New Build: Tweed...thing

[Splattle101]

I planned to do a quick build earlier this year to 'warm up' for a more complicated job sitting on my bench. This is the supposedly quick build.

It's kind of based on a tweed Champ, and it looks like this:

The cab is from Mojo, and it's a tweed Princeton with some mods. Here's one of the mods right here:

The cut away for the controls is sized for a Champ.

The concept with this amp was to build an amp based on a Champ, but corrected and modified.

'Corrected?!' I hear you cry. Yes, corrected. The fixes include:

* using a planned earth system instead of having it earth all over the place;
* modifying the layout to accommodate the earth plan;
* using a grid stopper on the output valve and the second gain stage in the preamp;
* using a screen stopper on the output valve; and
* using sensible filter capacitors rather than the 16 uF and 8 uF caps specified in the ancient schematic.

And of course the power supply is wired up to comply with modern safety standards (i.e., no death cap, three prong lead, dedicated safety earth, fuse on active rather than neutral, etc).

The mods, however, are a different story.

 

[Splattle101]

So what are the mods? To be honest, they're not that earth-shattering. I just wanted this to be a little different to your off-the-shelf tweed Champ.

Firstly, I wanted to get rid of the traditional 68 k grid load resistors in the standard Fender input network (and the amp 'design' game being what it is, because those 68 k resistors are standard in the Fenders, they've become industry standard ). I figured I'd go with 10 k to retain a bit more treble and sparkle when I turn the guitar vol down.

Secondly, I wanted to voice the preamp a bit like a post-JTM Marshall. That means a 680 nF bypass cap on the cathode resistor of the first gain stage. Simple enough.

But I also thought about the second low sensitivity input. I am never going to use it. Who actually does? I never do on any two-hole amp, and I only use it on four-holers to jump the inputs. So why have it on a two-holer at all? And if not an input, what else could I put in that hole (assuming a standard 5F1 chassis)?

For mine, the answer was: a two-position, single-throw switch connected
to a 10 uF cap in parallel with the 680 nF bypass cap.

With the 10 uF switched in, the 6 dB frequency of the first gain stage would be about 10 Hz (ala Fender). With it switched out and only the 680 nF cap bypassing the cathode resistor, the 6 dB gain point is about 150 Hz (ala Marshall JMP). So it's got Fender or Marshall voicing on a toggle.

 

[Splattle101]

The next mod was kind of fortuitous. I decided to go with a pair of transformers from Mangetic Components for this project. When I checked out their output transformer, I noticed it had two taps on the primary winding: 8 k Ohm and 5 k Ohm. I figured the 5 k tap was for using a 6L6 instead of the traditional 6V6. But why choose between them? Why not put these taps on a switch so the amp could safely run either kind of valve? Kind of like this:

After some research and asking around (thanks Don ) I picked on the right kind of switch for the job (a Garling jobbie). Setting this up, I ran one end of the OT primary to the anode of the valve, and the two taps to the switch. Then I ran a lead from the switch to the filter.

And thar she is:

You can see the blue lead of the 8 kOhm tap connected at the near corner of the switch, and the red cloth covered wire going to the filter on the far terminal at the near end. This seemed the easiest way to achieve the result, but I noticed that it did reverse the polarity of the transformer. If I'd thought a little more about this I could've saved myself some aggravation later, but you get that on the big jobs...and on the small, quick and dirty ones, too, apparently!

 

[Splattle101]

And while I'm on the subject of stupidity, this is the safety earth:

This thing is connected to the earth cable on the power cord. Nothing else is connected to this earth, ever. It is bonded to the chassis with a toothed earth tab that bites the metal. The chassis was sanded first to ensure good metal to metal contact. The tab is held down by a spring washer, which is fastened with a nut which in turn is held by a nyloc nut. It's not moving any time soon.

So what's stupid about it? Nothing. Nothing at all. It's required by law in most places.

What's stupid is that I had to drill a hole in the chassis to mount this bolt. Since this kind of earth connection is mandatory in most places, why the f**k don't the manufacturers of chassis start including a hole for it, for goodness sake??!!??!!

 

[Quill]

... why did the switch reverse the polarity of the transformer? That's a surprising problem ... ???

 

[Splattle101]

... why did the switch reverse the polarity of the transformer? That's a surprising problem ... ???

Well, normally one would connect the common end of the OT primary to the filter, and the appropriate tap to the anode. In this instance I've connected the common (red) to the anode so that I can connect the two taps to the switch. I then connected the switch to the filter. Simple.

It became a problem after I'd finished the build because this amp's topology is based on the 5F1, and the 5F1 has negative feedback. So I had the neg feedback hooked up as one normally does...except that because I'd reversed the usual phase of the OT, that negative feedback was now in-phase. It was positive feedback. Doh!

The effect when I switched on was lots of squealing and distortion from the amp and lots of head scratching from me as I tried to figure out what the f**k I had done wrong!

 

[Splattle101]

The next mod I wanted was a bigger speaker. I know many people think of the 8" speaker in the Champ as part of its character, but I just can't abide it for anything other than recording. In my experience it's not got enough cojones to hang in a three piece when it's in the room with the drums. I want a 10" speaker, minimum.

So when I ordered my cab from Mojo I asked them to put a 10" baffle in it. I must say, they did a really nice job of that cab. It's nicely made, and beautifully finished. I also had them lacquer it.

For a speaker I was going to use an old Celestion G10D-25. So I put it in the cab when it arrived, and hooked it up to my blackface Champ clone. Yuck. It sounded small, shrill and boxy. So I next fished out an old Eminence blue frame job that was put in my Bassman by a warranty tech when one of the re-issue Jensen P10Rs fractured. It's a lot like a P10R in appearance and sound, meaning it's nice and articulate but bright verging on brittle, with little to no real bass reproduction. It sounds ok in this cab, but I plan to fit something better soon.

 

[Splattle101]

I bought a 5F1 chassis online on fleabay. Some dude in Hong Kong. It was cheap and not too nasty. I had to re-size the holes for the chassis bolts, and make an extra hole for the safety earth, but apart from that it would've been suitable for a standard 5F1 straight off the shelf.

I also bought a turret board for a 5F1 from the same dealer. This was a less happy purchase. It wasn't very nice cosmetically and some of the turrets on opposite sides of the board didn't really line up. It also had some turret placements that were just wrong. Luckily this wasn't much of a problem because I intended to change the input circuit and the layout of the power filters, so I'd have to make some changes to the board anyway. But gawd it looks horrible.

The build went smoothly enough, but you can see the butt-ugliness of the board in the pic below:

In the pic above the filter for the preamp has been moved to the right of the cathode resistor and bypass cap for the power valve. This is to facilitate the earth scheme.

 

[Splattle101]

Which brings me to the discussion of earthing. In the trad Fender, all the filter caps are at one end of the board, and all of their negative ends are connected first together, and then to earth. My limited but slowly expanding reading on this tells me that the parts of the amp powered by a particular filter cap should be earthed at the negative end of that filter. So for example, the preamp earths should go to the neg end of the filter that powers the preamp. The filter should also be close to the thing it powers, thus reducing the length of leads. Then the various negative ends should ideally go to a star earth.

I didn't try to do this. Instead I had two earths (not counting the safety earth where the power cord comes into the chassis): one for the power amp, and one for the preamp. The power amp earth is a tag on one of the transformer bolts, and the negative ends of the power valve anode and screen supplies go there, along with the cathode of the output valve. The preamp earth is at the input. The preamp cathodes go to the neg end of the supply cap for the preamp, which then goes to a turret up near the input jack. Other preamp-related earths star to that turret. That turret then connects to the earth tab on the input jack.

I don't know if I've done this right, but it was my first attempt at a sensible, planned earth layout on the basis of modern best practice. As it turns out, the amp is quiet. I can hope that's because of the earthing.

 

[MiniB]

Cool stuff. I did a similar thing on my 5E3 with a push-pull switch and different cap values first on the V1 bypass cap, then the bright channel coupling cap. A 'fat' switch if you will. But really didn't end up using it and was fine with all the ultimate values I settled on.

I also did all the power section grounds to one bolt of the power tranny, and all the preamp earths to another. Amp is quite as all get-out. Bit this didn't work well with a Princeton Reverb build and went with multiple connections to the grounding bus wire on the pots....wheres in the 5E3 there were maybe a few common 'threads' that went to one point.

 

[Splattle101]

Anyway, after I got it built the bloody thing didn't work. This pissed me off no end because I had to travel interstate the next day and again the following week. This meant that I wasn't going to get it fixed any time in the next couple of weeks.

But I finally got to it yesterday. Armed with a schematic or three, a multimeter and a good beer (Little Creatures Pale Ale ) I retraced the circuit for the millionth time and found the bastard negative feedback loop. Simply moving it over to the other terminal solved the issue.

The amp worked just fine after that. The switches work as advertised. There's a noticeable thickening of the tone when the preamp switch is set to 'Fender'. On the Marshall setting it's bright and aggressive, and really makes the neck pickup of my Les Paul sing.

On the matter of glassware, I dug out another old AWA 12AX7 for the preamp:

For the power amp I chose an old, severely flogged JAN Tung-Sol 6L6WGB with a GE 5V4GA rectifier:

This combination puts about 385 V on the anode of the 6L6, with about 27 V across the cathode resistor at idle. So it's well inside its comfort zone. The switch between the power valve and recto is the OT switch. In this pic it's set to the 5 kOhm 6L6 tap.

BTW, that Tung-Sol 6L6 tests way better than new. It is just mongo strong, for all of the heat damage visible on the base. It's been sitting around in my stash in the forlorn hope I might eventually find a match for it. Fat friggin' chance!

 

[Splattle101]

Part of the attraction of having this amp capable of running 6L6s was to use the small horde of singleton 6L6s I've accumulated over the years from running larger amps that use pairs of them. When a 6L6 in my Super Reverb pops its clogs it'll usually leave a partner that is otherwise good to go. This amp is now a happy home for bereaved 6L6s, 5881s and KT66s. Yes, KT66s. The PT in this animal has enough heater current to push a KT66.

 

[Splattle101]

An aside.

Once this thing was running on my bench, I ran a whole lot of recto and power valves through it, testing voltages to see if all would be happy. The menagerie included 5Y3GT, 5V4GA and 5AR4 rectos, and Tung-Solski 6V6GT, Sylvania 5881, Tung-Sol 6L6WGB, TAD 6L6WGC and a Valve Art KT66.

I can report that it will happily run all of the above...except that friggin' VA KT66. Regardless of recto, I found the screen voltage to be consistently 4-8 V higher than the anode. Yes, the anode. No, I don't get it.

I checked with another VA KT66 and got the same result. Then, and here's the kicker, I checked with one of those old grey glass MOV KT66s out of my Bassman and lo! The bloody thing seemed happy!!

I don't know what this means or how it can happen. In addition to the dropping resistor between the filter stages, there's also an 810 Ohm resistor on the screen, right on the base. I have no idea how the screen can be reading higher than the anode. But it can't end well.

So it looks like no Chinese KT66s in this amp just yet. At least, not until I can figure out what on earth is happening.

 

[bilbarstow]

Splattle,

I love your tweed threads. So much info and an extra helping of amp porn.

The boxes are beautiful, but just like women, so much better unwrapped !

Thanks for sharing - I wish I had some insight to provide on your screen/anode issue, but it seems like you have useable combinations anyway.

 

[Splattle101]

Update

I've been playing this thing for a little while and I'm considering some tweaks, but I wanted to hear it with a proper speaker first. Well, the 'proper' speaker arrived yesterday.

It's a Weber 10A100, so it's a 10" speaker with an AlNiCo motor a 1" voice coil, allegedly voiced like an old Jensen. I don't know if that last bit is true but...

...holy snap-frozen grannies, but what a difference it makes!

Depth, smoothness of harmonics. Just so, so, VERY much better than that blue frame Eminence job. Simply superb.

So without further ado, here's a quick clip of it this singled-ended 6L6 thing. It's running a GE 5V5GA rectifier, a Tung-Sol JAN 6L6WGB, and an AWA 12AX7. The guitars were my old USA Telecaster with Tonerider Vintage Hot pickups and my Les Paul Classic with P-90s by Perry Ormsby. The playing and the flubs are all my own:

SoundClick artist: splattle - page with MP3 music downloads

00.00 - 00.46
Telecaster, amp vol 50%, strumming hardness.

00.47 - 01.40
Telecaster, amp vol 75%. I played a piece with the bypass set to Fender, then repeated it in Marshall mode. I did this with a clean bit and an overdriving part.

01.40 - 02.18
Les Paul with P-90s, amp vol 95%, playing similar licks at three different volume settings on the guitar.

02.22 - 04.19
Les Paul w P-90s, amp vol 75%, treble booster, varying vol on lick in E, exploring distortion voicing. I turned on the delay somewhere in here, too (Mooer, analogue echo).

04.20 - 05.10
Les Paul w P-90s, amp vol 75%, fuzz, echo as above, exploring different distortion levels with single notes and double stops.

It looked a bit like this:

 

[Batman]

Holy shite Splat, that sounds awesome!

My favorite is 3:30 to 4:20 especially when you really dig in. Very nice harmonic content to the grind!

 

[Soul Tramp]

Great little amp, Splatt! Excellent idea to use a Princeton cab to solve the speaker size problem of a Champ.

Very clever idea with the OT toggle.

Nice demo. Stills sounds like a Champ.

 

[Splattle101]

Thank you, gents.

I'm liking the sound of it a lot now. I look forward to the speaker breaking in so I can hear how it's really going to sound.

A point of interest for me in that recording is the little passage from 01.40 to 02.20. In there the amp is nearly maxed out and I played a series of licks with increasing vol at the guitar. Once the guitar is maxed out (01.57) there's a whole lot of compression and EQ effects happening that I find interesting and intriguing. That's why I put that part in the demo (goodness knows it wasn't for the artistic quality of what I was playing! ).

First, it fattens up like a bastage. Thick is the word that springs to mind.

Secondly, and more surprisingly, the top end rolls off. I've heard this before but always thought it was a speaker effect from the amp driving the speakers so hard they can't reproduce the little details of the high end. But that can't really be happening here. It's a 15 Watt speaker and I would guess that 6L6WGB is putting out about 7 Watts.

So it's an amp effect.

Thirdly, the compression to the notes sounds like...yup, rectifier sag. But it can't be. It's a single ended amp. So what is it? Is it purely the movement of the power valve's operating point? Or is there some combination of events that can happen in a single ended amp that cause the voltage across the recto to drop under load and cause this characteristic sound?

Hmmm.


...or is this just blocking distortion creeping in at the second gain stage because the vol is so high that the total grid resistance is really low? I've put a 10 kOhm grid stopper on that triode, but maybe it's just not enough...I see a 100 k grid stopper in my future.

Ah yes, let the tweaking begin.

 

[Splattle101]

...oh yeah, one for you Quill, if you should drop by this thread. That sound at 01.57 is what I was talking about a little while ago when I said that fuzz sounds a lot like a fully cranked amp. It does to me, anyway.

 

[Splattle101]

But on a more serious note, check out the ancient chicken head this thing sports:

I don't know if it's period correct, but it's distinctive. Just needs some parsley.