5F6A build

[Zippy]

Whew, I'm still recovering from an INTENSE concert last night.

Buddy Guy playing through two Bassman amps (Fender Replica and "Chicago Blues Box") - burning like Marshalls aspire to.

If you can nail that kinda stuff, Cliff, that's all you'll ever need.

I'm not often left stunned for words but ... oh, my, that ol' man can play.

[Cliff Schecht]

I used to not like Buddy Guy for whatever reason but over the years I've grown quite fond of him, his playing and his overall attitude. A great performer, lotsa charisma and holy shit he can make a clean note sustain for days. I hope I get to see him soon, I already missed one chance this last summer.


Also, for this build, I'm at a bit of a conundrum. The PT I used was labeled at 570V CT for the HV secondary (not on the bell, just taped on the leads). I measure almost exactly this when running off of my houses wall voltage (usually 123-125V). With no load and a full-wave SS rectifier I will be seeing about 400V on the plates. I expect this to drop a good 20-30V after loading which puts my B+ down at the 380-370V range. Now I know Fender schematics are labeled +/- 20% on the voltages labeled but 20% down puts us at 344V from 430V. Since I usually aim to hit the Fender schematic voltages without a few percent, this egregious oversight on my part has me very annoyed. I checked the xfmr on my function gen and got the ratio telling me that I should be expecting 570V from a 117V input, but obviously my math or thinking was off somewhere.

This build is completely wired up and fires up with no tubes just fine. Should I just deal with the lower B+ and adjust the bias/dropping resistors accordingly? Do I switch to 6V6's? I'm not sure what I want to do at this point but one option is to replace the PT, but this would be my last ditch effort as doing so would be a lot of work at this point, especially with my goofy heater wiring (6L6's and 12A*7's run on 12V).

[Zippy]

Cliff,

I'd fire it up and see what it sounds like after a few hours of play time. Then- IF there's an issue of clean/break up, I'd consider whether it's the pre- or power- that's at issue,ie, whether you need to raise voltage to the first stage and, if so, tweak the dropping string.

I've become partial to the KT- series tubes and would certainly consider dropping in a set of KT-66 to taste what that does to the output before I'd rush into swapping out the PT.

Bottom line: I'd listen to the amp and feel the response before I'd start making random changes - but maybe you've a better sense of what it'll do from just looking at it than I do.

Have a loud Sunday and enjoy!

[rp]

Cliff, you like to put amps together from found parts, an adventurous if not a virtuous trait, but you are going to have to put up with the compromises and the serendipity of it all and not expect specific outcomes. I do a sort of a mix of old an new, but I've learned to not compromise on the PT. You really need the PT to be as close as possible to the original voltages, current, and regulation too, ie not over spec'd or underpec's if you want a clone that sound anything close to the real thing.

IMO you just need to bite the bullet and get a 5F6A PT, which is very easy (Mojo, Allen) and you can use any JTM45 PT too, like the Marstran with the dual secs. You can also get Heybore to wind one for you for ~$90 that might drop into the holes you already have, I've had to do this. What do you want more an extra $100 or a good 5F6A?

Other than that since you're an experimenter anyway, why not try it as a 6V6GT amp and see what you get, though the 4.3K isn't optimal, you never know, and you can swap output taps too, if you like it it and you have room and the current it could become a proper 4x6V6 Bassman. Or just temporally convert it to a Tweed Super and see what you get till you find a better PT.

BTW FYI my 5F6A (w/ JTM45 power string) has a 330-0-330V PT that gives me 400V on the plates. I actually wanted ~430V.

OR how about a voltage doubler and a 280V dropping zener!

And hurry up already I waiting to know how that Stancor A3851 sounds

[Cliff Schecht]

Great advice guys. I think I'll fire it up and see how it sounds. I usually don't put in a PT unless I know it's going to work but with this build I've only been able to work on it a few hours at a time and was hoping I could make this PT work. I have considered a doubler and voltage dropper but knocking down 280V, even with an amplified Zener circuit (the one with a power BJT or FET) is going to create a LOT of heat. I might set this build off to the side for now and wait until I either find a proper PT or just deal with it as is. I'm on the low side of Fender's 20% numbers but I'm still in spec .

I have one of my A3851's in that 5E8A build and I love it in there. I took the amp to a guy who has owned a few tweed Twins as well as Supers, Bandmasters and the like (they all had the same circuit for a while) and he said it sounded like a real one. To me that was affirmation enough that this amp (and therefore the transformer) nails the sound I was after. In other words, these transformers sound frickin great in a Fender style amp!

[rp]

I have considered a doubler and voltage dropper but knocking down 280V, even with an amplified Zener circuit (the one with a power BJT or FET) is going to create a LOT of heat.

I was seriously kidding! You're dangerous.

Possible brainstorm: 340V = cathode bias class A 5F6A w/ NOS 5881???

Mine has a cathode bias switch, sounds real sweet and twangy at modest volume. Loud I prefer the fixed bias, but you know more and might be able to tweak it up really nice like this.

[Cliff Schecht]

It's actually not that dangerous. In my tweed Twin I'm knocking a 560V B+ down to 420V with a Zener amplifier circuit. If you heatsink the BJT and give it enough room to breath then it'll handle the voltage drop just fine. But this requires a large heatsink to meet the W/degC dissipation requirements (can be calculated to stay within a safe range) and lots of room, neither of which I have. I got away with it in my 5E8A because I can use the chassis as a heatsink and the rest of the circuit is quite small so it can be tucked in a corner or added in on the side.

I've got some NOS 5881's that I'm itching to throw in this build. Maybe I'll just leave the amp alone for now and run the 5881's hotter (current-wise) to compensate for the lost B+ and still hit the dissipation numbers that Fender does. While IME you can get the best sound out of an old Fender schematic by staying within maybe 5% of the schematic voltages, the Bassman might not sound bad with a lower B+. Haven't ever tried this before but I do know what a proper 5F6A should sound like as I have built a few and had a BMRI for a long time.

What should I drop the bias to with the 5881's? I'm feeling too lazy to plot load lines and calculate my bias (I do this enough at school with MOS and BJT designs). Anybody feel like doing me a small favor? I still need to clean the house and get up to school..

[diagrammatiks]

shoulda pm'd me about caps cliff!

I probably have the values to replace those cde brown molds.

[rp]

I don't know how to read load lines, and my math never works out right anyway, but I can reference a chart. This is from the Torres book. I have no idea how he derived this info so I don't know how valid it is, but here you go anyway.

Also, it's my understanding that true '50s Tung Sol 5881/Sylvania 6L6WGB and most 6l6G/B/C are slightly different animals and might not share the same curves.

If you cathode bias remember to subtract that out from the Vp.

w/ respect to Dan Torres. And here's the orig 5881 data sheet to save you or someone some time:

http://tubedata.tubes.se/sheets/127/5/5881.pdf

[Cliff Schecht]

Aww yeah damn, I forgot! FWIW I like the brown molded mouse turd guys and didn't mind using them in this build though. I had good luck with them in my 5E8A build and every one that I've tested (quite a few now) has tested good for leakage and value. I'm still interested in blue molded caps though as I am building a brown Vibroverb soon and would love to use some authentic caps in that build. I still have lotsa PIO stuff too that I'm definitely interested in trading. I actually have a few bags of PIO's that I've been meaning to post in the For Sale section. I just need to clean my bench/work area (I was recently relocated to the living room by my girlfriend!) and I will PM you.

I don't buy too much into the different caps completely changing the sound of an amp but I do enjoy sourcing old components, testing them and repurposing them for my needs.

I did recently open up my '54 Silvertone 1331 and noticed it was loaded (relatively speaking, its a 6SJ7 into a 6V6) with yellow Astrons. While I've owned this amp for years now, I didn't really know much about Fender tone caps and what era used what until relatively recently. I debated only for a few minutes whether to rip out the caps and use them in a Fender build but I decided against it. The amp is nearly original and while it doesn't have a lot of monetary value (like a poor mans TV front Champ with a tiny speaker), the amp screams blooz and is a great studio amp, cab rattles and all.

[diagrammatiks]

those yellow astrons are beautiful.

Those CDE's are pretty standard for the time. I think they are polyester although they might have been from the time that a lot more manu's were experimenting with diflim.

Some had more success then others though.

but if you have pio's you should just use the pio's in your builds man.

[Cliff Schecht]

Also thanks for the links rp. I will run the numbers later when I have my tablet and One Note open. I do run the curves when I'm biasing because I want to see where the load line crosses the knee (for non-triode strapped pentodes) or how much swing to expect from triode strapped output stages.

What I do is pull up One Note and do a selectable screen shot over the graph of the curves (windows key+S and highlight just what you want to capture, very useful feature!). Then off to the side of the picture I start a little text box with all of my calculations. Again OneNote is nice here because you can do simple math just by typing formulas ending with an equals sign and push space to solve. Then I figure out my load line slope and dissipation numbers, draw them directly on the graph and slide the load line around til I'm happy (another OneNote feature). From here you can usually eyeball the bias and run a few more numbers to verify that you have the right value. This makes things pretty quick as far as getting tangible results that are well organized/archived.

I also use the Duncan PSU designer dealy to predict my operating B+ and other simulation software to build the fixed bias network and get the right values before I build it (especially that resistor before the backwards diode, that one is harder to calculate by hand). This takes a lot of guesswork out of firing up my first builds. Lastly before I fire up with 120V, I hook up the power plug to a function generator (amplitude maxed out, freq set at 60 Hz) and check that I have all of the proper voltage ratios and no shorts. I like my builds to work properly (at least B+ and bias-wise) the first time I fire them up.

[Cliff Schecht]

those yellow astrons are beautiful.

Those CDE's are pretty standard for the time. I think they are polyester although they might have been from the time that a lot more manu's were experimenting with diflim.

Some had more success then others though.

but if you have pio's you should just use the pio's in your builds man.

The only PIO's I don't have a lot of are the 0.022uF guys. I did buy some this summer though so I have enough for a few builds. But I also have a small collection of older type caps (di-films and similar from the 50's-60's) that I like to put in my Fender builds to be slightly more period correct. The PIO's I save for pretty much anything else, like the 2203 or GA-100 builds I have coming up. Another reason I don't like the PIO's I have for Fender builds is the boards I make at home tend to be long and thin and don't accompany the long tubular nature of 600V/1000V PIO's very well (at least the 0.022uF ones I have, 0.1uF ones are shorter and fatter).

[diagrammatiks]

you're talking about the sangamo types with the regular molded outsides right?

there seems to be a lot more of the sealed metal encased ones showing up these days and they are generally a little smaller.

[Cliff Schecht]

Nah all my PIO's are metal encased, epoxy sealed types which I believe are mil-spec. They never go bad and stay well within tolerance but they are physically large IME and must be planned for if used. I'll have the room in my Vibroverb chassis but I'm still going to use up what I can of my non-PIO caps in my Fender stuff so that I can use the PIO stuff in amps that aren't meant to be as authentic.