Low power Musing 40 with parallel/cascade idea

[dwinstonwood]

Hey everyone,
The itch to build has returned, and I'm casting around for ideas. It's a lot of fun building a loud, high powered amp, but I just don't get enough opportunities to play one cranked.

I've started reading through this older tubenit & DummyLoad thread that explores using low power tubes in push/pull: https://el34world.com/Forum/index.php?topic=12526.0 and I'm brainstorming about a low power version of the Musing 40 that I built last year with a bunch of guidance from sluckey: http://sluckeyamps.com/misc/AA864_Musings.pdf

This amp has a "wasted" triode in the tube right after the gain pot, and before the PI. So, I'm thinking, why can't that unused triode be wired in parallel with the other one in the same tube, or have the two triodes cascade, or wire it with a switch to have both options?

I haven't even chosen power tubes (6DX8, 6BM8, etc...), or started drawing a schematic yet. I need to do a lot more reading, too. But, I thought I'd post this "concept" and see what kinds of comments I get.

Thanks for any input!

[dwinstonwood]

Admittedly, a weird idea.

I'm going to buy 6V6 iron - which, for me, means 365-ish VDC on the plates and not AB763 DR voltages - and then parallel that dual triode before the PI a little out of balance to thicken things a bit. This way I can use my Weber MiniMASS to push it at home. And, I'll still have the TMB TS and the Gain knob to fiddle with. I can build it in a Princeton Reverb chassis/cabinet. GCD sells a 12" PR cabinet, so I'll have my choice of speakers...

I guess I'll build it here on the forum. Though, I seem to come up with odd "morphed" projects that only appeal to me. 

[dwinstonwood]

The Hammond 290AX PR replacement PT is only $70.50 at Hawk. So, I'm going with that.
https://www.hammfg.com/files/parts/pdf/290AX.pdf

True to the original PR PT, the 290AX has no separate bias tap. I've modified the Musing 40 layout replacing the 470 3W bias resistor with a 100K 3W.

Please take a look at these two images and let me know if my changes are correct. One image is the original, and the other shows my mods.
Thanks!

[sluckey]

 

[dwinstonwood]

Thanks!! (and that was supposed to be 470 Ohm, not 470K - I fixed the goof in the post above)

[dwinstonwood]

OK, I don't want to try and reinvent the wheel here, and since I already bought all of the iron for a 6V6 amp I think I'm going with the Hoffman Plexi 6V6, but with his single input/channel "Plexi 50 Hot Switch Gain Mod" preamp.

So, everything to the right of the .022uF cap feeding the PI will be straight Plexi 6V6; everything to the left of that cap will be the Hot Switch circuit.

Plexi 6V6:
https://el34world.com/Hoffman/files/Hoffman_Plexi_6V6-V2.pdf

Plexi 50 HSGM:
https://el34world.com/Hoffman/files/Hoffman_Plexi50.pdf

It would be great to hear from anyone who's built this. Thanks.

[sluckey]

I've built a Plexi 6V6 but with my own board and chassis layout. I didn't do a hot switch and I didn't use Hoffman's stout chassis. I don't care for the tubes on the same side as the pots.

Better do a lot of planning if you're gonna put this in that stout 12" chassis. I'm sure it can be done but you will be the first. The hot switch board will likely not fit the 12" chassis. The hot switch plexi 50 board is 11-1/4" long. Measure twice, cut once.   

[dwinstonwood]

Thanks Steve. That's timely advice because I've been looking at the JTM45 chassis - or, similar - so I can use a readily available Marshall head cabinet. Thank you for sharing your experience!

[ac427v]

Still using the Hammond 290AX? I used one for a Princeton Reverb and liked the results with the low voltage secondary taps. Settled on a 5AR4 which gave me a B+ of 377. With a 5U4GB I got 362, 5R4GYB gave 356, 5Y3GT gave 348. 100k dropping resistor was perfect for bias. I wanted lower power so the 100ma rating was fine. If you want big Plexi 6V6 sound you might want more current capacity.

[dwinstonwood]

Thanks ac427v. Yes, I bought the 290AX. I'm hoping to land somewhere around 365 to 375 B+ using the 275-0-275 secondaries. And, I'm also paying attention to Hammond's 100ma rating. With three 12AX7's and two 6V6's I guestimate about 84ma of current draw?

[dwinstonwood]

I put the order in at Hawk Sunday night; the iron arrived this morning. $134.44 for all three parts including shipping seems like a good deal to me, especially with the fast, 3-day shipping. My last pair of trail running shoes cost more than that.

[dwinstonwood]

I'm combining Doug's Plexi 6V6 and Plexi 50 Hot Switch layouts. But, I'm going with a different B+ filtering schema. Instead of a 40/20/20/20 can and one 16, I'm using one 32/32 can, two 22's, and one 16. This way I can separate the grounding and put the single caps on the board. I also deleted the 100 Ohm heater CT resistor's.

All I have so far is the board, the power connection and the grounding. There's probably no need to copy the rest of Doug's layout. I did this much to work out the filtering.

PDF attached below...

[sluckey]

It's much easier to understand what you're saying if you post a schematic. Layouts are mostly useless for this kind of discussion.

[tubeswell]

My 2cw

[dwinstonwood]

I should know that by now. 
Here's what I drew up. All I did was copy Doug's schematics. But, re-drawing a schematic really helps me to see and understand the circuits much better than just looking at it. I have to actually think about each component and connection when I do this, and it all starts to actually make some sense to me. It's a way of studying it that works for me.

PDF: Oops! The first one was wrong. I've uploaded the right file.

[dwinstonwood]

Thanks tubeswell! I'm studying your comments now. I'm not sure what you mean with the galactic ground buss tie? Is that good or bad?

[Willabe]

I'm not sure what you mean with the galactic ground buss tie? Is that good or bad?

Here;

http://www.valvewizard.co.uk/Grounding.html

[sluckey]

You may find this simple hot gain pot useful...

[dwinstonwood]

Thank Willabe. I've never tried a single bus ground before - I've always separated preamp and power amp with their own grounds on opposite sides of the chassis. So, it looks like as long as everything is attached in the proper sequence it works well, like tubeswell was showing me.

That's really interesting sluckey. Now, I'm wondering if I'd even need the switch? If it was "on" all the time (no switch) your Hot Gain pot could turn down the V1-B gain to where it would probably closely resemble the switch in the off position, yes? I'm thinking I would still keep the Gain pot after the tonestack/pre-PI.
Thanks.

[Willabe]

I've never tried a single bus ground before - I've always separated preamp and power amp with their own grounds on opposite sides of the chassis. So, it looks like as long as everything is attached in the proper sequence it works well, like tubeswell was showing me.

Galactic ground refers to separate star ground points chained together to form a galaxy. Keeping each star ground loop as short as within reason and eliminating chassis random grounds.

You make a star for each power supply filter cap node. What ever circuit(s) a single filter cap node feeds, you tie all the grounds from that/those circuits together with that filter caps ground lead. That forms the star. Then you run a wire to the next ground star and so forth.

You can still make the stars for the power amp and preamp and ground the 2 separately. But you don't need to, if you wire up the stars and chain them together right.

You keep the filter cap close to the circuit it powers/feeds, no dog house on the back of the chassis, no can caps with multiple filter cap nodes, because each cap can has only 1 ground, all the caps in the can share the same ground. That lets the current flow round and round in a short loop, without disturbing any other circuit.

Galactic grounding is also a wired ground, not a random chassis ground system. Random grounding has return/ground current flowing through the chassis. Your using the chassis as a wire. Those grounds can cross each other randomly in the chassis, causing problems. You want to get rid of as many chassis grounds as possible. 

The noisiest ground is the 1st filter cap ground because it has the most ripple and most current. That's the most important ground to wire up right. PT CT wire directly to that 1st filter caps ground lead, nothing else. Then run a wire to the rest of the power tube grounds, including the OT ground wire.

Merlin goes through all this in that link I gave you.       

[Willabe]

More on grounding from Aiken amps;

https://www.aikenamps.com/index.php/grounding

[dwinstonwood]

Thank you Willabe! I found the Merlin passage that illuminated the aha lightbulb: 

"Note that no part of this ground system is yet connected to chassis / earth. The whole audio circuit is still effectively floating inside the chassis. The next section specifies the one-and-only connection to be made between the circuit ground and chassis."

I was stuck on thinking that the individual Stars were each bolted to the chassis.

Now I see how it's done. I'm going to redraw the grounds in my schematic by grouping them into individual stars for each B+ node, "daisy chain" them together and make one chassis connection at the input.

Thanks for your help! It's great to break through the confusion and reach that point of clarity.

[dwinstonwood]

OK, a revised schematic with Merlin's grounding system attached. Which is basically what tubeswell posted.

[dwinstonwood]

And a revised layout where I've attempted to follow Merlin's grounding guide. He does say to ground the speaker jack with the NFP/LTP node if global NFB is used. I'm sure there are errors.

I haven't settled on a chassis. But, the BFPR chassis will work. It actually has exactly the right number of control holes in the front; a place for the cap can; and my 290AX would drop right in. I just want a break from Fender chassis. The other options are the JTM45-type, or an aluminum Hammond 17x7x3 that I'd have to drill and cut.

It's time to start ordering parts.

[sluckey]

Nice drawing. Neat and logical flow. I did not look for errors.

[dwinstonwood]

Thanks! 

[ac427v]

To prevent inadvertent errors when working from a layout drawing, I include the red B+ connection from the reservoir cap to the OT. And the black wire from the power transformer center tap to the cap can ground.

[dwinstonwood]

Thanks ac427v, I added them in. I'm also finishing up the rest of the chassis components like the PT, power switch, fuse, AC inlet, etc.
One thing I've started doing is to use a terminal strip, or other insulated point, to anchor the input grid stopper to as it comes off of the socket pin. Plus, it's a lot easier to solder (and, unsolder) RG174 to a terminal than to a resistor lead. So, I added that, too. I think sluckey gave me that tip.

[dwinstonwood]

Since this thread title mentions the Musing 40, I have a question about it. Instead of using the fourth unused triode in parallel or cascaded, would it make sense to use it as a DC-coupled cathode follower to drive the tone stack?
In other words, V1-A would be the same input triode, and it would be DC-coupled to V1-B. I'd then use V2-A&B for the other two triodes (see attached schematic). How would that effect the character of the amp?
Thanks.

Here's the schematic that actually represents the original amp as I built it last year:

[dwinstonwood]

The Rubicon bridge is burned. Parts ordered.

[dwinstonwood]

Thanks for the super-fast shipping Doug!

I ordered an aluminum .08" blank, undrilled JTM45 chassis from zachmdhunter; should get here next week.

[dwinstonwood]

Hey all, I need some extra eyes to check my bias circuit. I'm using Hoffman's "variable resistor" circuit with the two 10uF caps.

Below is a detail of my connections and a link to his bias circuit page. I'm pretty sure I have it right, but I want to be 100% certain.

Also, should I go up to a 50K pot? I've used 10K's before and they seemed to work fine.

Thanks!

Scroll to bottom of page: https://el34world.com/charts/Biascircuits.htm

[sluckey]

Layout looks fine. I prefer to use a 50K pot and a 47K resistor. This gives a wider range of adjustment than a 10K pot.

     http://sluckeyamps.com/phoenix/phoenix.pdf

[dwinstonwood]

Thanks sluckey. I bought some wrong type caps and need to re-order some stuff anyway, so I'll toss a 50K in the cart while I'm at it.

[dwinstonwood]

I was reading through the Hoffman info pages at work and found these cool pin-out/voltage recording sheets. So, I printed some out. Much better than jotting stuff down on a notepad. There's a sheet for probably every amp tube type:
https://el34world.com/charts/Misc/Tube_Pinouts3.pdf
Just thought I'd share this useful resource.

[dwinstonwood]

I had enough free time to get the chassis drilled and to build and lace the board. This .08" chassis took its toll on my cheap Harbor Freight step bits (no surprise). I could also use a bigger, more powerful drill press. But, I got it done, and it was fun.

So, at this point I basically have a "kit" to assemble.

[sluckey]

That is a good feeling to reach that point.   

[dwinstonwood]

Probably my least favorite step. Double checked for continuity and shorts.

[Willabe]

Nice! 

[dwinstonwood]

Thank you.

[dwinstonwood]

I'm finally ready to install the parts on the board. 
I'll wire up the pots after that.

[Willabe]

Looking real nice! 

Something to maybe think about for next time? I think most guys, including the old Fender amps, fill in the R's/C's on the eyelet/turret board 1st, with fly leads for the tube socket, controls and any thing else.

More free room that way, instead of being constricted in the chassis.   

[dwinstonwood]

Thank you Willabe! I did build my 6G3 that way, it worked out fine, and I definitely see what you mean.
But, on a previous build - at sluckey's suggestion, I think? - I followed Hoffman's method of building up the amp in layers. https://el34world.com/projects/Plexi6V6.htm
It allows me to plan things out a little easier and triple check my progress before adding the next layer.  Plus, it does make it easier to hook wires onto the turrets.

[Willabe]

It allows me to plan things out a little easier and triple check my progress before adding the next layer.
Plus, it does make it easier to hook wires onto the turrets.

You can do it how ever you want, it's your build. But all that can be done loading the board 1st.

There's no wires to hook up to the turrets, because you install those wires when loading the board and let them 'fly'. You trim them after the board is installed.

Even if you didn't put in the fly leads it would still be easier and faster to load the board 1st.

Sluckey don't you usually load the board 1st on your builds. Unless you only did that on a couple builds then changed your method?

Try installing the empty board 1st in a tweed chassis.   

[sluckey]

Sluckey don't you usually load the board 1st on your builds. Unless you only did that on a couple builds then changed your method?

Most of my wires pass through holes in the board so I always make the wire connections to the board first leaving the wires long. I usually test fit the board several times during this process just to be sure I don't put a 10" wire where only a 2" wire is needed. After all wires are attached, I fit the board in place and trim each wire to it's final length. Then I put the board back on the bench, strip/tin the ends. This is much easier outside the chassis. Remember, I use teflon wire which can be a bitch to strip in tight quarters.

What I do next varies.

For most of my builds I loaded all components while the board is still outside the chassis. I think having the board out of the chassis makes this easier, especially for a tweed style chassis (which I hate). Then I permanently install the board and make all wire connections.

However, on two of my builds, L'il Maggie and Phoenix, I deviated from the above procedure. I still connected all wires to the board first. But then I permanently installed the board, made all the wire connections, and then finally loaded the components on the board. The advantage of this method is that the component leads stay straight and neat and don't inadvertently get bent while soldering all the wires.

Either of these methods work fine for me. And there may be slight variations from one build to another. If I was using solid core wire with push back insulation, I'm sure I would use a different method.
 

[dwinstonwood]

To be totally honest I've been cheap when it comes to buying wire. I've always skimped on the amounts I buy. I wasted so much wire with my 6G3 build because I either attached wires to the board that were way too long, or just a tad two short. I was new to amp building - I still am.

I know the solution: I simply need to buy 100' spools, and not 48" lengths.

Also, one of the things about amp building that I enjoy the most is creating what I'll call "working" layout drawing that tends to morph as I progress with the build (weird, I know). I've already changed several things in my layout for this amp, and I'll probably change some more. It's a work in progress; I'm figuring things out and moving things around as I go. And I want to be able to change some component values if I want or need to. If I built a second, identical one I'd know exactly how long to cut each wire.

But, I have to say this. This is only the fifth tube guitar amp I've built (I built my first one in 5/20). And this has been, hands down, the most enjoyable and stress-free build yet. Here's what I've learned so far. The placement and location of parts is very important for a quiet and stable amp. The lead dress probably just as important. Certain component values much less so. Coupling caps, cathode bypass caps, NFB resistors, etc., can all be tweaked to taste. I'll more than likely have to adjust my B+ dropping resistors and maybe a bias resistor, too. I'm thinking about trying to learn Duncan's PSUD2 app to estimate the dropping resistor values I'll need, because I don't know what they need to be at this point.

By installing the components last, I can wait until the end to decide on that stuff. I'm still researching various Marshall amps. They used a lot of different component values and I want to keep my options open. In the mean time I can solder up everything that I won't be changing. 

[PRR]

> I have to say this. This is only the fifth tube guitar amp I've built (I built my first one in 5/20). And this has been, hands down, the most enjoyable and stress-free build yet. Here's what I've learned so far.

You should write a book.

[dwinstonwood]

You should write a book.

Thanks PRR 

I don't think this is right, but I'll try to figure out how to use this app.

If it's close, I'll need to lower the 10K resistors.

[acheld]

Well, your simulation prompts a question:  Does the simulation as drawn take into account the current draw through various components?  As it is drawn, the circuit consists of several RC filters and a final 6.6K resistive load.  Is this sufficient to model the amp?

[dwinstonwood]

Thanks acheld. No, I doubt it represents my amp. 

I will read up on how to use the app tonight and see what progress I can make, like how to add current draw, for example, which I predict is somewhere between 76ma and "some higher number" (3 12AX7's and 2 6V6GT's).