  Thanks, Tubenit

[Geezer]

Heinz,

GREAT ideas (as usual) ........... keep em coming!

Yea, Baby!!!   Now we're talkin'!  ;D

I feel another build comin' on!  :P

[SoundmasterG]

EF86 > CF > tone stack is a VERY good idea, and works very well giving good gain and good tone. Of course the devil is in the details. The key is the screen voltage and supply voltage of the EF86.

Greg

[Heinz]

Here comes another design. I think I found a way to connect the 'bump tone'. It's a parallel mini-tonestack that comes into action when the regular ts is switched out. Both ts stay in the signal path but only one becomes effective while it's grounded.

The bump level selector is the same as in the last design. Maybe it's not only switching bypass caps but also parallel cathode resistors. I don't know how that would change the tone and if it would be a step in the right direction.

I feel another build comin' on!

That solves the question who's gonna test all these ideas...

EDIT: the master volume is now switchable

[Geezer]

EF86 > CF > tone stack is a VERY good idea, and works very well giving good gain and good tone. Of course the devil is in the details. The key is the screen voltage and supply voltage of the EF86.

Yes, what I've read is that the "target" ballpark is keeping the screens @ 2/3 voltage of the plates or less. I expect alot of experimentation will be needed (as I'm I total novice when it comes it input pentodes).

I think I found a way to connect the 'bump tone'. It's a parallel mini-tonestack that comes into action when the regular ts is switched out. Both ts stay in the signal path but only one becomes effective while it's grounded.

Heinz, that seems to stay in line with what Peter said in the video.
I think we're getting close to having the "basic" topology in order!

[Heinz]

I expect alot of experimentation will be needed (as I'm I total novice when it comes it input pentodes).

Copying the input stage of an existing amp will probably be a good start. The pentode stage as drawn by tubenit is from a Matchless DC30, for an AC15 the plate resistor is 220k and the screen-to-ground resistor 1M. I have made tests with both configurations (with signal generator and scope) and they work equally well, but I cannot say anything about their tonal qualities. I don't expect much difference, but I'd start with the AC15 values because I like its tone.

The gain of that stage is around 200, so even with the tonestack loss the signal level may be too high for the following triode stage (V2B). I have inserted a 220k resistor before the volume pot to bring down the signal level, but that is just a guess.

Now someone's got to start experimenting and find out if this design actually points into the right direction. I have a power stage module and a brand new pentode preamp module (see http://dhost.info/jschem/t7 if you don't know what I'm talking about) that would make a great start and leave only the tonestack/bump section for prototyping. However, I won't be able to do any experiments this week due to my job workload. I'll keep you informed as soon as I start tinkering, but you'll have to be patient.

[Geezer]

I won't be able to do any experiments this week due to my job workload. I'll keep you informed as soon as I start tinkering, but you'll have to be patient.

Then you'll be well ahead of me...I wouldn't start for several weeks, so patience won't be a problem on my end!  ;)

Sounds like you have the right setup to prototype. Mine will just be a "build it & tweak" type thing, using parts I have around. I have some good quality 50's/60's PT's & OT's (7k to 10k primaries) that came out of hi-fi & PA stuff running 2x6V6's that should work well.

Tubenit has agreed to do a layout, as soon as we have nailed down the topology.

[Heinz]

There you go! Looks good so far and not too complicated. But then...who knows if it actually works...


EDIT:  Heinz was kind enough to edit a layout I posted to correct it. I deleted mine to avoid confusion.

Tubenit

[Geezer]

T,
look @ Heinz post right above yours....he has attached "#9" schem that (I think) fixes your drawing of the MV/Tone...

[tubenit]

Wow!  I missed seeing it.  Thanks Heinz!   

I'll review it and edit anything that needs it and repost it or affirm that it's done and looks right to me.   Maybe I should delete all the other schematics that aren't being used?  What do you think ........ delete or leave em?

With respect, Tubenit

[tubenit]

DaGeezer &  Heinz ,

Take a look at the schematic and layout and let me know if this is a good workable 1st DRAFT for the build?  Feel free to edit and change anything you wish.  I would think Hoffman's 18watt PT & OT could work well with this.

With respect, Tubenit

[tubenit]

Just stumbled across this drawing on an idea to limit EF-86 microphonics. Haven't used it but thought I'd share the idea.  Also saw where some individuals were using shrink tubing on half of the tube to cut down microphonics.

With respect, Tubenit

[tubesornothing]

I have seen a large hi temp O ring used as well.  John at Pacific T.V.sells isolation tube mounts: http://members.shaw.ca/pacifictv/socket.htm#specialp

[Dynaflow]

DaGeezer &  Heinz ,

Take a look at the schematic and layout and let me know if this is a good workable 1st DRAFT for the build?  Feel free to edit and change anything you wish.  I would think Hoffman's 18watt PT & OT could work well with this.

With respect, Tubenit

 Nice layout! So a 21 parallel turret board should work? Cool stuff, I just happen to have a head chassis with three noval sockets... Hmm, love the sound of the ef86, wonder if it sounds similar to Gab's sound garage Voxer?...

 You guys are amazing, from description to design, I'm impressed...

Regards,

Dyna

[ACDCG400]

weber has some bases that are "floating" and are suspended by strings.

i love that o-ring trick. it works so well. =] does anyone have the link for that online?  i forgot which ones are for this. =/

[Geezer]

Ted Weber has the o-rings>>>>>>

https://taweber.powweb.com/store/tubesktord.htm

Tubenit, I've been pressed for time today & will comment on your layout as soon as I can....Thx again,  Geez'r

[echuta13]

I've been ordering my o-rings from McMaster-Carr.  Can't remember the specific size at the moment, but they were hi-temp (400-500 degree F) orange silicon, and about $10 for a bag of 100.

[macula56]

I am really looking forward to trying this out. I don't have much experience with EF86's. What are the differences between running the EF86 into a 12AX7 and running a 12Ax7 into an EF86?

[Geezer]

I don't have much experience with EF86's

Apparently none of us (here) do! So "join the club"....but we're gonna' have some experience when the dust settles!  ;)

What are the differences between running the EF86 into a 12AX7 and running a 12Ax7 into an EF86?

Once again, not sure, but to me the diff is getting that fat pentode tone 1st, rather than "fatting up" a 12AX7 tone(?)

The Matchless Clubman does the latter, a 12AX7 paralleled, into a pentode....not EF86, but a 6SH7 (lower gain than an EF86, I think)

http://www.schematicheaven.com/newamps/matchless_clubman.pdf


FWIW, I'm going to use a 6CA4/EZ81 rectifier in mine, just because I have a nice NOS RCA in the bin. It's a bit "stiffer" than a 5Y3, & I won't need to have a 5v supply.
To help me speed things up a bit, I'm toying with the idea of modding one of my existing amps to this design (one that I never play anymore). The one I have in mind already has pretty much everything in place, except for the recto socket. But, it's a Epi VJr chassis, so I'll have to see if there's room across the front for all the controls (?)  :-\
I think I'd have to "squish'em" in, Trainwreck style!

[Geezer]

Oh....and I have some isolated/shock mount sockets from old 50's/60's military equipment. I'll have to see if I can get them out without destroying them!

[Geezer]

Tubenit,
the sch & layout look good to me.
As I said above, I'm going to be using an EZ81/6CA4 recto, so I'll look that pinout up & change it on one copy (as an option to the octal recto).
I think the closest 5v octal to the EZ81 (which is what they use in the Soho) would be the 5V4.
A 5Y3 is going to be much softer/saggy-er & drop more voltage.

Thx again, G

[macula56]

Thanks for that info Mr. Geezeer. I'll be building mine in a VJ chassis as well. I was going to build something with 6V6's but now i'll put EL84's in it. I have already installed a couple of 9 pin sockets for pre tubes so I guess I'll add a couple more. And an octal socket. I also have some concentric pots I bought a few months ago that should work nicely with the limited front panel space. This should be a very interesting build. Thanks again, JMac.

[Geezer]

I just went & measured the front exposed faceplate area of the VJ....~11" across (exposed). If I leave the on/off switch in place (I already have a standby switch beside it on the right side) that leaves ~8" to place the pots. If you cozy them side by side (almost touching) you can get 1 pot/inch, so there's room for 7, but we only need 6 (if we put the switches above or below the function/pot they control)....the input jack could be done that way, too, above the volume pot.
If I butt the pots right up against each other, I think I could even squeeeeze in another pot for a mid control, if I want (which I just might!

Below is a scale drawing of the faceplate. I'll have to use smaller knobs, but I really like the Davies style that are found on Trainwrecks anyway, so that's cool (no chicken heads need apply!)

[Geezer]

Here's the face with a Mid pot added....still plenty of room, even with 1/8" between each pot case.

[Heinz]

I couldn't resist...

This is the first prototype of tubenit's schematic. It has a 6BM8 power section (similar to the Little Wing) instead of the EL84/12AX7 combination.
I have not built a tonestack for the preamp (which is equal to 'bump' enabled) and V2B has a 10k cathode resistor without a cap. Even without the bypass cap the gain is way too high to be usable. The signal is either off or distorted, so a lot more attenuation before the volume pot is required (or may a lower pot value). I'll play with the values when I have more time & post the results.

[Geezer]

V2B has a 10k cathode resistor without a cap. Even without the bypass cap the gain is way too high to be usable.

I'm thinking that V2B is not needed at all.........many/most EF86 designs just go straight from the EF86 > to the tonestack > to the PI.
However, Peter clearly stated (in the video) that the tonestack is "12AX7 driven", so maybe we should try:
 
EF86 (V1) > to the 12AX7 cathode follower (V2A) > to the tonestack > to the PI (leaving off V2B altogether).

All you would need to do to try this is bypass the last triode on your prototype & take the TS directly into the PI.

If the TS seemed to need "a little more" drive, maybe we could parallel the V2 triodes for the CF(?)

Just thinkin' out loud....Thanks for the report Heinz!!  ;)

[Heinz]

You're probably right. The EF86 input stage does not need another gain stage. However, I doubt that the guys at 65amps would leave a triode unused. A double CF does not make much sense, either.

Thinking about it...a different idea comes to my mind. We could use two separate CFs after the EF86, one driving the tonestack (normal path), the other going directly to the volume pot (bump path). These can be enabled/disabled by a switch that disconnects the load resistor from ground (as with the tonestacks in the current design). An additional benefit of this design would be a full separation of the two signal paths.
But then we lose the bypass cap switching at V2B. That means that we now need a new idea how to implement the bump level...maybe switch bypass caps for the EF86?

I can't draw a new schematic right now (being at work) but I'll post one when I get home.