OT - Speaker Hook-up Question

[Paul Marossy]

I have a 120 watt 12" dual voice coil speaker that I want to use in a guitar cab that I am going to make. It has two independent voice coils that are 4 ohms each. I measured 3.9 ohms for each coil with my DMM.

Now, my understanding is that you can wire the coils in series to get an 8 ohm load. However, when I connect them in series, I measure 12 ohms with my DMM. Shouldn't I be measuring around 8 ohms? When connected in parallel, my measurements are what I would expect, about 2.4 ohms.  Anyone have any idea why I would measure 12 ohms when wired in series? I don't get it...  :?

[Ge_Whiz]

Well, DC resistance and impedance ain't quite the same thing. Impedance varies with frequency, and I would guess that each coil is quoted as having an impedance of 4 ohms near to a frequency appropriate to that coil. That can throw the measurement of DC resistance of the series combination, but the series impedance (which varies with frequency anyway) will be close enough to 8 ohms for rock 'n' roll.

Most 8 ohm speakers in my experience measure about 5 ohms DC resistance.

[Paul Marossy]

Yeah, I know that DC resistance and impedance aren't the same thing. I was just expecting to read ~8 ohms when wired in series if they meaure 3.9 ohms each.

I guess it'll be fine then.  

[csj]

What's the primary/secondry impedance ratings of your OT? Calculate what the difference would be for what the 12 ohm secondary will be reflecting back to the primary and see if it's in the ballpark of what your tube/s load needs to see - if it's close just go for it. What's ballpark? I don't know really. I've routinely run p/p 6L6s at anywhere from 5k to 7k primary into an 8 ohm load. I think the textbook says they would need anywhere from 8k to 10k or at least I've heard/read from others that say they do. Again, with this whole impedance thing... work with whatever you've got but at least know how you've mismatched or fudged on the benchmarks. That's what I do at least and this is just MHO. I never use my best tubes anyway except when I play live by which time I've put enough mileage on the setup to feel confident enough to use them. Honestly, I've never had a problem from mismatches. Most of the problems I've had are from my own poor power supply and/or biasing designs. Thank goodness for the "variac".

[Paul Marossy]

5.6K / 8 ohm
SE amp

I don't know exactly how to do that calc csj. If you could enlighten me, that would be mighty righteous of you.  

[Lonestarjohnny]

Paul, you can check the DC resitance of a speaker, not ohmage, if you measure across the voice coil for the RVEC there is a formula normally found in the front or back of every instruction manual for a fluke meter  that will equate the RVEC to Ohmage, and if you have a 8 ohm speaker, normally you will see 7.6 ohms or close to this, a lttle under the rateing of a speaker and as the omage reading gets bigger the spread of difference in the reading of your meter will increase.
JD

[Lonestarjohnny]

Paul, what kinda speaker you wiring up, not an Altec by chance, if you are those old Dual voice coiled speakers had Isalation transformers on them, that would be a reason for the high resistance your seeing, you need the tranny in line with some of those old speaker's to get the ohmage to read correctly.
JD

[csj]

Yeah Paul, no problem...
you're not worried at this point about current really so we can forget your plate voltage and we'll assume, because it's the only thing you can really do, that your OT is 100% effecient. All you want to do is find out for the time being what the reflected load of the 12 ohm speaker is going to show on the primary. That's what your tube will be feeding output into.

Mr. Keen says a very important thing about transformers on his sight. He says they are "pure ratio". When one side does one thing the other will always follow based on the proportions of the ratio.

We want to find the impedance ratio. (Be thankful you know the impedances...wait till you've got the unknown, mystery OT) We're going to use this as the multiplier for or 12 ohm connection. All we know is the impedances. 5.6k and 8 ohm.

This is the ratio -

5600ohms/8ohms = 700 (no units)

so what gets reflected back to the primary is whatever the load impedance is... in your case now, the new load will be 12 ohms

mulitplied by the newly found impedance ratio.

700 X 12 ohms = 8400 ohms.

Your new load will reflect back an impedance to the primary which will now be 8400 ohms instead of 5600 ohms.

You can see immediatiely that one of the overall differences will be a current reduction. Assuming a 350v plate supply you'd reduce the current capacity (not factoring for bias or muEg) in the primary down from

I = 350v/5600 ohms = 62ma

to

I = 350v/8400 ohms = 42ma

This will, of course because of the "pure ratio" fact, cut the current down proportionately on the secondary. What do you get?
The proverbial power loss.

This along with the change in the overall frequency response because of the shifted impedances.

Is it a bad thing?.
To me, no.

I'd plug in my crappy tubes and let her rip.

(Please anyone, correct me for errors)

Take care,
Clay

[Paul Marossy]

Thanks csj.
That was a lot simpler than I thought it would be! I understand the ratio thing, the primary has X amount of turns and the secondary X amount of turns. You end up with a ratio. I just didn't know how you would calc something like that.  :oops:
Well, now I have a much better idea.

JD, no it's not an Altec. But maybe it needs to have an isolation transformer to properly wire it in series. I hadn't considered that...
If there is one on it, it must be internally mounted (and really small).

[Lonestarjohnny]

Paul, did I send you the Transformer Ratio Chart that I sent Kleber, it takes all this guessing out of what ratio you have, if you need the chart let me know, I'll be glad to send it to you.
JD

[Paul Marossy]

No, I haven't seen that chart JD. If you still have my email address, would you mind sending me a copy of it?

[csj]

Paul,
Of course, real transformer circuit analysis is, as we already know, incredibly complex and fascinating. I only wanted to set up a basic  static situation (one that assumed all other factors, other than the ones emphasized, were equal) so we could find some simple, firm ground. I know my example is a gross oversimplification but it should generally suffice.

[Paul Marossy]

Oh yeah, I know that it is far more complex than you made it out to be. A rough estimate is enough for me. That little exercise really helped me to see that playing through a speaker that is rated for less than the output transformer is kind of dangerous. That current can shoot way, way up!