Single Channel Deluxe Head Build

[capohk]

I'm getting around to starting my next build - a single channel deluxe of some type. Each time the design and build process seems to get a bit smoother and more intentional, partly due to the community help here. I don't think I'll be breaking new ground, but it's reassuring to document a build here and perhaps have some company on the journey...

I am starting with a Hammond 14x6x3" aluminium chassis and a Classictone 40-18078 PT. I have a Triad OT S-146A 25w UL taps, 6.6k primary 4,8 and 16 secondaries (rating pic below) from my reclamation drawer as well as a pair of 'coke bottle' 6V6g tubes. All untested of course, but that's part of the fun...

I also have a huge 10h choke that came from a 30's Wards Airline amp. I am on the fence about using that or any other choke as a 5E3 didn't use one, but my 6L6 Proluxe build has one and is super quiet.

I think I will do something akin to a single channel 6G3 with a cathode bias arrangement on the 6V6s. The tried and tested Tweed Princeton preamp vol/tone for the first stage and then a LTPI. Instead of a second channel, I thought to make the first triode cathode bypass cap switched to have two flavours of gain.

From what I can glean, the 6.6k OT might be more suited to a fixed bias amp. Or it might not matter. Other considerations are filtering values and coupling cap values. I have a 32/16 cap can and a 8uF for preamp B+ on hand, as well as an assortment of vintage caps which I am testing for DC leakage. I'll draw out the schematic as it helps to understand the layout. Previously I have just built up the amp PtP as I go, making decisions based on grounding points and shortest runs. This seems to result in a quiet amp. This time I am thinking to use a tag board or even a pre-made 5E3 turret board I have, if enough of the turrets seem useful in the eventual layout.

Questions:

What is a good mac-based layout programme? DIYLC does not seem to want to work for me.

Is there already a schematic/layout for this. It's not a million miles from an 18w in the power section, I'm guessing.

 

[capohk]

Just seeing how everything fits together.

 

[pinchegil]

What is a good mac-based layout programme? DIYLC does not seem to want to work for me.

What version of DIYLC and OSx are you running, I am running it just fine. I have found that different versions are really sensitive to the OSx version you have. I am using BigSur 11.2.1 and DIYLC v.2.0 (1.0) When I upgraded to Big Sur I did have an issue where the components drop down menus stopped working, I had to delete the old DIYLC and reinstall it (make sure you move your projects somewhere safe) YMMV

 

[King Fan]

What version of DIYLC and OSx are you running, I am running it just fine. I have found that different versions are really sensitive to the OSx version you have. I am using BigSur 11.2.1 and DIYLC v.2.0 (1.0) When I upgraded to Big Sur I did have an issue where the components drop down menus stopped working, I had to delete the old DIYLC and reinstall it (make sure you move your projects somewhere safe) YMMV

Yeah, DIYLC works well on my Macs -- with one little wrinkle. Your first install *and* all your updates need permission to get around Mac's 'Gatekeeper' -- the thing that tries to prevent unknown-developer malware loads.

What Mac OS are you running? I admit I'm that guy who runs the newest OS (once it's been out a couple weeks) and downloads most of the DIYLC updates pretty soon, too. If you're mixing older OS or DIY versions, double check that good. note from @pinchegil. On ol-l-d Macs and OS versions, I had to do more workarounds -- Java and I don't recall what.

If you're just having Gatekeeper issues, of course DIYLC isn't malware, it's just from a small developer Apple hasn't certified. To bypass Gatekeeper, as Apple explains here:

1. Try to open DIYLC. You'll get the 'developer can't be verified' dialog -- but your Mac will note DIYLC is the program 'in the queue.'
2. Open System Prefs, go to Security and Privacy and look at the General pane. You may or may not have to click the lock and type your (machine) password. Since you just tried to install DIYLC, it will be listed instead of the "example" app below. Click on 'install anyway.'

3. A final warning appears. Click on 'Open'. Apple explains, "The app is now saved as an exception to your security settings, and you can open it in the future by double-clicking it, just as you can any authorized app."
4. Just note you'll have to do this little two-step for future DIYLC upgrades.

 

[capohk]

Brilliant - sorted out, I am neck-deep in learning yet another thing. The amp-building rabbit hole just keeps getting deeper...

Looking at the design possibilities, because of my physical layout - 3 octal, 2 noval tube sockets, I am trying to decide which way to skin the cat. I only need one channel - I use a pedalboard on stage and rarely mess with my amp once we've sound-checked - the house engineers don't like it.

My 'real' amps are a 54 wide panel deluxe and a 63 Brown deluxe, both of which I love but are really too valuable to take out regularly, so a compact head unit that I can plug the house speaker cab into is what I am after.

The 18w Lite design from my understanding runs the preamp triodes in parallel rather than series. I have a another set of iron and tubes (including a pair of vintage ef86) from an EL84 based projector amp that I think will be better suited to a marshall/vox design so am turning back to the Deluxe for inspiration.

If I just build a one-channel 5E3, I will leave a triode unused, so the 6G3 pre-amp into a dual triode PI might be better. I am thinking to do a cut-and-shut 6g3 into 5C3 with the paraphrase PI as a nod to my two favourite amps.

BUT, later amps, including the 6G3 use the long-tail pair PI. I can find a lot of threads comparing LTP with cathodyne PIs but not a direct comparison between the two dual triode setups. Any thoughts on that choice would be much appreciated.

With essentially 3 gain stages including the PI, I think it would be good to employ some sort of master volume - Robrob suggests replacing the 100k voltage divider after the second preamp stage with a 100k pot for pre-phase inverter MV. That seems like an elegant design. Again, any thoughts much appreciated. Attempt at a schematic to follow...

 

[capohk]

 

[YellowBoots]

View attachment 1103690

Maybe I’m missing something but that paraphase PI looks a little wonky (to me). Is it copied from another design? Seems like a bit of signal would cancel at the junction of the power tube grid leaks. Possibly the lower 270K needs to go directly to ground instead of 56K resistor?

 

[capohk]

Copied from the fender schematic.

 

[YellowBoots]

Copied from the fender schematic. View attachment 1103746

I have a vague recollection of seeing this before and thinking to myself it didn’t make sense. I’m not denying that it works, but it seems like you would have a strange mix of negative feedback and signal cancellation on those power tube grids. Anyone experimented with this and the more typical arrangement? I’m curious what the end result is.

 

[King Fan]

LOL, @andrewRneumann I was thinking I'd never noticed that before. Now I see it says *5D3*. Note to self: read, *then* think.

Here's one of my laughable guesses -- was it some kind of 'floating' paraphase? Like Fender was evolving the Deluxe stepwise before the 5E3? Only source I can find:

The Floating Paraphase Circuit

www.vias.org

 

[Snfoilhat]

View attachment 1103690

That 100k:15k voltage divider is the plate load and has high voltages on it. A coupling cap and maybe a few other parts may be missing from that area

 

[mountainhick]

Brilliant - sorted out, I am neck-deep in learning yet another thing. The amp-building rabbit hole just keeps getting deeper...

Looking at the design possibilities, because of my physical layout - 3 octal, 2 noval tube sockets, I am trying to decide which way to skin the cat. I only need one channel - I use a pedalboard on stage and rarely mess with my amp once we've sound-checked - the house engineers don't like it.

My 'real' amps are a 54 wide panel deluxe and a 63 Brown deluxe, both of which I love but are really too valuable to take out regularly, so a compact head unit that I can plug the house speaker cab into is what I am after.

The 18w Lite design from my understanding runs the preamp triodes in parallel rather than series. I have a another set of iron and tubes (including a pair of vintage ef86) from an EL84 based projector amp that I think will be better suited to a marshall/vox design so am turning back to the Deluxe for inspiration.

If I just build a one-channel 5E3, I will leave a triode unused, so the 6G3 pre-amp into a dual triode PI might be better. I am thinking to do a cut-and-shut 6g3 into 5C3 with the paraphrase PI as a nod to my two favourite amps.

BUT, later amps, including the 6G3 use the long-tail pair PI. I can find a lot of threads comparing LTP with cathodyne PIs but not a direct comparison between the two dual triode setups. Any thoughts on that choice would be much appreciated.

With essentially 3 gain stages including the PI, I think it would be good to employ some sort of master volume - Robrob suggests replacing the 100k voltage divider after the second preamp stage with a 100k pot for pre-phase inverter MV. That seems like an elegant design. Again, any thoughts much appreciated. Attempt at a schematic to follow...

Maybe irrelevant to you but just a little different approach; I have a Deluxe on the to do list, and to mix it up a little, I'm pretty seriously considering first channel 5e3 voicing and second 6G3 voiced channel with catodyne PI. ABY box up front takes care of switching channels on the fly. Shouldn't be a problem for sound engineers if channels are matched volume. Your 5C3 6SC7 would be a bit different challenge with the common cathode and doing different biasing scheme, might work to use 6SL7 and split the cathode and biasing schemes per channel.

 

[YellowBoots]

That 100k:15k voltage divider is the plate load and has high voltages on it. A coupling cap and maybe a few other parts may be missing from that area

Agreed! I skimmed over that and didn’t even catch the obvious.

 

[capohk]

Yes. Good catch, should be a coupling cap just after the NFB return. I have added it.

As regards Paraphase vs LTP, all I can find is that they are more or less equal in gain and that an imbalance in the former produces a rattier/grindier OD and the LTP is smoother. If you consider the cathodyne PI as requiring the previous gain stage, all three use one dual triode. So whichever PI I employ, I am adding a gain stage to the tweed designs. The 6G3 has the extra gain stage already, but also LTP, fixed bias and higher voltage via the GZ34 rectifier. So it's not easy to isolate the effect of just changing one of those things. Perhaps it just needs to be built to find out...

It is the 'floating' paraphrase design. Merlin describes it in his pages too. My 54 deluxe is a transition C/D circuit, it even has the tube chart crossed out and the new tubes written over in pen for the preamp tube designation! Lily must have just got the new design. It has the newer PI of the D but retains the NFB from the C.

 

[YellowBoots]

Thanks for pointing out the paraphase writing in Merlin’s pages.

You may have some trouble, not sure. In the 5D3 schematic, the cathode of the paraphase valve is unbypassed:

 

[YellowBoots]

Yes. Good catch, should be a coupling cap just after the NFB return. I have added it.

It really needed a coupling cap before the master pot and a separate plate feed from B+. Can you show us your revised design?

 

[capohk]

To avoid high voltage on the pot? Yes, that would be sensible.

 

[YellowBoots]

To avoid high voltage on the pot? Yes, that would be sensible.

That and also high voltage on the following tube grid and also lots of current flowing through NFB circuit etc. No bueno.

 

[capohk]

Please forgive the mess. I finally cribbed this off an old thread. I had a few goes at it before I found this. I’m not sure where the NFB should be inserted. The plate load resistor would normally be 100k, but I wonder if it needs the extra 15k as per the voltage divider in the original circuit.

 

[capohk]

Or another step towards the 6g3 and use the ltpi