IMPEDANCE MEGATHREAD FAQ

A sub forum for our most useful repair, maintenence & upgrade threads.
Post Reply
User avatar
øøøøøøø
PAT. # 2.972.923
PAT. # 2.972.923
Posts: 10710
Joined: Sun Jan 06, 2008 8:26 pm
Location: Brooklyn, NY
Contact:

IMPEDANCE MEGATHREAD FAQ

Post by øøøøøøø » Tue Apr 26, 2011 9:10 am

I find myself fairly often answering questions about speaker impedance in amplifiers. I figured now that we have the "technical threads" sub-forum, it might be a good time to try and collect all the info in one place so that we're not constantly re-hashing the same stuff over and over-- we can just refer to this thread. If the mods deem it worthy, they can move it there.

First of all, the answers to the most common questions:

Q: I have an amp that specifies it wants to see an 8 ohm load. Is it OK to hook a 4 ohm, 2 ohm, or 16 ohm load to it? Will I blow up anything? (substitute any numbers you want)
A: "It depends." Typically, with Fender amps, you are OK going either one step up or one rung down-- either half, or double. In an old Fender amp requesting 8 ohms, you should be fine with either 4 ohms or 16 ohms. By "fine," I mean "not exceedingly likely to blow anything up." As always, there are no guarantees. With the upward mismatch (i.e. the load is double what's specified), you're in slightly more danger. With half the requested load, your risk is very low. Going too far beyond the "half or double" rule of thumb is asking for trouble. This applies mainly to Fender amps, but to many others as well. Any amp with robust transformers (Traynor, Ampeg, etc) will usually be pretty tolerant of a reasonable (i.e. 100% either direction) mismatch.

Q: My dad is an electronics engineer, and he told me that upward mismatches are safer (i.e. 4 ohm output into 16 ohm load), and this 80-year-old TV repair guy I'm friends with says the opposite-- that downward mismatches (i.e. 8 ohm output into 2 ohm load) are safer. The old dude probably just has alzheimer's, right?
A: No. The answer to this question depends on whether you're talking about tube or solid-state technology, and this is the source of the confusion. Solid-state amps typically have a very LOW output impedance, and will easily drive a load many times higher than their output impedance. In fact, they're usually happiest this way, and having a load with too LOW an impedance is dangerous-- dead shorts will kill a solid-state amp, and open circuits will be fine.

Tube amps are the opposite. Tube amps prefer if the output impedance of the amp and the load are matched as close as possible; however, to some degree they are fairly tolerant of loads that are lower in impedance than the source. Loads whose impedance are way higher than the amp's output impedance will cause what's called flyback voltages in the output transformer. This can lead to arcing, and can burn the insulation off the transformer's windings, causing death to the transformer. Tube amps are the opposite of solid-state, as a rule: they'd rather see a dead short on the output than an open circuit. Fender amps have a shorting jack on their speaker outputs for this reason-- If the speaker comes unplugged, it shorts the output transformer secondary, which will 'buy you some time' until you figure out that there's no speaker connected.

Q: I have an amp that wants to see 4 ohms, but my only cab right now is 16 ohms. Can I hook it up just to see what the amp sounds like and make sure it works? Is there any way to minimize the chance of damage?
A: Do so at your own risk. If the numbers were reversed (16 ohm amp, 4 ohm speaker), you'd likely be fine if you were reasonable with the demands placed on the amp--your tubes would wear a bit faster, probably. However, upward mismatches like the one in this example are more dangerous for the output transformer. The greater the upward mismatch, the greater the flyback voltages. At low volumes you would be fine. The higher the volume, the more voltage swing you're causing at the output tubes, and the greater the flyback voltages will be. I won't say it's "never OK," or that it will cause instant death to every amp-- but proceed with caution.

Q: what are the effects of mismatching impedances?
A: they can range from premature output tube wear/failure to toasted output transformers. The further off you go, the more you lose efficiency/volume, as well. The timbre of the sound can change a little bit, and you can lose some volume. However this is NOT an effective way of getting "breakup at lower volumes." I recommend not fiddling with impedance as a sound-shaping variable.

Q: What ways can multiple speakers be wired up to get different impedances?
A: Assuming speakers of like impedance: wiring 2 speakers in series gets you double the impedance of the individual drivers. Wiring 2 in parallel halves it. 4 in series quadruples it, 4 in parallel gives you 1/4 the impedance of one driver alone. 4 drivers can also be wired as two parallel pairs that are then wired in series with one another, or two series pairs that are then wired in parallel with one another. In either case, the total impedance is the same as one driver alone. Odd numbers of drivers (say, 3 speakers) are additive in series, and can be calculated in parallel by dividing one driver's impedance by the number of drivers.

Examples: two 8 ohm speakers in series are 16 ohms total, in parallel are 4 ohms total. Four 8 ohm speakers in series makes 32 ohms, and in parallel makes 2 ohms. Four 8 ohm speakers in series/parallel or parallel/series gives you 8 ohms in either case. Three 8 ohm speakers in series makes 24 ohms, and in parallel makes 2.667 ohms (8/3).

There are no other mathematical combinations possible. So don't ask. :)

Q: So... is there any way to get 8 ohms from two 8 ohm speakers?
A: (sigh...) no.

Q: Can I use two speakers of different impedances to get odd impedance numbers?
A: Yes, but... the speaker with the lower impedance will take the most power. If you have one 8 ohm and one 16 ohm speaker, most of the sound will come from the 8 ohm speaker as it will draw twice as much current as the 16 ohm speaker. Not a very efficient setup, and the 8 ohm speaker must be able to handle more power. Kinda defeats the purpose of having 2 speakers in a lot of ways.

Q: I still want to run two speakers of different impedances. How do I calculate the total load?
A: By using the "product over the sum" formula: Multiply the two impedances together, then divide that by the sum of the two impedances. For an 8 and 16 ohm mismatched pair, it would be 8*16 (128) divided by 8+16 (24) for a total of 5.333 ohms.

Q: What is "impedance" anyway?
A: it is the total opposition to electrical flow in a circuit

Q: How does that differ from resistance?
A: Resistance is one component of impedance. Reactance is the other. Your speaker has a coil of wire that has a resistance in its "at rest" state. When it moves inside the voice coil gap, a magnetic field is induced which also opposes electrical flow. This is the "reactance" component.

Q: How does that apply to speakers and amps?
A: This is a bit beyond the scope of this FAQ, but understanding Ohm's Law will take you a long way toward understanding this. Ohm's law is a relationship between three things: resistance (or impedance in our case), current, and voltage. Understanding the way these things relate, and what they are, is an essential pre-requisite for gaining an intuitive understanding of impedance relationships in amplifiers. So start searching, and see how much you can learn for yourself.

But here's some food for thought-- "loading" something with the proper impedance isn't unique to electronics-- it happens in everyday life all the time. If you talk, the sound coming out is "loaded" by the atmospheric pressure in the room. A megaphone held in front of your mouth can help "match" the source impedance with the "load," and make sound travel more efficiently. For another example, think of an acoustic guitar. The string is the sound source, and the guitar top is the 'load.' If the top is too thick, the vibrating strings won't be able to excite it very much-- you'd have to play very, very hard to pull any sound out, and the guitar won't be very loud. However, if the top was very thin (say, thinner than you can realistically make a guitar top), and you played really, really hard, then you would get that "choking" sound, and the guitar would NOT get any louder-- you'd hit a 'brick wall' and the guitar's volume output would stop tracking the dynamics of your playing input, if that makes sense.

This is all matching the load to the source energy, and is very related to what we're talking about with amps.

Q: My speaker says 8 ohms, but when I measured it with my ohmmeter, it only says 6.7 ohms. Is the speaker actually a 6.7 ohm speaker, or is it defective?
A: Neither. You are measuring the resistance component only... when the speaker is at rest the magnet wire has a resistance of 6.7 ohms. When the voice coil is charged with electricity, the reactance adds additional impedance on top of that.

Q: is impedance always constant?
A: No. Reactance (and therefore impedance) varies with frequency. It's a very 'ballpark' figure, and high degrees of precision are not necessary. Your speaker's impedance will never go below the voice coil's resistance, but can vary much higher depending on the signal put into the amp. For this reason, slight mismatches (5.3 ohms loading a 4 ohm amp output) are usually completely inconsequential and fine.

Q: I want to use the ext. speaker jack of my amp in addition to the main speaker out. What does this do to the impedance?
A: on most tube amps, the ext. speaker and main speaker jacks are wired in parallel. If you have a Deluxe Reverb with an onboard 8 ohm speaker, and you plug another 8 ohm speaker into the "ext speaker" jack, then you have a total load of 4 ohms presented to the amp. Fender amps are designed for this, and will be fine.

On some Ampeg amps (and maybe some others), the ext. speaker jack actually switches taps on the output transformer. On my B-25B, with one cabinet plugged in to the speaker jack, the output impedance is 16 ohms. However, plugging something into the extension speaker jack switches taps on the OT to make the amp's output impedance 8 ohms. That way, impedance is always matched regardless of whether 1 or 2 16 ohm cabs are connected-- a clever design. You can also 'cheat' by plugging a dummy plug into "ext speaker" and an 8 ohm cab into "speaker." I do this sometimes if I have to take the head to a venue with an 8 ohm cab. A George L plug works well. Bear in mind, this is kind of an "Ampeg thing" and is an unusual feature-- it's also not true of ALL Ampegs.

... that's enough for now. If we think of any more, I can add them to the list.

User avatar
øøøøøøø
PAT. # 2.972.923
PAT. # 2.972.923
Posts: 10710
Joined: Sun Jan 06, 2008 8:26 pm
Location: Brooklyn, NY
Contact:

Re: IMPEDANCE MEGATHREAD FAQ

Post by øøøøøøø » Tue Apr 26, 2011 11:00 am

Mods, there's one thing I'd like to make more clear but I've lost the ability to edit that post. If you deem this useful, would it be possible for me to get editing privileges for this post?

Want to replace the last sentence in the paragraph about the acoustic guitar/impedance analogy with this: "However, if the top was very thin (say, thinner than you can realistically make a guitar top), and you played really, really hard, then you would get that "choking" sound, and the guitar would NOT get any louder-- you'd hit a 'brick wall' and the guitar's volume output would stop tracking the dynamics of your playing input, if that makes sense."

User avatar
pocaloc
PAT. # 2.972.923
PAT. # 2.972.923
Posts: 1637
Joined: Tue Aug 04, 2009 2:40 pm
Location: Spokane, WA

Re: IMPEDANCE MEGATHREAD FAQ

Post by pocaloc » Tue Apr 26, 2011 11:56 am

I think you all have me to thank, in part, for this handy thead. :) I've been all over his ass with annoying questions. Thanks for being such a helpful/knowledgable resource.
Last edited by pocaloc on Fri Sep 09, 2011 3:26 pm, edited 1 time in total.

User avatar
zhivago
Mods
Mods
Posts: 20312
Joined: Tue Sep 12, 2006 6:18 am
Location: London, UK

Re: IMPEDANCE MEGATHREAD FAQ

Post by zhivago » Tue Apr 26, 2011 12:08 pm

øøøøøøø wrote:Mods, there's one thing I'd like to make more clear but I've lost the ability to edit that post. If you deem this useful, would it be possible for me to get editing privileges for this post?

Want to replace the last sentence in the paragraph about the acoustic guitar/impedance analogy with this: "However, if the top was very thin (say, thinner than you can realistically make a guitar top), and you played really, really hard, then you would get that "choking" sound, and the guitar would NOT get any louder-- you'd hit a 'brick wall' and the guitar's volume output would stop tracking the dynamics of your playing input, if that makes sense."

I'm not sure how to change privileges specific to the forum, so I just made the edit for you, Brad :)

great writeup, by the way 8)
Resident Spartan.

User avatar
øøøøøøø
PAT. # 2.972.923
PAT. # 2.972.923
Posts: 10710
Joined: Sun Jan 06, 2008 8:26 pm
Location: Brooklyn, NY
Contact:

Re: IMPEDANCE MEGATHREAD FAQ

Post by øøøøøøø » Tue Apr 26, 2011 12:29 pm

functionally as good! thanks for the edit.

pocaloc, you were but one of many. Your questions were not at all unique, hence the idea for this post. It really is kind of an opaque/mysterious concept. It's one of the most asked-about things having to do with amps.

In other words, the questions are not annoying at all on their own, however the mysteries are so consistent that repeating the same thing to different people many times can get a bit on the repetitious side after awhile. :)

User avatar
pocaloc
PAT. # 2.972.923
PAT. # 2.972.923
Posts: 1637
Joined: Tue Aug 04, 2009 2:40 pm
Location: Spokane, WA

Re: IMPEDANCE MEGATHREAD FAQ

Post by pocaloc » Tue Apr 26, 2011 12:36 pm

When I tested the 70's bandmaster reverb that I just bought on my 16 ohm 1x12 cabinet, it was very quiet, and you could smell it heating up. When I brought it to a local store and tested it on his 4 ohm '65 Bandmaster 2x12 cabinet, I think with Oxford speakers, it almost blew the roof off and didn't seem to be heating up. So that 4 to 16 ohm difference really affected the amp.

User avatar
øøøøøøø
PAT. # 2.972.923
PAT. # 2.972.923
Posts: 10710
Joined: Sun Jan 06, 2008 8:26 pm
Location: Brooklyn, NY
Contact:

Re: IMPEDANCE MEGATHREAD FAQ

Post by øøøøøøø » Tue Apr 26, 2011 12:44 pm

thanks for posting that. It's likely some damage would've occurred if you had been tempted to crank it up higher for any length of time.

User avatar
jimboyogi
PAT. # 2.972.923
PAT. # 2.972.923
Posts: 2053
Joined: Tue May 27, 2008 2:38 am
Location: Melbourne
Contact:

Re: IMPEDANCE MEGATHREAD FAQ

Post by jimboyogi » Tue Apr 26, 2011 1:11 pm

Great work Brad, an excellent write-up.

Is this a good place for information on power sharing in a multi-speaker array? If you have 4 speakers wired in 2 series pairs, then each pair paralleled, what proportion of the total power does each speaker see?

User avatar
Dave
PAT. # 2.972.923
PAT. # 2.972.923
Posts: 1068
Joined: Wed Mar 07, 2007 6:32 am
Location: ON

Re: IMPEDANCE MEGATHREAD FAQ

Post by Dave » Tue Apr 26, 2011 2:09 pm

:-[ :-[

Is my recent Traynor thread the straw that broke the camel's back?

User avatar
i love sharin foo
Mods
Mods
Posts: 5124
Joined: Tue Sep 12, 2006 8:26 am
Location: Pittsburgh, PA
Contact:

Re: IMPEDANCE MEGATHREAD FAQ

Post by i love sharin foo » Tue Apr 26, 2011 2:14 pm

øøøøøøø wrote:Mods, there's one thing I'd like to make more clear but I've lost the ability to edit that post. If you deem this useful, would it be possible for me to get editing privileges for this post?
There's some ghost in the machine right now making it impossible for members to edit posts. Jay said he'll have a crack at sorting it as soon as he has some time.

PS, thanks for another superb write up Brad!
"There are twists of time and space, of vision and reality, which only a dreamer can divine..."

User avatar
andrewdoeshair
PAT. # 2.972.923
PAT. # 2.972.923
Posts: 6808
Joined: Fri Oct 16, 2009 10:16 pm
Contact:

Re: IMPEDANCE MEGATHREAD FAQ

Post by andrewdoeshair » Tue Apr 26, 2011 2:47 pm

All this knowledge is giving me a raging brainer.
If you wish to make an apple pie from scratch, you must first invent the universe. - Carl Sagan

User avatar
Professor Plum
PAT. # 2.972.923
PAT. # 2.972.923
Posts: 2365
Joined: Tue Feb 26, 2008 7:31 pm
Location: behind my computer screen. in dirty jersey.
Contact:

Re: IMPEDANCE MEGATHREAD FAQ

Post by Professor Plum » Tue Apr 26, 2011 3:57 pm

andrewdoeshair wrote:All this knowledge is giving me a raging brainer.
:D 8) :w00t:
make feedback, not war. pick it up! pick it up! up, up!

User avatar
Dave
PAT. # 2.972.923
PAT. # 2.972.923
Posts: 1068
Joined: Wed Mar 07, 2007 6:32 am
Location: ON

Re: IMPEDANCE MEGATHREAD FAQ

Post by Dave » Wed Apr 27, 2011 10:48 am

Brad, can you include some info about choosing the right speakers in terms of wattage? Do you add up all of the wattage ratings on all speakers being used to determine a proper amount?

User avatar
sookwinder
Mods
Mods
Posts: 10378
Joined: Mon Sep 11, 2006 6:47 pm
Location: Melbourne Australia

Re: IMPEDANCE MEGATHREAD FAQ

Post by sookwinder » Wed Apr 27, 2011 2:01 pm

Fender amps have a shorting jack on their speaker outputs for this reason-- If the speaker comes unplugged, it shorts the output transformer secondary, which will 'buy you some time' until you figure out that there's no speaker connected.

This is something I had not realised. When I went and looked at the schematics, I can now see the shorting jack (other than on the VC schematic - the VC has a RCA plug)

Thanks Brad for doing this thread.
relaxing alternative to doing actual work ...

User avatar
øøøøøøø
PAT. # 2.972.923
PAT. # 2.972.923
Posts: 10710
Joined: Sun Jan 06, 2008 8:26 pm
Location: Brooklyn, NY
Contact:

Re: IMPEDANCE MEGATHREAD FAQ

Post by øøøøøøø » Thu Apr 28, 2011 12:00 pm

Dave wrote:Brad, can you include some info about choosing the right speakers in terms of wattage? Do you add up all of the wattage ratings on all speakers being used to determine a proper amount?
Sure. People will debate the finer points of this issue, but here's some background info that will inform your own decisions. I seem to think in bullet points, so forgive the ordered list.

1) Amps are rated for clean, undistorted output. So if your amp is rated for "50 watts," that means 50 watts with no distortion at all (or a very small number, like 1%). Remember that tube amps can show measurable distortion quite a bit before they show readily audible distortion. Even a slight bit of compression/fattening is still distortion when you measure it. The less negative feedback an amp has, the more gradual the onset of distortion, as well. So your 50w amp is 50w not distorted at all-- turned up to the point just before the onset of measurable (not necessarily readily audible) harmonic distortion. Tube amps will keep on getting louder past the onset of distortion, so if you crank your amp up to the point where it's "just almost breaking up," it will be putting out significantly more power than it's rated for. If you dime everything and use a booster pedal in front, you can easily exceed double the amp's rated output, in some cases. For this reason, sticking a single 50w speaker in a 50w amp is NOT recommended. It may not blow right away, but it certainly will not be reliably safe.

2) A good 'rule of thumb' is to use a total speaker load with a power rating of roughly double the amp's rated clean output. In other words, 100w total speaker power handling capability with a 50w amp. This is generally safe even for relatively aggressive volume/playing levels/styles.

3) Speaker power handling is additive assuming like impedance of all drivers. In other words, two 25w speakers in either series or parallel is functionally the same as a single 50w speaker. In the unlikely event that your speakers have different impedances, the speaker with the lower impedance will draw more than its share of the power.

4) Speakers generally fail because of heat. Failure modes can be many, but with the exception of a coil jumping all the way out of the gap, most failures are heat-related. Heat buildup is caused by losses inherent in the speaker's design. Higher-rated speakers are better at getting this heat away from crucial parts. A speaker's voice coil is a coil of thin wire (having painted-on insulation), wrapped around a (generally plastic) cylinder called a voice-coil former. The VC former has to remain perfectly round. If it gets de-formed, the voice coil will get stuck in the magnet gap and the speaker won't work. heat can also melt the painted-on insulation and cause shorts, or can burn through the wire entirely and cause open coils.

The louder you crank your amp, the more heat is generated in the speakers. You can often push speakers past their rated amounts for a time, and as long as they then get some 'rest' to cool down before punishing them again, they'll survive. In other words, they're not fuses-- 51 watts won't instantly kill a 50w speaker. The heat has to build up and remain high enough to cause damage. With experience, you can kind of use judgement calls... say you have a 35w amp and want to use a 50W speaker. You'll know that you can push it for awhile, but you would want to show some circumspection before you decide to dime the amp and run a synth bass through 3 fuzzboxes for a whole set. You get the idea.

I think of it kind of like the red line on your car's tachometer. If you rev up your engine past the red line for a second here or there, it won't kill the engine or even damage it. But if you put the car in park and floor it and hold the needle past the red line, eventually your engine will burn up. It's kinda like that with speakers. You can flirt with the 'red line' from time to time, but just don't sustain that level of demand.

Post Reply