Speaker impedance matching and An Eccentric Idea - What if?

Most of us know that optimal speaker/power amp impedance matching is a touch-and-go situation. It's never  simply "8 ohms into 8 ohms". Speakers vary drastically over frequency - and I suspect that they are measured with a simple, isolated sine wave sweep. What happens as real music is played is anyone's guess. Some amp/speaker configurations just sound nicer when they buck the conventional wisdom - usually in the direction of lower amp/ higher speaker impedances.
Further complications arise when we consider ( or, rather, when I experience) speaker break-in adjustments. My new Tyler Halos sounded best on the 4 ohm taps at first - something I anticipated from previous experience - but after 1000 hours I had the hunch to try the (nominal) 8 ohm taps. Everything opened up. There was no feeling of "out-of-control" which I got at first.
Here's the $100,000 Question: what happens when one uses multiple taps on the OPT simultaneously - as is done with the power tranny? Perhaps as a bi-wire arrangement. Or even in parallel? What happens? It's not something I wish to try as an experiment, but I'd love to hear what, theoretically, would be the result.

Are you suggesting connecting the woofers to the 8 ohm outputs and the upper frequency range to the 4 ohm jacks or visa versa? The only issue I think would be the relative volume levels which could be adjusted for if you have attenuators on your amps?

The situation is more theoretical than practical. Is there any reason why one shouldn't experiment? And what happens if, say, one paralleled the 16 and 8 ohm taps? What effective impedance is the result, if any? Is the final result a 911 call?

OK - Let's assume that you do this. And let's assume that you have a 3-tap (4-8-16) OPT. Let's assume both speakers are nominal 8-ohm. And, you connect 0-8 & 0-16.

Ohm's Law applies.

Probably not a good idea. Worse if 0-4 is chosen.

Peter W - That answers my question.
Here's a potential application. Let's say I have a set of speaks which are bi-wireable. The hi end feeds only the tweeters (a common arrangement). The nominal impedance is 8 ohms total, though I see that the actual impedance curve varies widely down to 4 ohms at certain frequencies in the mid-bass region.  The tweeters are too bright in my room. I connect the tweet terminals to the 8 ohm posts and the low/mids to the 16 ohm posts. What I could get is a more balanced tone without mismatching the amp's output any more than I would had I attempted to solve the problem by using the four ohm taps for the entire speaker load. Does this make any sense? Or am I backwards? My brain hurts.
Or, in fact, will both sets of amp terminals give me about six ohms? Which could be nice in the case of some modern speakers ???

Sounds like a good way to produce some magic smoke.

That is the "what if?" I'm afraid of, Corndog. Thank Bob it's simply pure conjecture and not some harebrained scheme I'd actually try.....{famous final words}.

Kentley wrote:That is the "what if?" I'm afraid of, Corndog. Thank Bob it's simply pure conjecture and not some harebrained scheme I'd actually try.....{famous final words}.

https://www.tubecad.com/2011/12/03/UBT-1.png

Here is a picture of a 3-tap audio output transformer.
By various rules-of-thumb, it is typically no problem for the transformer to put a low-impedance speaker on a high(er) impedance tap. In fact, it puts less stress on the transformer as the output energy is dissipated across a larger coil. All you are doing is reducing impedance across the entire transformer, but splitting the load in an unusual way. So:

In the case you posit, you get a not-quite 4-ohm nominal impedance with the speakers connected as you describe. Given that the lowest frequencies will show (typically) an actual impedance of 1/2 the nominal, you could be running as low as <2 ohms at some frequencies. You would have to look at the specifications of your particular amp to see if it is good for a <2-ohm, albeit short-term, load.

I have two solid-state amps that purport to be 100% stable down to 1 ohm. Not something I would care to try.

But, cutting to the chase, it is unlikely you would damage the amp with the speakers at 8 ohms (4 (to <1 ohm) ohms would be a different story, perhaps). What you would do is create all sorts of distortion products.

I work with a lot of the new amp configurations in live music and impedance seems to be much more important with the Class D amps etc...they claim "800 watts into 2 ohms, 600 into 4 ohms, 300 into 8 ohms" etc. so the actual power seems to be dependent on a proper impedance for that performance. I never really worried about impedance in my tube amps, as long as there was a load of some kind. Heck in the old days I may have even hooked up like eight Radio Shack speakers of various sizes and loads in parallel and plugged 'em in and blasted away. They were pretty tolerant. Now I am extra careful to not even have a braid of wire short between the speaker taps for fear of a catastrophe. I will leave the experimentation to Kentley!

The experiments will remain in the confines of my little skull, thank you!

Correct and in accordance with theory. Modern transistorized amplifiers operate as constant-voltage devices. Ohm's Law says that as such they will flow twice as much current if the load impedance is halved. This appiles until the ability of the device to deliver current is reached, at which point either protection circuitry limits output, or the amplifier blows up due to parts failure.

wildiowa wrote:I work with a lot of the new amp configurations in live music and impedance seems to be much more important with the Class D amps etc...they claim "800 watts into 2 ohms, 600 into 4 ohms, 300 into 8 ohms" etc. so the actual power seems to be dependent on a proper impedance for that performance. I never really worried about impedance in my tube amps, as long as there was a load of some kind. Heck in the old days I may have even hooked up like eight Radio Shack speakers of various sizes and loads in parallel and plugged 'em in and blasted away. They were pretty tolerant. Now I am extra careful to not even have a braid of wire short between the speaker taps for fear of a catastrophe. I will leave the experimentation to Kentley!

I've never quite understood the idea of attaching different ohm loads to the different taps (4/8/16) on tube amps). People say they play around with it, I just attach the speakers to the proper tap(s).

I do know if you hook up 2 sets of speakers to the same output on a solid state, you can fry the amp. I did it. Most of the cheap solid state receivers were 4 ohm, but said you could drive two sets of 8 ohm speakers. Many of them just had switches that paralleled the speakers, but some used some kind of load resistive arrangement (like a speaker switcher).

corndog-it's not producing magic smoke, it's when you let the magic smoke out of the equipment it quits working At least that's what I was told by an old Ham buddy of mine. He said it's a booger to get it back in....

Jason

Before we leave this topic...

I have an original crossover in a speaker that where the crossover point is 3k and the high end driver is 16 ohms.

I can get a replacement 3k crossover, but the only ones available are for an 8 ohm load.

I understand that using a 16 ohm driver on an 8 ohm crossover means the crossover point would be dropped in half...to 1.5k.

Is this theory correct? Does that mean if I want to replace the crossover, and can only get an 8 ohm unit, that I would need to order a 5k to 6k crossover point to get to my original 3k?

Hope I explained this correctly. This crossover is blown but I can only get one for 8 ohms, and my driver is 16 ohms. I need it to cross at 3k.

Thanks.

IA.

wildiowa wrote:Before we leave this topic...

I have an original crossover in a speaker that where the crossover point is 3k and the high end driver is 16 ohms.

I can get a replacement 3k crossover, but the only ones available are for an 8 ohm load.

I understand that using a 16 ohm driver on an 8 ohm crossover means the crossover point would be dropped in half...to 1.5k.

Is this theory correct? Does that mean if I want to replace the crossover, and can only get an 8 ohm unit, that I would need to order a 5k to 6k crossover point to get to my original 3k?

Hope I explained this correctly. This crossover is blown but I can only get one for 8 ohms, and my driver is 16 ohms. I need it to cross at 3k.

Thanks.

IA.

"crossover blown" ?? What is blown ? Why not repair it ?

Yes. why Not? Unless one (or more) of the inductors is open and where the break cannot be gotten, there are no exotic or unusual parts in a crossover such as cannot be had at Mouser, or DigiKey. Nor will the parts be very expensive.

Hope this is not a duplicate post...i got screwed up earlier...

I have tried and tried to fix this. There is no high end with the crossover hooked up; but tweeter works when taken out of crossover circuit so driver is OK; I also get a reading across the coil; woofer works with crossover.

I found a 10w power resistor that was knocked clear off the crossover board and thought that was it. Replaced it, still no high end. Inspected crossover (mounted on PC board) and found no broken tracers or loose solder. This is a very basic crossover out of a Carvin stage monitor. Only thing I can think of is the coil you refer to or it also has one of those LED light deals for an overload but I know nothing about those....

I am tired of messing with it! For $20 or $25 I can get a replacement from Parts Express or Simply Speakers and be done with it. But I did wonder about the crossover point if the high end driver was a mismatch. If the impedance change cuts it in half, I reasoned that I would order a 6k crossover so the end result was about 3k...as I remind you, just enough knowledge to be dangerous...not a tech-y theory guy. Thx. IA

Do you have a picture? Front and back?

It is going to be something simple more often than not.

Absolutely terrible at photos...will try later. I did try to measure across the coil and get no reading that I could determine. So maybe that is burned out. Also, it has one of the LED lightbulb limiters...do those things actually blow, or just light up and then return to normal?

wildiowa wrote:Absolutely terrible at photos...will try later. I did try to measure across the coil and get no reading that I could determine. So maybe that is burned out. Also, it has one of the LED lightbulb limiters...do those things actually blow, or just light up and then return to normal?

If the coil is open, that is the explanation.
LEDs can blow - and more easily than thought. They are cheap and easy to replace. Color matters, however.

If it the coil remains open, try removing it from the board and looking underneath it - 90% of such failures are right at the connections.

wildiowa - not a simple question, as there are too many variables. First, are BOTH the woofer and tweeter 16 ohms? Many speaker designs have a mix of speaker impedances.
Is the crossover a simple 6db/octave, with only one cap and one inductor on each speaker? If so, you may be in luck and can use a Xover for twice the original value.
But if 12db/octave (or even higher) then it simply can't be done, as the capacitors and inductors double and half not in unison with each other.
It should be easy to compare the Xover from a good speaker to the Xover from a bad speaker, then find the bad part, and replace it.