Looking for feedback on my design please?

[Amped]

Hi, I've made an attempt to design a 15W amp.  I started with a known-good design and modified it based on research and examples.  It may be a bit "out there" so I'd rally appreciate it if someone with a lot more knowledge than me could cast their eye over it to see if there's any major issues with it before I get too deep into buying bits (I've started getting some smaller items already).  I'm expecting that there will be some...

I'll post schematics (separate diagrams for PS, Preamp-including reverb and tremolo, and power section), also a layout and PSUD 2 screen grab (based on the values in the diagrams - I'm expecting to have to develop the power supply more on the bench before adding the rest...)

Happy to answer any "Why did you do that??" questions, all criticism/advice is gratefully received.

Thanks guys.

[Amped]

Here's the layout I was working on...

Heater wires don't look to be in a great position but it was a compromise given how I thought the earth would work best.  Might be noisy?

[Amped]

...and here's the PSUD2 screen grab.

Does the transformer look too small - maybe a 300-0-300?  Hard to find with the current I'll need to drive the tubes.

[Amped]

Some more information:
Layout diagram:
Its to scale as best I can, its a flattened box - the two horizontal lines show the fold points (fold towards you).  Was tentatively planning on putting the heater wires in a 1x1cm aluminum channel (the long rectangle) and use grommets or eyelets for the wires to come out to the sockets.  I guess I'll see how feasible it will be in the chassis (pretty tight).  The OT is the blob under the main smoothing caps.  This is the best layout for the shortest wire lengths I could come up with, not sure heater wires under the controls is a great idea though.  The component layout on the board is a bit "out there" but it seemed to keep things separated/tidy.  Not sure if being so close if coupling will be an issue or not(?).
Schematics:
The red bits are things I've added to an existing design.  The PT and OT are different, the whole Power Supply is new.   
PSUD2: 
I tried to keep the PS voltages in the same ball-park as the original design, seemed like a smart idea.  Not sure if I need more starting voltage though with all the filters.  It's going to take some trial-and-error I think, I'll build the PS first before the rest.  I added a R/C section to separate the Reverb Pentode, not sure if that was necessary or not but at ~10mA draw isolating it seemed like a good idea at the time. It is further up the PS though considering it is just after V1a in the signal chain.

[tubenit]

I personally would not want the reverb so early in the signal chain with an amp that I suspect will be pretty high gain.

And I would NOT want a .02 cap going into the dwell resistor.  Usually you have a .0005 to .001 there.   

I would place the reverb between the master volume pot after V2b (send)  and (recovery) V4b  around a resistor value of 220k or more.

I think you're likely to have your reverb pretty "gained up" with your design having the reverb so early in the signal chain ?

I don't know anything about the 6BX6 tube?  So, I am presuming that you know this reverb circuit will work OK?  IF you are not confidant of this as a "proven" design, then you can consider one of the 6BM8 circuits in the "one tube reverb" thread in ARCHIVES.

Having said all that,  I think you "should" build this however you want to.

With respect, Tubenit

[Amped]

Thanks - exactly what I'm after - the right place to put this was a point of contention and some late nights.  The reason I thought to put it down that end was due to the original design - it is actually a stand-alone effect pedal (schematic attached) so I'm thinking "get it early in the signal chain" https://www.valveheaven.com/2016/10/the-lamington-reverb-an-easy-to-build-valve-reverb-unit/.  I hear what you are saying though, I think an earlier version I had it further your way but couldn't rationalize the fact it was designed to go before the amp (and around a 220K resistor makes sense and I've seen that before in other Fender schematics off the top of my head).  This is probably my inexperience coming through though.  The design uses the 6BX6 valve, apparently was quite common here in Australia in black and white TV's so while they aren't made anymore there's a few still kicking around.
The design I copied was inspired from an article about pentode driven transformerless constant current reverb.  Its actually quite a compelling read (made sense to me anyway) http://www.channelroadamps.com/articles/reverb_driver/

re: cap values, thanks for your advice (and mark-up of the schematic), the original didn't have any coupling cap here at all.  I tried to make the layout as modular as possible so I can change things if I get it wrong, this is a "lift and shift" and I recon I'll make this change.

Thanks again mate.  Appreciate the input.

[tubenit]

Again, I think you should build it however you wish and seems best to you.    Please remember some of the early Fender amps (like the Tweed Twin or Tweed Super) that used the stand alone reverb were fairly clean tone amps compared to some of the higher gain EL84 amps.

with respect, Tubenit

[pdf64]

The topology is a bit unusual, to have

gain stage
Volume
Gain stage/CF
Tone stack
Gain stage
Master vol

the 3rd gain stage will likely get overdriven and mush up the response from the tone controls. Generally, with an overdrive preamp, the tone stack gets put after the last clipping stage.

Also with a highish gain preamp I'd be concerned about the layout; putting the tube sockets under the control panel (with lots of high impedance signal circuits) may lead to hum and also parasitic coupling, perhaps instability. What's the benefit in doing things that way? Generally, amp layouts have turned out the standard way we're used to because it generally works out for the best.

[Amped]

Thanks tubenit, you think it might be too high gain for the reverb?  Would removing the follower help do you think?

[Amped]

Thanks for looking at my design pdf64.  That does look a bit weird now that you mention it.  Hmmm...
If I moved the tone stack and CF down to the last gain stage (maybe remove the CF), that would be more like a Marshall I guess.  From your experience would that be more likely to work better?  How about removing the follower (just a spare pentode sitting there, what to do with it?)?

Layout - yea...
I get what your saying.  Layout's hard, so much to think about.  Reasoning is complicated; short answer this was to keep the wires as short as possible but maintain layout.  Getting everything lining up is a trick.  I was concerned more about the heater wires than the sockets but, yea they are right there aren't they?.  Yea I guess I am trying to re-invent the wheel, not a good look.

I am tempted to start from scratch.  I'll move everything around and see how it looks before I make up my mind.  Ah well, back to the drawing board.

Thanks again for your help.  This might take a while...

[pdf64]

The Marshall 2204 etc overdrive preamp topology has been copied and tweaked a gazillion times, so offers an excellent basis for your journey with this.
Regarding the reverb circuit placement, I think it would probably work out better after the master volume. Trying to incorporate reverb into an overdrive preamp is a real headache; overdrive compresses the signal and makes the reverb level apparently much louder. And that's without even starting on the acoustic feedback issues from the tank.
Look how the Mesa Boogie Mk series developed; by the MkIV they gave up and moved the reverb to the end of the preamp.

[tubeswell]

For a 15W amp, there are many tried and true designs. A tweed deluxe is pretty excellent bang for buck, or other tweed Fender push-pull amps using 6V6s (5F10 Harvard or 5F11 Vibrolux, or my favourite - 5G9 Tremolux).


If you like things cleaner, maybe a one-channel AB763 circuit with 2 x 6V6s (or for simplicity, ditch the reverb and 'Vibrato').


Or if you like EL84s, Matchless Tornado or Spitfire, or Marshall 18Wer.

[Amped]

@pdf34
Cheers again for the feedback.  I've reorganised my schematic to move things around as you have suggested [hopefully].  Does this look kinda what you were thinking?  If I was to build this (and it is an "if" at the moment), I'd probably start with the main amp then develop the additions as I go.

@tubeswell
Thanks for your suggestions, you can see me struggling a bit eh? ;-)  I'm going to have a look over them, I had a quick look at that spitfire and from what I saw it looked pretty awesome.

[tubenit]

Quick glance that looks fine.  Except I think you need some resistors for a voltage divider after the second gain stage between V1b & V2a?  Not sure what value?  Maybe 220k to 470k?

Look at the voltage divide in Sluckey's Dual Marshall design on the JCM800 side.  He has a 470k/470k

You also might want to change the post phase invertor caps from .047 to .02 which is more typical/common (from what I've seen) with EL84's?

Don't know if it was a typo or not, but I'd be much more amenable to using 47uf cathode cap for the power tubes then 470uf (which I don't remember ever seeing before with EL84's?).  I think the most I've ever used with power tube cathode caps has been 100uf?  As an FYI, the higher value cathode cap (in my experience) has moved the tone to sound more like fixed biased then cathode biased. 

Last thought to consider …………..  IF I were building the amp, I'd use 6V6's not EL84's.  That's just a personal preference thing and not something you "should" do. 


Tubenit

[sluckey]

V2-2 must have a resistor to ground. Tubenit suggests a voltage divider which will probably be necessary. The voltage divider would satisfy the need for resistor to ground also.

[ac427v]

Getting there?
You might add a cathode bypass cap on V4a, the reverb recovery stage. The current design uses no reverb mix stage and a cathodyne (no gain) PI. That may result in a wimpy reverb signal even at maximum setting. My cathodyne experience is with a Princeton Reverb which has a reverb recovery stage and another stage to amplify the mixed reverb/dry signal before the cathodyne PI. Your original design from a reverb pedal only amplifies the reverb signal to the level of a guitar pickup. You need enough voltage to drive a EL84 power amp. Adding a 4.7-22 uf cathode bypass cap on V4a will maximize the reverb signal. Others on this forum who know more than I do may comment on whether this will be enough gain without a separate "mix stage".

[Amped]

Woops - thanks good spotting (grid stopper/grid leak = added).

[Amped]

Updated diagram:

[Amped]

@tubenit et al.
I had wondered about values also - compared to other schematics.  The bones of this design were around the Lamington amplifier design (Valve Heaven), my reasoning was that if I used that as a starting point in theory at least that was a solid foundation (i.e. I wouldn't have to second-guess the maths for that part anyway).  I wanted to add a reverb - so had to update the power supply (wanted a valve rectifier anyway), then I had a spare triode (why not a Tremolo?) and down the rabbit hole I went.
The design shows a 470uF cap and 47nF caps for the PI - not saying that its right but if I do change them it rocks the foundation of the design (kinda already been done with the recent changes to stages I guess).  So I guess I'm at a cross roads, stay on the path trying to make something a bit "different" at the risk of lots of troubleshooting and potentially failure, or find a tried-and-tested design and put my efforts into - and getting satisfaction out of - the physical design and build. 
I have to say the Matchless stuff does look pretty awesome, the DC-30 with it's two channels looks particularly intriguing, but 30W was a bit bigger than I was going for (trying to keep volume and budget at sane levels). 
I guess - if you're willing to - what I'm asking is for some advice.  Given your experience and knowledge of what I'm about to try to start, what would you do in my shoes?  I don't know what I don't know yet if that makes sense(?).  Any input welcomed.

Apologies for the novel.

[Amped]

@ac427v
Thanks mate.  Yea the reverb...  Reading the theory it sounds good, I like the fact this one didn't need a transformer, but as it's been converted from a pedal to on-board it was always going to be a voyage of discovery.  Originally I had it injected further down the preamp stages so that it would be amplified with the signal but subsequent advice suggested it would be better in the commonly located FX loop send/receive position after the preamp - which makes a lot of sense to me.
Interesting I hadn't thought that the addition of a cathode bypass cap may increase the signal enough, definitely will be worth a try.  I guess another triode here would also help (LOL more valves, this is getting out of hand! )
Really appreciate your help, you're right more thought is needed for the reverb...

[tubenit]

what I'm asking is for some advice.  Given your experience and knowledge of what I'm about to try to start, what would you do in my shoes?

I am presuming you are familiar with the tone of the Lamington amp?  Hopefully, in person and not just YouTube?

IF you like the Lamington amp, then use that design and simply be prepared to tweak it and change values where & if needed. 

I did listen to a couple of YouTubes of this amp and got on the website to look at the original schematic for the 15w one which appears to have a James tone stack and no cathode follower in the 3rd gain stage?   

With respect, Tubenit

[Amped]

I've only ever heard web samples, never heard or used one or used in person.  This will be the same for most amps - the Matchless ones included.  Really I could only choose a design based on reviews, and vague idea about what I would like. 
Yes - I did change the tone stack, and added the cathode follower (spare triode, seemed a waste).  100% that could all go back to original.

[pdf64]

With only 1 gain stage before the power tubes, I don't think the reverb wet signal will be anywhere near strong enough for a decent effect.
A BF Fender wet signal has 3 gain stages (though none are wide open).
For best signal to noise (hum, hiss) ratio, an input stage cathode should be fully bypassed; input stage includes the reverb recovery in this case.

[Amped]

I recon you're more than likely right.  The original design of the reverb is as an effect between the guitar and amp, not after the preamp. 
I've sent a message to Grant at Valve Heaven (he developed the amp and reverb) and asked if he thinks it will work where I think it should go.  I'll post the results back here.
Thanks again.

[Amped]

Hi guys.  Grant got back to me and (surprise) he's not confident in the outcome.  As it is designed to be an external pedal effect, he feels it's unlikely to work. 
Thanks for all your advice and help, I'm calling "time-of-death" on this idea of mine... 

I am still going to build an amp, got an idea but I'm going to take a step back and start again from some existing drawings.  I've found some that are intriguing, and would be interested in some advice about a starting point so I'll start a new post to discuss.

Thanks again for your help.

[tubenit]

Thank you for taking the time to let us know the rest of the story on this!  It is appreciated.

With respect, Tubenit

[Amped]

Thanks - no problem, hopefully it will be useful for someone else who has the same crazy ideas, and they don't waste a month trying to figure out how to make it work 

Lesson learned; you can't plug random stuff together and expect it to work.  Needs some maths (and experience to know how/where to apply it).

[shooter]

Needs some maths (and experience to know how/where to apply it).

that's why you came here 
keep reading everything, keep the calc batteries charged, pretty soon you'll find some kool circuit from circa '39 that you just have to mate up with.......