5F6A bias (and now feedback)

[Cliff Schecht]

I built a 5F6A to replace the Bassman RI that was stolen from me. This was another freebee build so I guess it evens out in a way.

My only deviations from the original schematic are adjustable fixed bias and a 300-0-300V, SS rectifier configuration to give a 430V like the original amp sees. I was running some calculations on the biasing (stated at -48V) and found that with my HT of 430V, the plates are only dissipating about 14W. My calculations set me at a bias of -40V with a plate dissipation of 21W (70% of 30W for a 6L6GC) and -30V for a set of EL34's (17.5W at -30V).

So why is the Bassman output stage biased so cold in the original design? Does this help with the characteristic "clean" sound of the Bassman or do people usually bias these hotter?

[FYL]

The 5F6A was designed for 5881S (6L6GBs) with only 23 W plate dissipation.

[Cliff Schecht]

Makes sense. I guess I need to get out of the habit of always biasing using the pentode curves. The screens are close enough to the plate to treat the tube like a triode and bias accordingly. I need to redo my calculations..

[Cliff Schecht]

With the triode curves I get about -43V for the 6L6GC (21W plate diss.) and the same for EL34's (17.5W plate diss.). Pretty awesome that I can swap tube sets without having to touch the bias!

[martin manning]

Cliff, can you describe how you're using the triode curves to find the bias point?

[Cliff Schecht]

Feedback question: The original schematic uses a 27k feedback resistor off a 2 Ohm tap. What should I be using for the 8 Ohm tap? I have 100k in there right now.


Also, I've read a bunch of stuff about how the commonly available schematic isn't what is considered the "holy grail" 5F6A. How true is this? Is it worth it to deviate away from the original schematic?

[tubeswell]

Feedback question: The original schematic uses a 27k feedback resistor off a 2 Ohm tap. What should I be using for the 8 Ohm tap? I have 100k in there right now.

Assuming a Pr:Sec turns (/voltage) ratio of 44.72:1 ( based on 2000:1) for a stock tweed bassman OT, where the NFB voltage divider is 5k/(27k+5k), an input voltage off the secondary of say 12.63VAC RMS (based on a 565VAC RMS, or 800V PP swing on the primary) results in 1.97VAC at the top of the 5k pot.

Your OT would have a voltage ratio of 22.36:1 (based on 500:1 imp ratio). Therefore the voltage swing on your secondary would be 25.27VAC with the same 565VAC RMS swing on the primary. Therefore to get the same ~1.97VAC at the top of the 5k pot, the upper resistor in the voltage divider should be about 56k to get 5k/(56k+5k) = 0.082 (x 25.27 = 2VAC.) - According to my rough science

[Richie]

Also, I've read a bunch of stuff about how the commonly available schematic isn't what is considered the "holy grail" 5F6A. How true is this? Is it worth it to deviate away from the original schematic?

I would think the" commonly available schematic" is of the later amp.
All the 'Real" early ones i've had open on my bench had different tone circuit cap and slope resistor values,and PI was different in the values used.
Probably too, they used the Tirad transformers,Astron caps,and the parts from that era.. Remember, it was a bass amp,that guitar players stole from their bass players..

The early ones leaned more to bass amp values,and used a 6.8k tail of the PI instead of 10k in the "commonly available schematic" The cab baffle also changed.

Oh, i would also add. Holy Grail.. when you service amps,you see ones that need work done. If you take 2 identicle amps, one working fine.against one with leaky caps,old worn tubes, CC resistors way off spec, bass played through the speakers for years.. to whatever. Who knows why one might say one is the holy grail compared to one next to it. Until the one was serviced and repaired,then it might be the holy grail.

If your going to play guitar through it,i'd probably go half and half. Or look at the JTM 45 would be much easier.

[sunnydaze]

This spreadsheet is supposed to do the NFB calcs.

I think I got off this site about a year ago.

mike

[martin manning]

Feedback question: The original schematic uses a 27k feedback resistor off a 2 Ohm tap. What should I be using for the 8 Ohm tap? I have 100k in there right now.

Simple NFB/Presence scaling rules for changing OT Zsec at same Zpri:

Voltage Ratio = (New Zsec/Old Zsec)^0.5, so the 8-Ohm output has 2x the voltage of the 2-Ohm output. Change the divider to reduce FB voltage by half... the feedback (series) resistor should be ~twice as big = 56k with the same 5k to ground.

The presence cap might need to be adjusted to get the same frequency roll-off. It's value should change in inverse proportion to the parallel combination of the series and shunt resistors. In this case there is not much difference, (56k//5k)/(27k//5k) = 1.09, suggesting use of 0.09uF vs. the original 0.1uF.

If you want to reconfigure the network to avoid "scratchy presence pot" syndrome, replace the pot with a fixed shunt resistor, and place a pot wired as a variable resistor in series with the presence cap from the top of the shunt resistor to ground. The pot should be 5x the size of the shunt resistor so the response is essentially flat with the presence turned down all the way.

Also, I've read a bunch of stuff about how the commonly available schematic isn't what is considered the "holy grail" 5F6A. How true is this? Is it worth it to deviate away from the original schematic?

Well, the "commonly available " schematic (Pittman book, e.g.) is dated 1957, and the "holy grail" 5F6-A is said to be the 1959, so I can imagine that there were some component tweaks, quite possibly undocumented, in that time frame.

Edit: Looking at the 5F6-A (September 1957) vs. 5F6 (June 1957) , the only differences I can see are the change to a GZ34 rectifier (from 83), the screen resistors increasing to 470R (from 100R), and the elimination of the 1k5 grid stoppers on the power tubes. Transformer and choke part numbers are all the same.

[Richie]

and the A version had the 47pf cap between the PI plates..

I found a pic of one of the amps. You can see what was used. I've seen a few like this. I don't think you'll see a schem or layout that shows this as it is.Most all show it as having the .022 and 56k slope 10k tail etc..
As you can see those are original solder joints.

[martin manning]

and the A version had the 47pf cap between the PI plates..

I found a pic of one of the amps. You can see what was used. I've seen a few like this. I don't think you'll see a schem or layout that shows this as it is.Most all show it as having the .022 and 56k slope 10k tail etc..
As you can see those are original solder joints.

Both the 5F6 and 5F6-A schematics show the 47pF across the PI plates, but now that I've looked again, I noticed the FB goes to the top of the Mid pot on the 5F6, which makes it positive FB!

Re the tone stack and PI tail values, I found the same in the gut photos I have. I see 100k slope, 0.02 mid cap, 0.1 bass cap, still 250pF treble cap(?), and 6k8 PI tail. Is that it?

[martin manning]

Here's the '59 reissue. Feedback resistor is 27k, no-scratch presence wiring, small grid stoppers (47R), and tone stack and PI tail are like the photos, not the old schematic. This has been interesting.

[Cliff Schecht]

I appreciate the help with the feedback resistor guys. I'll pop in a 56k resistor and call it done.

I used 1.5k grid stoppers (2W CC's for the strong leads) on my build just out of habit. Would there be an audible difference if I removed these or should just I leave them in there?

The 5F6A schematic I built off of has a 56k/.022/.022 configuration. The RI has a 100k/.1/.022 configuration which gives a more scooped mid, bass focused sound. The JMP50 has a 33k/.022/.022 setup with a 500pF treble cap giving more pronounced mids and highs. I think the setup I used, the 56k/.022/.022 configuration, gives a nice mix between the two extremes and should allow for enough wiggle room to get something between a RI and a JMP50 type sound.

I could always add a DPDT switch to choose between the JMP50 and 5F6A style values too, but this seems unnecessary really for my needs. I'm a set and forget type of player so if I can get a good sound with my current setup, I'll probably leave it.

Once again I appreciate all of the help guys .

[rp]

Noting the bit of success Jim Marshall had with his clones I'd say he had one of those guitar player holy grail Bassmans in his hands. Then the JMP50s had their changes and were 47K then 100K to the 8 ohm tap. The guitar players just got gladder and gladder. So, lots to try.

I was in your shoes and decided to copy the earliest JTM45 with my build (2x.022, 56k, 250, 10K, 27K, 16ohm tap) and used 5881s. Such great fun I've no desire to change it, but I'm itching to build a JMP50.

Best would be to use an OT with 2/4/8/16 ohm taps to play with - maybe have Heyboer make it. I used old Traynor iron and only have 8/16. I'm curious what the NFB on the 2 ohm tap would sound like with 2x12s.