Fender 6G6-C Bassman Build

[dwinstonwood]

Hello.
Not counting rebuilds of earlier builds, this will be my ninth guitar amp project. It's actually the amp I wanted to build first. It's the one amp I've always wanted to own above all overs.

I'm building it in the chassis that was originally used for my 6V6 Plexi that I never used. So, I've completely redrawn the Fender layout to work in this chassis. It's also All-Octal. I'm using the 325-0-325 winding of my Hammond 290AX with JJ 6V6S's, so the plate voltage under load will probably be higher that a DR, even above 420DCV. I want headroom.

Here's my schematic and a photo of my progress (I like to solder in components after all of the underlying wiring is finished).

Note: The 6SL7 datasheet indicates a sort of low maximum Cathode to Grid voltage of 90V. This amp has a Cathode Follower. So, I'm elevating the Heater voltage on top of ~60VDC tapped from a voltage divider + filter cap right after the SS rectifier.

[dwinstonwood]

I'm plugging it in and testing things as I go. I had to fiddle around with the bias range resistor; this PT does not have a bias tap. I got the negative voltage to range from -26VDC to -46VDC; so, I'm at about -36VDC with the pot wiper centered. I seem to remember that this range is good for 6V6's.

I'm still waiting on my filter caps to arrive. But, I tack soldered a 30uF/500V cap into the reservoir location on the board and I measured 482VDC unloaded at that point (B+ node A). That looks good.

My 6.3VAC heater supply is riding on top of 59.8VDC created by the voltage divider circuit, so that's darn close the the 60VDC I guesstimated.

[dwinstonwood]

Almost done. I still need to add the RG174 grid/input wires. And, go over everything one more time.

[Esquirefreak]

Very neat!

/Max

[passaloutre]

I don’t know enough about the different 6G6 versions to say much, so I’ll ask, why the “C”? I feel like I hear more people talk about the B.

It looks like a very clean build nonetheless. What speakers do you plan to pair with it?

[Wonder2121]

What value bias range resistor got you to those voltages? Just asking cus I'm also coming of secondry...Be handy if you could post updates schem..

[Platefire]

I second it on the "neat work" 

[dwinstonwood]

Thanks everyone!
This was a challenging build because there was not a lot of room to get in with the iron.

My voltages are good. I will post an updated schematic tomorrow.

I have a microphonic tube. I think it might be the 6SC7 phase inverter. If so, I'll rewire the socket for two cathodes and use a 6SL7.

The Bass channel is very bright and Fendery with the Treble and Bass at noon. The Normal channel is "jazz" with tones at noon, but setting the Treble at 1 O'clock and the Bass at 11 O'clock sounds great!

Here are my preamp voltages compared to the Fender 6G6-B schematic numbers in parentheses:

V1-A Plate - 120 (135), Cathode - 1.4 (1.1)
V1-B Plate - 234 (230), Cathode - 122 (136)

V2-A Plate - 141 (150), Cathode - 1.4 (1.3)
V2-AB Plate - 134 (150), Cathode - 1.2 (1.2)

V3-A Plate - 183 (190), Cathode - 2 (1.8)
V3-B Plate - 220 (230), Cathode - 2.1 (1.9)

V4 A Plate - 271 (280)
V4-B Plate - 288 (300)
Cathode - 36 (30)

And, B+ voltages:

Bass channel B+ - 235.5 (230)
Normal channel B+ - 365 (355)
Phase inverter B+ - 416 (410)

The JJ 6V6S's are running with 438V on the plates. The schematic shows 5881's with 428V on the plates. I have them at about 55% dissipation with 17mA.

Once I get the microphony problem solved it's going to be a very nice build. At lower volumes it already sounds good.

[dwinstonwood]

I found the microphonic culprit. It was a supposedly NOS 6SL7 branded for IBM. I put an RCA in that socket and the weirdness stopped.

Here is a photo of the Normal channel settings that sound insanely good. It's a Marshall chassis, so right to left: Volume, Treble, Bass, Presence.

[dwinstonwood]

Wonder2121, the wall voltage was a bit higher this morning when I took this set of measurements. Anyway, here's the schematic with the final resistor values.

passaloutre, I guess I chose the "C" variant just to do something a bit different. But, Fender must have thought that there some little tweaks they should make; I did read somewhere that the "C" revision represented changes made to the Export version, but I don't know if that's factual. I'd love to know what version was in Abbey Road Studios; most references seem to claim it was a '64, possibly a "C"?

[DummyLoad]

I found the microphonic culprit. It was a supposedly NOS 6SL7 branded for IBM. I put an RCA in that socket and the weirdness stopped.

Here is a photo of the Normal channel settings that sound insanely good. It's a Marshall chassis, so right to left: Volume, Treble, Bass, Presence.

Maybe IBM branded tube was likely not checked for microphonics, probably was destined to be used as logic element?


--Pete

[dwinstonwood]

The Normal channel is flawless as far as early 60's preamps are concerned. I can't imagine changing anything. Maybe this is why Mr. Setzer built a sound and a career on it, and why George insistently held on to the one "they used in the studio" until his death.

But, the "C" version of the Bass channel is sad. No Bass, no low Mids with a guitar. I no longer own an electric bass, but maybe this channel conformed to the bright, clicky/chirpy sound that was popular on those variety TV shows back "in the day."

I'm changing it back to the "B" Bass channel. One pot, two resistors and a cap. Four parts. I'll report back.

[DummyLoad]

Seems to me the one to build is the 6G6 except with SS PS instead of the GZ34 rectifier. The 6G6 bias PS would be the way I would go, but with 16-22uF caps instead of the 8uF.

Also, with the 6G6, all four triodes are not on the same power rail as with the 6G6A,B & C revs - Perhaps made simpler construction of A,B, & C revs.

If I were to build one, I'd probably build a 6G6 with a dual rectifier arrangement, both have their merits. Perhaps the Abbey Rd. amp is/was a 6G6 modified with a SS PS and why it was so unique?


--Pete

[Wonder2121]

The 10k spike resistor in series with the 1N4407 on the CF isn't needed according to Merlins latest posts.
The 100pf snubber across the PI plates seems high, ppl usualy opt for 47pf unless you were trying to cut some highs.
Guessing your in Europe you might want to switch live & nuteral on the mains...
Raising screen res to 1k might be best to drop voltage below your plates..

[UncleRuckus]

Noob here - but the full-wave bridge to capacitor style of power supply has a peak initial voltage around 916V?
Not sure if the 500V initial cap is going to like that for long - maybe a thermistor would cut that down, IDK.
J 

[DummyLoad]

Noob here - but the full-wave bridge to capacitor style of power supply has a peak initial voltage around 916V?
Not sure if the 500V initial cap is going to like that for long - maybe a thermistor would cut that down, IDK.
J 

Not a FWB, it's a full wave center tapped arrangement. Secondary is 650V end to end, or 325V-0V-325V for each 1/2 primary with CT grounded, so each 1/2 primary is 1.414.x 325V. Operating each 1/2 secondary 180deg out of phase, so pulsed DC output freq. is 100/120Hz depending on line frequency.


--Pete

https://www.hammfg.com/electronics/transformers/rectifier

[dwinstonwood]

Thanks for the comments/interest.

I converted it back to a 6G6-B.

But, I ended up having two more microphonic tubes. So, I just installed all new JJ's purchased from Eurotubes. I'm using a JJ 6SN7 instead of the old metal 6SC7 for the PI. The other three preamp tubes are JJ 6SL7's. And, two matched JJ 6V6S's.

I can bang my finger on the chassis and there's no ringing. I can also tap my fingernail on all of the tubes with no microphonic noises.

With the old tubes there was a noticeable hum and also some hissing. With the six new JJ's the hum has almost disappeared; same with the hiss.

I guess I learned a lesson that so-called "NOS" tubes are hit or miss. 

Anyway, the Bass channel now sounds much, much better. I didn't change anything on the Normal channel, except one of the mixing resistors before the PI.

[Loomer]

Extremely neat looking build and one that definitely piques my interest as I've been pondering a Normal channel-only, 6V6 version of the 6G6-B!

How do you find the bass channel on guitar? Worth including or makes sense that I might sacrifice it in the name of simplicity?

[dwinstonwood]

Extremely neat looking build and one that definitely piques my interest as I've been pondering a Normal channel-only, 6V6 version of the 6G6-B!

How do you find the bass channel on guitar? Worth including or makes sense that I might sacrifice it in the name of simplicity?

Thanks Loomer! I say build both channels. It's not that much more work/expense, and you will probably wish you did down the road.

Now that I converted back to the "B" version I really like the Bass channel. I need to play the amp more at higher volumes to arrive at a valid opinion, but I'm pretty certain I'll use both channels. So, I'm glad I built both.

Here's a final "gut-shot" and my PDF drawings for this project. I'm really happy with the voltages and the JJ tubes.  It's a good, solid, and quiet amp.

[dwinstonwood]

There's one thing that's got my attention, and I'd appreciate opinions:

The B+ node for the Phase inverter - "node C" - measures 412VDC.

The 6SN7 plates measure 180VDC and 176VDC.

That's a drop of 232 volts across the 82K 1/2W plate resistor, and a drop of 236 volts across the 100K 1/2W plate resistor.

If my calculations are correct, that's around 2/3 of a watt for each of those 1/2W resistors.

I'm thinking I should swap in 1W resistors. It also seems like the 6SN7 draws more current that typical PI tubes.

What do you all think?

[Loomer]

Thanks Loomer! I say build both channels. It's not that much more work/expense, and you will probably wish you did down the road.

Now that I converted back to the "B" version I really like the Bass channel. I need to play the amp more at higher volumes to arrive at a valid opinion, but I'm pretty certain I'll use both channels. So, I'm glad I built both.

Here's a final "gut-shot" and my PDF drawings for this project. I'm really happy with the voltages and the JJ tubes.  It's a good, solid, and quiet amp.

Thanks for the feedback! That's my hunch, just do the work and don't look back, but seen as I'm very much a newbie in the amp building game the temptation to simplify is there!

Thanks a lot for your documentation, extremely well laid out - almost Sluckey-esque, dare I say! While I'll be looking to use ECC83's in the pre's I expect the core changes will pertain wiring and any value adjustment you might have made to accommodate the octals should be a straight value swap anwyay...

Lastly, if I were to go the whole hog, I'd also be tempted to add the bright and deep switches from the following models... Values are different around the eq but the topology the same so it'd be an easy addition...

[DummyLoad]

There's one thing that's got my attention, and I'd appreciate opinions:

The B+ node for the Phase inverter - "node C" - measures 412VDC.

The 6SN7 plates measure 180VDC and 176VDC.

That's a drop of 232 volts across the 82K 1/2W plate resistor, and a drop of 236 volts across the 100K 1/2W plate resistor.

If my calculations are correct, that's around 2/3 of a watt for each of those 1/2W resistors.

I'm thinking I should swap in 1W resistors. It also seems like the 6SN7 draws more current that typical PI tubes.

What do you all think?

Yes to 1W - If you're going to use the 6SN7 permanent, even with 7025 you want 1W plate resistors, 6SN7, however, with that much much current flow, (≈5.1mA | 4.2V/820Ω) that explains the higher Vrp and Vrp'. Assume for simplicity that 5mA is split equally for each triode, so then 2.5mA flowing through 82KΩ, that's 205V across that resistor, .

Typically low mu tubes used as a LTPI perform best with lower Rp and Rp' - 20-50ish KΩ with shared Rk bias set to 3-4mA. With your bias and Ra set to 82K, for best blance Ra' wants to be ≈120K should give best output swing symmetry. You'd do better to rewire for 6SL7 socket with your current operating values.

In short, a 6SN7 in place of the 6SC7/SL7, your LTPI is operating kind of wonky, but sometimes wonky sounds good, if you like it, leave it as is. In my experience, a better balanced LTPI sounds better overall. Nother note to consider, the 6G6 uses tubes with mu of 100 so your overall gain structure is not nearly the same, and with a 6SN7 as a LTPI, even less so. For a quick 6SN7 balance fix try a 820Ω > 2.2KΩ for shared Rk, for Rp' 100KΩ > 120KΩ.

IME, for US made NOS, the best microphonic characteristics are typically with the RCA 5691 sub for 6SL7, 5692 sub for the 6SN7, the Sylvania green & some yellow logo chrome top worked better in my experiments. Another tube that works good with audio that was designed for low microphonics and extreme duty is the Raytheon 2C52, however, it employs a 12.6V filament. Prices for NOS 5691 and 5692 are insane, so maybe try the Sylvanias, they're still "affordable"?

--Pete

[dwinstonwood]

Thanks for all that info Pete.

I put the least microphonic 6SL7 I have - an RCA blackplate - inthe PI socket and the voltages are darn close to the Fender schematic using a 7025.

Inverting plate: mine - 295; Fender - 300
Non-inverting plate: mine - 289; Fender - 280
Cathode: mine - 36; Fender - 30

So, now I have a 117V drop across the 82K for 1.4mA and about 0.17W and a 132V drop across the 100K for 1.4mA and about 0.19W.

I think the amp breaks up sooner with the 6SN7. Maybe I'll hook up the scope tonight and see if and where that breakup is happening.

Anyway, I just went ahead and ordered another JJ 6SL7, and I'll find a use for the 6SN7 down the road.

[dwinstonwood]

In the meantime I'm reading/working through Aiken's white paper on designing an LTP for a 6SL7:
https://www.aikenamps.com/index.php/designing-long-tail-pairs-the-load-line-approach

[dwinstonwood]

Finally, I got the last bits to finish this project, including a 6SL7 PI.

I've been playing it all afternoon through the Warehouse G12C/S in my 6G3 cab. I think it's an amazing sounding amp. Both channels are totally Fender-Great, but different.

The Normal channel is bright with tone knobs at noon, but in no way painful. Exactly what you'd want for Rockabilly.
The Bass channel is fantastic with tone knobs at noon, volume at 11 O'clock (this is at home, circa dinnertime), and presence dimed. In fact, this is may be the most useful Presence control I've built.

The G12C/S is right at 100db efficiency, so at lower bedroom volumes it lets you hear everything. But, it's a quiet amp at idle with single coils.

I'm glad I went with ~450VDC on the JJ 6V6S plates. I think this design wants to be somewhat high-strung to really bark.

[DummyLoad]

Finally, I got the last bits to finish this project, including a 6SL7 PI.

I've been playing it all afternoon through the Warehouse G12C/S in my 6G3 cab. I think it's an amazing sounding amp. Both channels are totally Fender-Great, but different.

The Normal channel is bright with tone knobs at noon, but in no way painful. Exactly what you'd want for Rockabilly.
The Bass channel is fantastic with tone knobs at noon, volume at 11 O'clock (this is at home, circa dinnertime), and presence dimed. In fact, this is may be the most useful Presence control I've built.

The G12C/S is right at 100db efficiency, so at lower bedroom volumes it lets you hear everything. But, it's a quiet amp at idle with single coils.

I'm glad I went with ~450VDC on the JJ 6V6S plates. I think this design wants to be somewhat high-strung to really bark.

What did you end up with for component values for the 6SL7 LTPI?


--Pete

[dwinstonwood]

Pete, I tried my best to follow Aiken's load line guide (that I linked to above). I basically ended up with cathode and tail resistor values very close to his: 680R for the cathode and 47K for the tail.

So, that's what I went with. 82K plate load resistors, and 680 cathode, and 47K tail.

I'll post the new PI voltages and an updated schematic when I get a chance to measure them...

David

[dwinstonwood]

So, Aiken's method is intended to produce a closely balanced LTP. He sets it up to drop between 25% and 30% of the plate supply voltage across the Tail resistor.

My B+ PI node is ~432VDC. So, when I measured the voltage drop across the Tail I was surprised at first to see 131 volts since Fender was only dropping 30-something volts on the 6G6-* schematics. But, 432 * .3 = 129.6 volts. I'm dropping ~30% and the LTP halves are well balanced even though I'm using two 82K's and not the typical 100K/82K plates resistors. Inverting side is 317.3V and the non-inverting bside is 323.1V, a 5.8V difference.

Here's the updated schematic PDF:

[dwinstonwood]

I can't stop futzing with it!

I increased the PI B+ dropping resistor from 4.7K to 6.8K which brought the PI voltages down a bit.

Now: 470 Cathode resistor and 22K Tail (like the AA864 and AB763).

Both plate load resistors are still 82K.

Inverting plate: 272.6V
Non-inverting plate: 274.2
That's only 1.6V difference.

89.4V dropped across Tail resistor. That's just a bit less than 22% of the 412V B+ PI node.

I'm going to live with this for a while. Sorry for yet another schematic, but here's what I'm calling the "final version" for now.

[dwinstonwood]

I've spent the last three hours playing it through a Celestion Creamback in an open-back cabinet.

The Bass channel is so good with my Tele's, especially with the Presence dimed. So full and rich sounding across lots of different knob settings. It almost feels like a tube rectifier.

The Normal channel is brighter, tighter, and harder. It would cut better in a band, I think.

I love the Bass channel at home and at night.

Next is to use some pedals.

[shooter]

 
when you get a build to the point you can't stop playing, it's time to button it up, call it good, and break out a clean sheet of paper to start noodle'n ….that next build!

[dwinstonwood]

...it's time to button it up, call it good, and break out a clean sheet of paper to start noodle'n ….that next build!

Agreed. I'm calling this project done. No disappointments with the B-version I ended up building.

If interested, here are the drawings including schematic, layout and turret board drill template.

Thanks for all the input and interest in this project!

David

[bmccowan]

Thanks for sharing the drawings David, they, like the amp, are nicely done. Since you are on an octal kick and I know you will be looking for a next project. A couple ideas that turned out well.
An all octal 6v6 Clubman.

And a sort-of Ampeg circuit I built in a an old Knight chassis: Peggy Knight

[dwinstonwood]

Ohhh, another Octal Pentode! Cool:
https://frank.pocnet.net/sheets/093/6/6SH7.pdf

Yes, I'll need to use one of those in a project. I used a 5693 Octal Pentode in a the amp I built before this one; a very cool looking bright red metal tube.

Thanks for the ideas bmccowan. I have a big, heavy 360-0-360 PT out of a Kalamazoo Bass 30, its matching 6.6K OPT, and its pair of 7591-A's. So, my next project will more than likely be some kind of Octal preamp (with a pentode!) for that power amp.

David