Quick speaker question

[surfsup]

I'm gonna get a new speaker. My amp calculations show two el84s outputting approx 25W

Would a 25W speaker be okay, or should I get a higher rating?

[gearhead]

If it was me, higher. I crank em up.

[Firestorm]

Depends on whether you want to really push the speaker or have it stay clean and controlled. To get 25W, I think you'd have to use 7189s or 7189As.

[surfsup]

Here's my readings EL84:
Pin 1, OPEN
Pin 2, .6mv
Pin 3, 10.62
Pin 4, heater
Pin 5, heater
Pin 6, OPEN
Pin 7, 326
Pin 8, OPEN
Pin 9, 329.8

100R on the screens
125R on the cathode

Pin9, 100R, 329.8V, 330.2V, 4.000mA
Pin3, 125R, 10.62V, 0V, 84.960mA

Plate Current = 80.960mA
Va-Vk = 329.8-10.62 = 319.18
Power = 319.18 * 0.08096 = 25.8408128

What am I doing wrong?

[solderstain]

Depends on WHICH 25w speaker, of course.

I take a big-picture view of it - Speaker manufacturers like Celestion, for example, design in a safety margin. A 25w speaker isn't going to crap out the instant you pump 26w into it. You'd have pump significantly more into a speaker, continuously, over time, to kill it. Heat is the enemy here.

An example of speakers handling more than they're 'rated' for: The early-'60s JMI Vox AC30s - the ones that came with two 15w-rated Celestion Alnico speakers.

Two 15w speakers = a theoretical 30w of power handling capability, right?

Now, factor in that the AC30 actually produces closer to 36w, at idle. I have a '63 AC30 with added Top Boost, and with a Variac set at 120VAC and a 5V4 rectifier, the calculated power is 35.7w. My amp still has its original May 1963 Celestion Blues in it. And I play that thing dimed. The speakers live just fine.

If it were ME, I'd have NO problem playing an amp that produces a calculated 25w through a (good quality, good condition) 25w speaker.

[tubeswell]

Po = Zo'*(Vg*gm)^2

where:
Po = Power output, Watts(avg)
Zo = Effective reflected load impedance, Ohms.
Zo' = Zo*loading factor; Zo = Zpp/4 (in this case I'm guessing the Pri Z is 8k?, so Zo' = 2000)
Vg = PI drive signal to each power tube, V(rms) (in this case bias voltage is around -10, so the RMS grid swing clean would be 7VRMS)
gm = Power tube avg transconductance, Ap-per-Vg (in this case a EL84 transconductance is around 11,300umhos or 11.3mA/V)

So putting it al together we have

= 8000R/4 x (7VRMS x .0113A/V)^2
= 12.5W clean

Well that's what I make it? I think I in-putted the correct numbers anyhow - I will take corrections on the chin as usual but it seems realistic for 2 x cathode biased EL84s. (I'm only assuming the Pri Z of your OT is 8k?)

Edit - Also I am only guessing the PI drive at the grid will be 7VRMS, based on my hunch about the EL84 bias voltage, but if the grid input from the PI is higher than 7VRMS, it will change the result.

[surfsup]

solderstain and anyone else, from the scumnico guy himself:

Surfsup,
Saw your TGP post, but figured it was easier to email you rather than update a 2 year old thread. You’ll want the 65w model for that amp since you’d push the 30w too hard and likely fry the voice coil. Hope that helps!

I am going to get this speaker or the Celestion G12M Greenback

tubeswell, whoa. thanks for that math! I never saw that calculation before. Seems my method (same that many seem to recommend) is not very accurate at all, and to be honest was wondering how to calculate clean output. Now I know. Thanks so much!

(yes 8k on the OT impedance
25 Watts
Primary: 8000 Ohms,
Screen Voltage 40%
Secondary: 4, 8 & 16 Ohms.)

[martin manning]

Here's my readings EL84:...

Under what conditions are these readings taken under?

[surfsup]

Under what conditions are these readings taken under?

Just sitting there with the amp on.

[martin manning]

So this is just idle dissipation. Power output by definition needs to be calculated or measured under signal conditions, usually at clipping. The speaker chosen should be capable of handling about twice that power level.

[surfsup]

Let me ask this then, what is the advantage of getting a 25W speaker if there is the same speaker in 65W format available from Scumback for the same price?

[martin manning]

Let me ask this then, what is the advantage of getting a 25W speaker if there is the same speaker in 65W format available from Scumback for the same price?

25W would be a minimum. No reason you can't go higher, and that would probably be a good idea if you want to play with the amp maxed out.

[pdf64]

Jim Seavall could best answer that. My guess would be that the 30 watt voice coil will be lighter than the 65 watt voice coil, which would probably lead to better treble response, all else being equal. Pete.

[solderstain]

solderstain and anyone else, from the scumnico guy himself:

Surfsup,
Saw your TGP post, but figured it was easier to email you rather than update a 2 year old thread. You’ll want the 65w model for that amp since you’d push the 30w too hard and likely fry the voice coil. Hope that helps!

Jim is certainly a good source, and he's been a long-time friend (known him since 2003 or so... - probably 3/4 of my vintage Celestions were bought from him - almost all of my vintage original G12M-25s, and all but one of my vintage original G12-65s... Yum! Thanks, Jim! ).

I'm just sharing my personal experience. I (and innumerable other owners of early JMI AC30-6 amps) have an amp that has speakers 'rated' at a total of 30w, and I (and innumerable other owners of early JMI AC30-6 amps) routinely pump more power through them than they're 'rated' for, and it's been going on with my particular amp for 48 years, and they sound like Gawd, despite being pummeled... How can that be? I still say speaker designers add a safety margin to their power ratings.

I, too, have a 2-EL84 amp at the house (my project mule) and occasionally plug it into a 1990s British-built G12M-25 reissue with no ill effects.

As with anything, the bottom-line advice is to do what you feel comfortable with.

[Bob S]

Most lead players I know would run away screaming if you put a 15W amp to a 100W cab. My ears tell me it sounds best if you at least get close to being balanced.
But I am partially deaf.