Speakers and cabinets

[dl5mhk]

I've not investigated parphase PI's to any great extent, only built & modded long tail PIs. The theory is that you always get some grind because the signal going to the 2nd half has already been amplified by the first. Harp amps have been built using both styles.

Well, i'll see tonight - i'm playing on a jam session, there can i crank it a bit. Here @ home is no chance to let it really sing.
At first, i had some switches built in, for a different cathode bypass cap, with or without NFB, and some more. Meanwhile i eliminated all aof them except one: for switching the plate voltage of the 1st preamp stage, 150 volts or 40 volts. Nice headroom with 150, and lots of sweet distortion with 40 - then the amp has that "Rod Piazza-Sound" on some of his CD's, especially when played with a chromatic...sounds kinda broken. Love it!

cu, Ralf

[MWJB]

That is exactly what I am saying, but only in the context of harp, where acoustic feedback imposes a limit on the max power output (a guitarist would simply turn up his guitar and the differences in power available would pretty well even out, though the tone would change slightly).

The point you seem to be picking up on is that the individual speaker's efficiency remains unchanged (agreed), but when arranged in a series array it's a harder signal for harp to push, compared to parallel. Volume drops off, feedback starts earlier plus you sometimes get unpleasant harmonics and a "stodgy" tone. It's not a subtle difference. Similar listening tests by a trusted 3rd party came to the same conclusion.

My basis for this conclusion was A/B testing the same amp with 2x8 ohms in series (16ohm tap) and 2x16ohms in parallel (8ohm tap).

This is a good trick for harp playing Blues DeVille owners, the OT already has a 2ohm tap, rewiring the speakers in parallel will help (you can junk the stock 4/8ohm switching jack assy), as will increasing the series resistance on the fixed resistor at the top of the volume pot (430K-570K total).

This is why I am surprised at the number of "harp" amps that are on the scene with series wired speakers. Builders tend to make the numbers work out with the parts on hand, assuming that either scenario will work the same.

[Bruce / Mission Amps]

The point you seem to be picking up on is that the individual speaker's efficiency remains unchanged (agreed), but when arranged in a series array it's a harder signal for harp to push, compared to parallel. Volume drops off, feedback starts earlier plus you sometimes get unpleasant harmonics and a "stodgy" tone.

Wow...
OK, I'll have to experiment with that too because electronically (in a technical sense) and intellectually, this should make zero defference in anything you've encountered or just said!
However, I think I can hear a slight improvement in overall tone when running a guitar amps OT at its highest speaker output load...
ie.,
a multi tap OT running at 16 ohms with a 16 ohm load if the OT is built with 4, 8 and 16 ohm taps.
A NFB issue could pop up by running different loads on an amp where the 8 ohm tap or 2 ohm tap changes the amount of NFB to a previous stage.

[MWJB]

Hi Bruce,

Your skepticism is understandable. However, from a logical viewpoint does it not make sense that, if nothing else, the lower load of all speakers in parallel makes it easier for the signal to push, compared to 4x the load running 4 speakers in series-parallel? Plus, any current only has to push half the diaphragm mass?

[clyde1]

If you're pushing 4 speakers, doesn't diaphragm mass stay the same no matter what impedance wiring configuration you use? If the OPT is made for 8 ohms impedance, running a 2 ohm speaker setup will not yield optimum power from your amp and is this the tone difference you're hearing?

[MWJB]

I'm talking about running matched loads in all scenarios. I do know people who like twice the rated speaker load, but as you say there is an audible drop in efficiency. For this reason I prefer matched loads.

Thinking about it further, the theory does back up what I'm saying...

Lets take 2 otherwise identical amps (or one amp with 2ohm & 8ohm taps) running 4x8ohm speakers. Let's say we're getting 30W at the speaker in both scenarios (you won't, but just for arguments sake...).

With a parallel 2 ohm load you get 3.87A with 4 paths to ground, all signal only has to push a speaker cone once.

8 ohm load load gives 1.94A with 2 paths to ground, signal must go through one speaker cone, then immediately push a second cone before going to ground (like pushing a speaker with twice the diaphragm, mass)...twice the work.

[clyde1]

I think you're confused, same # of speakers, same mass pushed in both configurations. I don't see any difference electrically. You can argue there is a tone difference, but no difference in the math I'd say.

[dl5mhk]

8 ohm load load gives 1.94A with 2 paths to ground, signal must go through one speaker cone, then immediately push a second cone before going to ground (like pushing a speaker with twice the diaphragm, mass)...twice the work.

That doesn't matter, since the current is the same in both speakers of a series. It's not that the current would become weaker after driving the 1st one, and then would be left less to drive the 2nd one.
The reason for choosing a 2 Ohms OT winding instead for i.e. an 8 ohms configuration with 2x2 speakers in the Fender Bassman and other 4x10" amps was rather to save copper in the OT -> saving money.

cu,
Ralf

[MWJB]

I'm not confused. The math is clearly as I explained it. As I have said (more than once), in the context of electric guitar you would simply adjust the volume control, if necessary, to achieve the same result volume-wise. However. with the limitations imposed by acoustic feedback when cupping a harmonica to a mic you will not achieve the same result in the 2 scenarios I describe.

It takes less time to rewire 4x speakers than it does to read this thread, perhaps you should try doing that, and investigate the practicalities, rather than "suggesting" that there "may" not be a difference, because you don't "see" it.

[MWJB]

But the current is less in the series scenario.

It only doesn't matter if it doesn't sound any different, or doesn't work any better or worse. 8 ohms sounds different and doesn't work as well. Rather than arguing perhaps you should try it?

The reason for going 2 ohms was the fact that they switched to 4x10" to stop the bass farting out on the low E, compared to the original 1x15" format of the bassman & Fender used parallel wiring & 8ohm Jensens. The reason the 4x10" tweed bassman had 4x10" RCA jacks for the speaker wiring was to stop a potential current bottleneck in the speaker harness...for the same reasons I stated earlier.

I would have thought that there was MORE copper in an OT with a 44:1 turns ratio (approx for 2ohm) than in an OT with a turns ratio of around 22:1 (8ohm)?

[clyde1]

I'm sorry MJ, I didn't mean to suggest there is no difference and I'm sure you've heard it. I just don't understand the mathematical calculations you're
using to support your theory. I would indeed like to try rewiring the speakers in my Maestro, however then I would have to change the OPT to a 2-ohm
piece, which I am loath to do.

[MWJB]

Hi Clyde1,

Indeed, I wouldn't change the OT in the Maestro either (unless you could find one that would fit exactly without drilling the chassis) especially if it sounds good as is...however, if I was scratch building a stage amp, specifically for harp, I would aim to keep multiple speakers in parallel.

The power calculations I was referring to are I=E/R:

30W into 2 ohms:
I= 7.75v/2ohms
I= 3.87A

30W into 8 ohms:
I= 15.49v/8ohms
I= 1.94A

The voltages in the equations can either be measured at the speaker (as in the case of the 2ohm example) or arrived at by the square root of (30W*8ohms) divided by 8ohms = 15.49v...if we assume that both amps would hit a realistic 30W with a mic & speakers NOT a signal generator (no acoustic component to cause feedback) & a dummy load (which can't feedback either).

Cheers, Mark.

[dl5mhk]

But the current is less in the series scenario.

Yep. But the voltage is higher...

I would have thought that there was MORE copper in an OT with a 44:1 turns ratio (approx for 2ohm) than in an OT with a turns ratio of around 22:1 (8ohm)?

I would calculate the primary side of the OT first...and then the secondary, because the inductance of the primary (and the core size) has a big influence on the frequency range of the whole thing. After that i would do the math for the secondary - and that means less copper...

cu,
Ralf

[MWJB]

The current is the important number in this instance.

Yes, I see what you are saying about the OTs.