Speaker impedance preferences

[greiswig]

I'm kind of curious what other people's experiences are with changing speaker impedances. I have a D'Lite with one of Brandon's OT's in it...with four different output taps: 4-32 ohms!

I run a two-speaker cabinet, with a pair of 8 ohm speakers in it. I could either end up with a 16 ohm cab or a 4 ohm cab, obviously. And I've got ample taps for either one. I don't run another cabinet, so leaving things high impedance to allow another one in parallel isn't an issue.

What have people found the effect to be of changing impedances on a 2x cabinet when they can match impedance (or equivalent mismatch) with the OT?

[heisthl]

Rule of thumb is to use the most winding you can for best tone - so 16 ohms. That said, many find the purposeful mismatch of an 8 ohm speaker on the 4 ohm tap to sound good.

[greiswig]

Thanks, heisthl. I'd heard that before, too. Have most of you found this to be the case?

Okay, here's a related question: in PA's, I was taught that the rule of thumb is that you want the power amp to see a lower impedance load, because it will "pull the most power" out of the amp. For someone like the poster in another thread who finds his 40W amp too quiet, would rewiring a 16 ohm cabinet to be a 4 ohm cabinet make any difference in the perceived loudness when running the amp flat out?

[doveman]

Same experience here.

My D'Lite 22/33 sounds slightly better at 16 ohms (two 8 ohm cabs in series) but is slightly louder at 4 ohms (two 8 ohm cabs in parallel).

When I say better, it's tighter ... that may be the best description.

[greiswig]

Fascinating. I'd be curious as to whether people who prefer the "tighter" 16 ohm setup have the jumper between the bass pot lugs?

I *think* I'm preferring it set up at 4 ohms on the speaker, running from either the 4 ohm or 8 ohm taps. It sounds fuller, but still has nice treble to it. I have that jumper set up on a switch between dead short, 68k ohms, and 120k ohms. Running with that shorted, I can see how you would want to tighten up the bass. At those higher resistances, running the cabinet lower impedance seems to deepen where the bass sits, but it still isn't sloppy.

[talbany]

I agree with Heisthl.. use the tap with the most winds.. to my ears produces a thicker meatier tone as well well as enhanced harmonic content.. generates those fragile harmonics you can't get from a crystal lattice.. Whatever!!!


Tony VVT

[heisthl]

Fender type output iron has a reputation for being able to withstand being off by one notch i.e. 4 ohm tap 8 ohm spk, 8 ohm tap 4 ohm spk. Other brands may not be so robust. If you drive a 8 ohm speaker off the 4 ohm tap keep checking the output transformer for heat buildup. If you drive a 4 ohm speaker off the 8 ohm tap be prepared to replace your output tubes more often.
Regarding your PA, if it's solid state do an exact match - direct coupled output transistors are not known to be forgiving like tubes and that extra current flow to a lower impedance load is a real heat generator.

[greiswig]

Regarding your PA, if it's solid state do an exact match - direct coupled output transistors are not known to be forgiving like tubes and that extra current flow to a lower impedance load is a real heat generator.

Not quite what I was asking about. In a lot of PA amps, the amp will either have an impedance switch, or will be able to accept, say, 8 or 4 ohm loads. What I was told was that, given that choice, try to put the lower impedance load on the amp and it will have more headroom to work with, more overall volume.

So what I'm asking about with my question is just whether that same logic applies to guitar amps, tube amps in particular.

[ayan]

Regarding your PA, if it's solid state do an exact match - direct coupled output transistors are not known to be forgiving like tubes and that extra current flow to a lower impedance load is a real heat generator.

Not quite what I was asking about. In a lot of PA amps, the amp will either have an impedance switch, or will be able to accept, say, 8 or 4 ohm loads. What I was told was that, given that choice, try to put the lower impedance load on the amp and it will have more headroom to work with, more overall volume.

So what I'm asking about with my question is just whether that same logic applies to guitar amps, tube amps in particular.

I know there are two schools of thought on that regarding tubes: one says that putting a higher impedance load into the amp is safer -- as is the case with solid state -- because the amp will not try to put out more power (as would be the case with solid state); however, the problem seems to be with the reflective impedance that occurs when you have a transformer in the picture and thus some, like Mike Soldano, will say that it is WORSE to put a higher impedance load than it is to put a lower one.

For what it's worth, the first many ODS style amps I built used Marshall style OT. They are said to be less tolerant of impedance mismatches, but I never had any problem, and deliberately mismatching the impedances resulted in hardly any audible tonal difference at nominal volumes. The last two amps I built use Heyboer multi-tap twin style OTs, which are supposedly more tolerant of impedance mismatches. I have never had any problems with them either, but the one thing that makes a night and day difference is plugging in an 8-ohm speaker and switching the impedance selector down to 4-ohms. The amp will lose a significant amount of high end and sound a lot sweeter, great for low volume playing, I prefer the thing matched for gigging volumes.

Hope this helps,

Gil

[dogears]

Two days ago, I wired up my Jenkins 2X12 with a set of 8 ohm Celestion G1265 speakers in series. 16 ohms.

The low end has the exact "splatty" quality you hear in Rugged Road for instance. The series wiring is not as dampened and the texture is definately enhanced to my ear. Not as smooth and controlled, a very cool alternative tone.

I do still like my parallel cabs as well. Just saying that series vs parallel wiring is quite noticeable!

[Aurora]

For a well designed SS PA ( or any power amp) , which "shouldn't" give a hoot if you serve it 4 or 8 , or even down to 2 ohms, provided it can deliver the voltage and/or current requested, the logic goes as follows:
Lowering the impedance from 8 to 4 ohms, by parallelling speakers, will require the amp to deliver twice the current - voltage amplification remains the same. Adding an extra 8 ohm speaker ( or any doubling) will double your output power, as the speakers share the extra current. Using two speakers in series will have the both speakers see the same current, but half the voltage each - power remains the same as with one speaker. Poorly or marginally designed SS amps may not be able to supply the voltage or current needed when changing load impedance - all amps have a design impedance as basis for the design.

With tube amps things may be a bit different - and by all means - I'm not the very skilled tube designer...
For the typical ODS, the design impedance or load line is set around 4 k.ohms to the plates. Loading the OT with half the normal impedance will of course change the plate load to 2K total, which will certainly affect the behaviour of the OT - and most probably the tone. Same is loading the amp with 8 ohms speakers, and impedance set to 4 ohms -the plate load is now 8k.ohms, - a much lighter load on the OT with possible similar change in tone, but as the voltage supplied is now only half of the nominal voltage, as should be selected at the impedance switch - power is also now halved. OTOH- real speakers are highly unlinear impedances and the real result may deviate from the theory...............

The simple calcs above wil also hold for series /parallel fiddling with 1-2 or 2-4 speakers, provided the power supply is able to deliver, and the OT can handle the extra current when loading the amp with "half impedance" ( as set with the switch).
Some of you most certainly know these simple calc's, but not all
( and almost being a " tube noob" myself, I'm absolutely not trying to patronize anyone!)

[dogears]

Actually power is not a linear thing. I believe the change in power is logarithmic when moving a step in either direction.

For a well designed SS PA ( or any power amp) , which "shouldn't" give a hoot if you serve it 4 or 8 , or even down to 2 ohms, provided it can deliver the voltage and/or current requested, the logic goes as follows:
Lowering the impedance from 8 to 4 ohms, by parallelling speakers, will require the amp to deliver twice the current - voltage amplification remains the same. Adding an extra 8 ohm speaker ( or any doubling) will double your output power, as the speakers share the extra current. Using two speakers in series will have the both speakers see the same current, but half the voltage each - power remains the same as with one speaker. Poorly or marginally designed SS amps may not be able to supply the voltage or current needed when changing load impedance - all amps have a design impedance as basis for the design.

With tube amps things may be a bit different - and by all means - I'm not the very skilled tube designer...
For the typical ODS, the design impedance or load line is set around 4 k.ohms to the plates. Loading the OT with half the normal impedance will of course change the plate load to 2K total, which will certainly affect the behaviour of the OT - and most probably the tone. Same is loading the amp with 8 ohms speakers, and impedance set to 4 ohms -the plate load is now 8k.ohms, - a much lighter load on the OT with possible similar change in tone, but as the voltage supplied is now only half of the nominal voltage, as should be selected at the impedance switch - power is also now halved. OTOH- real speakers are highly unlinear impedances and the real result may deviate from the theory...............

The simple calcs above wil also hold for series /parallel fiddling with 1-2 or 2-4 speakers, provided the power supply is able to deliver, and the OT can handle the extra current when loading the amp with "half impedance" ( as set with the switch).
Some of you most certainly know these simple calc's, but not all
( and almost being a " tube noob" myself, I'm absolutely not trying to patronize anyone!)

[Aurora]

Theoretically and ideally, power is indeed linear - according to Ohms law. Almost all audio amps are designed with a set, fixed voltage amplification factor - current is drawn as requested by the load impedance, provided the amp is able to give it. In the real world, amps behave somewhat here and there....depending how well they are designed.

[dogears]

From Aiken's site:

Note that the tap voltage is not linear with respect to the impedance, it varies linearly with the square root of the impedance. That is, the voltage on the 8 ohm tap is not half the voltage on the 16 ohm tap, rather, the voltage on the 4 ohm tap is half the voltage on the 16 ohm tap. It helps if you think of the equation for power: P = V^2/R. If you have 100W into 16 ohms, the voltage is V = sqrt(100*16) = 40V RMS. If you have 100W into 8 ohms, the voltage is V = sqrt(100*8) = 28.28V RMS. If you have 100W into 4 ohms, the voltage is V = sqrt(100*4) = 20V RMS.

Theoretically and ideally, power is indeed linear - according to Ohms law. Almost all audio amps are designed with a set, fixed voltage amplification factor - current is drawn as requested by the load impedance, provided the amp is able to give it. In the real world, amps behave somewhat here and there....depending how well they are designed.

[ayan]

Actually power is not a linear thing. I believe the change in power is logarithmic when moving a step in either direction.

To determine whether a function is linear or not, one needs to specify with respect to what, and what is fixed and what varies. Power is the produc of voltage time current, so expressed that way, for a fixed output voltage, power is linear with respect to current. Since current and voltage are related to impedance by Ohm's law, you can get all of the following expressions, in which power will be related to the 2nd power of either voltage or current.

P = V x I; P = V^2/Z; P = I^2 x Z

There are no logarithmic relationships involved per se. The dB is an artifact, a convenient way to express large quantities, with the added feature that the log of a product is the sum of the logs and that the log of the nth power of something is n times the logarithm of the something. Piece of cake to add numbers or multiply them by 2, so handling powers in dBs is very convenient.

Gil