How Doubling Guitar Amp Power Output Affects Speaker Performance?

[Helmholtz]

Curiosity Satisfied…

First, I want to thank you all for your insights…

I guess I ended-up answering my own question, after thinking for a couple of days on what Enzo posted about power & speaker efficiency.

I think I approached the 2x 6V6's - parallel, single-ended design inquiry from the wrong point of view…
it is not about the amp… is all about the speaker.

The speaker field coil requires some minimum power to produce enough magnetic flux, for the speaker to be able to work. The Gibson amps BR-9 & early GA-9 models, use the Jensen F10-U field-coil speaker, which is what was available at the time these amps were manufactured.

Based on various articles from the web, I estimate that the F10-U speaker field-coil requires about 5.5W to work, or approx. 68V @ 81mA (= 850 ohms not 1000 ohms indicated in the schematic).

An amp with (1-6SJ7 and 1-6V6) OR (1-6SN7 and 1-6V6) will idle at approx. 46~48 mA, not enough for the speaker to work properly. If we used a 6L6 power tube instead, then the B+ would be way too high for the available 450V electrolytic filter caps… 2x 6V6's fit the bill perfectly… lower operational voltage, 80 mA idle current.

I would appreciate if anybody has any specs or info on the F10-U to confirm (or not) my estimates. Thanks.

Sorry, I absolutely don't understand what all this has to do with your original question.

- I thought you were going to use a PM speaker, so why worry about the field coil speaker and its energizing current?
- It is true that the efficiency of a field coil speaker increases with the field current. The generation of the magnetic field doesn't require power though, rather there are power losses caused by the DCR of the coil wire. All this power gets dissipated as heat. I assume there is an optimum field current for the speaker.
- There is no reason why a 6L6 would increase the B+ or need higher B+. Typical plate current in SE class A is around 50mA.
- Nothing wrong with using 2x6V6 in parallel if the PT can supply enough current.

[K Teacher]

Hi Helmholtz,

The original question was "How Doubling Guitar Amp Power Output Affects Speaker Performance?"...
I was trying to understand the benefit(s) (reason) of doubling the output power (2x 6V6 parallel vs 1x 6V6) and supposed there was some benefit other than loudness.
According to theory, an increased of 3dB in loudness is barely perceived by the average listener...

However this has nothing to do with increasing speaker performance... the Jensen F10-U needs around 80mA to work. Like you pointed out, the typical plate current in SE class A is only around 50mA... the easiest/ more economical way to increase the total current to 80mA is to add another 6V6 wired in parallel to the first one...

[K Teacher]

So, I guess my initial hypotheses didn't worked-out...

[K Teacher]

Of course, you can use the same configuration for PM speakers, but the benefit is not there... perhaps this is why this guitar amp output configuration is not used anymore...

[g1]

Some of the earlier Gibsons used a ballast resistor (to ground) after the FC to pull extra current if that was required. (early EH150 for example)
So that would be a much cheaper option. Not saying your theory is wrong, but Gibson did a lot of weird things and changed up mid-stream often. Maybe they thought it was saving the cost of a PI compared to a push-pull 2x6V6 version.

[Chuck H]

Sorry, no specs on the f10-U, but...

The parameters you're questioning regarding field coil speakers are entirely ratio dependent. As to impedance and frequency response relative to inductance these parameters were all able to be satisfied by even a single valve single ended amp in the era referenced. There are a few hi fi designs that demonstrate this.

It's more likely the economy and lack of stringencies of the market that dictated available designs.If the stringency wasn't there, they didn't bother to meet it. But there's another aspect...

considering guitar amplifiers we have to keep in mind that they are tone PRODUCERS, and not tone replicators. Anything coming out of a guitar amp would have been recorded and reproduced by circuits that are purpose built for THAT aspect. Since only studios and aficionados needed the highest reproduction accuracy this luxury would have been fudged or even omitted from instrument amplifier criteria.

That isn't to say that guitar amps from any era don't sound good. This is the very thing they are purpose built for. They evolved with guitars and electronic music. But they AREN'T reproduction amplifiers. So...

Don't worry so much about what is "proper" and maybe focus research on how players like the sound of any given design. How well the speaker is accommodated for absolute accuracy is a problem for recording and reproduction. Which aren't guitar amp tone considerations.

[pdf64]

... the Jensen F10-U needs around 80mA to work... add another 6V6 wired in parallel to the first one...

If you find a site with relevant info, especially if it’s somewhat arcane, please share a web link

A 2nd power tube could have been in push-pull, rather than SE parallel. For a given amp performance, that could save cost on the OT and HT filtering.

[K Teacher]

Hi g1,
Thanks for the info... I never paid attention to the EH-150 before...

Hi pdf64,
I have been looking for info on the Jensen F10-U speaker...
I even wrote to Jensen, but they said that they do not have information on speakers this old on their files anymore.
If I find anything I will share here...

Looking on several places for info on the F10-U speaker, I noticed that, at that time, Jensen had the same field-coil speaker type with different field-coils (more current, less volts and vice-versa, all requiring the same power) to accommodate the amp manufacturer's design.

What I got on the Jensen F10-U speaker used on Gibson's BR-9 & earlier GA-9, so far, is this (not sure if it is accurate):

- 10 in field coil
- Smooth cone (like current production P10Q)
- 8W handling power
- 3~4 ohm VC (voice-coil) impedance
- VC DC resistance ??? ohms
- Sensitivity ??? dB (as far as I know, they did not pubish sensitivity back them...)
- FC (field-coil) DC resistance -- 850 ohm +/- 5% (NOT 1000 ohms as annotated on the schematic, that is for the F12-N)
- FC power -- 5.5W (for Jensen F12-N is 14W) (this implies minimum required current for FC to operate)

NOTE: Confined resistors need to have a de-rate of at least 5x, that is, in this case (F10-U), 850 ohm - 30W FC replacement resistor.

About the design configuration... you are right... the Gibson's BR-9 circuit is very similar but is push-pull and uses transformer PI (phase-inverter)...
additionally there is an earlier GA-8, also push-pull self-split PI.
To my surprise, Gibson's latest GA-8/ GA-9 design is PARALLEL SE !!! Improvement ???

[K Teacher]

I did not get any info on the F10-U VC diameter either... maybe 3/4"?

BTW, F12-N power handling is only 18W, NOT 50W like modern P12N...
Also, want to point out that all of these power figures are rated power, that is RMS...
I estimate musical power (instantaneous peak) to be about 2x rated power...

[K Teacher]

Thinking a little bit more on the subject... what limits the VC power handling is heat...
If the VC is wrapped on a heating generating element, like the FC, its power handling capability will be diminished compared to a PM (permanent-magnet) speaker...

[Helmholtz]

BTW, cathode biased (no bypass cap) class A push-pull stages can be operated in self-split mode, thus avoiding the need for a separate PI and allowing for a smaller, more efficient OT.
The operating principle is the same as with a LTPI or differential amp. In its simplest configuration only the upper tube is driven by the preamp while the grid of the lower tube is grounded.

[pdf64]

BTW, cathode biased (no bypass cap) class A push-pull stages can be operated in self-split mode, thus avoiding the need for a separate PI and allowing for a smaller, more efficient OT.
The operating principle is the same as with a LTPI or differential amp. In its simplest configuration only the upper tube is driven by the preamp while the grid of the lower tube is grounded.

Have you ever tried that? I wonder how the arrangement would cope with being overdriven
The concern being that paraphase splitters that share an unbypassed cathode can have a positive feedback mechanism that can cause them to overdrive horribly.
I've got a suitable amp chassis / PT to try such experiments out on; one day I'll have to get it actually wired up and try these things out

[Helmholtz]

Have you ever tried that?

No, just remembered some old time amp circuit.
How would you explain a positive feedback effect?

[pdf64]

Crikey, working that stuff out is tough and I’m outta practice. I worked it out previously as being the same mechanism as V4 on a Fender reverb channel; if the shared bypass cap goes high esr, high reverb settings cause electronic (rather than mechanical) oscillation.
But I’m stuck trying rationalise that at the moment

[Helmholtz]

Crikey, working that stuff out is tough and I’m outta practice. I worked it out previously as being the same mechanism as V4 on a Fender reverb channel; if the shared bypass cap goes high esr, high reverb settings cause electronic (rather than mechanical) oscillation.
But I’m stuck trying rationalise that at the moment

Unbypassed cathode resistance is common practice with LTPIs. I don't think V4 in a Fender compares to a differential amp.

Some literature: http://www.valvewizard.co.uk/SelfSplit.pdf