5f2a Planning -- Whadda ya got?

[JuneauMike]

So I'm mostly posting here as a way of thinking out loud. I'm trying to get motivated to build a 5f2a out of leftover parts with new parts that I've not yet bought. Just picked up a chassis from Weber to help get me committed to the project. I'm finding that transformers are the Wild West these days with a lot of old standards no longer around; so I'm going to go with some Hammond iron (P-T290AX for power and P-T1760CP for output).

I've got a really nice 5F1 scratch built that I've enjoyed over the years as my "couch amp." My son will be leaving the dorms next year and living in his own apartment so I feel comfortable about giving him the Champ and a guitar to occupy his down time. So I think a Princeton would be a nice replacement for dad. (I have never been crazy about the voicing of the Champ and I had thought about moding it to GA-5 specs to see if I liked that better, but then I fell into a JHS Haunting Mids and that turned out to be all the mods I needed since it gave me more control over the midrange output -- which can be kinda annoying for an amp that is EQ'd so flat.)

I've seen @King Fan's excellent layout that I think I will use with two main modifications. I'd like to move the output bias resistor and cap to the end of the power section in the same way as a 5e3 (it just makes sense) and I'm going to install an adjustable NFB on the underside of the chassis as a set-it-and-forget-it option.

Anyway, that's the approximate plan. Phase 1 is to buy some lumber when the weather starts to improve enough to work in the garage and build the cabinet first. I'll just use Mojotone's dimensions for a 5f2 with 10-inch speaker. I've got a bunch of tweed on-hand so that should cost next to nothing. And I'll use a Weber alnico speaker for the build (I've used them on my builds).

It's been a few years since I've built a whole amp so hopefully this will be fun. Anyone got helpful thoughts on this circuit? What range of 5w resistors do you like to have on hand for final biasing? etc. etc.

 

[chas.wahl]

I'm currently building a 5F2-A in a head cabinet, with the following "special features":

Ground bus topology: a single on-board bus from PT's HT center tap all the way to circuit ground to chassis near input jack, with modified star nodes at smoothing caps, per Merlin Blencowe (ValveWizard) "Grounding" chapter. Along with this, I'm following his Fig 15.5 schematic, putting the preamp smoothing caps into the circuit close to the elements they serve, and vice-versa -- for instance, the location of the 6V6 cathode resistor and cap.

Adjustable NFB, per Uncle Doug (10k fixed resistor, 50k linear pot), with some nail polish on the end of the pot track to totally negate NFB at the "high resistance" end

Switchable 1uF for stock 25uF capacitor on V1a cathode bypass, per Dave Hunter "Two Stroke" amp design, but a bit higher-value cap (1 uF rather than 0.47). Rather than muck too much with the chassis control side, I'm using the #2 input hole for the (SPDT) switch, and moving the 2nd input to the bottom of the chassis -- using that jack's lugs to anchor the two grid resistors at the tube socket pin.

Two voltage supplies from power transformer: 550 VAC or 650 VAC, using DPDT switch (need to remember to turn the amp off and de-power filter caps before using this)

An SPDT switch next to the above, that inserts or bypasses an aluminum-clad 25 W resistor, to knock ~10 to 20 V off the B+ (ditto)

Choke between the first two (16uF) filter caps, like 5F2: 4 H, 50 mA, 167 DCR
(see https://el34world.com/Forum/index.php?topic=19378.0)

Screen grid resistor (470R, 3W) from 6V6, pin 4 to B+ node
(see https://el34world.com/Forum/index.php?topic=21489.0)

Slightly larger preamp smoothing caps: 10uF rather than 8uF, because that's what was available from F&T

Output transformer with 3 impedances: 3.2-8-16 ohms and rotary impedance selector switch to single output jack

Blencowe "humdinger" center tap on heater circuit for hum-canceling; grounded to 6V6 cathode to raise heater ground reference (to about 19 V)

Diodes protecting B+ bus from rectifier failing shorted, per RG Keen "Immortal Amp" series.

Provision for adding resistors to both 6.3 V and 5V heater circuits, to lower voltage to spec, per Merlin Blencowe "Heater Supplies" article.
(See http://www.valvewizard.co.uk/heater.html)

I was going to add cap discharge bleeder and open speaker circuit protection resistors, but on reflection I think I'll omit them.

Chose a 5 W Ohmite wirewound for the 6v6 cathode resistor

I can provide a schematic and layout, if you'd like, such as they are.

 

[JuneauMike]

What ranges for the 5w would you buy to have onhand? I think the schematic calls for 470ohm, but I'm not sure where the voltages will fall at. The PT has 275 secondaries as an option so I shouldn't have any runaway B+ to deal with (knock on wood).

I think I'm going to do a single bus bar that I can snip to separate power/pre sections if I get noise. But I'm kinda done with laboring over grounding schemes anymore. I've got too many grey hairs worrying about grounding schemes and at the end of the day, none of them work and all of them work, depending on who you talk to.

 

[King Fan]

Great project. Good for you. I'm glad someone is finding some use for all those 5F2a layouts I tried to draw.

I'd like to move the output bias resistor and cap to the end of the power section in the same way as a 5e3 (it just makes sense)

I like your idea to do this mostly to talk it through in your head, so I'm not gonna try to talk *you* out of anything. In fact, my drawings were intended to try to improve grounding and other details *while still using a standard board* -- to make 'em accessible to beginning builders. As you know, the bias ground, done right, should return to the second, screens, node, a long reach on the 5F2a board.

But the classic 5E3 layout, though more logical, still has wiring issues. If I were gonna do a custom 5F2a board, I'd consider putting the bias just downstream of the reservoir cap, like Rob does on the optimized 5E3.

This makes it easier to put the HV CT right on the reservoir negative, and to put those guys first on the bus (whether single or split); the bias is still close to the output socket. And hmm, I'm sure you've considered a unified 32uF reservoir cap. There may be other 5F2a board tweaks that'd be nice if not needed.

 

[JuneauMike]

Great project. Good for you. I'm glad someone is finding some use for all those 5F2a layouts I tried to draw.



I like your idea to do this mostly to talk it through in your head, so I'm not gonna try to talk *you* out of anything. In fact, my drawings were intended to try to improve grounding and other details *while still using a standard board* -- to make 'em accessible to beginning builders. As you know, the bias ground, done right, should return to the second, screens, node, a long reach on the 5F2a board.

But the classic 5E3 layout, though more logical, still has wiring issues. If I were gonna do a custom 5F2a board, I'd consider putting the bias just downstream of the reservoir cap, like Rob does on the optimized 5E3.

View attachment 1079627

This makes it easier to put the HV CT right on the reservoir negative, and to put those guys first on the bus (whether single or split); the bias is still close to the output socket. And hmm, I'm sure you've considered a unified 32uF reservoir cap. There may be other 5F2a board tweaks that'd be nice if not needed.

Yeah, I am going to stick with twin filter caps at B+1 just like the layout. I only see F&T 16ufs in 475v but I think that will be enough. Using two 22uf filter caps at B+1 would probably stress a 5Y3 at startup, wouldn't it? (or does it only see the first filter cap? Hmm.)

I really haven't given much thought to moving the output bias R yet. I really am just going to plunk it down at the end of the board (just like the 5y3) and sort it out on DIYLC later (the output grid resistor has to be accounted for so I'd probably run a jumper to ground it at the B+2 negative to keep it within the power section's ground group).

 

[chas.wahl]

The 5Y3 will see two filter caps without either a choke or resistor between them as a single big capacitor. @Lowerleftcoast suggested, in a thread I started awhile back, that putting a pair (for balance, also suggested by Merlin Blencowe) of fairly small resistors between two caps would isolate the 2nd from the first -- I guess what you're trying to do is buffer the farther cap from being part of the inrush "suck" with something resistive/inductive.

The 6V6 cathode resistor I have is 588-45F470E at Mouser (lop off the "588-" for Ohmite's product number). The Ohmite 200-series Brown Devil (used on orig. 5E3s) is cute too, and available in 5-1/4 W. Both are more expensive than a Xicon cement type, though.

 

[JuneauMike]

I guess what you're trying to do is buffer the farther cap from being part of the inrush "suck".

Exactly what I'm trying to do. Though maybe it won't be an issue with this small of an amp. Like I said, I haven't thought through all of this build yet. Just spit-ballin' right now.

 

[chas.wahl]

Well, common wisdom seems to be that a 5Y3 should be able to tolerate 40 uF connected, but don't quote me on that if things go south.

Here's muh layout I'm working with, such as it is, warts & all:

Chassis is a standard 5F2-A chromed steel item from AES, modified with much gnashing of teeth and the assistance of a drill press, a step bit, a Dremel tool and worrying with files and other caveman tools. Board is also from AES, one of those perforated epoxy ones (holes slightly too small for eyelets, but fine when drilled out to 1/8" ø), similarly modified.

 

[King Fan]

It's a wonderful, simple amp, and it'd actually still sound great the way Fender built it. Don't let me complicate things; enjoy your choices. I agree the output grid resistor needs some thought -- would it be possible to move it to the socket, like we do on the 5E3? Don't follow me, though, that's beyond my pay grade.

Just sitting around the campfire telling cowboy stories, now, I was thinking a ~32uF first cap would work exactly like 2x16 and take up less space and wiring. And Chas is right -- a 5Y3 can handle 40uF. The 30-32uF figures on data sheets are "typical", not max. OTOH, 44 is not 40 -- and you don't need that much, so I wouldn't go there.

I'd want @Lowerleftcoast fully explain the logic of the balanced first filter, but Merlin is pretty clear: "If there is an additional stage of smoothing immediately after the reservoir capacitor then it is beneficial to split the dropping resistance into two parts to create a balanced filter, which will retard ripple current from entering the audio ground." OTOH, this all may be like gilding the hood ornament on a SuperSport Dodge...

 

[Mr Ridesglide]

I think you can get a princeton chassis and cut one hole in the rear for a small 7 pin tube and also one more jack in the front and make a killer Harvard like amp. i did it and did some research for the "correct" lower voltage PT and it was the two best amps i ever made. (people heard them and bought them from me, so i can't answer too many questions about them now)
super cool amps for certain with 3 diff inputs to choose from even!

found the information:
- Chassis by Mojotone - It was ordered as a 5F2A Princeton chassis, then two extra holes were punched to accommodate the extra power tube and extra pre-amp tube.

- Cabinet also by Mojotone - they do a great job of making a finger jointed pine cabinet and applying the tweed. Three coats of shellac cut with denatured alcohol to give it protection and add an amber hue.

- Power Transformer is a Hammond 290CAX; Nearly spot on vintage voltage supply for this amp. Gives it a very "authentic" tone and feel.

- Output Transformer specs: Equivalent to #125A10B & #022913.
Audio Power 12 W (RMS)
Primary Impedance 8,000 Ω
Secondary Impedance 8 Ω

- Filter Caps - F&T
- Other caps are Blue Sozo Molded, Orange Drop, Mallory, and a Sprague Electrolytic for the bias circuit.
- Resistors are mostly carbon comp, with a couple of larger ones being metal oxide for the power rail.
-Jacks are switchcraft.
-Pots are Alpha.
- 5Y3GT Rectifier Tube
- 2 each 6V6GT Power Tubes
- 1 12AX7 Phase Inverter Tube
- 1 6AT6 Preamp Tube

I was largely true to the circuit, but added:
- Bias Pot; makes changing power tubes easier.
- 1 Ohm Resistors from cathode to ground on power tube sockets to make it easier to get the current on the power tubes for biasing.


Speaker is V.O.R. 10" Alnico; I prefer this to the Weber 10A125 or 10S in this amp.

 

[King Fan]

Wait, wait, if we're gonna talk Mike into evolving his 5F2a, I'm gonna suggest the GA-5 path is straighter and shorter!!! But hmm, a Harvard-like amp would be pretty cool...

 

[Lowerleftcoast]

Well, common wisdom seems to be that a 5Y3 should be able to tolerate 40 uF connected, but don't quote me on that if things go south.

FYI The AB764 Bronco sports a 40uF reservoir cap.

fully explain the logic of the balanced first filter, but Merlin is pretty clear

Hey, Merlin said it not me! I'll let Merlin 'splain the *balanced* filter...

I realize *reservoir* caps are set apart in the filtering dialog. There is resistance from the rectifier so it is a First Order Low Pass RC filter. That being the case, when there is not a load in between, the next RC filter is cascaded with the first RC filter making a two pole filter. If both RC filters have the same cutoff frequency it is a Second Order Low Pass RC filter. The advantage is a Second Order filter is the 12dB per octave roll-off vs. 6dB of a First Order filter. Chances are our two Low Pass filters will not have the same cutoff frequency but at some point this two pole filter will roll-off 12dB per octave. Imo, this is a good thing for a SE amp.

(It is not talked about much but the Princeton Reverb circuit has a Second Order filter. Yep, it is that *unused* node.)

 

[tele_savales]

Well, common wisdom seems to be that a 5Y3 should be able to tolerate 40 uF connected, but don't quote me on that if things go south.

Here's muh layout I'm working with, such as it is, warts & all:

View attachment 1079697

Chassis is a standard 5F2-A chromed steel item from AES, modified with much gnashing of teeth and the assistance of a drill press, a step bit, a Dremel tool and worrying with files and other caveman tools. Board is also from AES, one of those perforated epoxy ones (holes slightly too small for eyelets, but fine when drilled out to 1/8" ø), similarly modified.

Really cool layout, good ideas. I think I'm also going to use the single bus on my next build, after thinking heavily on your grounding thread. What program did you do your layout on? Also, will you really ever use the low input on the bottom panel? I get why, it's just that I built an AA764 w a single jack and and a hi/lo switch and it pretty much stays on hi all the time.

 

[chas.wahl]

. . . 'splain the *balanced* filter...

I realize *reservoir* caps are set apart in the filtering dialog. There is resistance from the rectifier so it is a First Order Low Pass RC filter. That being the case, when there is not a load in between, the next RC filter is cascaded with the first RC filter making a two pole filter. If both RC filters have the same cutoff frequency it is a Second Order Low Pass RC filter. The advantage is a Second Order filter is the 12dB per octave roll-off vs. 6dB of a First Order filter. Chances are our two Low Pass filters will not have the same cutoff frequency but at some point this two pole filter will roll-off 12dB per octave. Imo, this is a good thing for a SE amp.

Whoa! Stuff I've never bothered to learn or understand; could we have some references to graphic aids?

(It is not talked about much but the Princeton Reverb circuit has a Second Order filter. Yep, it is that *unused* node.)

I'll have to pay more attention -- haven't developed much rapport with that one. But maybe that will help with my first request.

What program did you do your layout on?

I'm an architect by training (like George Costanza) and work things out in AutoCAD, since it's just more comfortable to me -- not for everyone, for sure.

Also, will you really ever use the low input on the bottom panel?

Mostly will be using single-coils, but the LO input for humbuckers?

I get why, it's just that I built an AA764 w a single jack and and a hi/lo switch and it pretty much stays on hi all the time.

Just trying to keep things so that with everything set a certain way, it's a stock 5F2-A.

 

[theprofessor]

I don't really have much to add, as I generally like to keep circuits close to stock, with modifications being mostly about more filtering and better grounding and such. That said,

- In my opinion, these amps don't need two inputs. You're not using a microphone straight in while you play (I'm guessing). So find something fun to do there. A mini switch for something like a boost or switchable cathode bypass cap.
- If you don't use a humdinger pot (see below), elevate your 6.3V center tap on that 290AX at the cathode of the octal tube.
- I like to increase filtering, but, as you say, not to put too much stress on a 5Y3. I like the reservoir cap bigger, at the least. I used 22uF on the last Champ I made and liked that.
- Since you have room in the floor of a 5f2a chassis, you might consider putting a humdinger pot there. Plenty of room, and why not? This will obviate the need to use your 6.3V heater center tap.
- I really like the transformers you're using. I have used the 290AX a lot, as it has plenty of current, a 6.3V center tap, and two high voltage secondaries at good voltage points. I'd think you'd use the 275-0-275 one. The 1760CP is a great output transformer too. I think it's the paper bobbin version they made.
- Related to voltages: On that Champ I mentioned, I used a 290AX and the 1760CP. I used the 275-0-275 secondary. With a 5Y3, I recorded 348 VDC for B+1 on my startup chart at 124.7 VAC at the wall. I may have messed with the bias some after that, but it should give you ballpark. The Fender schematic gives 340 VDC, I believe, so that's pretty darn close.
- I'd think you might want a range from 300R to 470R on hand for your bias resistor, just in case. A lot depends on your voltages and how strong your power tube is. I ended up with a 360R on my last Champ.
- I don't really like the bias resistor at the far left end of the board, because it can get underneath the lamp. And the bias resistor is the one part you're likely to change. And I don't want to try to get my soldering iron underneath the lamp assembly. The reservoir cap, on the other hand, tends not to need to be accessed as often.
- use a 10" AlNiCo speaker with the amp, as God intended.

 

[chas.wahl]

OTOH, this all may be like gilding the hood ornament on a SuperSport Dodge...

Erm, cultural correction, presuming the reference is to the little old lady from Pasadena: it was a "Superstock Dodge" (426, Max Wedge or Hemi). Don't ask me why I remember that; ain't memory a b***h? Supersport was an upscale version of several Chevy models, around the same time.

 

[JuneauMike]

Wait, wait, if we're gonna talk Mike into evolving his 5F2a, I'm gonna suggest the GA-5 path is straighter and shorter!!! But hmm, a Harvard-like amp would be pretty cool...

I've got a 5f11 Vibrolux which turns into a Harvard whenever I turn the trem circuit down. Ha. This is to replace a 5v1 Champ, that I love.

 

[jsnwhite619]

The higher voltage tap on the 290 can put a 6v6 pretty high even with 5y3. My go-to Tweed Princeton changes are

30-22-22 cap lineup - skip the double first section.

Decrease the 22k dropping resistor to 10k if you want more out of it.

If you stick to single coils, increase the first from right .022uf cap to .033 or .047 if you want a little more dirt. Can get too much with humbuckers.

56k resistor on the negative feedback more of a rowdy 5E3 vibe.

Stick to the lower voltage tap and you can swap any rectifier you want with the 290AX to increase the horsepower, but 5y3 will get you pretty close to 100% vintage Champ/Princeton level on the 275v option.

 

[King Fan]

Thinking out loud, not suggesting. Though I eventually converted my 5F2a to GA-5 resistor values -- easy and to me worthwhile -- I do miss the sweet tweed / SE drive it'd get around 6-8 on the dial. Other than a really subtle footswitch NFB cut with cap, it was pretty much stock. If I built another, I'd try to make sure any sonic mods would be switchable-offable...

 

[Jeru]

What ranges for the 5w would you buy to have onhand? I think the schematic calls for 470ohm, but I'm not sure where the voltages will fall at.

I don't know if this is "right" or not -- but on a couple of ~5w SE builds I've had good luck
wiring one of those 5W ceramic resistors at a value of roughly 50ohm below spec in series
with a 100 ohm pot -- that way I can dial in any value from -50 / +50.