Ascertaining speaker impedance without desoldering anything

[pbreyfogle]

I have an output transformer/speaker form an old hi-fi in my 5e3 and I would like to figure out the speaker impedance to build an L-pad for it. Seems like this can be done without desoldering anything:

V across pins 3 of the 6v6's: 416
V across speaker terminals: 7.4

so, turns ratio: 1:56
impedance ratio: 1:3160

If the typical 6v6 impedance is 6.5k then my speaker is about 2ohms?

Does that logic work? Does the input center tap have an effect I'm not accounting for?

[Satamax]

Hi Pbreyfogle.

Well iirc you don't account for the center tap, it just divides the ref imp that goes to each tube. But your numbers seem high, thought that may be nearing class A. Cos i doubt the speaker would be 2Ω. http://www.radioremembered.org/outimp.htm

[Enzo]

Are you not going to have to disconnect the wires between the transformer and the speaker to insert the L-Pad anyway?

All you would need to do is remove one of the two speaker wires anyway to measure. Is that really so much trouble?

[pbreyfogle]

Yeah, Satamax, I thought 2Ω was low as well; but the numbers work out so well, and I get the impression that back in the day speakers were more often 2Ω. I'll check the readings again and see what I get.

The speaker is a Capehart -though that may be the radio/record player brand since the rest of the unit is unmarked- if that is any indication.

Enzo, Given a multimeter (and, yes, a soldering iron) how would you go about finding speaker impedence on an unknown unit?

[Enzo]

I have been in audio for over 50 years, and while I have seen 2 ohm speakers, I cannot recall a time when they would have ever been called common. Table radios and small phonographs often used 3.2 ohm speakers, but real full sized speakers were always pretty much 4/8/16. Of course there are exceptions to most anything.

The reason we have to unsolder at least one wire from the speaker is that the transformer winding would otherwise be in parallel with the speaker voice coil. That transformer winding is quite likely to have a much lower resistance than the speaker coil.

Using an ohm meter, we are measuring the DC resistance of the voice coil rather than impedance. And impedance really is only a nominal rating since it varies widely with frequency. A general rule of thumb is that the DC resistance will usually be a little lower than the impedance. An 8 ohm speaker will typically measure 6-7 ohms, and a 4 ohm about 3. I often have wondered if those table radios had 4 ohm impedance speakers that measured 3.2 on a service man's meter and were thus described that way in the literature.

Or if you really want to get into it and have a couple hundred dollars to spend, go here and check stock number 72-6947:

www.mcminone.com

or even more serious, stock number 72-7420 and 72-7900

What those units are are AC ohm meters. Instead of DC current, they send out AC current at 1kHz or other freqs. The nominal impedance of a speaker is generally specified at 1kHz. The more basic one above is often on the monthly sale flyer about $20 cheaper.

[Steve Conner]

It really is easy. You just unsolder one of the wires from the speaker and use your multimeter to measure the resistance between the two terminals on the speaker.

Like Enzo said, the value you measure is a little less than what the nominal impedance will be. But I've never seen it low enough that you would mistake the impedance for the next lower value. So if you read 6 or 7 ohms, you can be pretty sure it's an 8 ohm speaker.

As an aside, I heard that horn loading can actually increase the impedance of a speaker a lot. An 8 ohm unit can end up 10-12 ohms.

[pbreyfogle]

Thanks

Thanks for the clear answers. I had no idea the impedence and the resistance were so close... I get 3.4 ohms.